diff --git a/checkpoints/checkpoint-1000/chat_template.jinja b/checkpoints/checkpoint-1000/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1000/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1000/config.json b/checkpoints/checkpoint-1000/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1000/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1000/generation_config.json b/checkpoints/checkpoint-1000/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1000/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1000/model.safetensors b/checkpoints/checkpoint-1000/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..16de07ff6e8fd9f8f76867325789c47db6063811
--- /dev/null
+++ b/checkpoints/checkpoint-1000/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b60b9721e8ef1f3891450933ec64142bc9cfe67de9222e6e595cbaf93615cc89
+size 37983421000
diff --git a/checkpoints/checkpoint-1000/tokenizer.json b/checkpoints/checkpoint-1000/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1000/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1000/tokenizer_config.json b/checkpoints/checkpoint-1000/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1000/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1000/trainer_state.json b/checkpoints/checkpoint-1000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..bd4a6f174ad568d089b536e4b4515171088383e6
--- /dev/null
+++ b/checkpoints/checkpoint-1000/trainer_state.json
@@ -0,0 +1,7034 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.5246589716684156,
+ "eval_steps": 500,
+ "global_step": 1000,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 2.6474172555198464e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1050/chat_template.jinja b/checkpoints/checkpoint-1050/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1050/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1050/config.json b/checkpoints/checkpoint-1050/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1050/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1050/generation_config.json b/checkpoints/checkpoint-1050/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1050/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1050/model.safetensors b/checkpoints/checkpoint-1050/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..3345309ba3de6fd75045026c9c35e886d4e2a385
--- /dev/null
+++ b/checkpoints/checkpoint-1050/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0af4b5c857bfa01ef912e3e33ffa17d4eba301aeac7bfa182d732e3c71456f7c
+size 37983421000
diff --git a/checkpoints/checkpoint-1050/tokenizer.json b/checkpoints/checkpoint-1050/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1050/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1050/tokenizer_config.json b/checkpoints/checkpoint-1050/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1050/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1050/trainer_state.json b/checkpoints/checkpoint-1050/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..ce4cb5dca9d3fe1166f49b5e570edb13149029c7
--- /dev/null
+++ b/checkpoints/checkpoint-1050/trainer_state.json
@@ -0,0 +1,7384 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.5508919202518363,
+ "eval_steps": 500,
+ "global_step": 1050,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 2.7807034424426496e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1100/chat_template.jinja b/checkpoints/checkpoint-1100/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1100/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1100/config.json b/checkpoints/checkpoint-1100/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1100/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1100/generation_config.json b/checkpoints/checkpoint-1100/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1100/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1100/model.safetensors b/checkpoints/checkpoint-1100/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..50034c01a58ffa8f24265f561689eecd0792e377
--- /dev/null
+++ b/checkpoints/checkpoint-1100/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5719b84236c151af82dbd75e10e2d4c39897920531bb242223570329c2bc7f29
+size 37983421000
diff --git a/checkpoints/checkpoint-1100/tokenizer.json b/checkpoints/checkpoint-1100/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1100/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1100/tokenizer_config.json b/checkpoints/checkpoint-1100/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1100/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1100/trainer_state.json b/checkpoints/checkpoint-1100/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..577f5ec7626595ae8e8fc9c8d2cd56dafc9b09b0
--- /dev/null
+++ b/checkpoints/checkpoint-1100/trainer_state.json
@@ -0,0 +1,7734 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.577124868835257,
+ "eval_steps": 500,
+ "global_step": 1100,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 2.913533231628288e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1150/chat_template.jinja b/checkpoints/checkpoint-1150/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1150/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1150/config.json b/checkpoints/checkpoint-1150/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1150/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1150/generation_config.json b/checkpoints/checkpoint-1150/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1150/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1150/model.safetensors b/checkpoints/checkpoint-1150/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..d347f31a788b7f2027e2edad6e9072e036983c89
--- /dev/null
+++ b/checkpoints/checkpoint-1150/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9b903d41c62ea1fcacaa9211f97521b9e7d61835a1b9bc3a262f17661159167d
+size 37983421000
diff --git a/checkpoints/checkpoint-1150/tokenizer.json b/checkpoints/checkpoint-1150/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1150/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1150/tokenizer_config.json b/checkpoints/checkpoint-1150/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1150/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1150/trainer_state.json b/checkpoints/checkpoint-1150/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..c26bad796dbad270e690cf48ec8813f87d993d17
--- /dev/null
+++ b/checkpoints/checkpoint-1150/trainer_state.json
@@ -0,0 +1,8084 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.6033578174186779,
+ "eval_steps": 500,
+ "global_step": 1150,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ },
+ {
+ "epoch": 0.5776495278069255,
+ "grad_norm": 0.1862053032854098,
+ "learning_rate": 2.2065553964033126e-05,
+ "loss": 0.4816875457763672,
+ "step": 1101
+ },
+ {
+ "epoch": 0.5781741867785939,
+ "grad_norm": 0.16996015894363117,
+ "learning_rate": 2.20295734045532e-05,
+ "loss": 0.4471197724342346,
+ "step": 1102
+ },
+ {
+ "epoch": 0.5786988457502623,
+ "grad_norm": 0.16889227008390692,
+ "learning_rate": 2.1993607706997725e-05,
+ "loss": 0.468391478061676,
+ "step": 1103
+ },
+ {
+ "epoch": 0.5792235047219307,
+ "grad_norm": 0.16670931681562373,
+ "learning_rate": 2.195765696907745e-05,
+ "loss": 0.44934162497520447,
+ "step": 1104
+ },
+ {
+ "epoch": 0.5797481636935992,
+ "grad_norm": 0.16328418741223716,
+ "learning_rate": 2.1921721288462482e-05,
+ "loss": 0.4558270275592804,
+ "step": 1105
+ },
+ {
+ "epoch": 0.5802728226652676,
+ "grad_norm": 0.17364660408703936,
+ "learning_rate": 2.188580076278203e-05,
+ "loss": 0.4597751796245575,
+ "step": 1106
+ },
+ {
+ "epoch": 0.580797481636936,
+ "grad_norm": 0.1628618453800601,
+ "learning_rate": 2.18498954896241e-05,
+ "loss": 0.4521859288215637,
+ "step": 1107
+ },
+ {
+ "epoch": 0.5813221406086044,
+ "grad_norm": 0.16813817951871482,
+ "learning_rate": 2.1814005566535302e-05,
+ "loss": 0.4581005573272705,
+ "step": 1108
+ },
+ {
+ "epoch": 0.5818467995802729,
+ "grad_norm": 0.16825923629726025,
+ "learning_rate": 2.1778131091020516e-05,
+ "loss": 0.4433916211128235,
+ "step": 1109
+ },
+ {
+ "epoch": 0.5823714585519413,
+ "grad_norm": 0.17194743467883944,
+ "learning_rate": 2.1742272160542654e-05,
+ "loss": 0.47296425700187683,
+ "step": 1110
+ },
+ {
+ "epoch": 0.5828961175236097,
+ "grad_norm": 0.16947730546039969,
+ "learning_rate": 2.170642887252241e-05,
+ "loss": 0.4815937876701355,
+ "step": 1111
+ },
+ {
+ "epoch": 0.583420776495278,
+ "grad_norm": 0.1657449928668473,
+ "learning_rate": 2.167060132433798e-05,
+ "loss": 0.4479164779186249,
+ "step": 1112
+ },
+ {
+ "epoch": 0.5839454354669464,
+ "grad_norm": 0.1658952516607364,
+ "learning_rate": 2.1634789613324778e-05,
+ "loss": 0.4637463688850403,
+ "step": 1113
+ },
+ {
+ "epoch": 0.5844700944386149,
+ "grad_norm": 0.16777002188998236,
+ "learning_rate": 2.1598993836775218e-05,
+ "loss": 0.457524836063385,
+ "step": 1114
+ },
+ {
+ "epoch": 0.5849947534102833,
+ "grad_norm": 0.15837566447341322,
+ "learning_rate": 2.1563214091938405e-05,
+ "loss": 0.44938087463378906,
+ "step": 1115
+ },
+ {
+ "epoch": 0.5855194123819517,
+ "grad_norm": 0.17162206617035314,
+ "learning_rate": 2.1527450476019896e-05,
+ "loss": 0.4393450617790222,
+ "step": 1116
+ },
+ {
+ "epoch": 0.5860440713536201,
+ "grad_norm": 0.16224101921433814,
+ "learning_rate": 2.1491703086181426e-05,
+ "loss": 0.451913982629776,
+ "step": 1117
+ },
+ {
+ "epoch": 0.5865687303252886,
+ "grad_norm": 0.1736253504575144,
+ "learning_rate": 2.145597201954066e-05,
+ "loss": 0.4611290395259857,
+ "step": 1118
+ },
+ {
+ "epoch": 0.587093389296957,
+ "grad_norm": 0.16308408333737234,
+ "learning_rate": 2.1420257373170903e-05,
+ "loss": 0.45898568630218506,
+ "step": 1119
+ },
+ {
+ "epoch": 0.5876180482686254,
+ "grad_norm": 0.16948634794175524,
+ "learning_rate": 2.1384559244100848e-05,
+ "loss": 0.45982882380485535,
+ "step": 1120
+ },
+ {
+ "epoch": 0.5881427072402938,
+ "grad_norm": 0.16676366538565252,
+ "learning_rate": 2.1348877729314336e-05,
+ "loss": 0.44826576113700867,
+ "step": 1121
+ },
+ {
+ "epoch": 0.5886673662119623,
+ "grad_norm": 0.16431125630712295,
+ "learning_rate": 2.1313212925750037e-05,
+ "loss": 0.4616445302963257,
+ "step": 1122
+ },
+ {
+ "epoch": 0.5891920251836307,
+ "grad_norm": 0.1631659061616955,
+ "learning_rate": 2.1277564930301252e-05,
+ "loss": 0.45651188492774963,
+ "step": 1123
+ },
+ {
+ "epoch": 0.5897166841552991,
+ "grad_norm": 0.15805950502689237,
+ "learning_rate": 2.1241933839815604e-05,
+ "loss": 0.4486045837402344,
+ "step": 1124
+ },
+ {
+ "epoch": 0.5902413431269675,
+ "grad_norm": 0.16530288038603502,
+ "learning_rate": 2.1206319751094773e-05,
+ "loss": 0.4485066533088684,
+ "step": 1125
+ },
+ {
+ "epoch": 0.5907660020986358,
+ "grad_norm": 0.16061903194239047,
+ "learning_rate": 2.1170722760894295e-05,
+ "loss": 0.4565330743789673,
+ "step": 1126
+ },
+ {
+ "epoch": 0.5912906610703043,
+ "grad_norm": 0.1643346999353627,
+ "learning_rate": 2.1135142965923195e-05,
+ "loss": 0.45560717582702637,
+ "step": 1127
+ },
+ {
+ "epoch": 0.5918153200419727,
+ "grad_norm": 0.17484644315215753,
+ "learning_rate": 2.109958046284382e-05,
+ "loss": 0.4795891046524048,
+ "step": 1128
+ },
+ {
+ "epoch": 0.5923399790136411,
+ "grad_norm": 0.1623762678622651,
+ "learning_rate": 2.1064035348271528e-05,
+ "loss": 0.47395604848861694,
+ "step": 1129
+ },
+ {
+ "epoch": 0.5928646379853095,
+ "grad_norm": 0.16888824241669015,
+ "learning_rate": 2.1028507718774456e-05,
+ "loss": 0.4601455330848694,
+ "step": 1130
+ },
+ {
+ "epoch": 0.593389296956978,
+ "grad_norm": 0.16291744568806302,
+ "learning_rate": 2.0992997670873193e-05,
+ "loss": 0.46248236298561096,
+ "step": 1131
+ },
+ {
+ "epoch": 0.5939139559286464,
+ "grad_norm": 0.16185598691236988,
+ "learning_rate": 2.0957505301040603e-05,
+ "loss": 0.45085301995277405,
+ "step": 1132
+ },
+ {
+ "epoch": 0.5944386149003148,
+ "grad_norm": 0.1678000806117814,
+ "learning_rate": 2.0922030705701518e-05,
+ "loss": 0.44939014315605164,
+ "step": 1133
+ },
+ {
+ "epoch": 0.5949632738719832,
+ "grad_norm": 0.16104474282288994,
+ "learning_rate": 2.088657398123246e-05,
+ "loss": 0.4313976764678955,
+ "step": 1134
+ },
+ {
+ "epoch": 0.5954879328436516,
+ "grad_norm": 0.1631128968199669,
+ "learning_rate": 2.085113522396141e-05,
+ "loss": 0.4508278965950012,
+ "step": 1135
+ },
+ {
+ "epoch": 0.5960125918153201,
+ "grad_norm": 0.16307688195844533,
+ "learning_rate": 2.0815714530167557e-05,
+ "loss": 0.4467576742172241,
+ "step": 1136
+ },
+ {
+ "epoch": 0.5965372507869885,
+ "grad_norm": 0.16456832304532207,
+ "learning_rate": 2.0780311996080976e-05,
+ "loss": 0.4621703624725342,
+ "step": 1137
+ },
+ {
+ "epoch": 0.5970619097586569,
+ "grad_norm": 0.16346217748325367,
+ "learning_rate": 2.0744927717882435e-05,
+ "loss": 0.4589449465274811,
+ "step": 1138
+ },
+ {
+ "epoch": 0.5975865687303253,
+ "grad_norm": 0.16180691501076994,
+ "learning_rate": 2.0709561791703097e-05,
+ "loss": 0.43698400259017944,
+ "step": 1139
+ },
+ {
+ "epoch": 0.5981112277019937,
+ "grad_norm": 0.15695459976898835,
+ "learning_rate": 2.067421431362427e-05,
+ "loss": 0.452271044254303,
+ "step": 1140
+ },
+ {
+ "epoch": 0.5986358866736621,
+ "grad_norm": 0.15930893553721054,
+ "learning_rate": 2.063888537967713e-05,
+ "loss": 0.44868963956832886,
+ "step": 1141
+ },
+ {
+ "epoch": 0.5991605456453305,
+ "grad_norm": 0.15963802284017928,
+ "learning_rate": 2.0603575085842485e-05,
+ "loss": 0.4376864731311798,
+ "step": 1142
+ },
+ {
+ "epoch": 0.5996852046169989,
+ "grad_norm": 0.16610266900507178,
+ "learning_rate": 2.05682835280505e-05,
+ "loss": 0.44676727056503296,
+ "step": 1143
+ },
+ {
+ "epoch": 0.6002098635886673,
+ "grad_norm": 0.15978253236669612,
+ "learning_rate": 2.053301080218044e-05,
+ "loss": 0.434773713350296,
+ "step": 1144
+ },
+ {
+ "epoch": 0.6007345225603358,
+ "grad_norm": 0.16305141803526324,
+ "learning_rate": 2.0497757004060403e-05,
+ "loss": 0.4529780149459839,
+ "step": 1145
+ },
+ {
+ "epoch": 0.6012591815320042,
+ "grad_norm": 0.1649737864101215,
+ "learning_rate": 2.046252222946707e-05,
+ "loss": 0.4571434259414673,
+ "step": 1146
+ },
+ {
+ "epoch": 0.6017838405036726,
+ "grad_norm": 0.15738824968326598,
+ "learning_rate": 2.0427306574125427e-05,
+ "loss": 0.4446289539337158,
+ "step": 1147
+ },
+ {
+ "epoch": 0.602308499475341,
+ "grad_norm": 0.1731139535185187,
+ "learning_rate": 2.039211013370853e-05,
+ "loss": 0.4551090598106384,
+ "step": 1148
+ },
+ {
+ "epoch": 0.6028331584470095,
+ "grad_norm": 0.15857251900043037,
+ "learning_rate": 2.0356933003837237e-05,
+ "loss": 0.4644010663032532,
+ "step": 1149
+ },
+ {
+ "epoch": 0.6033578174186779,
+ "grad_norm": 0.1589582861134211,
+ "learning_rate": 2.032177528007992e-05,
+ "loss": 0.45741650462150574,
+ "step": 1150
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 3.0452041839017984e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1200/chat_template.jinja b/checkpoints/checkpoint-1200/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1200/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1200/config.json b/checkpoints/checkpoint-1200/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1200/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1200/generation_config.json b/checkpoints/checkpoint-1200/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1200/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1200/model.safetensors b/checkpoints/checkpoint-1200/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..21c82cc0481b4f7084ce59281d14a24e66078e33
--- /dev/null
+++ b/checkpoints/checkpoint-1200/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b0006208cea4cd9220c7db90badede83ad4ac9516dc0ed0296d25e5b33fdf16
+size 37983421000
diff --git a/checkpoints/checkpoint-1200/tokenizer.json b/checkpoints/checkpoint-1200/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1200/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1200/tokenizer_config.json b/checkpoints/checkpoint-1200/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1200/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1200/trainer_state.json b/checkpoints/checkpoint-1200/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..bf744dea340766d5e0f279bb0733354e7d91d105
--- /dev/null
+++ b/checkpoints/checkpoint-1200/trainer_state.json
@@ -0,0 +1,8434 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.6295907660020986,
+ "eval_steps": 500,
+ "global_step": 1200,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ },
+ {
+ "epoch": 0.5776495278069255,
+ "grad_norm": 0.1862053032854098,
+ "learning_rate": 2.2065553964033126e-05,
+ "loss": 0.4816875457763672,
+ "step": 1101
+ },
+ {
+ "epoch": 0.5781741867785939,
+ "grad_norm": 0.16996015894363117,
+ "learning_rate": 2.20295734045532e-05,
+ "loss": 0.4471197724342346,
+ "step": 1102
+ },
+ {
+ "epoch": 0.5786988457502623,
+ "grad_norm": 0.16889227008390692,
+ "learning_rate": 2.1993607706997725e-05,
+ "loss": 0.468391478061676,
+ "step": 1103
+ },
+ {
+ "epoch": 0.5792235047219307,
+ "grad_norm": 0.16670931681562373,
+ "learning_rate": 2.195765696907745e-05,
+ "loss": 0.44934162497520447,
+ "step": 1104
+ },
+ {
+ "epoch": 0.5797481636935992,
+ "grad_norm": 0.16328418741223716,
+ "learning_rate": 2.1921721288462482e-05,
+ "loss": 0.4558270275592804,
+ "step": 1105
+ },
+ {
+ "epoch": 0.5802728226652676,
+ "grad_norm": 0.17364660408703936,
+ "learning_rate": 2.188580076278203e-05,
+ "loss": 0.4597751796245575,
+ "step": 1106
+ },
+ {
+ "epoch": 0.580797481636936,
+ "grad_norm": 0.1628618453800601,
+ "learning_rate": 2.18498954896241e-05,
+ "loss": 0.4521859288215637,
+ "step": 1107
+ },
+ {
+ "epoch": 0.5813221406086044,
+ "grad_norm": 0.16813817951871482,
+ "learning_rate": 2.1814005566535302e-05,
+ "loss": 0.4581005573272705,
+ "step": 1108
+ },
+ {
+ "epoch": 0.5818467995802729,
+ "grad_norm": 0.16825923629726025,
+ "learning_rate": 2.1778131091020516e-05,
+ "loss": 0.4433916211128235,
+ "step": 1109
+ },
+ {
+ "epoch": 0.5823714585519413,
+ "grad_norm": 0.17194743467883944,
+ "learning_rate": 2.1742272160542654e-05,
+ "loss": 0.47296425700187683,
+ "step": 1110
+ },
+ {
+ "epoch": 0.5828961175236097,
+ "grad_norm": 0.16947730546039969,
+ "learning_rate": 2.170642887252241e-05,
+ "loss": 0.4815937876701355,
+ "step": 1111
+ },
+ {
+ "epoch": 0.583420776495278,
+ "grad_norm": 0.1657449928668473,
+ "learning_rate": 2.167060132433798e-05,
+ "loss": 0.4479164779186249,
+ "step": 1112
+ },
+ {
+ "epoch": 0.5839454354669464,
+ "grad_norm": 0.1658952516607364,
+ "learning_rate": 2.1634789613324778e-05,
+ "loss": 0.4637463688850403,
+ "step": 1113
+ },
+ {
+ "epoch": 0.5844700944386149,
+ "grad_norm": 0.16777002188998236,
+ "learning_rate": 2.1598993836775218e-05,
+ "loss": 0.457524836063385,
+ "step": 1114
+ },
+ {
+ "epoch": 0.5849947534102833,
+ "grad_norm": 0.15837566447341322,
+ "learning_rate": 2.1563214091938405e-05,
+ "loss": 0.44938087463378906,
+ "step": 1115
+ },
+ {
+ "epoch": 0.5855194123819517,
+ "grad_norm": 0.17162206617035314,
+ "learning_rate": 2.1527450476019896e-05,
+ "loss": 0.4393450617790222,
+ "step": 1116
+ },
+ {
+ "epoch": 0.5860440713536201,
+ "grad_norm": 0.16224101921433814,
+ "learning_rate": 2.1491703086181426e-05,
+ "loss": 0.451913982629776,
+ "step": 1117
+ },
+ {
+ "epoch": 0.5865687303252886,
+ "grad_norm": 0.1736253504575144,
+ "learning_rate": 2.145597201954066e-05,
+ "loss": 0.4611290395259857,
+ "step": 1118
+ },
+ {
+ "epoch": 0.587093389296957,
+ "grad_norm": 0.16308408333737234,
+ "learning_rate": 2.1420257373170903e-05,
+ "loss": 0.45898568630218506,
+ "step": 1119
+ },
+ {
+ "epoch": 0.5876180482686254,
+ "grad_norm": 0.16948634794175524,
+ "learning_rate": 2.1384559244100848e-05,
+ "loss": 0.45982882380485535,
+ "step": 1120
+ },
+ {
+ "epoch": 0.5881427072402938,
+ "grad_norm": 0.16676366538565252,
+ "learning_rate": 2.1348877729314336e-05,
+ "loss": 0.44826576113700867,
+ "step": 1121
+ },
+ {
+ "epoch": 0.5886673662119623,
+ "grad_norm": 0.16431125630712295,
+ "learning_rate": 2.1313212925750037e-05,
+ "loss": 0.4616445302963257,
+ "step": 1122
+ },
+ {
+ "epoch": 0.5891920251836307,
+ "grad_norm": 0.1631659061616955,
+ "learning_rate": 2.1277564930301252e-05,
+ "loss": 0.45651188492774963,
+ "step": 1123
+ },
+ {
+ "epoch": 0.5897166841552991,
+ "grad_norm": 0.15805950502689237,
+ "learning_rate": 2.1241933839815604e-05,
+ "loss": 0.4486045837402344,
+ "step": 1124
+ },
+ {
+ "epoch": 0.5902413431269675,
+ "grad_norm": 0.16530288038603502,
+ "learning_rate": 2.1206319751094773e-05,
+ "loss": 0.4485066533088684,
+ "step": 1125
+ },
+ {
+ "epoch": 0.5907660020986358,
+ "grad_norm": 0.16061903194239047,
+ "learning_rate": 2.1170722760894295e-05,
+ "loss": 0.4565330743789673,
+ "step": 1126
+ },
+ {
+ "epoch": 0.5912906610703043,
+ "grad_norm": 0.1643346999353627,
+ "learning_rate": 2.1135142965923195e-05,
+ "loss": 0.45560717582702637,
+ "step": 1127
+ },
+ {
+ "epoch": 0.5918153200419727,
+ "grad_norm": 0.17484644315215753,
+ "learning_rate": 2.109958046284382e-05,
+ "loss": 0.4795891046524048,
+ "step": 1128
+ },
+ {
+ "epoch": 0.5923399790136411,
+ "grad_norm": 0.1623762678622651,
+ "learning_rate": 2.1064035348271528e-05,
+ "loss": 0.47395604848861694,
+ "step": 1129
+ },
+ {
+ "epoch": 0.5928646379853095,
+ "grad_norm": 0.16888824241669015,
+ "learning_rate": 2.1028507718774456e-05,
+ "loss": 0.4601455330848694,
+ "step": 1130
+ },
+ {
+ "epoch": 0.593389296956978,
+ "grad_norm": 0.16291744568806302,
+ "learning_rate": 2.0992997670873193e-05,
+ "loss": 0.46248236298561096,
+ "step": 1131
+ },
+ {
+ "epoch": 0.5939139559286464,
+ "grad_norm": 0.16185598691236988,
+ "learning_rate": 2.0957505301040603e-05,
+ "loss": 0.45085301995277405,
+ "step": 1132
+ },
+ {
+ "epoch": 0.5944386149003148,
+ "grad_norm": 0.1678000806117814,
+ "learning_rate": 2.0922030705701518e-05,
+ "loss": 0.44939014315605164,
+ "step": 1133
+ },
+ {
+ "epoch": 0.5949632738719832,
+ "grad_norm": 0.16104474282288994,
+ "learning_rate": 2.088657398123246e-05,
+ "loss": 0.4313976764678955,
+ "step": 1134
+ },
+ {
+ "epoch": 0.5954879328436516,
+ "grad_norm": 0.1631128968199669,
+ "learning_rate": 2.085113522396141e-05,
+ "loss": 0.4508278965950012,
+ "step": 1135
+ },
+ {
+ "epoch": 0.5960125918153201,
+ "grad_norm": 0.16307688195844533,
+ "learning_rate": 2.0815714530167557e-05,
+ "loss": 0.4467576742172241,
+ "step": 1136
+ },
+ {
+ "epoch": 0.5965372507869885,
+ "grad_norm": 0.16456832304532207,
+ "learning_rate": 2.0780311996080976e-05,
+ "loss": 0.4621703624725342,
+ "step": 1137
+ },
+ {
+ "epoch": 0.5970619097586569,
+ "grad_norm": 0.16346217748325367,
+ "learning_rate": 2.0744927717882435e-05,
+ "loss": 0.4589449465274811,
+ "step": 1138
+ },
+ {
+ "epoch": 0.5975865687303253,
+ "grad_norm": 0.16180691501076994,
+ "learning_rate": 2.0709561791703097e-05,
+ "loss": 0.43698400259017944,
+ "step": 1139
+ },
+ {
+ "epoch": 0.5981112277019937,
+ "grad_norm": 0.15695459976898835,
+ "learning_rate": 2.067421431362427e-05,
+ "loss": 0.452271044254303,
+ "step": 1140
+ },
+ {
+ "epoch": 0.5986358866736621,
+ "grad_norm": 0.15930893553721054,
+ "learning_rate": 2.063888537967713e-05,
+ "loss": 0.44868963956832886,
+ "step": 1141
+ },
+ {
+ "epoch": 0.5991605456453305,
+ "grad_norm": 0.15963802284017928,
+ "learning_rate": 2.0603575085842485e-05,
+ "loss": 0.4376864731311798,
+ "step": 1142
+ },
+ {
+ "epoch": 0.5996852046169989,
+ "grad_norm": 0.16610266900507178,
+ "learning_rate": 2.05682835280505e-05,
+ "loss": 0.44676727056503296,
+ "step": 1143
+ },
+ {
+ "epoch": 0.6002098635886673,
+ "grad_norm": 0.15978253236669612,
+ "learning_rate": 2.053301080218044e-05,
+ "loss": 0.434773713350296,
+ "step": 1144
+ },
+ {
+ "epoch": 0.6007345225603358,
+ "grad_norm": 0.16305141803526324,
+ "learning_rate": 2.0497757004060403e-05,
+ "loss": 0.4529780149459839,
+ "step": 1145
+ },
+ {
+ "epoch": 0.6012591815320042,
+ "grad_norm": 0.1649737864101215,
+ "learning_rate": 2.046252222946707e-05,
+ "loss": 0.4571434259414673,
+ "step": 1146
+ },
+ {
+ "epoch": 0.6017838405036726,
+ "grad_norm": 0.15738824968326598,
+ "learning_rate": 2.0427306574125427e-05,
+ "loss": 0.4446289539337158,
+ "step": 1147
+ },
+ {
+ "epoch": 0.602308499475341,
+ "grad_norm": 0.1731139535185187,
+ "learning_rate": 2.039211013370853e-05,
+ "loss": 0.4551090598106384,
+ "step": 1148
+ },
+ {
+ "epoch": 0.6028331584470095,
+ "grad_norm": 0.15857251900043037,
+ "learning_rate": 2.0356933003837237e-05,
+ "loss": 0.4644010663032532,
+ "step": 1149
+ },
+ {
+ "epoch": 0.6033578174186779,
+ "grad_norm": 0.1589582861134211,
+ "learning_rate": 2.032177528007992e-05,
+ "loss": 0.45741650462150574,
+ "step": 1150
+ },
+ {
+ "epoch": 0.6038824763903463,
+ "grad_norm": 0.15639092706575122,
+ "learning_rate": 2.0286637057952257e-05,
+ "loss": 0.4375547766685486,
+ "step": 1151
+ },
+ {
+ "epoch": 0.6044071353620147,
+ "grad_norm": 0.16335605581337187,
+ "learning_rate": 2.0251518432916933e-05,
+ "loss": 0.4636140465736389,
+ "step": 1152
+ },
+ {
+ "epoch": 0.6049317943336832,
+ "grad_norm": 0.15437437113359356,
+ "learning_rate": 2.0216419500383372e-05,
+ "loss": 0.44632142782211304,
+ "step": 1153
+ },
+ {
+ "epoch": 0.6054564533053515,
+ "grad_norm": 0.15522039198698664,
+ "learning_rate": 2.0181340355707545e-05,
+ "loss": 0.449049174785614,
+ "step": 1154
+ },
+ {
+ "epoch": 0.6059811122770199,
+ "grad_norm": 0.1599951917909323,
+ "learning_rate": 2.0146281094191604e-05,
+ "loss": 0.44395962357521057,
+ "step": 1155
+ },
+ {
+ "epoch": 0.6065057712486883,
+ "grad_norm": 0.16866621620527716,
+ "learning_rate": 2.011124181108373e-05,
+ "loss": 0.45693716406822205,
+ "step": 1156
+ },
+ {
+ "epoch": 0.6070304302203567,
+ "grad_norm": 0.15508935650135075,
+ "learning_rate": 2.0076222601577815e-05,
+ "loss": 0.4525151252746582,
+ "step": 1157
+ },
+ {
+ "epoch": 0.6075550891920252,
+ "grad_norm": 0.16508998118005971,
+ "learning_rate": 2.004122356081321e-05,
+ "loss": 0.466006875038147,
+ "step": 1158
+ },
+ {
+ "epoch": 0.6080797481636936,
+ "grad_norm": 0.15817418864022237,
+ "learning_rate": 2.0006244783874466e-05,
+ "loss": 0.4496110677719116,
+ "step": 1159
+ },
+ {
+ "epoch": 0.608604407135362,
+ "grad_norm": 0.16127270259749824,
+ "learning_rate": 1.997128636579109e-05,
+ "loss": 0.45918628573417664,
+ "step": 1160
+ },
+ {
+ "epoch": 0.6091290661070304,
+ "grad_norm": 0.1702599924159677,
+ "learning_rate": 1.9936348401537292e-05,
+ "loss": 0.47218555212020874,
+ "step": 1161
+ },
+ {
+ "epoch": 0.6096537250786989,
+ "grad_norm": 0.15690176248846127,
+ "learning_rate": 1.9901430986031695e-05,
+ "loss": 0.4516097903251648,
+ "step": 1162
+ },
+ {
+ "epoch": 0.6101783840503673,
+ "grad_norm": 0.1713614405614144,
+ "learning_rate": 1.9866534214137085e-05,
+ "loss": 0.45030856132507324,
+ "step": 1163
+ },
+ {
+ "epoch": 0.6107030430220357,
+ "grad_norm": 0.1670865563026229,
+ "learning_rate": 1.9831658180660195e-05,
+ "loss": 0.45603489875793457,
+ "step": 1164
+ },
+ {
+ "epoch": 0.6112277019937041,
+ "grad_norm": 0.16897539113401952,
+ "learning_rate": 1.9796802980351392e-05,
+ "loss": 0.46314537525177,
+ "step": 1165
+ },
+ {
+ "epoch": 0.6117523609653726,
+ "grad_norm": 0.15832868870697644,
+ "learning_rate": 1.976196870790446e-05,
+ "loss": 0.44784611463546753,
+ "step": 1166
+ },
+ {
+ "epoch": 0.612277019937041,
+ "grad_norm": 0.16595365551891114,
+ "learning_rate": 1.9727155457956315e-05,
+ "loss": 0.4409624934196472,
+ "step": 1167
+ },
+ {
+ "epoch": 0.6128016789087093,
+ "grad_norm": 0.1642630411212017,
+ "learning_rate": 1.969236332508677e-05,
+ "loss": 0.46120351552963257,
+ "step": 1168
+ },
+ {
+ "epoch": 0.6133263378803777,
+ "grad_norm": 0.16481796135868812,
+ "learning_rate": 1.9657592403818253e-05,
+ "loss": 0.45544904470443726,
+ "step": 1169
+ },
+ {
+ "epoch": 0.6138509968520461,
+ "grad_norm": 0.1629832779852364,
+ "learning_rate": 1.962284278861557e-05,
+ "loss": 0.4578462839126587,
+ "step": 1170
+ },
+ {
+ "epoch": 0.6143756558237146,
+ "grad_norm": 0.15916219613603688,
+ "learning_rate": 1.9588114573885664e-05,
+ "loss": 0.4546983540058136,
+ "step": 1171
+ },
+ {
+ "epoch": 0.614900314795383,
+ "grad_norm": 0.1660520165166075,
+ "learning_rate": 1.95534078539773e-05,
+ "loss": 0.4601942300796509,
+ "step": 1172
+ },
+ {
+ "epoch": 0.6154249737670514,
+ "grad_norm": 0.16924794578577768,
+ "learning_rate": 1.9518722723180877e-05,
+ "loss": 0.45997846126556396,
+ "step": 1173
+ },
+ {
+ "epoch": 0.6159496327387198,
+ "grad_norm": 0.16597646133045382,
+ "learning_rate": 1.9484059275728127e-05,
+ "loss": 0.4594605565071106,
+ "step": 1174
+ },
+ {
+ "epoch": 0.6164742917103883,
+ "grad_norm": 0.16964243792519748,
+ "learning_rate": 1.9449417605791876e-05,
+ "loss": 0.4638690650463104,
+ "step": 1175
+ },
+ {
+ "epoch": 0.6169989506820567,
+ "grad_norm": 0.1758591426791696,
+ "learning_rate": 1.9414797807485796e-05,
+ "loss": 0.4666447639465332,
+ "step": 1176
+ },
+ {
+ "epoch": 0.6175236096537251,
+ "grad_norm": 0.16217661227891753,
+ "learning_rate": 1.9380199974864126e-05,
+ "loss": 0.44252535700798035,
+ "step": 1177
+ },
+ {
+ "epoch": 0.6180482686253935,
+ "grad_norm": 0.1576446184499019,
+ "learning_rate": 1.934562420192142e-05,
+ "loss": 0.4499641954898834,
+ "step": 1178
+ },
+ {
+ "epoch": 0.618572927597062,
+ "grad_norm": 0.16231008092636912,
+ "learning_rate": 1.9311070582592327e-05,
+ "loss": 0.4561542570590973,
+ "step": 1179
+ },
+ {
+ "epoch": 0.6190975865687304,
+ "grad_norm": 0.16790382249577043,
+ "learning_rate": 1.9276539210751292e-05,
+ "loss": 0.4484325647354126,
+ "step": 1180
+ },
+ {
+ "epoch": 0.6196222455403987,
+ "grad_norm": 0.18794729932605855,
+ "learning_rate": 1.9242030180212323e-05,
+ "loss": 0.44169026613235474,
+ "step": 1181
+ },
+ {
+ "epoch": 0.6201469045120671,
+ "grad_norm": 0.16948051243889814,
+ "learning_rate": 1.920754358472872e-05,
+ "loss": 0.4592876434326172,
+ "step": 1182
+ },
+ {
+ "epoch": 0.6206715634837355,
+ "grad_norm": 0.16762558710029374,
+ "learning_rate": 1.917307951799286e-05,
+ "loss": 0.4402095079421997,
+ "step": 1183
+ },
+ {
+ "epoch": 0.621196222455404,
+ "grad_norm": 0.18170280116387003,
+ "learning_rate": 1.9138638073635885e-05,
+ "loss": 0.4633026123046875,
+ "step": 1184
+ },
+ {
+ "epoch": 0.6217208814270724,
+ "grad_norm": 0.1718051709136108,
+ "learning_rate": 1.9104219345227486e-05,
+ "loss": 0.4510066509246826,
+ "step": 1185
+ },
+ {
+ "epoch": 0.6222455403987408,
+ "grad_norm": 0.16829066326345576,
+ "learning_rate": 1.906982342627565e-05,
+ "loss": 0.4632919430732727,
+ "step": 1186
+ },
+ {
+ "epoch": 0.6227701993704092,
+ "grad_norm": 0.16669141808548274,
+ "learning_rate": 1.9035450410226395e-05,
+ "loss": 0.455544650554657,
+ "step": 1187
+ },
+ {
+ "epoch": 0.6232948583420777,
+ "grad_norm": 0.16679996559722232,
+ "learning_rate": 1.90011003904635e-05,
+ "loss": 0.459994375705719,
+ "step": 1188
+ },
+ {
+ "epoch": 0.6238195173137461,
+ "grad_norm": 0.17379754713278164,
+ "learning_rate": 1.8966773460308284e-05,
+ "loss": 0.4576749801635742,
+ "step": 1189
+ },
+ {
+ "epoch": 0.6243441762854145,
+ "grad_norm": 0.17088539733389246,
+ "learning_rate": 1.8932469713019335e-05,
+ "loss": 0.44853052496910095,
+ "step": 1190
+ },
+ {
+ "epoch": 0.6248688352570829,
+ "grad_norm": 0.15556883889201015,
+ "learning_rate": 1.8898189241792248e-05,
+ "loss": 0.44060447812080383,
+ "step": 1191
+ },
+ {
+ "epoch": 0.6253934942287513,
+ "grad_norm": 0.1640093917439215,
+ "learning_rate": 1.8863932139759388e-05,
+ "loss": 0.4531911611557007,
+ "step": 1192
+ },
+ {
+ "epoch": 0.6259181532004198,
+ "grad_norm": 0.16444284826591576,
+ "learning_rate": 1.882969849998964e-05,
+ "loss": 0.44176363945007324,
+ "step": 1193
+ },
+ {
+ "epoch": 0.6264428121720882,
+ "grad_norm": 0.1696462520995642,
+ "learning_rate": 1.8795488415488143e-05,
+ "loss": 0.44516652822494507,
+ "step": 1194
+ },
+ {
+ "epoch": 0.6269674711437565,
+ "grad_norm": 0.16810724705191796,
+ "learning_rate": 1.8761301979196032e-05,
+ "loss": 0.4542563855648041,
+ "step": 1195
+ },
+ {
+ "epoch": 0.6274921301154249,
+ "grad_norm": 0.17334813313855824,
+ "learning_rate": 1.8727139283990204e-05,
+ "loss": 0.47769248485565186,
+ "step": 1196
+ },
+ {
+ "epoch": 0.6280167890870934,
+ "grad_norm": 0.16308212035303235,
+ "learning_rate": 1.8693000422683043e-05,
+ "loss": 0.4407799243927002,
+ "step": 1197
+ },
+ {
+ "epoch": 0.6285414480587618,
+ "grad_norm": 0.16853319139928977,
+ "learning_rate": 1.8658885488022204e-05,
+ "loss": 0.4565497040748596,
+ "step": 1198
+ },
+ {
+ "epoch": 0.6290661070304302,
+ "grad_norm": 0.16694968781284703,
+ "learning_rate": 1.8624794572690328e-05,
+ "loss": 0.44522374868392944,
+ "step": 1199
+ },
+ {
+ "epoch": 0.6295907660020986,
+ "grad_norm": 0.16418579468339103,
+ "learning_rate": 1.8590727769304788e-05,
+ "loss": 0.4441048502922058,
+ "step": 1200
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 3.176764234583245e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1250/chat_template.jinja b/checkpoints/checkpoint-1250/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1250/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1250/config.json b/checkpoints/checkpoint-1250/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1250/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1250/generation_config.json b/checkpoints/checkpoint-1250/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1250/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1250/model.safetensors b/checkpoints/checkpoint-1250/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..6f2be952569226ac8445df192ca0492511fa5196
--- /dev/null
+++ b/checkpoints/checkpoint-1250/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c2c80a9ace96fcf0f17215405c55527d064e929770dd74ef9ff3bed955f2358
+size 37983421000
diff --git a/checkpoints/checkpoint-1250/tokenizer.json b/checkpoints/checkpoint-1250/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1250/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1250/tokenizer_config.json b/checkpoints/checkpoint-1250/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1250/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1250/trainer_state.json b/checkpoints/checkpoint-1250/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..0f22f6d120fe16233ceb475459e29668e514c0ab
--- /dev/null
+++ b/checkpoints/checkpoint-1250/trainer_state.json
@@ -0,0 +1,8784 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.6558237145855194,
+ "eval_steps": 500,
+ "global_step": 1250,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ },
+ {
+ "epoch": 0.5776495278069255,
+ "grad_norm": 0.1862053032854098,
+ "learning_rate": 2.2065553964033126e-05,
+ "loss": 0.4816875457763672,
+ "step": 1101
+ },
+ {
+ "epoch": 0.5781741867785939,
+ "grad_norm": 0.16996015894363117,
+ "learning_rate": 2.20295734045532e-05,
+ "loss": 0.4471197724342346,
+ "step": 1102
+ },
+ {
+ "epoch": 0.5786988457502623,
+ "grad_norm": 0.16889227008390692,
+ "learning_rate": 2.1993607706997725e-05,
+ "loss": 0.468391478061676,
+ "step": 1103
+ },
+ {
+ "epoch": 0.5792235047219307,
+ "grad_norm": 0.16670931681562373,
+ "learning_rate": 2.195765696907745e-05,
+ "loss": 0.44934162497520447,
+ "step": 1104
+ },
+ {
+ "epoch": 0.5797481636935992,
+ "grad_norm": 0.16328418741223716,
+ "learning_rate": 2.1921721288462482e-05,
+ "loss": 0.4558270275592804,
+ "step": 1105
+ },
+ {
+ "epoch": 0.5802728226652676,
+ "grad_norm": 0.17364660408703936,
+ "learning_rate": 2.188580076278203e-05,
+ "loss": 0.4597751796245575,
+ "step": 1106
+ },
+ {
+ "epoch": 0.580797481636936,
+ "grad_norm": 0.1628618453800601,
+ "learning_rate": 2.18498954896241e-05,
+ "loss": 0.4521859288215637,
+ "step": 1107
+ },
+ {
+ "epoch": 0.5813221406086044,
+ "grad_norm": 0.16813817951871482,
+ "learning_rate": 2.1814005566535302e-05,
+ "loss": 0.4581005573272705,
+ "step": 1108
+ },
+ {
+ "epoch": 0.5818467995802729,
+ "grad_norm": 0.16825923629726025,
+ "learning_rate": 2.1778131091020516e-05,
+ "loss": 0.4433916211128235,
+ "step": 1109
+ },
+ {
+ "epoch": 0.5823714585519413,
+ "grad_norm": 0.17194743467883944,
+ "learning_rate": 2.1742272160542654e-05,
+ "loss": 0.47296425700187683,
+ "step": 1110
+ },
+ {
+ "epoch": 0.5828961175236097,
+ "grad_norm": 0.16947730546039969,
+ "learning_rate": 2.170642887252241e-05,
+ "loss": 0.4815937876701355,
+ "step": 1111
+ },
+ {
+ "epoch": 0.583420776495278,
+ "grad_norm": 0.1657449928668473,
+ "learning_rate": 2.167060132433798e-05,
+ "loss": 0.4479164779186249,
+ "step": 1112
+ },
+ {
+ "epoch": 0.5839454354669464,
+ "grad_norm": 0.1658952516607364,
+ "learning_rate": 2.1634789613324778e-05,
+ "loss": 0.4637463688850403,
+ "step": 1113
+ },
+ {
+ "epoch": 0.5844700944386149,
+ "grad_norm": 0.16777002188998236,
+ "learning_rate": 2.1598993836775218e-05,
+ "loss": 0.457524836063385,
+ "step": 1114
+ },
+ {
+ "epoch": 0.5849947534102833,
+ "grad_norm": 0.15837566447341322,
+ "learning_rate": 2.1563214091938405e-05,
+ "loss": 0.44938087463378906,
+ "step": 1115
+ },
+ {
+ "epoch": 0.5855194123819517,
+ "grad_norm": 0.17162206617035314,
+ "learning_rate": 2.1527450476019896e-05,
+ "loss": 0.4393450617790222,
+ "step": 1116
+ },
+ {
+ "epoch": 0.5860440713536201,
+ "grad_norm": 0.16224101921433814,
+ "learning_rate": 2.1491703086181426e-05,
+ "loss": 0.451913982629776,
+ "step": 1117
+ },
+ {
+ "epoch": 0.5865687303252886,
+ "grad_norm": 0.1736253504575144,
+ "learning_rate": 2.145597201954066e-05,
+ "loss": 0.4611290395259857,
+ "step": 1118
+ },
+ {
+ "epoch": 0.587093389296957,
+ "grad_norm": 0.16308408333737234,
+ "learning_rate": 2.1420257373170903e-05,
+ "loss": 0.45898568630218506,
+ "step": 1119
+ },
+ {
+ "epoch": 0.5876180482686254,
+ "grad_norm": 0.16948634794175524,
+ "learning_rate": 2.1384559244100848e-05,
+ "loss": 0.45982882380485535,
+ "step": 1120
+ },
+ {
+ "epoch": 0.5881427072402938,
+ "grad_norm": 0.16676366538565252,
+ "learning_rate": 2.1348877729314336e-05,
+ "loss": 0.44826576113700867,
+ "step": 1121
+ },
+ {
+ "epoch": 0.5886673662119623,
+ "grad_norm": 0.16431125630712295,
+ "learning_rate": 2.1313212925750037e-05,
+ "loss": 0.4616445302963257,
+ "step": 1122
+ },
+ {
+ "epoch": 0.5891920251836307,
+ "grad_norm": 0.1631659061616955,
+ "learning_rate": 2.1277564930301252e-05,
+ "loss": 0.45651188492774963,
+ "step": 1123
+ },
+ {
+ "epoch": 0.5897166841552991,
+ "grad_norm": 0.15805950502689237,
+ "learning_rate": 2.1241933839815604e-05,
+ "loss": 0.4486045837402344,
+ "step": 1124
+ },
+ {
+ "epoch": 0.5902413431269675,
+ "grad_norm": 0.16530288038603502,
+ "learning_rate": 2.1206319751094773e-05,
+ "loss": 0.4485066533088684,
+ "step": 1125
+ },
+ {
+ "epoch": 0.5907660020986358,
+ "grad_norm": 0.16061903194239047,
+ "learning_rate": 2.1170722760894295e-05,
+ "loss": 0.4565330743789673,
+ "step": 1126
+ },
+ {
+ "epoch": 0.5912906610703043,
+ "grad_norm": 0.1643346999353627,
+ "learning_rate": 2.1135142965923195e-05,
+ "loss": 0.45560717582702637,
+ "step": 1127
+ },
+ {
+ "epoch": 0.5918153200419727,
+ "grad_norm": 0.17484644315215753,
+ "learning_rate": 2.109958046284382e-05,
+ "loss": 0.4795891046524048,
+ "step": 1128
+ },
+ {
+ "epoch": 0.5923399790136411,
+ "grad_norm": 0.1623762678622651,
+ "learning_rate": 2.1064035348271528e-05,
+ "loss": 0.47395604848861694,
+ "step": 1129
+ },
+ {
+ "epoch": 0.5928646379853095,
+ "grad_norm": 0.16888824241669015,
+ "learning_rate": 2.1028507718774456e-05,
+ "loss": 0.4601455330848694,
+ "step": 1130
+ },
+ {
+ "epoch": 0.593389296956978,
+ "grad_norm": 0.16291744568806302,
+ "learning_rate": 2.0992997670873193e-05,
+ "loss": 0.46248236298561096,
+ "step": 1131
+ },
+ {
+ "epoch": 0.5939139559286464,
+ "grad_norm": 0.16185598691236988,
+ "learning_rate": 2.0957505301040603e-05,
+ "loss": 0.45085301995277405,
+ "step": 1132
+ },
+ {
+ "epoch": 0.5944386149003148,
+ "grad_norm": 0.1678000806117814,
+ "learning_rate": 2.0922030705701518e-05,
+ "loss": 0.44939014315605164,
+ "step": 1133
+ },
+ {
+ "epoch": 0.5949632738719832,
+ "grad_norm": 0.16104474282288994,
+ "learning_rate": 2.088657398123246e-05,
+ "loss": 0.4313976764678955,
+ "step": 1134
+ },
+ {
+ "epoch": 0.5954879328436516,
+ "grad_norm": 0.1631128968199669,
+ "learning_rate": 2.085113522396141e-05,
+ "loss": 0.4508278965950012,
+ "step": 1135
+ },
+ {
+ "epoch": 0.5960125918153201,
+ "grad_norm": 0.16307688195844533,
+ "learning_rate": 2.0815714530167557e-05,
+ "loss": 0.4467576742172241,
+ "step": 1136
+ },
+ {
+ "epoch": 0.5965372507869885,
+ "grad_norm": 0.16456832304532207,
+ "learning_rate": 2.0780311996080976e-05,
+ "loss": 0.4621703624725342,
+ "step": 1137
+ },
+ {
+ "epoch": 0.5970619097586569,
+ "grad_norm": 0.16346217748325367,
+ "learning_rate": 2.0744927717882435e-05,
+ "loss": 0.4589449465274811,
+ "step": 1138
+ },
+ {
+ "epoch": 0.5975865687303253,
+ "grad_norm": 0.16180691501076994,
+ "learning_rate": 2.0709561791703097e-05,
+ "loss": 0.43698400259017944,
+ "step": 1139
+ },
+ {
+ "epoch": 0.5981112277019937,
+ "grad_norm": 0.15695459976898835,
+ "learning_rate": 2.067421431362427e-05,
+ "loss": 0.452271044254303,
+ "step": 1140
+ },
+ {
+ "epoch": 0.5986358866736621,
+ "grad_norm": 0.15930893553721054,
+ "learning_rate": 2.063888537967713e-05,
+ "loss": 0.44868963956832886,
+ "step": 1141
+ },
+ {
+ "epoch": 0.5991605456453305,
+ "grad_norm": 0.15963802284017928,
+ "learning_rate": 2.0603575085842485e-05,
+ "loss": 0.4376864731311798,
+ "step": 1142
+ },
+ {
+ "epoch": 0.5996852046169989,
+ "grad_norm": 0.16610266900507178,
+ "learning_rate": 2.05682835280505e-05,
+ "loss": 0.44676727056503296,
+ "step": 1143
+ },
+ {
+ "epoch": 0.6002098635886673,
+ "grad_norm": 0.15978253236669612,
+ "learning_rate": 2.053301080218044e-05,
+ "loss": 0.434773713350296,
+ "step": 1144
+ },
+ {
+ "epoch": 0.6007345225603358,
+ "grad_norm": 0.16305141803526324,
+ "learning_rate": 2.0497757004060403e-05,
+ "loss": 0.4529780149459839,
+ "step": 1145
+ },
+ {
+ "epoch": 0.6012591815320042,
+ "grad_norm": 0.1649737864101215,
+ "learning_rate": 2.046252222946707e-05,
+ "loss": 0.4571434259414673,
+ "step": 1146
+ },
+ {
+ "epoch": 0.6017838405036726,
+ "grad_norm": 0.15738824968326598,
+ "learning_rate": 2.0427306574125427e-05,
+ "loss": 0.4446289539337158,
+ "step": 1147
+ },
+ {
+ "epoch": 0.602308499475341,
+ "grad_norm": 0.1731139535185187,
+ "learning_rate": 2.039211013370853e-05,
+ "loss": 0.4551090598106384,
+ "step": 1148
+ },
+ {
+ "epoch": 0.6028331584470095,
+ "grad_norm": 0.15857251900043037,
+ "learning_rate": 2.0356933003837237e-05,
+ "loss": 0.4644010663032532,
+ "step": 1149
+ },
+ {
+ "epoch": 0.6033578174186779,
+ "grad_norm": 0.1589582861134211,
+ "learning_rate": 2.032177528007992e-05,
+ "loss": 0.45741650462150574,
+ "step": 1150
+ },
+ {
+ "epoch": 0.6038824763903463,
+ "grad_norm": 0.15639092706575122,
+ "learning_rate": 2.0286637057952257e-05,
+ "loss": 0.4375547766685486,
+ "step": 1151
+ },
+ {
+ "epoch": 0.6044071353620147,
+ "grad_norm": 0.16335605581337187,
+ "learning_rate": 2.0251518432916933e-05,
+ "loss": 0.4636140465736389,
+ "step": 1152
+ },
+ {
+ "epoch": 0.6049317943336832,
+ "grad_norm": 0.15437437113359356,
+ "learning_rate": 2.0216419500383372e-05,
+ "loss": 0.44632142782211304,
+ "step": 1153
+ },
+ {
+ "epoch": 0.6054564533053515,
+ "grad_norm": 0.15522039198698664,
+ "learning_rate": 2.0181340355707545e-05,
+ "loss": 0.449049174785614,
+ "step": 1154
+ },
+ {
+ "epoch": 0.6059811122770199,
+ "grad_norm": 0.1599951917909323,
+ "learning_rate": 2.0146281094191604e-05,
+ "loss": 0.44395962357521057,
+ "step": 1155
+ },
+ {
+ "epoch": 0.6065057712486883,
+ "grad_norm": 0.16866621620527716,
+ "learning_rate": 2.011124181108373e-05,
+ "loss": 0.45693716406822205,
+ "step": 1156
+ },
+ {
+ "epoch": 0.6070304302203567,
+ "grad_norm": 0.15508935650135075,
+ "learning_rate": 2.0076222601577815e-05,
+ "loss": 0.4525151252746582,
+ "step": 1157
+ },
+ {
+ "epoch": 0.6075550891920252,
+ "grad_norm": 0.16508998118005971,
+ "learning_rate": 2.004122356081321e-05,
+ "loss": 0.466006875038147,
+ "step": 1158
+ },
+ {
+ "epoch": 0.6080797481636936,
+ "grad_norm": 0.15817418864022237,
+ "learning_rate": 2.0006244783874466e-05,
+ "loss": 0.4496110677719116,
+ "step": 1159
+ },
+ {
+ "epoch": 0.608604407135362,
+ "grad_norm": 0.16127270259749824,
+ "learning_rate": 1.997128636579109e-05,
+ "loss": 0.45918628573417664,
+ "step": 1160
+ },
+ {
+ "epoch": 0.6091290661070304,
+ "grad_norm": 0.1702599924159677,
+ "learning_rate": 1.9936348401537292e-05,
+ "loss": 0.47218555212020874,
+ "step": 1161
+ },
+ {
+ "epoch": 0.6096537250786989,
+ "grad_norm": 0.15690176248846127,
+ "learning_rate": 1.9901430986031695e-05,
+ "loss": 0.4516097903251648,
+ "step": 1162
+ },
+ {
+ "epoch": 0.6101783840503673,
+ "grad_norm": 0.1713614405614144,
+ "learning_rate": 1.9866534214137085e-05,
+ "loss": 0.45030856132507324,
+ "step": 1163
+ },
+ {
+ "epoch": 0.6107030430220357,
+ "grad_norm": 0.1670865563026229,
+ "learning_rate": 1.9831658180660195e-05,
+ "loss": 0.45603489875793457,
+ "step": 1164
+ },
+ {
+ "epoch": 0.6112277019937041,
+ "grad_norm": 0.16897539113401952,
+ "learning_rate": 1.9796802980351392e-05,
+ "loss": 0.46314537525177,
+ "step": 1165
+ },
+ {
+ "epoch": 0.6117523609653726,
+ "grad_norm": 0.15832868870697644,
+ "learning_rate": 1.976196870790446e-05,
+ "loss": 0.44784611463546753,
+ "step": 1166
+ },
+ {
+ "epoch": 0.612277019937041,
+ "grad_norm": 0.16595365551891114,
+ "learning_rate": 1.9727155457956315e-05,
+ "loss": 0.4409624934196472,
+ "step": 1167
+ },
+ {
+ "epoch": 0.6128016789087093,
+ "grad_norm": 0.1642630411212017,
+ "learning_rate": 1.969236332508677e-05,
+ "loss": 0.46120351552963257,
+ "step": 1168
+ },
+ {
+ "epoch": 0.6133263378803777,
+ "grad_norm": 0.16481796135868812,
+ "learning_rate": 1.9657592403818253e-05,
+ "loss": 0.45544904470443726,
+ "step": 1169
+ },
+ {
+ "epoch": 0.6138509968520461,
+ "grad_norm": 0.1629832779852364,
+ "learning_rate": 1.962284278861557e-05,
+ "loss": 0.4578462839126587,
+ "step": 1170
+ },
+ {
+ "epoch": 0.6143756558237146,
+ "grad_norm": 0.15916219613603688,
+ "learning_rate": 1.9588114573885664e-05,
+ "loss": 0.4546983540058136,
+ "step": 1171
+ },
+ {
+ "epoch": 0.614900314795383,
+ "grad_norm": 0.1660520165166075,
+ "learning_rate": 1.95534078539773e-05,
+ "loss": 0.4601942300796509,
+ "step": 1172
+ },
+ {
+ "epoch": 0.6154249737670514,
+ "grad_norm": 0.16924794578577768,
+ "learning_rate": 1.9518722723180877e-05,
+ "loss": 0.45997846126556396,
+ "step": 1173
+ },
+ {
+ "epoch": 0.6159496327387198,
+ "grad_norm": 0.16597646133045382,
+ "learning_rate": 1.9484059275728127e-05,
+ "loss": 0.4594605565071106,
+ "step": 1174
+ },
+ {
+ "epoch": 0.6164742917103883,
+ "grad_norm": 0.16964243792519748,
+ "learning_rate": 1.9449417605791876e-05,
+ "loss": 0.4638690650463104,
+ "step": 1175
+ },
+ {
+ "epoch": 0.6169989506820567,
+ "grad_norm": 0.1758591426791696,
+ "learning_rate": 1.9414797807485796e-05,
+ "loss": 0.4666447639465332,
+ "step": 1176
+ },
+ {
+ "epoch": 0.6175236096537251,
+ "grad_norm": 0.16217661227891753,
+ "learning_rate": 1.9380199974864126e-05,
+ "loss": 0.44252535700798035,
+ "step": 1177
+ },
+ {
+ "epoch": 0.6180482686253935,
+ "grad_norm": 0.1576446184499019,
+ "learning_rate": 1.934562420192142e-05,
+ "loss": 0.4499641954898834,
+ "step": 1178
+ },
+ {
+ "epoch": 0.618572927597062,
+ "grad_norm": 0.16231008092636912,
+ "learning_rate": 1.9311070582592327e-05,
+ "loss": 0.4561542570590973,
+ "step": 1179
+ },
+ {
+ "epoch": 0.6190975865687304,
+ "grad_norm": 0.16790382249577043,
+ "learning_rate": 1.9276539210751292e-05,
+ "loss": 0.4484325647354126,
+ "step": 1180
+ },
+ {
+ "epoch": 0.6196222455403987,
+ "grad_norm": 0.18794729932605855,
+ "learning_rate": 1.9242030180212323e-05,
+ "loss": 0.44169026613235474,
+ "step": 1181
+ },
+ {
+ "epoch": 0.6201469045120671,
+ "grad_norm": 0.16948051243889814,
+ "learning_rate": 1.920754358472872e-05,
+ "loss": 0.4592876434326172,
+ "step": 1182
+ },
+ {
+ "epoch": 0.6206715634837355,
+ "grad_norm": 0.16762558710029374,
+ "learning_rate": 1.917307951799286e-05,
+ "loss": 0.4402095079421997,
+ "step": 1183
+ },
+ {
+ "epoch": 0.621196222455404,
+ "grad_norm": 0.18170280116387003,
+ "learning_rate": 1.9138638073635885e-05,
+ "loss": 0.4633026123046875,
+ "step": 1184
+ },
+ {
+ "epoch": 0.6217208814270724,
+ "grad_norm": 0.1718051709136108,
+ "learning_rate": 1.9104219345227486e-05,
+ "loss": 0.4510066509246826,
+ "step": 1185
+ },
+ {
+ "epoch": 0.6222455403987408,
+ "grad_norm": 0.16829066326345576,
+ "learning_rate": 1.906982342627565e-05,
+ "loss": 0.4632919430732727,
+ "step": 1186
+ },
+ {
+ "epoch": 0.6227701993704092,
+ "grad_norm": 0.16669141808548274,
+ "learning_rate": 1.9035450410226395e-05,
+ "loss": 0.455544650554657,
+ "step": 1187
+ },
+ {
+ "epoch": 0.6232948583420777,
+ "grad_norm": 0.16679996559722232,
+ "learning_rate": 1.90011003904635e-05,
+ "loss": 0.459994375705719,
+ "step": 1188
+ },
+ {
+ "epoch": 0.6238195173137461,
+ "grad_norm": 0.17379754713278164,
+ "learning_rate": 1.8966773460308284e-05,
+ "loss": 0.4576749801635742,
+ "step": 1189
+ },
+ {
+ "epoch": 0.6243441762854145,
+ "grad_norm": 0.17088539733389246,
+ "learning_rate": 1.8932469713019335e-05,
+ "loss": 0.44853052496910095,
+ "step": 1190
+ },
+ {
+ "epoch": 0.6248688352570829,
+ "grad_norm": 0.15556883889201015,
+ "learning_rate": 1.8898189241792248e-05,
+ "loss": 0.44060447812080383,
+ "step": 1191
+ },
+ {
+ "epoch": 0.6253934942287513,
+ "grad_norm": 0.1640093917439215,
+ "learning_rate": 1.8863932139759388e-05,
+ "loss": 0.4531911611557007,
+ "step": 1192
+ },
+ {
+ "epoch": 0.6259181532004198,
+ "grad_norm": 0.16444284826591576,
+ "learning_rate": 1.882969849998964e-05,
+ "loss": 0.44176363945007324,
+ "step": 1193
+ },
+ {
+ "epoch": 0.6264428121720882,
+ "grad_norm": 0.1696462520995642,
+ "learning_rate": 1.8795488415488143e-05,
+ "loss": 0.44516652822494507,
+ "step": 1194
+ },
+ {
+ "epoch": 0.6269674711437565,
+ "grad_norm": 0.16810724705191796,
+ "learning_rate": 1.8761301979196032e-05,
+ "loss": 0.4542563855648041,
+ "step": 1195
+ },
+ {
+ "epoch": 0.6274921301154249,
+ "grad_norm": 0.17334813313855824,
+ "learning_rate": 1.8727139283990204e-05,
+ "loss": 0.47769248485565186,
+ "step": 1196
+ },
+ {
+ "epoch": 0.6280167890870934,
+ "grad_norm": 0.16308212035303235,
+ "learning_rate": 1.8693000422683043e-05,
+ "loss": 0.4407799243927002,
+ "step": 1197
+ },
+ {
+ "epoch": 0.6285414480587618,
+ "grad_norm": 0.16853319139928977,
+ "learning_rate": 1.8658885488022204e-05,
+ "loss": 0.4565497040748596,
+ "step": 1198
+ },
+ {
+ "epoch": 0.6290661070304302,
+ "grad_norm": 0.16694968781284703,
+ "learning_rate": 1.8624794572690328e-05,
+ "loss": 0.44522374868392944,
+ "step": 1199
+ },
+ {
+ "epoch": 0.6295907660020986,
+ "grad_norm": 0.16418579468339103,
+ "learning_rate": 1.8590727769304788e-05,
+ "loss": 0.4441048502922058,
+ "step": 1200
+ },
+ {
+ "epoch": 0.630115424973767,
+ "grad_norm": 0.1661930777124156,
+ "learning_rate": 1.8556685170417486e-05,
+ "loss": 0.457327663898468,
+ "step": 1201
+ },
+ {
+ "epoch": 0.6306400839454355,
+ "grad_norm": 0.16536630027039814,
+ "learning_rate": 1.852266686851452e-05,
+ "loss": 0.462494432926178,
+ "step": 1202
+ },
+ {
+ "epoch": 0.6311647429171039,
+ "grad_norm": 0.16611474088307454,
+ "learning_rate": 1.8488672956016013e-05,
+ "loss": 0.4557701647281647,
+ "step": 1203
+ },
+ {
+ "epoch": 0.6316894018887723,
+ "grad_norm": 0.16759343079749697,
+ "learning_rate": 1.8454703525275813e-05,
+ "loss": 0.45069995522499084,
+ "step": 1204
+ },
+ {
+ "epoch": 0.6322140608604407,
+ "grad_norm": 0.16186676294485336,
+ "learning_rate": 1.8420758668581255e-05,
+ "loss": 0.44954800605773926,
+ "step": 1205
+ },
+ {
+ "epoch": 0.6327387198321092,
+ "grad_norm": 0.15687159835368553,
+ "learning_rate": 1.8386838478152916e-05,
+ "loss": 0.4490429759025574,
+ "step": 1206
+ },
+ {
+ "epoch": 0.6332633788037776,
+ "grad_norm": 0.16242951976115577,
+ "learning_rate": 1.8352943046144372e-05,
+ "loss": 0.45819586515426636,
+ "step": 1207
+ },
+ {
+ "epoch": 0.633788037775446,
+ "grad_norm": 0.16516087266517232,
+ "learning_rate": 1.831907246464191e-05,
+ "loss": 0.44769883155822754,
+ "step": 1208
+ },
+ {
+ "epoch": 0.6343126967471143,
+ "grad_norm": 0.16709475871512644,
+ "learning_rate": 1.8285226825664332e-05,
+ "loss": 0.4699128568172455,
+ "step": 1209
+ },
+ {
+ "epoch": 0.6348373557187827,
+ "grad_norm": 0.16288623150595774,
+ "learning_rate": 1.8251406221162636e-05,
+ "loss": 0.45322558283805847,
+ "step": 1210
+ },
+ {
+ "epoch": 0.6353620146904512,
+ "grad_norm": 0.1675722848054237,
+ "learning_rate": 1.821761074301986e-05,
+ "loss": 0.4752347469329834,
+ "step": 1211
+ },
+ {
+ "epoch": 0.6358866736621196,
+ "grad_norm": 0.163285963514504,
+ "learning_rate": 1.818384048305074e-05,
+ "loss": 0.441958487033844,
+ "step": 1212
+ },
+ {
+ "epoch": 0.636411332633788,
+ "grad_norm": 0.17078102287557764,
+ "learning_rate": 1.8150095533001517e-05,
+ "loss": 0.4551333487033844,
+ "step": 1213
+ },
+ {
+ "epoch": 0.6369359916054564,
+ "grad_norm": 0.16188540746097138,
+ "learning_rate": 1.811637598454966e-05,
+ "loss": 0.4470779597759247,
+ "step": 1214
+ },
+ {
+ "epoch": 0.6374606505771249,
+ "grad_norm": 0.15235628218958286,
+ "learning_rate": 1.808268192930364e-05,
+ "loss": 0.43604153394699097,
+ "step": 1215
+ },
+ {
+ "epoch": 0.6379853095487933,
+ "grad_norm": 0.1604601949959104,
+ "learning_rate": 1.8049013458802665e-05,
+ "loss": 0.4445713758468628,
+ "step": 1216
+ },
+ {
+ "epoch": 0.6385099685204617,
+ "grad_norm": 0.15511797131215901,
+ "learning_rate": 1.8015370664516427e-05,
+ "loss": 0.4455301761627197,
+ "step": 1217
+ },
+ {
+ "epoch": 0.6390346274921301,
+ "grad_norm": 0.16815669546944095,
+ "learning_rate": 1.798175363784487e-05,
+ "loss": 0.46135014295578003,
+ "step": 1218
+ },
+ {
+ "epoch": 0.6395592864637986,
+ "grad_norm": 0.161558177688922,
+ "learning_rate": 1.7948162470117917e-05,
+ "loss": 0.4519304037094116,
+ "step": 1219
+ },
+ {
+ "epoch": 0.640083945435467,
+ "grad_norm": 0.16677842130048232,
+ "learning_rate": 1.7914597252595265e-05,
+ "loss": 0.4550550878047943,
+ "step": 1220
+ },
+ {
+ "epoch": 0.6406086044071354,
+ "grad_norm": 0.1597331480423138,
+ "learning_rate": 1.7881058076466095e-05,
+ "loss": 0.441120445728302,
+ "step": 1221
+ },
+ {
+ "epoch": 0.6411332633788038,
+ "grad_norm": 0.15951106545246824,
+ "learning_rate": 1.7847545032848844e-05,
+ "loss": 0.4471382200717926,
+ "step": 1222
+ },
+ {
+ "epoch": 0.6416579223504721,
+ "grad_norm": 0.16157787674667565,
+ "learning_rate": 1.7814058212790947e-05,
+ "loss": 0.4550648331642151,
+ "step": 1223
+ },
+ {
+ "epoch": 0.6421825813221406,
+ "grad_norm": 0.16989186763352415,
+ "learning_rate": 1.7780597707268583e-05,
+ "loss": 0.4515089988708496,
+ "step": 1224
+ },
+ {
+ "epoch": 0.642707240293809,
+ "grad_norm": 0.15952153223099508,
+ "learning_rate": 1.7747163607186468e-05,
+ "loss": 0.4658638834953308,
+ "step": 1225
+ },
+ {
+ "epoch": 0.6432318992654774,
+ "grad_norm": 0.15205926699647218,
+ "learning_rate": 1.7713756003377567e-05,
+ "loss": 0.4330868422985077,
+ "step": 1226
+ },
+ {
+ "epoch": 0.6437565582371458,
+ "grad_norm": 0.15784806480936794,
+ "learning_rate": 1.7680374986602865e-05,
+ "loss": 0.4506068527698517,
+ "step": 1227
+ },
+ {
+ "epoch": 0.6442812172088143,
+ "grad_norm": 0.16006709280686687,
+ "learning_rate": 1.7647020647551105e-05,
+ "loss": 0.446941077709198,
+ "step": 1228
+ },
+ {
+ "epoch": 0.6448058761804827,
+ "grad_norm": 0.16064631158222062,
+ "learning_rate": 1.761369307683856e-05,
+ "loss": 0.4418525695800781,
+ "step": 1229
+ },
+ {
+ "epoch": 0.6453305351521511,
+ "grad_norm": 0.16454136480444392,
+ "learning_rate": 1.7580392365008767e-05,
+ "loss": 0.46268391609191895,
+ "step": 1230
+ },
+ {
+ "epoch": 0.6458551941238195,
+ "grad_norm": 0.15401246535771462,
+ "learning_rate": 1.754711860253233e-05,
+ "loss": 0.4372989535331726,
+ "step": 1231
+ },
+ {
+ "epoch": 0.646379853095488,
+ "grad_norm": 0.1544567975028383,
+ "learning_rate": 1.7513871879806603e-05,
+ "loss": 0.44806575775146484,
+ "step": 1232
+ },
+ {
+ "epoch": 0.6469045120671564,
+ "grad_norm": 0.16411225112730965,
+ "learning_rate": 1.748065228715549e-05,
+ "loss": 0.43782058358192444,
+ "step": 1233
+ },
+ {
+ "epoch": 0.6474291710388248,
+ "grad_norm": 0.1552767832972093,
+ "learning_rate": 1.7447459914829183e-05,
+ "loss": 0.44531378149986267,
+ "step": 1234
+ },
+ {
+ "epoch": 0.6479538300104932,
+ "grad_norm": 0.16214695257123513,
+ "learning_rate": 1.741429485300392e-05,
+ "loss": 0.47399842739105225,
+ "step": 1235
+ },
+ {
+ "epoch": 0.6484784889821616,
+ "grad_norm": 0.16684861111484112,
+ "learning_rate": 1.7381157191781766e-05,
+ "loss": 0.4654282331466675,
+ "step": 1236
+ },
+ {
+ "epoch": 0.64900314795383,
+ "grad_norm": 0.15901864178630273,
+ "learning_rate": 1.7348047021190334e-05,
+ "loss": 0.45128414034843445,
+ "step": 1237
+ },
+ {
+ "epoch": 0.6495278069254984,
+ "grad_norm": 0.15360991961742707,
+ "learning_rate": 1.7314964431182523e-05,
+ "loss": 0.4303375482559204,
+ "step": 1238
+ },
+ {
+ "epoch": 0.6500524658971668,
+ "grad_norm": 0.1633151574385741,
+ "learning_rate": 1.728190951163633e-05,
+ "loss": 0.45568403601646423,
+ "step": 1239
+ },
+ {
+ "epoch": 0.6505771248688352,
+ "grad_norm": 0.16377853571058645,
+ "learning_rate": 1.724888235235457e-05,
+ "loss": 0.46503543853759766,
+ "step": 1240
+ },
+ {
+ "epoch": 0.6511017838405037,
+ "grad_norm": 0.1595023801275368,
+ "learning_rate": 1.7215883043064655e-05,
+ "loss": 0.46337294578552246,
+ "step": 1241
+ },
+ {
+ "epoch": 0.6516264428121721,
+ "grad_norm": 0.16324796182212292,
+ "learning_rate": 1.7182911673418313e-05,
+ "loss": 0.46197354793548584,
+ "step": 1242
+ },
+ {
+ "epoch": 0.6521511017838405,
+ "grad_norm": 0.16353734830879765,
+ "learning_rate": 1.714996833299138e-05,
+ "loss": 0.45741671323776245,
+ "step": 1243
+ },
+ {
+ "epoch": 0.6526757607555089,
+ "grad_norm": 0.15941587678263386,
+ "learning_rate": 1.7117053111283528e-05,
+ "loss": 0.4514503479003906,
+ "step": 1244
+ },
+ {
+ "epoch": 0.6532004197271774,
+ "grad_norm": 0.1642653230145008,
+ "learning_rate": 1.708416609771804e-05,
+ "loss": 0.4437655210494995,
+ "step": 1245
+ },
+ {
+ "epoch": 0.6537250786988458,
+ "grad_norm": 0.15814155519130152,
+ "learning_rate": 1.7051307381641587e-05,
+ "loss": 0.4314163327217102,
+ "step": 1246
+ },
+ {
+ "epoch": 0.6542497376705142,
+ "grad_norm": 0.16086078370768056,
+ "learning_rate": 1.701847705232394e-05,
+ "loss": 0.4526430666446686,
+ "step": 1247
+ },
+ {
+ "epoch": 0.6547743966421826,
+ "grad_norm": 0.16389360190715183,
+ "learning_rate": 1.698567519895775e-05,
+ "loss": 0.4278966188430786,
+ "step": 1248
+ },
+ {
+ "epoch": 0.655299055613851,
+ "grad_norm": 0.16303142287804262,
+ "learning_rate": 1.6952901910658316e-05,
+ "loss": 0.45778220891952515,
+ "step": 1249
+ },
+ {
+ "epoch": 0.6558237145855194,
+ "grad_norm": 0.1675209956904048,
+ "learning_rate": 1.6920157276463318e-05,
+ "loss": 0.4489403963088989,
+ "step": 1250
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 3.308425305705677e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1300/chat_template.jinja b/checkpoints/checkpoint-1300/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1300/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1300/config.json b/checkpoints/checkpoint-1300/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1300/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1300/generation_config.json b/checkpoints/checkpoint-1300/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1300/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1300/model.safetensors b/checkpoints/checkpoint-1300/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..ce999841f30be605806caeb45ebdaf1f9e2ddb36
--- /dev/null
+++ b/checkpoints/checkpoint-1300/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2dfe532c536ee712f5c34443bc95cd63fa6f0351c203680753251b26fc88a2fb
+size 37983421000
diff --git a/checkpoints/checkpoint-1300/tokenizer.json b/checkpoints/checkpoint-1300/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1300/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1300/tokenizer_config.json b/checkpoints/checkpoint-1300/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1300/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1300/trainer_state.json b/checkpoints/checkpoint-1300/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..2189d5c9f7eeb6045d8b49166f4783b240639a47
--- /dev/null
+++ b/checkpoints/checkpoint-1300/trainer_state.json
@@ -0,0 +1,9134 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.6820566631689402,
+ "eval_steps": 500,
+ "global_step": 1300,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ },
+ {
+ "epoch": 0.5776495278069255,
+ "grad_norm": 0.1862053032854098,
+ "learning_rate": 2.2065553964033126e-05,
+ "loss": 0.4816875457763672,
+ "step": 1101
+ },
+ {
+ "epoch": 0.5781741867785939,
+ "grad_norm": 0.16996015894363117,
+ "learning_rate": 2.20295734045532e-05,
+ "loss": 0.4471197724342346,
+ "step": 1102
+ },
+ {
+ "epoch": 0.5786988457502623,
+ "grad_norm": 0.16889227008390692,
+ "learning_rate": 2.1993607706997725e-05,
+ "loss": 0.468391478061676,
+ "step": 1103
+ },
+ {
+ "epoch": 0.5792235047219307,
+ "grad_norm": 0.16670931681562373,
+ "learning_rate": 2.195765696907745e-05,
+ "loss": 0.44934162497520447,
+ "step": 1104
+ },
+ {
+ "epoch": 0.5797481636935992,
+ "grad_norm": 0.16328418741223716,
+ "learning_rate": 2.1921721288462482e-05,
+ "loss": 0.4558270275592804,
+ "step": 1105
+ },
+ {
+ "epoch": 0.5802728226652676,
+ "grad_norm": 0.17364660408703936,
+ "learning_rate": 2.188580076278203e-05,
+ "loss": 0.4597751796245575,
+ "step": 1106
+ },
+ {
+ "epoch": 0.580797481636936,
+ "grad_norm": 0.1628618453800601,
+ "learning_rate": 2.18498954896241e-05,
+ "loss": 0.4521859288215637,
+ "step": 1107
+ },
+ {
+ "epoch": 0.5813221406086044,
+ "grad_norm": 0.16813817951871482,
+ "learning_rate": 2.1814005566535302e-05,
+ "loss": 0.4581005573272705,
+ "step": 1108
+ },
+ {
+ "epoch": 0.5818467995802729,
+ "grad_norm": 0.16825923629726025,
+ "learning_rate": 2.1778131091020516e-05,
+ "loss": 0.4433916211128235,
+ "step": 1109
+ },
+ {
+ "epoch": 0.5823714585519413,
+ "grad_norm": 0.17194743467883944,
+ "learning_rate": 2.1742272160542654e-05,
+ "loss": 0.47296425700187683,
+ "step": 1110
+ },
+ {
+ "epoch": 0.5828961175236097,
+ "grad_norm": 0.16947730546039969,
+ "learning_rate": 2.170642887252241e-05,
+ "loss": 0.4815937876701355,
+ "step": 1111
+ },
+ {
+ "epoch": 0.583420776495278,
+ "grad_norm": 0.1657449928668473,
+ "learning_rate": 2.167060132433798e-05,
+ "loss": 0.4479164779186249,
+ "step": 1112
+ },
+ {
+ "epoch": 0.5839454354669464,
+ "grad_norm": 0.1658952516607364,
+ "learning_rate": 2.1634789613324778e-05,
+ "loss": 0.4637463688850403,
+ "step": 1113
+ },
+ {
+ "epoch": 0.5844700944386149,
+ "grad_norm": 0.16777002188998236,
+ "learning_rate": 2.1598993836775218e-05,
+ "loss": 0.457524836063385,
+ "step": 1114
+ },
+ {
+ "epoch": 0.5849947534102833,
+ "grad_norm": 0.15837566447341322,
+ "learning_rate": 2.1563214091938405e-05,
+ "loss": 0.44938087463378906,
+ "step": 1115
+ },
+ {
+ "epoch": 0.5855194123819517,
+ "grad_norm": 0.17162206617035314,
+ "learning_rate": 2.1527450476019896e-05,
+ "loss": 0.4393450617790222,
+ "step": 1116
+ },
+ {
+ "epoch": 0.5860440713536201,
+ "grad_norm": 0.16224101921433814,
+ "learning_rate": 2.1491703086181426e-05,
+ "loss": 0.451913982629776,
+ "step": 1117
+ },
+ {
+ "epoch": 0.5865687303252886,
+ "grad_norm": 0.1736253504575144,
+ "learning_rate": 2.145597201954066e-05,
+ "loss": 0.4611290395259857,
+ "step": 1118
+ },
+ {
+ "epoch": 0.587093389296957,
+ "grad_norm": 0.16308408333737234,
+ "learning_rate": 2.1420257373170903e-05,
+ "loss": 0.45898568630218506,
+ "step": 1119
+ },
+ {
+ "epoch": 0.5876180482686254,
+ "grad_norm": 0.16948634794175524,
+ "learning_rate": 2.1384559244100848e-05,
+ "loss": 0.45982882380485535,
+ "step": 1120
+ },
+ {
+ "epoch": 0.5881427072402938,
+ "grad_norm": 0.16676366538565252,
+ "learning_rate": 2.1348877729314336e-05,
+ "loss": 0.44826576113700867,
+ "step": 1121
+ },
+ {
+ "epoch": 0.5886673662119623,
+ "grad_norm": 0.16431125630712295,
+ "learning_rate": 2.1313212925750037e-05,
+ "loss": 0.4616445302963257,
+ "step": 1122
+ },
+ {
+ "epoch": 0.5891920251836307,
+ "grad_norm": 0.1631659061616955,
+ "learning_rate": 2.1277564930301252e-05,
+ "loss": 0.45651188492774963,
+ "step": 1123
+ },
+ {
+ "epoch": 0.5897166841552991,
+ "grad_norm": 0.15805950502689237,
+ "learning_rate": 2.1241933839815604e-05,
+ "loss": 0.4486045837402344,
+ "step": 1124
+ },
+ {
+ "epoch": 0.5902413431269675,
+ "grad_norm": 0.16530288038603502,
+ "learning_rate": 2.1206319751094773e-05,
+ "loss": 0.4485066533088684,
+ "step": 1125
+ },
+ {
+ "epoch": 0.5907660020986358,
+ "grad_norm": 0.16061903194239047,
+ "learning_rate": 2.1170722760894295e-05,
+ "loss": 0.4565330743789673,
+ "step": 1126
+ },
+ {
+ "epoch": 0.5912906610703043,
+ "grad_norm": 0.1643346999353627,
+ "learning_rate": 2.1135142965923195e-05,
+ "loss": 0.45560717582702637,
+ "step": 1127
+ },
+ {
+ "epoch": 0.5918153200419727,
+ "grad_norm": 0.17484644315215753,
+ "learning_rate": 2.109958046284382e-05,
+ "loss": 0.4795891046524048,
+ "step": 1128
+ },
+ {
+ "epoch": 0.5923399790136411,
+ "grad_norm": 0.1623762678622651,
+ "learning_rate": 2.1064035348271528e-05,
+ "loss": 0.47395604848861694,
+ "step": 1129
+ },
+ {
+ "epoch": 0.5928646379853095,
+ "grad_norm": 0.16888824241669015,
+ "learning_rate": 2.1028507718774456e-05,
+ "loss": 0.4601455330848694,
+ "step": 1130
+ },
+ {
+ "epoch": 0.593389296956978,
+ "grad_norm": 0.16291744568806302,
+ "learning_rate": 2.0992997670873193e-05,
+ "loss": 0.46248236298561096,
+ "step": 1131
+ },
+ {
+ "epoch": 0.5939139559286464,
+ "grad_norm": 0.16185598691236988,
+ "learning_rate": 2.0957505301040603e-05,
+ "loss": 0.45085301995277405,
+ "step": 1132
+ },
+ {
+ "epoch": 0.5944386149003148,
+ "grad_norm": 0.1678000806117814,
+ "learning_rate": 2.0922030705701518e-05,
+ "loss": 0.44939014315605164,
+ "step": 1133
+ },
+ {
+ "epoch": 0.5949632738719832,
+ "grad_norm": 0.16104474282288994,
+ "learning_rate": 2.088657398123246e-05,
+ "loss": 0.4313976764678955,
+ "step": 1134
+ },
+ {
+ "epoch": 0.5954879328436516,
+ "grad_norm": 0.1631128968199669,
+ "learning_rate": 2.085113522396141e-05,
+ "loss": 0.4508278965950012,
+ "step": 1135
+ },
+ {
+ "epoch": 0.5960125918153201,
+ "grad_norm": 0.16307688195844533,
+ "learning_rate": 2.0815714530167557e-05,
+ "loss": 0.4467576742172241,
+ "step": 1136
+ },
+ {
+ "epoch": 0.5965372507869885,
+ "grad_norm": 0.16456832304532207,
+ "learning_rate": 2.0780311996080976e-05,
+ "loss": 0.4621703624725342,
+ "step": 1137
+ },
+ {
+ "epoch": 0.5970619097586569,
+ "grad_norm": 0.16346217748325367,
+ "learning_rate": 2.0744927717882435e-05,
+ "loss": 0.4589449465274811,
+ "step": 1138
+ },
+ {
+ "epoch": 0.5975865687303253,
+ "grad_norm": 0.16180691501076994,
+ "learning_rate": 2.0709561791703097e-05,
+ "loss": 0.43698400259017944,
+ "step": 1139
+ },
+ {
+ "epoch": 0.5981112277019937,
+ "grad_norm": 0.15695459976898835,
+ "learning_rate": 2.067421431362427e-05,
+ "loss": 0.452271044254303,
+ "step": 1140
+ },
+ {
+ "epoch": 0.5986358866736621,
+ "grad_norm": 0.15930893553721054,
+ "learning_rate": 2.063888537967713e-05,
+ "loss": 0.44868963956832886,
+ "step": 1141
+ },
+ {
+ "epoch": 0.5991605456453305,
+ "grad_norm": 0.15963802284017928,
+ "learning_rate": 2.0603575085842485e-05,
+ "loss": 0.4376864731311798,
+ "step": 1142
+ },
+ {
+ "epoch": 0.5996852046169989,
+ "grad_norm": 0.16610266900507178,
+ "learning_rate": 2.05682835280505e-05,
+ "loss": 0.44676727056503296,
+ "step": 1143
+ },
+ {
+ "epoch": 0.6002098635886673,
+ "grad_norm": 0.15978253236669612,
+ "learning_rate": 2.053301080218044e-05,
+ "loss": 0.434773713350296,
+ "step": 1144
+ },
+ {
+ "epoch": 0.6007345225603358,
+ "grad_norm": 0.16305141803526324,
+ "learning_rate": 2.0497757004060403e-05,
+ "loss": 0.4529780149459839,
+ "step": 1145
+ },
+ {
+ "epoch": 0.6012591815320042,
+ "grad_norm": 0.1649737864101215,
+ "learning_rate": 2.046252222946707e-05,
+ "loss": 0.4571434259414673,
+ "step": 1146
+ },
+ {
+ "epoch": 0.6017838405036726,
+ "grad_norm": 0.15738824968326598,
+ "learning_rate": 2.0427306574125427e-05,
+ "loss": 0.4446289539337158,
+ "step": 1147
+ },
+ {
+ "epoch": 0.602308499475341,
+ "grad_norm": 0.1731139535185187,
+ "learning_rate": 2.039211013370853e-05,
+ "loss": 0.4551090598106384,
+ "step": 1148
+ },
+ {
+ "epoch": 0.6028331584470095,
+ "grad_norm": 0.15857251900043037,
+ "learning_rate": 2.0356933003837237e-05,
+ "loss": 0.4644010663032532,
+ "step": 1149
+ },
+ {
+ "epoch": 0.6033578174186779,
+ "grad_norm": 0.1589582861134211,
+ "learning_rate": 2.032177528007992e-05,
+ "loss": 0.45741650462150574,
+ "step": 1150
+ },
+ {
+ "epoch": 0.6038824763903463,
+ "grad_norm": 0.15639092706575122,
+ "learning_rate": 2.0286637057952257e-05,
+ "loss": 0.4375547766685486,
+ "step": 1151
+ },
+ {
+ "epoch": 0.6044071353620147,
+ "grad_norm": 0.16335605581337187,
+ "learning_rate": 2.0251518432916933e-05,
+ "loss": 0.4636140465736389,
+ "step": 1152
+ },
+ {
+ "epoch": 0.6049317943336832,
+ "grad_norm": 0.15437437113359356,
+ "learning_rate": 2.0216419500383372e-05,
+ "loss": 0.44632142782211304,
+ "step": 1153
+ },
+ {
+ "epoch": 0.6054564533053515,
+ "grad_norm": 0.15522039198698664,
+ "learning_rate": 2.0181340355707545e-05,
+ "loss": 0.449049174785614,
+ "step": 1154
+ },
+ {
+ "epoch": 0.6059811122770199,
+ "grad_norm": 0.1599951917909323,
+ "learning_rate": 2.0146281094191604e-05,
+ "loss": 0.44395962357521057,
+ "step": 1155
+ },
+ {
+ "epoch": 0.6065057712486883,
+ "grad_norm": 0.16866621620527716,
+ "learning_rate": 2.011124181108373e-05,
+ "loss": 0.45693716406822205,
+ "step": 1156
+ },
+ {
+ "epoch": 0.6070304302203567,
+ "grad_norm": 0.15508935650135075,
+ "learning_rate": 2.0076222601577815e-05,
+ "loss": 0.4525151252746582,
+ "step": 1157
+ },
+ {
+ "epoch": 0.6075550891920252,
+ "grad_norm": 0.16508998118005971,
+ "learning_rate": 2.004122356081321e-05,
+ "loss": 0.466006875038147,
+ "step": 1158
+ },
+ {
+ "epoch": 0.6080797481636936,
+ "grad_norm": 0.15817418864022237,
+ "learning_rate": 2.0006244783874466e-05,
+ "loss": 0.4496110677719116,
+ "step": 1159
+ },
+ {
+ "epoch": 0.608604407135362,
+ "grad_norm": 0.16127270259749824,
+ "learning_rate": 1.997128636579109e-05,
+ "loss": 0.45918628573417664,
+ "step": 1160
+ },
+ {
+ "epoch": 0.6091290661070304,
+ "grad_norm": 0.1702599924159677,
+ "learning_rate": 1.9936348401537292e-05,
+ "loss": 0.47218555212020874,
+ "step": 1161
+ },
+ {
+ "epoch": 0.6096537250786989,
+ "grad_norm": 0.15690176248846127,
+ "learning_rate": 1.9901430986031695e-05,
+ "loss": 0.4516097903251648,
+ "step": 1162
+ },
+ {
+ "epoch": 0.6101783840503673,
+ "grad_norm": 0.1713614405614144,
+ "learning_rate": 1.9866534214137085e-05,
+ "loss": 0.45030856132507324,
+ "step": 1163
+ },
+ {
+ "epoch": 0.6107030430220357,
+ "grad_norm": 0.1670865563026229,
+ "learning_rate": 1.9831658180660195e-05,
+ "loss": 0.45603489875793457,
+ "step": 1164
+ },
+ {
+ "epoch": 0.6112277019937041,
+ "grad_norm": 0.16897539113401952,
+ "learning_rate": 1.9796802980351392e-05,
+ "loss": 0.46314537525177,
+ "step": 1165
+ },
+ {
+ "epoch": 0.6117523609653726,
+ "grad_norm": 0.15832868870697644,
+ "learning_rate": 1.976196870790446e-05,
+ "loss": 0.44784611463546753,
+ "step": 1166
+ },
+ {
+ "epoch": 0.612277019937041,
+ "grad_norm": 0.16595365551891114,
+ "learning_rate": 1.9727155457956315e-05,
+ "loss": 0.4409624934196472,
+ "step": 1167
+ },
+ {
+ "epoch": 0.6128016789087093,
+ "grad_norm": 0.1642630411212017,
+ "learning_rate": 1.969236332508677e-05,
+ "loss": 0.46120351552963257,
+ "step": 1168
+ },
+ {
+ "epoch": 0.6133263378803777,
+ "grad_norm": 0.16481796135868812,
+ "learning_rate": 1.9657592403818253e-05,
+ "loss": 0.45544904470443726,
+ "step": 1169
+ },
+ {
+ "epoch": 0.6138509968520461,
+ "grad_norm": 0.1629832779852364,
+ "learning_rate": 1.962284278861557e-05,
+ "loss": 0.4578462839126587,
+ "step": 1170
+ },
+ {
+ "epoch": 0.6143756558237146,
+ "grad_norm": 0.15916219613603688,
+ "learning_rate": 1.9588114573885664e-05,
+ "loss": 0.4546983540058136,
+ "step": 1171
+ },
+ {
+ "epoch": 0.614900314795383,
+ "grad_norm": 0.1660520165166075,
+ "learning_rate": 1.95534078539773e-05,
+ "loss": 0.4601942300796509,
+ "step": 1172
+ },
+ {
+ "epoch": 0.6154249737670514,
+ "grad_norm": 0.16924794578577768,
+ "learning_rate": 1.9518722723180877e-05,
+ "loss": 0.45997846126556396,
+ "step": 1173
+ },
+ {
+ "epoch": 0.6159496327387198,
+ "grad_norm": 0.16597646133045382,
+ "learning_rate": 1.9484059275728127e-05,
+ "loss": 0.4594605565071106,
+ "step": 1174
+ },
+ {
+ "epoch": 0.6164742917103883,
+ "grad_norm": 0.16964243792519748,
+ "learning_rate": 1.9449417605791876e-05,
+ "loss": 0.4638690650463104,
+ "step": 1175
+ },
+ {
+ "epoch": 0.6169989506820567,
+ "grad_norm": 0.1758591426791696,
+ "learning_rate": 1.9414797807485796e-05,
+ "loss": 0.4666447639465332,
+ "step": 1176
+ },
+ {
+ "epoch": 0.6175236096537251,
+ "grad_norm": 0.16217661227891753,
+ "learning_rate": 1.9380199974864126e-05,
+ "loss": 0.44252535700798035,
+ "step": 1177
+ },
+ {
+ "epoch": 0.6180482686253935,
+ "grad_norm": 0.1576446184499019,
+ "learning_rate": 1.934562420192142e-05,
+ "loss": 0.4499641954898834,
+ "step": 1178
+ },
+ {
+ "epoch": 0.618572927597062,
+ "grad_norm": 0.16231008092636912,
+ "learning_rate": 1.9311070582592327e-05,
+ "loss": 0.4561542570590973,
+ "step": 1179
+ },
+ {
+ "epoch": 0.6190975865687304,
+ "grad_norm": 0.16790382249577043,
+ "learning_rate": 1.9276539210751292e-05,
+ "loss": 0.4484325647354126,
+ "step": 1180
+ },
+ {
+ "epoch": 0.6196222455403987,
+ "grad_norm": 0.18794729932605855,
+ "learning_rate": 1.9242030180212323e-05,
+ "loss": 0.44169026613235474,
+ "step": 1181
+ },
+ {
+ "epoch": 0.6201469045120671,
+ "grad_norm": 0.16948051243889814,
+ "learning_rate": 1.920754358472872e-05,
+ "loss": 0.4592876434326172,
+ "step": 1182
+ },
+ {
+ "epoch": 0.6206715634837355,
+ "grad_norm": 0.16762558710029374,
+ "learning_rate": 1.917307951799286e-05,
+ "loss": 0.4402095079421997,
+ "step": 1183
+ },
+ {
+ "epoch": 0.621196222455404,
+ "grad_norm": 0.18170280116387003,
+ "learning_rate": 1.9138638073635885e-05,
+ "loss": 0.4633026123046875,
+ "step": 1184
+ },
+ {
+ "epoch": 0.6217208814270724,
+ "grad_norm": 0.1718051709136108,
+ "learning_rate": 1.9104219345227486e-05,
+ "loss": 0.4510066509246826,
+ "step": 1185
+ },
+ {
+ "epoch": 0.6222455403987408,
+ "grad_norm": 0.16829066326345576,
+ "learning_rate": 1.906982342627565e-05,
+ "loss": 0.4632919430732727,
+ "step": 1186
+ },
+ {
+ "epoch": 0.6227701993704092,
+ "grad_norm": 0.16669141808548274,
+ "learning_rate": 1.9035450410226395e-05,
+ "loss": 0.455544650554657,
+ "step": 1187
+ },
+ {
+ "epoch": 0.6232948583420777,
+ "grad_norm": 0.16679996559722232,
+ "learning_rate": 1.90011003904635e-05,
+ "loss": 0.459994375705719,
+ "step": 1188
+ },
+ {
+ "epoch": 0.6238195173137461,
+ "grad_norm": 0.17379754713278164,
+ "learning_rate": 1.8966773460308284e-05,
+ "loss": 0.4576749801635742,
+ "step": 1189
+ },
+ {
+ "epoch": 0.6243441762854145,
+ "grad_norm": 0.17088539733389246,
+ "learning_rate": 1.8932469713019335e-05,
+ "loss": 0.44853052496910095,
+ "step": 1190
+ },
+ {
+ "epoch": 0.6248688352570829,
+ "grad_norm": 0.15556883889201015,
+ "learning_rate": 1.8898189241792248e-05,
+ "loss": 0.44060447812080383,
+ "step": 1191
+ },
+ {
+ "epoch": 0.6253934942287513,
+ "grad_norm": 0.1640093917439215,
+ "learning_rate": 1.8863932139759388e-05,
+ "loss": 0.4531911611557007,
+ "step": 1192
+ },
+ {
+ "epoch": 0.6259181532004198,
+ "grad_norm": 0.16444284826591576,
+ "learning_rate": 1.882969849998964e-05,
+ "loss": 0.44176363945007324,
+ "step": 1193
+ },
+ {
+ "epoch": 0.6264428121720882,
+ "grad_norm": 0.1696462520995642,
+ "learning_rate": 1.8795488415488143e-05,
+ "loss": 0.44516652822494507,
+ "step": 1194
+ },
+ {
+ "epoch": 0.6269674711437565,
+ "grad_norm": 0.16810724705191796,
+ "learning_rate": 1.8761301979196032e-05,
+ "loss": 0.4542563855648041,
+ "step": 1195
+ },
+ {
+ "epoch": 0.6274921301154249,
+ "grad_norm": 0.17334813313855824,
+ "learning_rate": 1.8727139283990204e-05,
+ "loss": 0.47769248485565186,
+ "step": 1196
+ },
+ {
+ "epoch": 0.6280167890870934,
+ "grad_norm": 0.16308212035303235,
+ "learning_rate": 1.8693000422683043e-05,
+ "loss": 0.4407799243927002,
+ "step": 1197
+ },
+ {
+ "epoch": 0.6285414480587618,
+ "grad_norm": 0.16853319139928977,
+ "learning_rate": 1.8658885488022204e-05,
+ "loss": 0.4565497040748596,
+ "step": 1198
+ },
+ {
+ "epoch": 0.6290661070304302,
+ "grad_norm": 0.16694968781284703,
+ "learning_rate": 1.8624794572690328e-05,
+ "loss": 0.44522374868392944,
+ "step": 1199
+ },
+ {
+ "epoch": 0.6295907660020986,
+ "grad_norm": 0.16418579468339103,
+ "learning_rate": 1.8590727769304788e-05,
+ "loss": 0.4441048502922058,
+ "step": 1200
+ },
+ {
+ "epoch": 0.630115424973767,
+ "grad_norm": 0.1661930777124156,
+ "learning_rate": 1.8556685170417486e-05,
+ "loss": 0.457327663898468,
+ "step": 1201
+ },
+ {
+ "epoch": 0.6306400839454355,
+ "grad_norm": 0.16536630027039814,
+ "learning_rate": 1.852266686851452e-05,
+ "loss": 0.462494432926178,
+ "step": 1202
+ },
+ {
+ "epoch": 0.6311647429171039,
+ "grad_norm": 0.16611474088307454,
+ "learning_rate": 1.8488672956016013e-05,
+ "loss": 0.4557701647281647,
+ "step": 1203
+ },
+ {
+ "epoch": 0.6316894018887723,
+ "grad_norm": 0.16759343079749697,
+ "learning_rate": 1.8454703525275813e-05,
+ "loss": 0.45069995522499084,
+ "step": 1204
+ },
+ {
+ "epoch": 0.6322140608604407,
+ "grad_norm": 0.16186676294485336,
+ "learning_rate": 1.8420758668581255e-05,
+ "loss": 0.44954800605773926,
+ "step": 1205
+ },
+ {
+ "epoch": 0.6327387198321092,
+ "grad_norm": 0.15687159835368553,
+ "learning_rate": 1.8386838478152916e-05,
+ "loss": 0.4490429759025574,
+ "step": 1206
+ },
+ {
+ "epoch": 0.6332633788037776,
+ "grad_norm": 0.16242951976115577,
+ "learning_rate": 1.8352943046144372e-05,
+ "loss": 0.45819586515426636,
+ "step": 1207
+ },
+ {
+ "epoch": 0.633788037775446,
+ "grad_norm": 0.16516087266517232,
+ "learning_rate": 1.831907246464191e-05,
+ "loss": 0.44769883155822754,
+ "step": 1208
+ },
+ {
+ "epoch": 0.6343126967471143,
+ "grad_norm": 0.16709475871512644,
+ "learning_rate": 1.8285226825664332e-05,
+ "loss": 0.4699128568172455,
+ "step": 1209
+ },
+ {
+ "epoch": 0.6348373557187827,
+ "grad_norm": 0.16288623150595774,
+ "learning_rate": 1.8251406221162636e-05,
+ "loss": 0.45322558283805847,
+ "step": 1210
+ },
+ {
+ "epoch": 0.6353620146904512,
+ "grad_norm": 0.1675722848054237,
+ "learning_rate": 1.821761074301986e-05,
+ "loss": 0.4752347469329834,
+ "step": 1211
+ },
+ {
+ "epoch": 0.6358866736621196,
+ "grad_norm": 0.163285963514504,
+ "learning_rate": 1.818384048305074e-05,
+ "loss": 0.441958487033844,
+ "step": 1212
+ },
+ {
+ "epoch": 0.636411332633788,
+ "grad_norm": 0.17078102287557764,
+ "learning_rate": 1.8150095533001517e-05,
+ "loss": 0.4551333487033844,
+ "step": 1213
+ },
+ {
+ "epoch": 0.6369359916054564,
+ "grad_norm": 0.16188540746097138,
+ "learning_rate": 1.811637598454966e-05,
+ "loss": 0.4470779597759247,
+ "step": 1214
+ },
+ {
+ "epoch": 0.6374606505771249,
+ "grad_norm": 0.15235628218958286,
+ "learning_rate": 1.808268192930364e-05,
+ "loss": 0.43604153394699097,
+ "step": 1215
+ },
+ {
+ "epoch": 0.6379853095487933,
+ "grad_norm": 0.1604601949959104,
+ "learning_rate": 1.8049013458802665e-05,
+ "loss": 0.4445713758468628,
+ "step": 1216
+ },
+ {
+ "epoch": 0.6385099685204617,
+ "grad_norm": 0.15511797131215901,
+ "learning_rate": 1.8015370664516427e-05,
+ "loss": 0.4455301761627197,
+ "step": 1217
+ },
+ {
+ "epoch": 0.6390346274921301,
+ "grad_norm": 0.16815669546944095,
+ "learning_rate": 1.798175363784487e-05,
+ "loss": 0.46135014295578003,
+ "step": 1218
+ },
+ {
+ "epoch": 0.6395592864637986,
+ "grad_norm": 0.161558177688922,
+ "learning_rate": 1.7948162470117917e-05,
+ "loss": 0.4519304037094116,
+ "step": 1219
+ },
+ {
+ "epoch": 0.640083945435467,
+ "grad_norm": 0.16677842130048232,
+ "learning_rate": 1.7914597252595265e-05,
+ "loss": 0.4550550878047943,
+ "step": 1220
+ },
+ {
+ "epoch": 0.6406086044071354,
+ "grad_norm": 0.1597331480423138,
+ "learning_rate": 1.7881058076466095e-05,
+ "loss": 0.441120445728302,
+ "step": 1221
+ },
+ {
+ "epoch": 0.6411332633788038,
+ "grad_norm": 0.15951106545246824,
+ "learning_rate": 1.7847545032848844e-05,
+ "loss": 0.4471382200717926,
+ "step": 1222
+ },
+ {
+ "epoch": 0.6416579223504721,
+ "grad_norm": 0.16157787674667565,
+ "learning_rate": 1.7814058212790947e-05,
+ "loss": 0.4550648331642151,
+ "step": 1223
+ },
+ {
+ "epoch": 0.6421825813221406,
+ "grad_norm": 0.16989186763352415,
+ "learning_rate": 1.7780597707268583e-05,
+ "loss": 0.4515089988708496,
+ "step": 1224
+ },
+ {
+ "epoch": 0.642707240293809,
+ "grad_norm": 0.15952153223099508,
+ "learning_rate": 1.7747163607186468e-05,
+ "loss": 0.4658638834953308,
+ "step": 1225
+ },
+ {
+ "epoch": 0.6432318992654774,
+ "grad_norm": 0.15205926699647218,
+ "learning_rate": 1.7713756003377567e-05,
+ "loss": 0.4330868422985077,
+ "step": 1226
+ },
+ {
+ "epoch": 0.6437565582371458,
+ "grad_norm": 0.15784806480936794,
+ "learning_rate": 1.7680374986602865e-05,
+ "loss": 0.4506068527698517,
+ "step": 1227
+ },
+ {
+ "epoch": 0.6442812172088143,
+ "grad_norm": 0.16006709280686687,
+ "learning_rate": 1.7647020647551105e-05,
+ "loss": 0.446941077709198,
+ "step": 1228
+ },
+ {
+ "epoch": 0.6448058761804827,
+ "grad_norm": 0.16064631158222062,
+ "learning_rate": 1.761369307683856e-05,
+ "loss": 0.4418525695800781,
+ "step": 1229
+ },
+ {
+ "epoch": 0.6453305351521511,
+ "grad_norm": 0.16454136480444392,
+ "learning_rate": 1.7580392365008767e-05,
+ "loss": 0.46268391609191895,
+ "step": 1230
+ },
+ {
+ "epoch": 0.6458551941238195,
+ "grad_norm": 0.15401246535771462,
+ "learning_rate": 1.754711860253233e-05,
+ "loss": 0.4372989535331726,
+ "step": 1231
+ },
+ {
+ "epoch": 0.646379853095488,
+ "grad_norm": 0.1544567975028383,
+ "learning_rate": 1.7513871879806603e-05,
+ "loss": 0.44806575775146484,
+ "step": 1232
+ },
+ {
+ "epoch": 0.6469045120671564,
+ "grad_norm": 0.16411225112730965,
+ "learning_rate": 1.748065228715549e-05,
+ "loss": 0.43782058358192444,
+ "step": 1233
+ },
+ {
+ "epoch": 0.6474291710388248,
+ "grad_norm": 0.1552767832972093,
+ "learning_rate": 1.7447459914829183e-05,
+ "loss": 0.44531378149986267,
+ "step": 1234
+ },
+ {
+ "epoch": 0.6479538300104932,
+ "grad_norm": 0.16214695257123513,
+ "learning_rate": 1.741429485300392e-05,
+ "loss": 0.47399842739105225,
+ "step": 1235
+ },
+ {
+ "epoch": 0.6484784889821616,
+ "grad_norm": 0.16684861111484112,
+ "learning_rate": 1.7381157191781766e-05,
+ "loss": 0.4654282331466675,
+ "step": 1236
+ },
+ {
+ "epoch": 0.64900314795383,
+ "grad_norm": 0.15901864178630273,
+ "learning_rate": 1.7348047021190334e-05,
+ "loss": 0.45128414034843445,
+ "step": 1237
+ },
+ {
+ "epoch": 0.6495278069254984,
+ "grad_norm": 0.15360991961742707,
+ "learning_rate": 1.7314964431182523e-05,
+ "loss": 0.4303375482559204,
+ "step": 1238
+ },
+ {
+ "epoch": 0.6500524658971668,
+ "grad_norm": 0.1633151574385741,
+ "learning_rate": 1.728190951163633e-05,
+ "loss": 0.45568403601646423,
+ "step": 1239
+ },
+ {
+ "epoch": 0.6505771248688352,
+ "grad_norm": 0.16377853571058645,
+ "learning_rate": 1.724888235235457e-05,
+ "loss": 0.46503543853759766,
+ "step": 1240
+ },
+ {
+ "epoch": 0.6511017838405037,
+ "grad_norm": 0.1595023801275368,
+ "learning_rate": 1.7215883043064655e-05,
+ "loss": 0.46337294578552246,
+ "step": 1241
+ },
+ {
+ "epoch": 0.6516264428121721,
+ "grad_norm": 0.16324796182212292,
+ "learning_rate": 1.7182911673418313e-05,
+ "loss": 0.46197354793548584,
+ "step": 1242
+ },
+ {
+ "epoch": 0.6521511017838405,
+ "grad_norm": 0.16353734830879765,
+ "learning_rate": 1.714996833299138e-05,
+ "loss": 0.45741671323776245,
+ "step": 1243
+ },
+ {
+ "epoch": 0.6526757607555089,
+ "grad_norm": 0.15941587678263386,
+ "learning_rate": 1.7117053111283528e-05,
+ "loss": 0.4514503479003906,
+ "step": 1244
+ },
+ {
+ "epoch": 0.6532004197271774,
+ "grad_norm": 0.1642653230145008,
+ "learning_rate": 1.708416609771804e-05,
+ "loss": 0.4437655210494995,
+ "step": 1245
+ },
+ {
+ "epoch": 0.6537250786988458,
+ "grad_norm": 0.15814155519130152,
+ "learning_rate": 1.7051307381641587e-05,
+ "loss": 0.4314163327217102,
+ "step": 1246
+ },
+ {
+ "epoch": 0.6542497376705142,
+ "grad_norm": 0.16086078370768056,
+ "learning_rate": 1.701847705232394e-05,
+ "loss": 0.4526430666446686,
+ "step": 1247
+ },
+ {
+ "epoch": 0.6547743966421826,
+ "grad_norm": 0.16389360190715183,
+ "learning_rate": 1.698567519895775e-05,
+ "loss": 0.4278966188430786,
+ "step": 1248
+ },
+ {
+ "epoch": 0.655299055613851,
+ "grad_norm": 0.16303142287804262,
+ "learning_rate": 1.6952901910658316e-05,
+ "loss": 0.45778220891952515,
+ "step": 1249
+ },
+ {
+ "epoch": 0.6558237145855194,
+ "grad_norm": 0.1675209956904048,
+ "learning_rate": 1.6920157276463318e-05,
+ "loss": 0.4489403963088989,
+ "step": 1250
+ },
+ {
+ "epoch": 0.6563483735571878,
+ "grad_norm": 0.16228311551074323,
+ "learning_rate": 1.6887441385332604e-05,
+ "loss": 0.4534342885017395,
+ "step": 1251
+ },
+ {
+ "epoch": 0.6568730325288562,
+ "grad_norm": 0.16408957851046913,
+ "learning_rate": 1.6854754326147916e-05,
+ "loss": 0.4549208879470825,
+ "step": 1252
+ },
+ {
+ "epoch": 0.6573976915005246,
+ "grad_norm": 0.15967922111432845,
+ "learning_rate": 1.682209618771268e-05,
+ "loss": 0.4451141357421875,
+ "step": 1253
+ },
+ {
+ "epoch": 0.657922350472193,
+ "grad_norm": 0.16440698722204208,
+ "learning_rate": 1.678946705875176e-05,
+ "loss": 0.454200804233551,
+ "step": 1254
+ },
+ {
+ "epoch": 0.6584470094438615,
+ "grad_norm": 0.16251120184938497,
+ "learning_rate": 1.675686702791117e-05,
+ "loss": 0.4580027461051941,
+ "step": 1255
+ },
+ {
+ "epoch": 0.6589716684155299,
+ "grad_norm": 0.1643717386150457,
+ "learning_rate": 1.6724296183757924e-05,
+ "loss": 0.4422311782836914,
+ "step": 1256
+ },
+ {
+ "epoch": 0.6594963273871983,
+ "grad_norm": 0.1644586414378722,
+ "learning_rate": 1.66917546147797e-05,
+ "loss": 0.4593179225921631,
+ "step": 1257
+ },
+ {
+ "epoch": 0.6600209863588667,
+ "grad_norm": 0.16751074186967685,
+ "learning_rate": 1.6659242409384657e-05,
+ "loss": 0.457930326461792,
+ "step": 1258
+ },
+ {
+ "epoch": 0.6605456453305352,
+ "grad_norm": 0.15398418609155334,
+ "learning_rate": 1.6626759655901182e-05,
+ "loss": 0.44002899527549744,
+ "step": 1259
+ },
+ {
+ "epoch": 0.6610703043022036,
+ "grad_norm": 0.16328113175790226,
+ "learning_rate": 1.659430644257763e-05,
+ "loss": 0.4500438868999481,
+ "step": 1260
+ },
+ {
+ "epoch": 0.661594963273872,
+ "grad_norm": 0.1694244340615106,
+ "learning_rate": 1.656188285758214e-05,
+ "loss": 0.4818427264690399,
+ "step": 1261
+ },
+ {
+ "epoch": 0.6621196222455404,
+ "grad_norm": 0.1578320701754368,
+ "learning_rate": 1.652948898900232e-05,
+ "loss": 0.4492436647415161,
+ "step": 1262
+ },
+ {
+ "epoch": 0.6626442812172089,
+ "grad_norm": 0.16538865917828469,
+ "learning_rate": 1.6497124924845064e-05,
+ "loss": 0.45070338249206543,
+ "step": 1263
+ },
+ {
+ "epoch": 0.6631689401888772,
+ "grad_norm": 0.15670817848256852,
+ "learning_rate": 1.6464790753036296e-05,
+ "loss": 0.45439285039901733,
+ "step": 1264
+ },
+ {
+ "epoch": 0.6636935991605456,
+ "grad_norm": 0.16184786585965885,
+ "learning_rate": 1.64324865614207e-05,
+ "loss": 0.446522980928421,
+ "step": 1265
+ },
+ {
+ "epoch": 0.664218258132214,
+ "grad_norm": 0.15719726462251984,
+ "learning_rate": 1.6400212437761554e-05,
+ "loss": 0.4495280981063843,
+ "step": 1266
+ },
+ {
+ "epoch": 0.6647429171038824,
+ "grad_norm": 0.1680804414134405,
+ "learning_rate": 1.6367968469740416e-05,
+ "loss": 0.4652596712112427,
+ "step": 1267
+ },
+ {
+ "epoch": 0.6652675760755509,
+ "grad_norm": 0.15919541846683566,
+ "learning_rate": 1.6335754744956937e-05,
+ "loss": 0.43842798471450806,
+ "step": 1268
+ },
+ {
+ "epoch": 0.6657922350472193,
+ "grad_norm": 0.16322634187139418,
+ "learning_rate": 1.630357135092859e-05,
+ "loss": 0.45242664217948914,
+ "step": 1269
+ },
+ {
+ "epoch": 0.6663168940188877,
+ "grad_norm": 0.15644867734734783,
+ "learning_rate": 1.6271418375090445e-05,
+ "loss": 0.44801655411720276,
+ "step": 1270
+ },
+ {
+ "epoch": 0.6668415529905561,
+ "grad_norm": 0.1514717416124423,
+ "learning_rate": 1.623929590479496e-05,
+ "loss": 0.4506194293498993,
+ "step": 1271
+ },
+ {
+ "epoch": 0.6673662119622246,
+ "grad_norm": 0.1740345085495889,
+ "learning_rate": 1.6207204027311683e-05,
+ "loss": 0.4818177819252014,
+ "step": 1272
+ },
+ {
+ "epoch": 0.667890870933893,
+ "grad_norm": 0.15797701900765457,
+ "learning_rate": 1.6175142829827067e-05,
+ "loss": 0.4639382064342499,
+ "step": 1273
+ },
+ {
+ "epoch": 0.6684155299055614,
+ "grad_norm": 0.15663812316653875,
+ "learning_rate": 1.6143112399444198e-05,
+ "loss": 0.46217501163482666,
+ "step": 1274
+ },
+ {
+ "epoch": 0.6689401888772298,
+ "grad_norm": 0.15474018977043758,
+ "learning_rate": 1.6111112823182592e-05,
+ "loss": 0.43691688776016235,
+ "step": 1275
+ },
+ {
+ "epoch": 0.6694648478488983,
+ "grad_norm": 0.1615634942529431,
+ "learning_rate": 1.6079144187977923e-05,
+ "loss": 0.4412601590156555,
+ "step": 1276
+ },
+ {
+ "epoch": 0.6699895068205667,
+ "grad_norm": 0.15440022842754444,
+ "learning_rate": 1.604720658068184e-05,
+ "loss": 0.4523273706436157,
+ "step": 1277
+ },
+ {
+ "epoch": 0.670514165792235,
+ "grad_norm": 0.1634985665776242,
+ "learning_rate": 1.6015300088061664e-05,
+ "loss": 0.46264398097991943,
+ "step": 1278
+ },
+ {
+ "epoch": 0.6710388247639034,
+ "grad_norm": 0.15037886370689527,
+ "learning_rate": 1.598342479680018e-05,
+ "loss": 0.451465904712677,
+ "step": 1279
+ },
+ {
+ "epoch": 0.6715634837355718,
+ "grad_norm": 0.16895242437746613,
+ "learning_rate": 1.595158079349542e-05,
+ "loss": 0.4466145634651184,
+ "step": 1280
+ },
+ {
+ "epoch": 0.6720881427072403,
+ "grad_norm": 0.1629548326999642,
+ "learning_rate": 1.5919768164660425e-05,
+ "loss": 0.45415782928466797,
+ "step": 1281
+ },
+ {
+ "epoch": 0.6726128016789087,
+ "grad_norm": 0.1563872102268991,
+ "learning_rate": 1.588798699672298e-05,
+ "loss": 0.4386737048625946,
+ "step": 1282
+ },
+ {
+ "epoch": 0.6731374606505771,
+ "grad_norm": 0.16038172773344717,
+ "learning_rate": 1.5856237376025403e-05,
+ "loss": 0.446260929107666,
+ "step": 1283
+ },
+ {
+ "epoch": 0.6736621196222455,
+ "grad_norm": 0.16189780714407623,
+ "learning_rate": 1.582451938882431e-05,
+ "loss": 0.45458072423934937,
+ "step": 1284
+ },
+ {
+ "epoch": 0.674186778593914,
+ "grad_norm": 0.15204521785049285,
+ "learning_rate": 1.5792833121290367e-05,
+ "loss": 0.4425617754459381,
+ "step": 1285
+ },
+ {
+ "epoch": 0.6747114375655824,
+ "grad_norm": 0.16237376704792778,
+ "learning_rate": 1.5761178659508063e-05,
+ "loss": 0.44631800055503845,
+ "step": 1286
+ },
+ {
+ "epoch": 0.6752360965372508,
+ "grad_norm": 0.1582063618466069,
+ "learning_rate": 1.57295560894755e-05,
+ "loss": 0.44280534982681274,
+ "step": 1287
+ },
+ {
+ "epoch": 0.6757607555089192,
+ "grad_norm": 0.16587032349847394,
+ "learning_rate": 1.5697965497104116e-05,
+ "loss": 0.4495997130870819,
+ "step": 1288
+ },
+ {
+ "epoch": 0.6762854144805877,
+ "grad_norm": 0.1557269262755563,
+ "learning_rate": 1.566640696821848e-05,
+ "loss": 0.4493359625339508,
+ "step": 1289
+ },
+ {
+ "epoch": 0.6768100734522561,
+ "grad_norm": 0.16038490357012256,
+ "learning_rate": 1.5634880588556043e-05,
+ "loss": 0.4447544813156128,
+ "step": 1290
+ },
+ {
+ "epoch": 0.6773347324239245,
+ "grad_norm": 0.15355201908546406,
+ "learning_rate": 1.560338644376692e-05,
+ "loss": 0.4468432664871216,
+ "step": 1291
+ },
+ {
+ "epoch": 0.6778593913955928,
+ "grad_norm": 0.15077009515589776,
+ "learning_rate": 1.5571924619413668e-05,
+ "loss": 0.44959044456481934,
+ "step": 1292
+ },
+ {
+ "epoch": 0.6783840503672612,
+ "grad_norm": 0.1583038146613965,
+ "learning_rate": 1.5540495200971012e-05,
+ "loss": 0.42450425028800964,
+ "step": 1293
+ },
+ {
+ "epoch": 0.6789087093389297,
+ "grad_norm": 0.15950687190598087,
+ "learning_rate": 1.5509098273825635e-05,
+ "loss": 0.4527238607406616,
+ "step": 1294
+ },
+ {
+ "epoch": 0.6794333683105981,
+ "grad_norm": 0.1526262208847808,
+ "learning_rate": 1.5477733923275968e-05,
+ "loss": 0.4379724860191345,
+ "step": 1295
+ },
+ {
+ "epoch": 0.6799580272822665,
+ "grad_norm": 0.15541579650021778,
+ "learning_rate": 1.5446402234531934e-05,
+ "loss": 0.4462166130542755,
+ "step": 1296
+ },
+ {
+ "epoch": 0.6804826862539349,
+ "grad_norm": 0.15483930734313786,
+ "learning_rate": 1.5415103292714718e-05,
+ "loss": 0.4390607178211212,
+ "step": 1297
+ },
+ {
+ "epoch": 0.6810073452256034,
+ "grad_norm": 0.1578005731149154,
+ "learning_rate": 1.5383837182856532e-05,
+ "loss": 0.44462254643440247,
+ "step": 1298
+ },
+ {
+ "epoch": 0.6815320041972718,
+ "grad_norm": 0.15830183884958823,
+ "learning_rate": 1.5352603989900406e-05,
+ "loss": 0.4563537538051605,
+ "step": 1299
+ },
+ {
+ "epoch": 0.6820566631689402,
+ "grad_norm": 0.15605472163445522,
+ "learning_rate": 1.5321403798699923e-05,
+ "loss": 0.43852412700653076,
+ "step": 1300
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 3.440896142999552e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1350/chat_template.jinja b/checkpoints/checkpoint-1350/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1350/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1350/config.json b/checkpoints/checkpoint-1350/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1350/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1350/generation_config.json b/checkpoints/checkpoint-1350/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1350/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1350/model.safetensors b/checkpoints/checkpoint-1350/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..5854be7f471783e22c46556e2a772e018faad973
--- /dev/null
+++ b/checkpoints/checkpoint-1350/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ffcf21d63ad779fd9373a3560a28cf218cefed581250fd6956df9e12527a1404
+size 37983421000
diff --git a/checkpoints/checkpoint-1350/tokenizer.json b/checkpoints/checkpoint-1350/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1350/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1350/tokenizer_config.json b/checkpoints/checkpoint-1350/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1350/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1350/trainer_state.json b/checkpoints/checkpoint-1350/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..4f900ac41b1161d4a61c6950a616b5642ac1ac6c
--- /dev/null
+++ b/checkpoints/checkpoint-1350/trainer_state.json
@@ -0,0 +1,9484 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.7082896117523609,
+ "eval_steps": 500,
+ "global_step": 1350,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ },
+ {
+ "epoch": 0.5776495278069255,
+ "grad_norm": 0.1862053032854098,
+ "learning_rate": 2.2065553964033126e-05,
+ "loss": 0.4816875457763672,
+ "step": 1101
+ },
+ {
+ "epoch": 0.5781741867785939,
+ "grad_norm": 0.16996015894363117,
+ "learning_rate": 2.20295734045532e-05,
+ "loss": 0.4471197724342346,
+ "step": 1102
+ },
+ {
+ "epoch": 0.5786988457502623,
+ "grad_norm": 0.16889227008390692,
+ "learning_rate": 2.1993607706997725e-05,
+ "loss": 0.468391478061676,
+ "step": 1103
+ },
+ {
+ "epoch": 0.5792235047219307,
+ "grad_norm": 0.16670931681562373,
+ "learning_rate": 2.195765696907745e-05,
+ "loss": 0.44934162497520447,
+ "step": 1104
+ },
+ {
+ "epoch": 0.5797481636935992,
+ "grad_norm": 0.16328418741223716,
+ "learning_rate": 2.1921721288462482e-05,
+ "loss": 0.4558270275592804,
+ "step": 1105
+ },
+ {
+ "epoch": 0.5802728226652676,
+ "grad_norm": 0.17364660408703936,
+ "learning_rate": 2.188580076278203e-05,
+ "loss": 0.4597751796245575,
+ "step": 1106
+ },
+ {
+ "epoch": 0.580797481636936,
+ "grad_norm": 0.1628618453800601,
+ "learning_rate": 2.18498954896241e-05,
+ "loss": 0.4521859288215637,
+ "step": 1107
+ },
+ {
+ "epoch": 0.5813221406086044,
+ "grad_norm": 0.16813817951871482,
+ "learning_rate": 2.1814005566535302e-05,
+ "loss": 0.4581005573272705,
+ "step": 1108
+ },
+ {
+ "epoch": 0.5818467995802729,
+ "grad_norm": 0.16825923629726025,
+ "learning_rate": 2.1778131091020516e-05,
+ "loss": 0.4433916211128235,
+ "step": 1109
+ },
+ {
+ "epoch": 0.5823714585519413,
+ "grad_norm": 0.17194743467883944,
+ "learning_rate": 2.1742272160542654e-05,
+ "loss": 0.47296425700187683,
+ "step": 1110
+ },
+ {
+ "epoch": 0.5828961175236097,
+ "grad_norm": 0.16947730546039969,
+ "learning_rate": 2.170642887252241e-05,
+ "loss": 0.4815937876701355,
+ "step": 1111
+ },
+ {
+ "epoch": 0.583420776495278,
+ "grad_norm": 0.1657449928668473,
+ "learning_rate": 2.167060132433798e-05,
+ "loss": 0.4479164779186249,
+ "step": 1112
+ },
+ {
+ "epoch": 0.5839454354669464,
+ "grad_norm": 0.1658952516607364,
+ "learning_rate": 2.1634789613324778e-05,
+ "loss": 0.4637463688850403,
+ "step": 1113
+ },
+ {
+ "epoch": 0.5844700944386149,
+ "grad_norm": 0.16777002188998236,
+ "learning_rate": 2.1598993836775218e-05,
+ "loss": 0.457524836063385,
+ "step": 1114
+ },
+ {
+ "epoch": 0.5849947534102833,
+ "grad_norm": 0.15837566447341322,
+ "learning_rate": 2.1563214091938405e-05,
+ "loss": 0.44938087463378906,
+ "step": 1115
+ },
+ {
+ "epoch": 0.5855194123819517,
+ "grad_norm": 0.17162206617035314,
+ "learning_rate": 2.1527450476019896e-05,
+ "loss": 0.4393450617790222,
+ "step": 1116
+ },
+ {
+ "epoch": 0.5860440713536201,
+ "grad_norm": 0.16224101921433814,
+ "learning_rate": 2.1491703086181426e-05,
+ "loss": 0.451913982629776,
+ "step": 1117
+ },
+ {
+ "epoch": 0.5865687303252886,
+ "grad_norm": 0.1736253504575144,
+ "learning_rate": 2.145597201954066e-05,
+ "loss": 0.4611290395259857,
+ "step": 1118
+ },
+ {
+ "epoch": 0.587093389296957,
+ "grad_norm": 0.16308408333737234,
+ "learning_rate": 2.1420257373170903e-05,
+ "loss": 0.45898568630218506,
+ "step": 1119
+ },
+ {
+ "epoch": 0.5876180482686254,
+ "grad_norm": 0.16948634794175524,
+ "learning_rate": 2.1384559244100848e-05,
+ "loss": 0.45982882380485535,
+ "step": 1120
+ },
+ {
+ "epoch": 0.5881427072402938,
+ "grad_norm": 0.16676366538565252,
+ "learning_rate": 2.1348877729314336e-05,
+ "loss": 0.44826576113700867,
+ "step": 1121
+ },
+ {
+ "epoch": 0.5886673662119623,
+ "grad_norm": 0.16431125630712295,
+ "learning_rate": 2.1313212925750037e-05,
+ "loss": 0.4616445302963257,
+ "step": 1122
+ },
+ {
+ "epoch": 0.5891920251836307,
+ "grad_norm": 0.1631659061616955,
+ "learning_rate": 2.1277564930301252e-05,
+ "loss": 0.45651188492774963,
+ "step": 1123
+ },
+ {
+ "epoch": 0.5897166841552991,
+ "grad_norm": 0.15805950502689237,
+ "learning_rate": 2.1241933839815604e-05,
+ "loss": 0.4486045837402344,
+ "step": 1124
+ },
+ {
+ "epoch": 0.5902413431269675,
+ "grad_norm": 0.16530288038603502,
+ "learning_rate": 2.1206319751094773e-05,
+ "loss": 0.4485066533088684,
+ "step": 1125
+ },
+ {
+ "epoch": 0.5907660020986358,
+ "grad_norm": 0.16061903194239047,
+ "learning_rate": 2.1170722760894295e-05,
+ "loss": 0.4565330743789673,
+ "step": 1126
+ },
+ {
+ "epoch": 0.5912906610703043,
+ "grad_norm": 0.1643346999353627,
+ "learning_rate": 2.1135142965923195e-05,
+ "loss": 0.45560717582702637,
+ "step": 1127
+ },
+ {
+ "epoch": 0.5918153200419727,
+ "grad_norm": 0.17484644315215753,
+ "learning_rate": 2.109958046284382e-05,
+ "loss": 0.4795891046524048,
+ "step": 1128
+ },
+ {
+ "epoch": 0.5923399790136411,
+ "grad_norm": 0.1623762678622651,
+ "learning_rate": 2.1064035348271528e-05,
+ "loss": 0.47395604848861694,
+ "step": 1129
+ },
+ {
+ "epoch": 0.5928646379853095,
+ "grad_norm": 0.16888824241669015,
+ "learning_rate": 2.1028507718774456e-05,
+ "loss": 0.4601455330848694,
+ "step": 1130
+ },
+ {
+ "epoch": 0.593389296956978,
+ "grad_norm": 0.16291744568806302,
+ "learning_rate": 2.0992997670873193e-05,
+ "loss": 0.46248236298561096,
+ "step": 1131
+ },
+ {
+ "epoch": 0.5939139559286464,
+ "grad_norm": 0.16185598691236988,
+ "learning_rate": 2.0957505301040603e-05,
+ "loss": 0.45085301995277405,
+ "step": 1132
+ },
+ {
+ "epoch": 0.5944386149003148,
+ "grad_norm": 0.1678000806117814,
+ "learning_rate": 2.0922030705701518e-05,
+ "loss": 0.44939014315605164,
+ "step": 1133
+ },
+ {
+ "epoch": 0.5949632738719832,
+ "grad_norm": 0.16104474282288994,
+ "learning_rate": 2.088657398123246e-05,
+ "loss": 0.4313976764678955,
+ "step": 1134
+ },
+ {
+ "epoch": 0.5954879328436516,
+ "grad_norm": 0.1631128968199669,
+ "learning_rate": 2.085113522396141e-05,
+ "loss": 0.4508278965950012,
+ "step": 1135
+ },
+ {
+ "epoch": 0.5960125918153201,
+ "grad_norm": 0.16307688195844533,
+ "learning_rate": 2.0815714530167557e-05,
+ "loss": 0.4467576742172241,
+ "step": 1136
+ },
+ {
+ "epoch": 0.5965372507869885,
+ "grad_norm": 0.16456832304532207,
+ "learning_rate": 2.0780311996080976e-05,
+ "loss": 0.4621703624725342,
+ "step": 1137
+ },
+ {
+ "epoch": 0.5970619097586569,
+ "grad_norm": 0.16346217748325367,
+ "learning_rate": 2.0744927717882435e-05,
+ "loss": 0.4589449465274811,
+ "step": 1138
+ },
+ {
+ "epoch": 0.5975865687303253,
+ "grad_norm": 0.16180691501076994,
+ "learning_rate": 2.0709561791703097e-05,
+ "loss": 0.43698400259017944,
+ "step": 1139
+ },
+ {
+ "epoch": 0.5981112277019937,
+ "grad_norm": 0.15695459976898835,
+ "learning_rate": 2.067421431362427e-05,
+ "loss": 0.452271044254303,
+ "step": 1140
+ },
+ {
+ "epoch": 0.5986358866736621,
+ "grad_norm": 0.15930893553721054,
+ "learning_rate": 2.063888537967713e-05,
+ "loss": 0.44868963956832886,
+ "step": 1141
+ },
+ {
+ "epoch": 0.5991605456453305,
+ "grad_norm": 0.15963802284017928,
+ "learning_rate": 2.0603575085842485e-05,
+ "loss": 0.4376864731311798,
+ "step": 1142
+ },
+ {
+ "epoch": 0.5996852046169989,
+ "grad_norm": 0.16610266900507178,
+ "learning_rate": 2.05682835280505e-05,
+ "loss": 0.44676727056503296,
+ "step": 1143
+ },
+ {
+ "epoch": 0.6002098635886673,
+ "grad_norm": 0.15978253236669612,
+ "learning_rate": 2.053301080218044e-05,
+ "loss": 0.434773713350296,
+ "step": 1144
+ },
+ {
+ "epoch": 0.6007345225603358,
+ "grad_norm": 0.16305141803526324,
+ "learning_rate": 2.0497757004060403e-05,
+ "loss": 0.4529780149459839,
+ "step": 1145
+ },
+ {
+ "epoch": 0.6012591815320042,
+ "grad_norm": 0.1649737864101215,
+ "learning_rate": 2.046252222946707e-05,
+ "loss": 0.4571434259414673,
+ "step": 1146
+ },
+ {
+ "epoch": 0.6017838405036726,
+ "grad_norm": 0.15738824968326598,
+ "learning_rate": 2.0427306574125427e-05,
+ "loss": 0.4446289539337158,
+ "step": 1147
+ },
+ {
+ "epoch": 0.602308499475341,
+ "grad_norm": 0.1731139535185187,
+ "learning_rate": 2.039211013370853e-05,
+ "loss": 0.4551090598106384,
+ "step": 1148
+ },
+ {
+ "epoch": 0.6028331584470095,
+ "grad_norm": 0.15857251900043037,
+ "learning_rate": 2.0356933003837237e-05,
+ "loss": 0.4644010663032532,
+ "step": 1149
+ },
+ {
+ "epoch": 0.6033578174186779,
+ "grad_norm": 0.1589582861134211,
+ "learning_rate": 2.032177528007992e-05,
+ "loss": 0.45741650462150574,
+ "step": 1150
+ },
+ {
+ "epoch": 0.6038824763903463,
+ "grad_norm": 0.15639092706575122,
+ "learning_rate": 2.0286637057952257e-05,
+ "loss": 0.4375547766685486,
+ "step": 1151
+ },
+ {
+ "epoch": 0.6044071353620147,
+ "grad_norm": 0.16335605581337187,
+ "learning_rate": 2.0251518432916933e-05,
+ "loss": 0.4636140465736389,
+ "step": 1152
+ },
+ {
+ "epoch": 0.6049317943336832,
+ "grad_norm": 0.15437437113359356,
+ "learning_rate": 2.0216419500383372e-05,
+ "loss": 0.44632142782211304,
+ "step": 1153
+ },
+ {
+ "epoch": 0.6054564533053515,
+ "grad_norm": 0.15522039198698664,
+ "learning_rate": 2.0181340355707545e-05,
+ "loss": 0.449049174785614,
+ "step": 1154
+ },
+ {
+ "epoch": 0.6059811122770199,
+ "grad_norm": 0.1599951917909323,
+ "learning_rate": 2.0146281094191604e-05,
+ "loss": 0.44395962357521057,
+ "step": 1155
+ },
+ {
+ "epoch": 0.6065057712486883,
+ "grad_norm": 0.16866621620527716,
+ "learning_rate": 2.011124181108373e-05,
+ "loss": 0.45693716406822205,
+ "step": 1156
+ },
+ {
+ "epoch": 0.6070304302203567,
+ "grad_norm": 0.15508935650135075,
+ "learning_rate": 2.0076222601577815e-05,
+ "loss": 0.4525151252746582,
+ "step": 1157
+ },
+ {
+ "epoch": 0.6075550891920252,
+ "grad_norm": 0.16508998118005971,
+ "learning_rate": 2.004122356081321e-05,
+ "loss": 0.466006875038147,
+ "step": 1158
+ },
+ {
+ "epoch": 0.6080797481636936,
+ "grad_norm": 0.15817418864022237,
+ "learning_rate": 2.0006244783874466e-05,
+ "loss": 0.4496110677719116,
+ "step": 1159
+ },
+ {
+ "epoch": 0.608604407135362,
+ "grad_norm": 0.16127270259749824,
+ "learning_rate": 1.997128636579109e-05,
+ "loss": 0.45918628573417664,
+ "step": 1160
+ },
+ {
+ "epoch": 0.6091290661070304,
+ "grad_norm": 0.1702599924159677,
+ "learning_rate": 1.9936348401537292e-05,
+ "loss": 0.47218555212020874,
+ "step": 1161
+ },
+ {
+ "epoch": 0.6096537250786989,
+ "grad_norm": 0.15690176248846127,
+ "learning_rate": 1.9901430986031695e-05,
+ "loss": 0.4516097903251648,
+ "step": 1162
+ },
+ {
+ "epoch": 0.6101783840503673,
+ "grad_norm": 0.1713614405614144,
+ "learning_rate": 1.9866534214137085e-05,
+ "loss": 0.45030856132507324,
+ "step": 1163
+ },
+ {
+ "epoch": 0.6107030430220357,
+ "grad_norm": 0.1670865563026229,
+ "learning_rate": 1.9831658180660195e-05,
+ "loss": 0.45603489875793457,
+ "step": 1164
+ },
+ {
+ "epoch": 0.6112277019937041,
+ "grad_norm": 0.16897539113401952,
+ "learning_rate": 1.9796802980351392e-05,
+ "loss": 0.46314537525177,
+ "step": 1165
+ },
+ {
+ "epoch": 0.6117523609653726,
+ "grad_norm": 0.15832868870697644,
+ "learning_rate": 1.976196870790446e-05,
+ "loss": 0.44784611463546753,
+ "step": 1166
+ },
+ {
+ "epoch": 0.612277019937041,
+ "grad_norm": 0.16595365551891114,
+ "learning_rate": 1.9727155457956315e-05,
+ "loss": 0.4409624934196472,
+ "step": 1167
+ },
+ {
+ "epoch": 0.6128016789087093,
+ "grad_norm": 0.1642630411212017,
+ "learning_rate": 1.969236332508677e-05,
+ "loss": 0.46120351552963257,
+ "step": 1168
+ },
+ {
+ "epoch": 0.6133263378803777,
+ "grad_norm": 0.16481796135868812,
+ "learning_rate": 1.9657592403818253e-05,
+ "loss": 0.45544904470443726,
+ "step": 1169
+ },
+ {
+ "epoch": 0.6138509968520461,
+ "grad_norm": 0.1629832779852364,
+ "learning_rate": 1.962284278861557e-05,
+ "loss": 0.4578462839126587,
+ "step": 1170
+ },
+ {
+ "epoch": 0.6143756558237146,
+ "grad_norm": 0.15916219613603688,
+ "learning_rate": 1.9588114573885664e-05,
+ "loss": 0.4546983540058136,
+ "step": 1171
+ },
+ {
+ "epoch": 0.614900314795383,
+ "grad_norm": 0.1660520165166075,
+ "learning_rate": 1.95534078539773e-05,
+ "loss": 0.4601942300796509,
+ "step": 1172
+ },
+ {
+ "epoch": 0.6154249737670514,
+ "grad_norm": 0.16924794578577768,
+ "learning_rate": 1.9518722723180877e-05,
+ "loss": 0.45997846126556396,
+ "step": 1173
+ },
+ {
+ "epoch": 0.6159496327387198,
+ "grad_norm": 0.16597646133045382,
+ "learning_rate": 1.9484059275728127e-05,
+ "loss": 0.4594605565071106,
+ "step": 1174
+ },
+ {
+ "epoch": 0.6164742917103883,
+ "grad_norm": 0.16964243792519748,
+ "learning_rate": 1.9449417605791876e-05,
+ "loss": 0.4638690650463104,
+ "step": 1175
+ },
+ {
+ "epoch": 0.6169989506820567,
+ "grad_norm": 0.1758591426791696,
+ "learning_rate": 1.9414797807485796e-05,
+ "loss": 0.4666447639465332,
+ "step": 1176
+ },
+ {
+ "epoch": 0.6175236096537251,
+ "grad_norm": 0.16217661227891753,
+ "learning_rate": 1.9380199974864126e-05,
+ "loss": 0.44252535700798035,
+ "step": 1177
+ },
+ {
+ "epoch": 0.6180482686253935,
+ "grad_norm": 0.1576446184499019,
+ "learning_rate": 1.934562420192142e-05,
+ "loss": 0.4499641954898834,
+ "step": 1178
+ },
+ {
+ "epoch": 0.618572927597062,
+ "grad_norm": 0.16231008092636912,
+ "learning_rate": 1.9311070582592327e-05,
+ "loss": 0.4561542570590973,
+ "step": 1179
+ },
+ {
+ "epoch": 0.6190975865687304,
+ "grad_norm": 0.16790382249577043,
+ "learning_rate": 1.9276539210751292e-05,
+ "loss": 0.4484325647354126,
+ "step": 1180
+ },
+ {
+ "epoch": 0.6196222455403987,
+ "grad_norm": 0.18794729932605855,
+ "learning_rate": 1.9242030180212323e-05,
+ "loss": 0.44169026613235474,
+ "step": 1181
+ },
+ {
+ "epoch": 0.6201469045120671,
+ "grad_norm": 0.16948051243889814,
+ "learning_rate": 1.920754358472872e-05,
+ "loss": 0.4592876434326172,
+ "step": 1182
+ },
+ {
+ "epoch": 0.6206715634837355,
+ "grad_norm": 0.16762558710029374,
+ "learning_rate": 1.917307951799286e-05,
+ "loss": 0.4402095079421997,
+ "step": 1183
+ },
+ {
+ "epoch": 0.621196222455404,
+ "grad_norm": 0.18170280116387003,
+ "learning_rate": 1.9138638073635885e-05,
+ "loss": 0.4633026123046875,
+ "step": 1184
+ },
+ {
+ "epoch": 0.6217208814270724,
+ "grad_norm": 0.1718051709136108,
+ "learning_rate": 1.9104219345227486e-05,
+ "loss": 0.4510066509246826,
+ "step": 1185
+ },
+ {
+ "epoch": 0.6222455403987408,
+ "grad_norm": 0.16829066326345576,
+ "learning_rate": 1.906982342627565e-05,
+ "loss": 0.4632919430732727,
+ "step": 1186
+ },
+ {
+ "epoch": 0.6227701993704092,
+ "grad_norm": 0.16669141808548274,
+ "learning_rate": 1.9035450410226395e-05,
+ "loss": 0.455544650554657,
+ "step": 1187
+ },
+ {
+ "epoch": 0.6232948583420777,
+ "grad_norm": 0.16679996559722232,
+ "learning_rate": 1.90011003904635e-05,
+ "loss": 0.459994375705719,
+ "step": 1188
+ },
+ {
+ "epoch": 0.6238195173137461,
+ "grad_norm": 0.17379754713278164,
+ "learning_rate": 1.8966773460308284e-05,
+ "loss": 0.4576749801635742,
+ "step": 1189
+ },
+ {
+ "epoch": 0.6243441762854145,
+ "grad_norm": 0.17088539733389246,
+ "learning_rate": 1.8932469713019335e-05,
+ "loss": 0.44853052496910095,
+ "step": 1190
+ },
+ {
+ "epoch": 0.6248688352570829,
+ "grad_norm": 0.15556883889201015,
+ "learning_rate": 1.8898189241792248e-05,
+ "loss": 0.44060447812080383,
+ "step": 1191
+ },
+ {
+ "epoch": 0.6253934942287513,
+ "grad_norm": 0.1640093917439215,
+ "learning_rate": 1.8863932139759388e-05,
+ "loss": 0.4531911611557007,
+ "step": 1192
+ },
+ {
+ "epoch": 0.6259181532004198,
+ "grad_norm": 0.16444284826591576,
+ "learning_rate": 1.882969849998964e-05,
+ "loss": 0.44176363945007324,
+ "step": 1193
+ },
+ {
+ "epoch": 0.6264428121720882,
+ "grad_norm": 0.1696462520995642,
+ "learning_rate": 1.8795488415488143e-05,
+ "loss": 0.44516652822494507,
+ "step": 1194
+ },
+ {
+ "epoch": 0.6269674711437565,
+ "grad_norm": 0.16810724705191796,
+ "learning_rate": 1.8761301979196032e-05,
+ "loss": 0.4542563855648041,
+ "step": 1195
+ },
+ {
+ "epoch": 0.6274921301154249,
+ "grad_norm": 0.17334813313855824,
+ "learning_rate": 1.8727139283990204e-05,
+ "loss": 0.47769248485565186,
+ "step": 1196
+ },
+ {
+ "epoch": 0.6280167890870934,
+ "grad_norm": 0.16308212035303235,
+ "learning_rate": 1.8693000422683043e-05,
+ "loss": 0.4407799243927002,
+ "step": 1197
+ },
+ {
+ "epoch": 0.6285414480587618,
+ "grad_norm": 0.16853319139928977,
+ "learning_rate": 1.8658885488022204e-05,
+ "loss": 0.4565497040748596,
+ "step": 1198
+ },
+ {
+ "epoch": 0.6290661070304302,
+ "grad_norm": 0.16694968781284703,
+ "learning_rate": 1.8624794572690328e-05,
+ "loss": 0.44522374868392944,
+ "step": 1199
+ },
+ {
+ "epoch": 0.6295907660020986,
+ "grad_norm": 0.16418579468339103,
+ "learning_rate": 1.8590727769304788e-05,
+ "loss": 0.4441048502922058,
+ "step": 1200
+ },
+ {
+ "epoch": 0.630115424973767,
+ "grad_norm": 0.1661930777124156,
+ "learning_rate": 1.8556685170417486e-05,
+ "loss": 0.457327663898468,
+ "step": 1201
+ },
+ {
+ "epoch": 0.6306400839454355,
+ "grad_norm": 0.16536630027039814,
+ "learning_rate": 1.852266686851452e-05,
+ "loss": 0.462494432926178,
+ "step": 1202
+ },
+ {
+ "epoch": 0.6311647429171039,
+ "grad_norm": 0.16611474088307454,
+ "learning_rate": 1.8488672956016013e-05,
+ "loss": 0.4557701647281647,
+ "step": 1203
+ },
+ {
+ "epoch": 0.6316894018887723,
+ "grad_norm": 0.16759343079749697,
+ "learning_rate": 1.8454703525275813e-05,
+ "loss": 0.45069995522499084,
+ "step": 1204
+ },
+ {
+ "epoch": 0.6322140608604407,
+ "grad_norm": 0.16186676294485336,
+ "learning_rate": 1.8420758668581255e-05,
+ "loss": 0.44954800605773926,
+ "step": 1205
+ },
+ {
+ "epoch": 0.6327387198321092,
+ "grad_norm": 0.15687159835368553,
+ "learning_rate": 1.8386838478152916e-05,
+ "loss": 0.4490429759025574,
+ "step": 1206
+ },
+ {
+ "epoch": 0.6332633788037776,
+ "grad_norm": 0.16242951976115577,
+ "learning_rate": 1.8352943046144372e-05,
+ "loss": 0.45819586515426636,
+ "step": 1207
+ },
+ {
+ "epoch": 0.633788037775446,
+ "grad_norm": 0.16516087266517232,
+ "learning_rate": 1.831907246464191e-05,
+ "loss": 0.44769883155822754,
+ "step": 1208
+ },
+ {
+ "epoch": 0.6343126967471143,
+ "grad_norm": 0.16709475871512644,
+ "learning_rate": 1.8285226825664332e-05,
+ "loss": 0.4699128568172455,
+ "step": 1209
+ },
+ {
+ "epoch": 0.6348373557187827,
+ "grad_norm": 0.16288623150595774,
+ "learning_rate": 1.8251406221162636e-05,
+ "loss": 0.45322558283805847,
+ "step": 1210
+ },
+ {
+ "epoch": 0.6353620146904512,
+ "grad_norm": 0.1675722848054237,
+ "learning_rate": 1.821761074301986e-05,
+ "loss": 0.4752347469329834,
+ "step": 1211
+ },
+ {
+ "epoch": 0.6358866736621196,
+ "grad_norm": 0.163285963514504,
+ "learning_rate": 1.818384048305074e-05,
+ "loss": 0.441958487033844,
+ "step": 1212
+ },
+ {
+ "epoch": 0.636411332633788,
+ "grad_norm": 0.17078102287557764,
+ "learning_rate": 1.8150095533001517e-05,
+ "loss": 0.4551333487033844,
+ "step": 1213
+ },
+ {
+ "epoch": 0.6369359916054564,
+ "grad_norm": 0.16188540746097138,
+ "learning_rate": 1.811637598454966e-05,
+ "loss": 0.4470779597759247,
+ "step": 1214
+ },
+ {
+ "epoch": 0.6374606505771249,
+ "grad_norm": 0.15235628218958286,
+ "learning_rate": 1.808268192930364e-05,
+ "loss": 0.43604153394699097,
+ "step": 1215
+ },
+ {
+ "epoch": 0.6379853095487933,
+ "grad_norm": 0.1604601949959104,
+ "learning_rate": 1.8049013458802665e-05,
+ "loss": 0.4445713758468628,
+ "step": 1216
+ },
+ {
+ "epoch": 0.6385099685204617,
+ "grad_norm": 0.15511797131215901,
+ "learning_rate": 1.8015370664516427e-05,
+ "loss": 0.4455301761627197,
+ "step": 1217
+ },
+ {
+ "epoch": 0.6390346274921301,
+ "grad_norm": 0.16815669546944095,
+ "learning_rate": 1.798175363784487e-05,
+ "loss": 0.46135014295578003,
+ "step": 1218
+ },
+ {
+ "epoch": 0.6395592864637986,
+ "grad_norm": 0.161558177688922,
+ "learning_rate": 1.7948162470117917e-05,
+ "loss": 0.4519304037094116,
+ "step": 1219
+ },
+ {
+ "epoch": 0.640083945435467,
+ "grad_norm": 0.16677842130048232,
+ "learning_rate": 1.7914597252595265e-05,
+ "loss": 0.4550550878047943,
+ "step": 1220
+ },
+ {
+ "epoch": 0.6406086044071354,
+ "grad_norm": 0.1597331480423138,
+ "learning_rate": 1.7881058076466095e-05,
+ "loss": 0.441120445728302,
+ "step": 1221
+ },
+ {
+ "epoch": 0.6411332633788038,
+ "grad_norm": 0.15951106545246824,
+ "learning_rate": 1.7847545032848844e-05,
+ "loss": 0.4471382200717926,
+ "step": 1222
+ },
+ {
+ "epoch": 0.6416579223504721,
+ "grad_norm": 0.16157787674667565,
+ "learning_rate": 1.7814058212790947e-05,
+ "loss": 0.4550648331642151,
+ "step": 1223
+ },
+ {
+ "epoch": 0.6421825813221406,
+ "grad_norm": 0.16989186763352415,
+ "learning_rate": 1.7780597707268583e-05,
+ "loss": 0.4515089988708496,
+ "step": 1224
+ },
+ {
+ "epoch": 0.642707240293809,
+ "grad_norm": 0.15952153223099508,
+ "learning_rate": 1.7747163607186468e-05,
+ "loss": 0.4658638834953308,
+ "step": 1225
+ },
+ {
+ "epoch": 0.6432318992654774,
+ "grad_norm": 0.15205926699647218,
+ "learning_rate": 1.7713756003377567e-05,
+ "loss": 0.4330868422985077,
+ "step": 1226
+ },
+ {
+ "epoch": 0.6437565582371458,
+ "grad_norm": 0.15784806480936794,
+ "learning_rate": 1.7680374986602865e-05,
+ "loss": 0.4506068527698517,
+ "step": 1227
+ },
+ {
+ "epoch": 0.6442812172088143,
+ "grad_norm": 0.16006709280686687,
+ "learning_rate": 1.7647020647551105e-05,
+ "loss": 0.446941077709198,
+ "step": 1228
+ },
+ {
+ "epoch": 0.6448058761804827,
+ "grad_norm": 0.16064631158222062,
+ "learning_rate": 1.761369307683856e-05,
+ "loss": 0.4418525695800781,
+ "step": 1229
+ },
+ {
+ "epoch": 0.6453305351521511,
+ "grad_norm": 0.16454136480444392,
+ "learning_rate": 1.7580392365008767e-05,
+ "loss": 0.46268391609191895,
+ "step": 1230
+ },
+ {
+ "epoch": 0.6458551941238195,
+ "grad_norm": 0.15401246535771462,
+ "learning_rate": 1.754711860253233e-05,
+ "loss": 0.4372989535331726,
+ "step": 1231
+ },
+ {
+ "epoch": 0.646379853095488,
+ "grad_norm": 0.1544567975028383,
+ "learning_rate": 1.7513871879806603e-05,
+ "loss": 0.44806575775146484,
+ "step": 1232
+ },
+ {
+ "epoch": 0.6469045120671564,
+ "grad_norm": 0.16411225112730965,
+ "learning_rate": 1.748065228715549e-05,
+ "loss": 0.43782058358192444,
+ "step": 1233
+ },
+ {
+ "epoch": 0.6474291710388248,
+ "grad_norm": 0.1552767832972093,
+ "learning_rate": 1.7447459914829183e-05,
+ "loss": 0.44531378149986267,
+ "step": 1234
+ },
+ {
+ "epoch": 0.6479538300104932,
+ "grad_norm": 0.16214695257123513,
+ "learning_rate": 1.741429485300392e-05,
+ "loss": 0.47399842739105225,
+ "step": 1235
+ },
+ {
+ "epoch": 0.6484784889821616,
+ "grad_norm": 0.16684861111484112,
+ "learning_rate": 1.7381157191781766e-05,
+ "loss": 0.4654282331466675,
+ "step": 1236
+ },
+ {
+ "epoch": 0.64900314795383,
+ "grad_norm": 0.15901864178630273,
+ "learning_rate": 1.7348047021190334e-05,
+ "loss": 0.45128414034843445,
+ "step": 1237
+ },
+ {
+ "epoch": 0.6495278069254984,
+ "grad_norm": 0.15360991961742707,
+ "learning_rate": 1.7314964431182523e-05,
+ "loss": 0.4303375482559204,
+ "step": 1238
+ },
+ {
+ "epoch": 0.6500524658971668,
+ "grad_norm": 0.1633151574385741,
+ "learning_rate": 1.728190951163633e-05,
+ "loss": 0.45568403601646423,
+ "step": 1239
+ },
+ {
+ "epoch": 0.6505771248688352,
+ "grad_norm": 0.16377853571058645,
+ "learning_rate": 1.724888235235457e-05,
+ "loss": 0.46503543853759766,
+ "step": 1240
+ },
+ {
+ "epoch": 0.6511017838405037,
+ "grad_norm": 0.1595023801275368,
+ "learning_rate": 1.7215883043064655e-05,
+ "loss": 0.46337294578552246,
+ "step": 1241
+ },
+ {
+ "epoch": 0.6516264428121721,
+ "grad_norm": 0.16324796182212292,
+ "learning_rate": 1.7182911673418313e-05,
+ "loss": 0.46197354793548584,
+ "step": 1242
+ },
+ {
+ "epoch": 0.6521511017838405,
+ "grad_norm": 0.16353734830879765,
+ "learning_rate": 1.714996833299138e-05,
+ "loss": 0.45741671323776245,
+ "step": 1243
+ },
+ {
+ "epoch": 0.6526757607555089,
+ "grad_norm": 0.15941587678263386,
+ "learning_rate": 1.7117053111283528e-05,
+ "loss": 0.4514503479003906,
+ "step": 1244
+ },
+ {
+ "epoch": 0.6532004197271774,
+ "grad_norm": 0.1642653230145008,
+ "learning_rate": 1.708416609771804e-05,
+ "loss": 0.4437655210494995,
+ "step": 1245
+ },
+ {
+ "epoch": 0.6537250786988458,
+ "grad_norm": 0.15814155519130152,
+ "learning_rate": 1.7051307381641587e-05,
+ "loss": 0.4314163327217102,
+ "step": 1246
+ },
+ {
+ "epoch": 0.6542497376705142,
+ "grad_norm": 0.16086078370768056,
+ "learning_rate": 1.701847705232394e-05,
+ "loss": 0.4526430666446686,
+ "step": 1247
+ },
+ {
+ "epoch": 0.6547743966421826,
+ "grad_norm": 0.16389360190715183,
+ "learning_rate": 1.698567519895775e-05,
+ "loss": 0.4278966188430786,
+ "step": 1248
+ },
+ {
+ "epoch": 0.655299055613851,
+ "grad_norm": 0.16303142287804262,
+ "learning_rate": 1.6952901910658316e-05,
+ "loss": 0.45778220891952515,
+ "step": 1249
+ },
+ {
+ "epoch": 0.6558237145855194,
+ "grad_norm": 0.1675209956904048,
+ "learning_rate": 1.6920157276463318e-05,
+ "loss": 0.4489403963088989,
+ "step": 1250
+ },
+ {
+ "epoch": 0.6563483735571878,
+ "grad_norm": 0.16228311551074323,
+ "learning_rate": 1.6887441385332604e-05,
+ "loss": 0.4534342885017395,
+ "step": 1251
+ },
+ {
+ "epoch": 0.6568730325288562,
+ "grad_norm": 0.16408957851046913,
+ "learning_rate": 1.6854754326147916e-05,
+ "loss": 0.4549208879470825,
+ "step": 1252
+ },
+ {
+ "epoch": 0.6573976915005246,
+ "grad_norm": 0.15967922111432845,
+ "learning_rate": 1.682209618771268e-05,
+ "loss": 0.4451141357421875,
+ "step": 1253
+ },
+ {
+ "epoch": 0.657922350472193,
+ "grad_norm": 0.16440698722204208,
+ "learning_rate": 1.678946705875176e-05,
+ "loss": 0.454200804233551,
+ "step": 1254
+ },
+ {
+ "epoch": 0.6584470094438615,
+ "grad_norm": 0.16251120184938497,
+ "learning_rate": 1.675686702791117e-05,
+ "loss": 0.4580027461051941,
+ "step": 1255
+ },
+ {
+ "epoch": 0.6589716684155299,
+ "grad_norm": 0.1643717386150457,
+ "learning_rate": 1.6724296183757924e-05,
+ "loss": 0.4422311782836914,
+ "step": 1256
+ },
+ {
+ "epoch": 0.6594963273871983,
+ "grad_norm": 0.1644586414378722,
+ "learning_rate": 1.66917546147797e-05,
+ "loss": 0.4593179225921631,
+ "step": 1257
+ },
+ {
+ "epoch": 0.6600209863588667,
+ "grad_norm": 0.16751074186967685,
+ "learning_rate": 1.6659242409384657e-05,
+ "loss": 0.457930326461792,
+ "step": 1258
+ },
+ {
+ "epoch": 0.6605456453305352,
+ "grad_norm": 0.15398418609155334,
+ "learning_rate": 1.6626759655901182e-05,
+ "loss": 0.44002899527549744,
+ "step": 1259
+ },
+ {
+ "epoch": 0.6610703043022036,
+ "grad_norm": 0.16328113175790226,
+ "learning_rate": 1.659430644257763e-05,
+ "loss": 0.4500438868999481,
+ "step": 1260
+ },
+ {
+ "epoch": 0.661594963273872,
+ "grad_norm": 0.1694244340615106,
+ "learning_rate": 1.656188285758214e-05,
+ "loss": 0.4818427264690399,
+ "step": 1261
+ },
+ {
+ "epoch": 0.6621196222455404,
+ "grad_norm": 0.1578320701754368,
+ "learning_rate": 1.652948898900232e-05,
+ "loss": 0.4492436647415161,
+ "step": 1262
+ },
+ {
+ "epoch": 0.6626442812172089,
+ "grad_norm": 0.16538865917828469,
+ "learning_rate": 1.6497124924845064e-05,
+ "loss": 0.45070338249206543,
+ "step": 1263
+ },
+ {
+ "epoch": 0.6631689401888772,
+ "grad_norm": 0.15670817848256852,
+ "learning_rate": 1.6464790753036296e-05,
+ "loss": 0.45439285039901733,
+ "step": 1264
+ },
+ {
+ "epoch": 0.6636935991605456,
+ "grad_norm": 0.16184786585965885,
+ "learning_rate": 1.64324865614207e-05,
+ "loss": 0.446522980928421,
+ "step": 1265
+ },
+ {
+ "epoch": 0.664218258132214,
+ "grad_norm": 0.15719726462251984,
+ "learning_rate": 1.6400212437761554e-05,
+ "loss": 0.4495280981063843,
+ "step": 1266
+ },
+ {
+ "epoch": 0.6647429171038824,
+ "grad_norm": 0.1680804414134405,
+ "learning_rate": 1.6367968469740416e-05,
+ "loss": 0.4652596712112427,
+ "step": 1267
+ },
+ {
+ "epoch": 0.6652675760755509,
+ "grad_norm": 0.15919541846683566,
+ "learning_rate": 1.6335754744956937e-05,
+ "loss": 0.43842798471450806,
+ "step": 1268
+ },
+ {
+ "epoch": 0.6657922350472193,
+ "grad_norm": 0.16322634187139418,
+ "learning_rate": 1.630357135092859e-05,
+ "loss": 0.45242664217948914,
+ "step": 1269
+ },
+ {
+ "epoch": 0.6663168940188877,
+ "grad_norm": 0.15644867734734783,
+ "learning_rate": 1.6271418375090445e-05,
+ "loss": 0.44801655411720276,
+ "step": 1270
+ },
+ {
+ "epoch": 0.6668415529905561,
+ "grad_norm": 0.1514717416124423,
+ "learning_rate": 1.623929590479496e-05,
+ "loss": 0.4506194293498993,
+ "step": 1271
+ },
+ {
+ "epoch": 0.6673662119622246,
+ "grad_norm": 0.1740345085495889,
+ "learning_rate": 1.6207204027311683e-05,
+ "loss": 0.4818177819252014,
+ "step": 1272
+ },
+ {
+ "epoch": 0.667890870933893,
+ "grad_norm": 0.15797701900765457,
+ "learning_rate": 1.6175142829827067e-05,
+ "loss": 0.4639382064342499,
+ "step": 1273
+ },
+ {
+ "epoch": 0.6684155299055614,
+ "grad_norm": 0.15663812316653875,
+ "learning_rate": 1.6143112399444198e-05,
+ "loss": 0.46217501163482666,
+ "step": 1274
+ },
+ {
+ "epoch": 0.6689401888772298,
+ "grad_norm": 0.15474018977043758,
+ "learning_rate": 1.6111112823182592e-05,
+ "loss": 0.43691688776016235,
+ "step": 1275
+ },
+ {
+ "epoch": 0.6694648478488983,
+ "grad_norm": 0.1615634942529431,
+ "learning_rate": 1.6079144187977923e-05,
+ "loss": 0.4412601590156555,
+ "step": 1276
+ },
+ {
+ "epoch": 0.6699895068205667,
+ "grad_norm": 0.15440022842754444,
+ "learning_rate": 1.604720658068184e-05,
+ "loss": 0.4523273706436157,
+ "step": 1277
+ },
+ {
+ "epoch": 0.670514165792235,
+ "grad_norm": 0.1634985665776242,
+ "learning_rate": 1.6015300088061664e-05,
+ "loss": 0.46264398097991943,
+ "step": 1278
+ },
+ {
+ "epoch": 0.6710388247639034,
+ "grad_norm": 0.15037886370689527,
+ "learning_rate": 1.598342479680018e-05,
+ "loss": 0.451465904712677,
+ "step": 1279
+ },
+ {
+ "epoch": 0.6715634837355718,
+ "grad_norm": 0.16895242437746613,
+ "learning_rate": 1.595158079349542e-05,
+ "loss": 0.4466145634651184,
+ "step": 1280
+ },
+ {
+ "epoch": 0.6720881427072403,
+ "grad_norm": 0.1629548326999642,
+ "learning_rate": 1.5919768164660425e-05,
+ "loss": 0.45415782928466797,
+ "step": 1281
+ },
+ {
+ "epoch": 0.6726128016789087,
+ "grad_norm": 0.1563872102268991,
+ "learning_rate": 1.588798699672298e-05,
+ "loss": 0.4386737048625946,
+ "step": 1282
+ },
+ {
+ "epoch": 0.6731374606505771,
+ "grad_norm": 0.16038172773344717,
+ "learning_rate": 1.5856237376025403e-05,
+ "loss": 0.446260929107666,
+ "step": 1283
+ },
+ {
+ "epoch": 0.6736621196222455,
+ "grad_norm": 0.16189780714407623,
+ "learning_rate": 1.582451938882431e-05,
+ "loss": 0.45458072423934937,
+ "step": 1284
+ },
+ {
+ "epoch": 0.674186778593914,
+ "grad_norm": 0.15204521785049285,
+ "learning_rate": 1.5792833121290367e-05,
+ "loss": 0.4425617754459381,
+ "step": 1285
+ },
+ {
+ "epoch": 0.6747114375655824,
+ "grad_norm": 0.16237376704792778,
+ "learning_rate": 1.5761178659508063e-05,
+ "loss": 0.44631800055503845,
+ "step": 1286
+ },
+ {
+ "epoch": 0.6752360965372508,
+ "grad_norm": 0.1582063618466069,
+ "learning_rate": 1.57295560894755e-05,
+ "loss": 0.44280534982681274,
+ "step": 1287
+ },
+ {
+ "epoch": 0.6757607555089192,
+ "grad_norm": 0.16587032349847394,
+ "learning_rate": 1.5697965497104116e-05,
+ "loss": 0.4495997130870819,
+ "step": 1288
+ },
+ {
+ "epoch": 0.6762854144805877,
+ "grad_norm": 0.1557269262755563,
+ "learning_rate": 1.566640696821848e-05,
+ "loss": 0.4493359625339508,
+ "step": 1289
+ },
+ {
+ "epoch": 0.6768100734522561,
+ "grad_norm": 0.16038490357012256,
+ "learning_rate": 1.5634880588556043e-05,
+ "loss": 0.4447544813156128,
+ "step": 1290
+ },
+ {
+ "epoch": 0.6773347324239245,
+ "grad_norm": 0.15355201908546406,
+ "learning_rate": 1.560338644376692e-05,
+ "loss": 0.4468432664871216,
+ "step": 1291
+ },
+ {
+ "epoch": 0.6778593913955928,
+ "grad_norm": 0.15077009515589776,
+ "learning_rate": 1.5571924619413668e-05,
+ "loss": 0.44959044456481934,
+ "step": 1292
+ },
+ {
+ "epoch": 0.6783840503672612,
+ "grad_norm": 0.1583038146613965,
+ "learning_rate": 1.5540495200971012e-05,
+ "loss": 0.42450425028800964,
+ "step": 1293
+ },
+ {
+ "epoch": 0.6789087093389297,
+ "grad_norm": 0.15950687190598087,
+ "learning_rate": 1.5509098273825635e-05,
+ "loss": 0.4527238607406616,
+ "step": 1294
+ },
+ {
+ "epoch": 0.6794333683105981,
+ "grad_norm": 0.1526262208847808,
+ "learning_rate": 1.5477733923275968e-05,
+ "loss": 0.4379724860191345,
+ "step": 1295
+ },
+ {
+ "epoch": 0.6799580272822665,
+ "grad_norm": 0.15541579650021778,
+ "learning_rate": 1.5446402234531934e-05,
+ "loss": 0.4462166130542755,
+ "step": 1296
+ },
+ {
+ "epoch": 0.6804826862539349,
+ "grad_norm": 0.15483930734313786,
+ "learning_rate": 1.5415103292714718e-05,
+ "loss": 0.4390607178211212,
+ "step": 1297
+ },
+ {
+ "epoch": 0.6810073452256034,
+ "grad_norm": 0.1578005731149154,
+ "learning_rate": 1.5383837182856532e-05,
+ "loss": 0.44462254643440247,
+ "step": 1298
+ },
+ {
+ "epoch": 0.6815320041972718,
+ "grad_norm": 0.15830183884958823,
+ "learning_rate": 1.5352603989900406e-05,
+ "loss": 0.4563537538051605,
+ "step": 1299
+ },
+ {
+ "epoch": 0.6820566631689402,
+ "grad_norm": 0.15605472163445522,
+ "learning_rate": 1.5321403798699923e-05,
+ "loss": 0.43852412700653076,
+ "step": 1300
+ },
+ {
+ "epoch": 0.6825813221406086,
+ "grad_norm": 0.15770470227141278,
+ "learning_rate": 1.5290236694019017e-05,
+ "loss": 0.4438081979751587,
+ "step": 1301
+ },
+ {
+ "epoch": 0.683105981112277,
+ "grad_norm": 0.16175339845012077,
+ "learning_rate": 1.5259102760531747e-05,
+ "loss": 0.46670734882354736,
+ "step": 1302
+ },
+ {
+ "epoch": 0.6836306400839455,
+ "grad_norm": 0.16359504716154477,
+ "learning_rate": 1.5228002082822036e-05,
+ "loss": 0.4590187668800354,
+ "step": 1303
+ },
+ {
+ "epoch": 0.6841552990556139,
+ "grad_norm": 0.15422879500254713,
+ "learning_rate": 1.5196934745383457e-05,
+ "loss": 0.45287445187568665,
+ "step": 1304
+ },
+ {
+ "epoch": 0.6846799580272823,
+ "grad_norm": 0.15929902143574254,
+ "learning_rate": 1.5165900832619015e-05,
+ "loss": 0.44286859035491943,
+ "step": 1305
+ },
+ {
+ "epoch": 0.6852046169989506,
+ "grad_norm": 0.14838868714016124,
+ "learning_rate": 1.5134900428840904e-05,
+ "loss": 0.45155346393585205,
+ "step": 1306
+ },
+ {
+ "epoch": 0.6857292759706191,
+ "grad_norm": 0.15494288640397672,
+ "learning_rate": 1.5103933618270283e-05,
+ "loss": 0.44010141491889954,
+ "step": 1307
+ },
+ {
+ "epoch": 0.6862539349422875,
+ "grad_norm": 0.1686262360598338,
+ "learning_rate": 1.5073000485037042e-05,
+ "loss": 0.446555495262146,
+ "step": 1308
+ },
+ {
+ "epoch": 0.6867785939139559,
+ "grad_norm": 0.16122110823982602,
+ "learning_rate": 1.5042101113179579e-05,
+ "loss": 0.44640374183654785,
+ "step": 1309
+ },
+ {
+ "epoch": 0.6873032528856243,
+ "grad_norm": 0.16849937797919867,
+ "learning_rate": 1.5011235586644572e-05,
+ "loss": 0.4465785026550293,
+ "step": 1310
+ },
+ {
+ "epoch": 0.6878279118572928,
+ "grad_norm": 0.16062971262800754,
+ "learning_rate": 1.498040398928674e-05,
+ "loss": 0.445321649312973,
+ "step": 1311
+ },
+ {
+ "epoch": 0.6883525708289612,
+ "grad_norm": 0.1608953185993034,
+ "learning_rate": 1.4949606404868638e-05,
+ "loss": 0.46518227458000183,
+ "step": 1312
+ },
+ {
+ "epoch": 0.6888772298006296,
+ "grad_norm": 0.15880255672830915,
+ "learning_rate": 1.491884291706041e-05,
+ "loss": 0.44582217931747437,
+ "step": 1313
+ },
+ {
+ "epoch": 0.689401888772298,
+ "grad_norm": 0.16715145200789822,
+ "learning_rate": 1.4888113609439563e-05,
+ "loss": 0.4435983896255493,
+ "step": 1314
+ },
+ {
+ "epoch": 0.6899265477439664,
+ "grad_norm": 0.15263740340694454,
+ "learning_rate": 1.485741856549075e-05,
+ "loss": 0.4405066967010498,
+ "step": 1315
+ },
+ {
+ "epoch": 0.6904512067156349,
+ "grad_norm": 0.16247034518813966,
+ "learning_rate": 1.4826757868605517e-05,
+ "loss": 0.4544215500354767,
+ "step": 1316
+ },
+ {
+ "epoch": 0.6909758656873033,
+ "grad_norm": 0.15649146705978162,
+ "learning_rate": 1.4796131602082142e-05,
+ "loss": 0.4454413056373596,
+ "step": 1317
+ },
+ {
+ "epoch": 0.6915005246589717,
+ "grad_norm": 0.16286718636058178,
+ "learning_rate": 1.4765539849125318e-05,
+ "loss": 0.4497183561325073,
+ "step": 1318
+ },
+ {
+ "epoch": 0.6920251836306401,
+ "grad_norm": 0.15812342085005582,
+ "learning_rate": 1.473498269284599e-05,
+ "loss": 0.445032000541687,
+ "step": 1319
+ },
+ {
+ "epoch": 0.6925498426023085,
+ "grad_norm": 0.1558101290511678,
+ "learning_rate": 1.4704460216261123e-05,
+ "loss": 0.45769619941711426,
+ "step": 1320
+ },
+ {
+ "epoch": 0.6930745015739769,
+ "grad_norm": 0.15642889059495207,
+ "learning_rate": 1.4673972502293428e-05,
+ "loss": 0.44788360595703125,
+ "step": 1321
+ },
+ {
+ "epoch": 0.6935991605456453,
+ "grad_norm": 0.15722513518113074,
+ "learning_rate": 1.4643519633771222e-05,
+ "loss": 0.449631929397583,
+ "step": 1322
+ },
+ {
+ "epoch": 0.6941238195173137,
+ "grad_norm": 0.15587633748081542,
+ "learning_rate": 1.4613101693428125e-05,
+ "loss": 0.4450187087059021,
+ "step": 1323
+ },
+ {
+ "epoch": 0.6946484784889821,
+ "grad_norm": 0.1577107074882479,
+ "learning_rate": 1.458271876390287e-05,
+ "loss": 0.45108890533447266,
+ "step": 1324
+ },
+ {
+ "epoch": 0.6951731374606506,
+ "grad_norm": 0.15649839486990696,
+ "learning_rate": 1.4552370927739073e-05,
+ "loss": 0.46430301666259766,
+ "step": 1325
+ },
+ {
+ "epoch": 0.695697796432319,
+ "grad_norm": 0.18573600420690636,
+ "learning_rate": 1.4522058267385008e-05,
+ "loss": 0.4701972007751465,
+ "step": 1326
+ },
+ {
+ "epoch": 0.6962224554039874,
+ "grad_norm": 0.16296497084037542,
+ "learning_rate": 1.4491780865193402e-05,
+ "loss": 0.4607710838317871,
+ "step": 1327
+ },
+ {
+ "epoch": 0.6967471143756558,
+ "grad_norm": 0.161124007217343,
+ "learning_rate": 1.446153880342117e-05,
+ "loss": 0.45095115900039673,
+ "step": 1328
+ },
+ {
+ "epoch": 0.6972717733473243,
+ "grad_norm": 0.1537884039166766,
+ "learning_rate": 1.4431332164229222e-05,
+ "loss": 0.424068808555603,
+ "step": 1329
+ },
+ {
+ "epoch": 0.6977964323189927,
+ "grad_norm": 0.16588834503456087,
+ "learning_rate": 1.4401161029682237e-05,
+ "loss": 0.4622839093208313,
+ "step": 1330
+ },
+ {
+ "epoch": 0.6983210912906611,
+ "grad_norm": 0.16100510604015544,
+ "learning_rate": 1.4371025481748435e-05,
+ "loss": 0.4660705626010895,
+ "step": 1331
+ },
+ {
+ "epoch": 0.6988457502623295,
+ "grad_norm": 0.15559888940590075,
+ "learning_rate": 1.4340925602299337e-05,
+ "loss": 0.4484976530075073,
+ "step": 1332
+ },
+ {
+ "epoch": 0.6993704092339978,
+ "grad_norm": 0.15503821966213488,
+ "learning_rate": 1.4310861473109616e-05,
+ "loss": 0.45636600255966187,
+ "step": 1333
+ },
+ {
+ "epoch": 0.6998950682056663,
+ "grad_norm": 0.15539074896109334,
+ "learning_rate": 1.4280833175856743e-05,
+ "loss": 0.45863068103790283,
+ "step": 1334
+ },
+ {
+ "epoch": 0.7004197271773347,
+ "grad_norm": 0.15664821009154872,
+ "learning_rate": 1.4250840792120896e-05,
+ "loss": 0.44043463468551636,
+ "step": 1335
+ },
+ {
+ "epoch": 0.7009443861490031,
+ "grad_norm": 0.15825478808520285,
+ "learning_rate": 1.4220884403384655e-05,
+ "loss": 0.46522051095962524,
+ "step": 1336
+ },
+ {
+ "epoch": 0.7014690451206715,
+ "grad_norm": 0.16004268568141544,
+ "learning_rate": 1.4190964091032843e-05,
+ "loss": 0.46114686131477356,
+ "step": 1337
+ },
+ {
+ "epoch": 0.70199370409234,
+ "grad_norm": 0.16413139949945635,
+ "learning_rate": 1.4161079936352234e-05,
+ "loss": 0.4457515478134155,
+ "step": 1338
+ },
+ {
+ "epoch": 0.7025183630640084,
+ "grad_norm": 0.1538890577972702,
+ "learning_rate": 1.4131232020531398e-05,
+ "loss": 0.4616093337535858,
+ "step": 1339
+ },
+ {
+ "epoch": 0.7030430220356768,
+ "grad_norm": 0.16240737854323284,
+ "learning_rate": 1.4101420424660434e-05,
+ "loss": 0.44631537795066833,
+ "step": 1340
+ },
+ {
+ "epoch": 0.7035676810073452,
+ "grad_norm": 0.15984708603406073,
+ "learning_rate": 1.4071645229730762e-05,
+ "loss": 0.4366717040538788,
+ "step": 1341
+ },
+ {
+ "epoch": 0.7040923399790137,
+ "grad_norm": 0.15149695695635904,
+ "learning_rate": 1.4041906516634946e-05,
+ "loss": 0.44129428267478943,
+ "step": 1342
+ },
+ {
+ "epoch": 0.7046169989506821,
+ "grad_norm": 0.15041167501768415,
+ "learning_rate": 1.4012204366166399e-05,
+ "loss": 0.43654775619506836,
+ "step": 1343
+ },
+ {
+ "epoch": 0.7051416579223505,
+ "grad_norm": 0.1596097627963756,
+ "learning_rate": 1.3982538859019217e-05,
+ "loss": 0.4652445614337921,
+ "step": 1344
+ },
+ {
+ "epoch": 0.7056663168940189,
+ "grad_norm": 0.15621704205027884,
+ "learning_rate": 1.3952910075787941e-05,
+ "loss": 0.44577041268348694,
+ "step": 1345
+ },
+ {
+ "epoch": 0.7061909758656874,
+ "grad_norm": 0.16073041098008783,
+ "learning_rate": 1.3923318096967341e-05,
+ "loss": 0.4391976594924927,
+ "step": 1346
+ },
+ {
+ "epoch": 0.7067156348373557,
+ "grad_norm": 0.14793853946372448,
+ "learning_rate": 1.3893763002952204e-05,
+ "loss": 0.4463799297809601,
+ "step": 1347
+ },
+ {
+ "epoch": 0.7072402938090241,
+ "grad_norm": 0.16176080358867354,
+ "learning_rate": 1.3864244874037097e-05,
+ "loss": 0.45183271169662476,
+ "step": 1348
+ },
+ {
+ "epoch": 0.7077649527806925,
+ "grad_norm": 0.16372771179026924,
+ "learning_rate": 1.3834763790416171e-05,
+ "loss": 0.44205576181411743,
+ "step": 1349
+ },
+ {
+ "epoch": 0.7082896117523609,
+ "grad_norm": 0.15135222516127125,
+ "learning_rate": 1.3805319832182937e-05,
+ "loss": 0.4390261173248291,
+ "step": 1350
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 3.573309495889101e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1400/chat_template.jinja b/checkpoints/checkpoint-1400/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1400/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1400/config.json b/checkpoints/checkpoint-1400/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1400/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1400/generation_config.json b/checkpoints/checkpoint-1400/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1400/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1400/model.safetensors b/checkpoints/checkpoint-1400/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..b0ceea5f1cc0f89476f1da3dcf473ecb15569e41
--- /dev/null
+++ b/checkpoints/checkpoint-1400/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26e9d888d161c13e7732ba9776d398271b5c8daef7c46cf1db20e772792ac081
+size 37983421000
diff --git a/checkpoints/checkpoint-1400/tokenizer.json b/checkpoints/checkpoint-1400/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1400/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1400/tokenizer_config.json b/checkpoints/checkpoint-1400/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1400/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1400/trainer_state.json b/checkpoints/checkpoint-1400/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..9e4e420940b63d8fb13500b57f833c26ad60539d
--- /dev/null
+++ b/checkpoints/checkpoint-1400/trainer_state.json
@@ -0,0 +1,9834 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.7345225603357818,
+ "eval_steps": 500,
+ "global_step": 1400,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ },
+ {
+ "epoch": 0.5776495278069255,
+ "grad_norm": 0.1862053032854098,
+ "learning_rate": 2.2065553964033126e-05,
+ "loss": 0.4816875457763672,
+ "step": 1101
+ },
+ {
+ "epoch": 0.5781741867785939,
+ "grad_norm": 0.16996015894363117,
+ "learning_rate": 2.20295734045532e-05,
+ "loss": 0.4471197724342346,
+ "step": 1102
+ },
+ {
+ "epoch": 0.5786988457502623,
+ "grad_norm": 0.16889227008390692,
+ "learning_rate": 2.1993607706997725e-05,
+ "loss": 0.468391478061676,
+ "step": 1103
+ },
+ {
+ "epoch": 0.5792235047219307,
+ "grad_norm": 0.16670931681562373,
+ "learning_rate": 2.195765696907745e-05,
+ "loss": 0.44934162497520447,
+ "step": 1104
+ },
+ {
+ "epoch": 0.5797481636935992,
+ "grad_norm": 0.16328418741223716,
+ "learning_rate": 2.1921721288462482e-05,
+ "loss": 0.4558270275592804,
+ "step": 1105
+ },
+ {
+ "epoch": 0.5802728226652676,
+ "grad_norm": 0.17364660408703936,
+ "learning_rate": 2.188580076278203e-05,
+ "loss": 0.4597751796245575,
+ "step": 1106
+ },
+ {
+ "epoch": 0.580797481636936,
+ "grad_norm": 0.1628618453800601,
+ "learning_rate": 2.18498954896241e-05,
+ "loss": 0.4521859288215637,
+ "step": 1107
+ },
+ {
+ "epoch": 0.5813221406086044,
+ "grad_norm": 0.16813817951871482,
+ "learning_rate": 2.1814005566535302e-05,
+ "loss": 0.4581005573272705,
+ "step": 1108
+ },
+ {
+ "epoch": 0.5818467995802729,
+ "grad_norm": 0.16825923629726025,
+ "learning_rate": 2.1778131091020516e-05,
+ "loss": 0.4433916211128235,
+ "step": 1109
+ },
+ {
+ "epoch": 0.5823714585519413,
+ "grad_norm": 0.17194743467883944,
+ "learning_rate": 2.1742272160542654e-05,
+ "loss": 0.47296425700187683,
+ "step": 1110
+ },
+ {
+ "epoch": 0.5828961175236097,
+ "grad_norm": 0.16947730546039969,
+ "learning_rate": 2.170642887252241e-05,
+ "loss": 0.4815937876701355,
+ "step": 1111
+ },
+ {
+ "epoch": 0.583420776495278,
+ "grad_norm": 0.1657449928668473,
+ "learning_rate": 2.167060132433798e-05,
+ "loss": 0.4479164779186249,
+ "step": 1112
+ },
+ {
+ "epoch": 0.5839454354669464,
+ "grad_norm": 0.1658952516607364,
+ "learning_rate": 2.1634789613324778e-05,
+ "loss": 0.4637463688850403,
+ "step": 1113
+ },
+ {
+ "epoch": 0.5844700944386149,
+ "grad_norm": 0.16777002188998236,
+ "learning_rate": 2.1598993836775218e-05,
+ "loss": 0.457524836063385,
+ "step": 1114
+ },
+ {
+ "epoch": 0.5849947534102833,
+ "grad_norm": 0.15837566447341322,
+ "learning_rate": 2.1563214091938405e-05,
+ "loss": 0.44938087463378906,
+ "step": 1115
+ },
+ {
+ "epoch": 0.5855194123819517,
+ "grad_norm": 0.17162206617035314,
+ "learning_rate": 2.1527450476019896e-05,
+ "loss": 0.4393450617790222,
+ "step": 1116
+ },
+ {
+ "epoch": 0.5860440713536201,
+ "grad_norm": 0.16224101921433814,
+ "learning_rate": 2.1491703086181426e-05,
+ "loss": 0.451913982629776,
+ "step": 1117
+ },
+ {
+ "epoch": 0.5865687303252886,
+ "grad_norm": 0.1736253504575144,
+ "learning_rate": 2.145597201954066e-05,
+ "loss": 0.4611290395259857,
+ "step": 1118
+ },
+ {
+ "epoch": 0.587093389296957,
+ "grad_norm": 0.16308408333737234,
+ "learning_rate": 2.1420257373170903e-05,
+ "loss": 0.45898568630218506,
+ "step": 1119
+ },
+ {
+ "epoch": 0.5876180482686254,
+ "grad_norm": 0.16948634794175524,
+ "learning_rate": 2.1384559244100848e-05,
+ "loss": 0.45982882380485535,
+ "step": 1120
+ },
+ {
+ "epoch": 0.5881427072402938,
+ "grad_norm": 0.16676366538565252,
+ "learning_rate": 2.1348877729314336e-05,
+ "loss": 0.44826576113700867,
+ "step": 1121
+ },
+ {
+ "epoch": 0.5886673662119623,
+ "grad_norm": 0.16431125630712295,
+ "learning_rate": 2.1313212925750037e-05,
+ "loss": 0.4616445302963257,
+ "step": 1122
+ },
+ {
+ "epoch": 0.5891920251836307,
+ "grad_norm": 0.1631659061616955,
+ "learning_rate": 2.1277564930301252e-05,
+ "loss": 0.45651188492774963,
+ "step": 1123
+ },
+ {
+ "epoch": 0.5897166841552991,
+ "grad_norm": 0.15805950502689237,
+ "learning_rate": 2.1241933839815604e-05,
+ "loss": 0.4486045837402344,
+ "step": 1124
+ },
+ {
+ "epoch": 0.5902413431269675,
+ "grad_norm": 0.16530288038603502,
+ "learning_rate": 2.1206319751094773e-05,
+ "loss": 0.4485066533088684,
+ "step": 1125
+ },
+ {
+ "epoch": 0.5907660020986358,
+ "grad_norm": 0.16061903194239047,
+ "learning_rate": 2.1170722760894295e-05,
+ "loss": 0.4565330743789673,
+ "step": 1126
+ },
+ {
+ "epoch": 0.5912906610703043,
+ "grad_norm": 0.1643346999353627,
+ "learning_rate": 2.1135142965923195e-05,
+ "loss": 0.45560717582702637,
+ "step": 1127
+ },
+ {
+ "epoch": 0.5918153200419727,
+ "grad_norm": 0.17484644315215753,
+ "learning_rate": 2.109958046284382e-05,
+ "loss": 0.4795891046524048,
+ "step": 1128
+ },
+ {
+ "epoch": 0.5923399790136411,
+ "grad_norm": 0.1623762678622651,
+ "learning_rate": 2.1064035348271528e-05,
+ "loss": 0.47395604848861694,
+ "step": 1129
+ },
+ {
+ "epoch": 0.5928646379853095,
+ "grad_norm": 0.16888824241669015,
+ "learning_rate": 2.1028507718774456e-05,
+ "loss": 0.4601455330848694,
+ "step": 1130
+ },
+ {
+ "epoch": 0.593389296956978,
+ "grad_norm": 0.16291744568806302,
+ "learning_rate": 2.0992997670873193e-05,
+ "loss": 0.46248236298561096,
+ "step": 1131
+ },
+ {
+ "epoch": 0.5939139559286464,
+ "grad_norm": 0.16185598691236988,
+ "learning_rate": 2.0957505301040603e-05,
+ "loss": 0.45085301995277405,
+ "step": 1132
+ },
+ {
+ "epoch": 0.5944386149003148,
+ "grad_norm": 0.1678000806117814,
+ "learning_rate": 2.0922030705701518e-05,
+ "loss": 0.44939014315605164,
+ "step": 1133
+ },
+ {
+ "epoch": 0.5949632738719832,
+ "grad_norm": 0.16104474282288994,
+ "learning_rate": 2.088657398123246e-05,
+ "loss": 0.4313976764678955,
+ "step": 1134
+ },
+ {
+ "epoch": 0.5954879328436516,
+ "grad_norm": 0.1631128968199669,
+ "learning_rate": 2.085113522396141e-05,
+ "loss": 0.4508278965950012,
+ "step": 1135
+ },
+ {
+ "epoch": 0.5960125918153201,
+ "grad_norm": 0.16307688195844533,
+ "learning_rate": 2.0815714530167557e-05,
+ "loss": 0.4467576742172241,
+ "step": 1136
+ },
+ {
+ "epoch": 0.5965372507869885,
+ "grad_norm": 0.16456832304532207,
+ "learning_rate": 2.0780311996080976e-05,
+ "loss": 0.4621703624725342,
+ "step": 1137
+ },
+ {
+ "epoch": 0.5970619097586569,
+ "grad_norm": 0.16346217748325367,
+ "learning_rate": 2.0744927717882435e-05,
+ "loss": 0.4589449465274811,
+ "step": 1138
+ },
+ {
+ "epoch": 0.5975865687303253,
+ "grad_norm": 0.16180691501076994,
+ "learning_rate": 2.0709561791703097e-05,
+ "loss": 0.43698400259017944,
+ "step": 1139
+ },
+ {
+ "epoch": 0.5981112277019937,
+ "grad_norm": 0.15695459976898835,
+ "learning_rate": 2.067421431362427e-05,
+ "loss": 0.452271044254303,
+ "step": 1140
+ },
+ {
+ "epoch": 0.5986358866736621,
+ "grad_norm": 0.15930893553721054,
+ "learning_rate": 2.063888537967713e-05,
+ "loss": 0.44868963956832886,
+ "step": 1141
+ },
+ {
+ "epoch": 0.5991605456453305,
+ "grad_norm": 0.15963802284017928,
+ "learning_rate": 2.0603575085842485e-05,
+ "loss": 0.4376864731311798,
+ "step": 1142
+ },
+ {
+ "epoch": 0.5996852046169989,
+ "grad_norm": 0.16610266900507178,
+ "learning_rate": 2.05682835280505e-05,
+ "loss": 0.44676727056503296,
+ "step": 1143
+ },
+ {
+ "epoch": 0.6002098635886673,
+ "grad_norm": 0.15978253236669612,
+ "learning_rate": 2.053301080218044e-05,
+ "loss": 0.434773713350296,
+ "step": 1144
+ },
+ {
+ "epoch": 0.6007345225603358,
+ "grad_norm": 0.16305141803526324,
+ "learning_rate": 2.0497757004060403e-05,
+ "loss": 0.4529780149459839,
+ "step": 1145
+ },
+ {
+ "epoch": 0.6012591815320042,
+ "grad_norm": 0.1649737864101215,
+ "learning_rate": 2.046252222946707e-05,
+ "loss": 0.4571434259414673,
+ "step": 1146
+ },
+ {
+ "epoch": 0.6017838405036726,
+ "grad_norm": 0.15738824968326598,
+ "learning_rate": 2.0427306574125427e-05,
+ "loss": 0.4446289539337158,
+ "step": 1147
+ },
+ {
+ "epoch": 0.602308499475341,
+ "grad_norm": 0.1731139535185187,
+ "learning_rate": 2.039211013370853e-05,
+ "loss": 0.4551090598106384,
+ "step": 1148
+ },
+ {
+ "epoch": 0.6028331584470095,
+ "grad_norm": 0.15857251900043037,
+ "learning_rate": 2.0356933003837237e-05,
+ "loss": 0.4644010663032532,
+ "step": 1149
+ },
+ {
+ "epoch": 0.6033578174186779,
+ "grad_norm": 0.1589582861134211,
+ "learning_rate": 2.032177528007992e-05,
+ "loss": 0.45741650462150574,
+ "step": 1150
+ },
+ {
+ "epoch": 0.6038824763903463,
+ "grad_norm": 0.15639092706575122,
+ "learning_rate": 2.0286637057952257e-05,
+ "loss": 0.4375547766685486,
+ "step": 1151
+ },
+ {
+ "epoch": 0.6044071353620147,
+ "grad_norm": 0.16335605581337187,
+ "learning_rate": 2.0251518432916933e-05,
+ "loss": 0.4636140465736389,
+ "step": 1152
+ },
+ {
+ "epoch": 0.6049317943336832,
+ "grad_norm": 0.15437437113359356,
+ "learning_rate": 2.0216419500383372e-05,
+ "loss": 0.44632142782211304,
+ "step": 1153
+ },
+ {
+ "epoch": 0.6054564533053515,
+ "grad_norm": 0.15522039198698664,
+ "learning_rate": 2.0181340355707545e-05,
+ "loss": 0.449049174785614,
+ "step": 1154
+ },
+ {
+ "epoch": 0.6059811122770199,
+ "grad_norm": 0.1599951917909323,
+ "learning_rate": 2.0146281094191604e-05,
+ "loss": 0.44395962357521057,
+ "step": 1155
+ },
+ {
+ "epoch": 0.6065057712486883,
+ "grad_norm": 0.16866621620527716,
+ "learning_rate": 2.011124181108373e-05,
+ "loss": 0.45693716406822205,
+ "step": 1156
+ },
+ {
+ "epoch": 0.6070304302203567,
+ "grad_norm": 0.15508935650135075,
+ "learning_rate": 2.0076222601577815e-05,
+ "loss": 0.4525151252746582,
+ "step": 1157
+ },
+ {
+ "epoch": 0.6075550891920252,
+ "grad_norm": 0.16508998118005971,
+ "learning_rate": 2.004122356081321e-05,
+ "loss": 0.466006875038147,
+ "step": 1158
+ },
+ {
+ "epoch": 0.6080797481636936,
+ "grad_norm": 0.15817418864022237,
+ "learning_rate": 2.0006244783874466e-05,
+ "loss": 0.4496110677719116,
+ "step": 1159
+ },
+ {
+ "epoch": 0.608604407135362,
+ "grad_norm": 0.16127270259749824,
+ "learning_rate": 1.997128636579109e-05,
+ "loss": 0.45918628573417664,
+ "step": 1160
+ },
+ {
+ "epoch": 0.6091290661070304,
+ "grad_norm": 0.1702599924159677,
+ "learning_rate": 1.9936348401537292e-05,
+ "loss": 0.47218555212020874,
+ "step": 1161
+ },
+ {
+ "epoch": 0.6096537250786989,
+ "grad_norm": 0.15690176248846127,
+ "learning_rate": 1.9901430986031695e-05,
+ "loss": 0.4516097903251648,
+ "step": 1162
+ },
+ {
+ "epoch": 0.6101783840503673,
+ "grad_norm": 0.1713614405614144,
+ "learning_rate": 1.9866534214137085e-05,
+ "loss": 0.45030856132507324,
+ "step": 1163
+ },
+ {
+ "epoch": 0.6107030430220357,
+ "grad_norm": 0.1670865563026229,
+ "learning_rate": 1.9831658180660195e-05,
+ "loss": 0.45603489875793457,
+ "step": 1164
+ },
+ {
+ "epoch": 0.6112277019937041,
+ "grad_norm": 0.16897539113401952,
+ "learning_rate": 1.9796802980351392e-05,
+ "loss": 0.46314537525177,
+ "step": 1165
+ },
+ {
+ "epoch": 0.6117523609653726,
+ "grad_norm": 0.15832868870697644,
+ "learning_rate": 1.976196870790446e-05,
+ "loss": 0.44784611463546753,
+ "step": 1166
+ },
+ {
+ "epoch": 0.612277019937041,
+ "grad_norm": 0.16595365551891114,
+ "learning_rate": 1.9727155457956315e-05,
+ "loss": 0.4409624934196472,
+ "step": 1167
+ },
+ {
+ "epoch": 0.6128016789087093,
+ "grad_norm": 0.1642630411212017,
+ "learning_rate": 1.969236332508677e-05,
+ "loss": 0.46120351552963257,
+ "step": 1168
+ },
+ {
+ "epoch": 0.6133263378803777,
+ "grad_norm": 0.16481796135868812,
+ "learning_rate": 1.9657592403818253e-05,
+ "loss": 0.45544904470443726,
+ "step": 1169
+ },
+ {
+ "epoch": 0.6138509968520461,
+ "grad_norm": 0.1629832779852364,
+ "learning_rate": 1.962284278861557e-05,
+ "loss": 0.4578462839126587,
+ "step": 1170
+ },
+ {
+ "epoch": 0.6143756558237146,
+ "grad_norm": 0.15916219613603688,
+ "learning_rate": 1.9588114573885664e-05,
+ "loss": 0.4546983540058136,
+ "step": 1171
+ },
+ {
+ "epoch": 0.614900314795383,
+ "grad_norm": 0.1660520165166075,
+ "learning_rate": 1.95534078539773e-05,
+ "loss": 0.4601942300796509,
+ "step": 1172
+ },
+ {
+ "epoch": 0.6154249737670514,
+ "grad_norm": 0.16924794578577768,
+ "learning_rate": 1.9518722723180877e-05,
+ "loss": 0.45997846126556396,
+ "step": 1173
+ },
+ {
+ "epoch": 0.6159496327387198,
+ "grad_norm": 0.16597646133045382,
+ "learning_rate": 1.9484059275728127e-05,
+ "loss": 0.4594605565071106,
+ "step": 1174
+ },
+ {
+ "epoch": 0.6164742917103883,
+ "grad_norm": 0.16964243792519748,
+ "learning_rate": 1.9449417605791876e-05,
+ "loss": 0.4638690650463104,
+ "step": 1175
+ },
+ {
+ "epoch": 0.6169989506820567,
+ "grad_norm": 0.1758591426791696,
+ "learning_rate": 1.9414797807485796e-05,
+ "loss": 0.4666447639465332,
+ "step": 1176
+ },
+ {
+ "epoch": 0.6175236096537251,
+ "grad_norm": 0.16217661227891753,
+ "learning_rate": 1.9380199974864126e-05,
+ "loss": 0.44252535700798035,
+ "step": 1177
+ },
+ {
+ "epoch": 0.6180482686253935,
+ "grad_norm": 0.1576446184499019,
+ "learning_rate": 1.934562420192142e-05,
+ "loss": 0.4499641954898834,
+ "step": 1178
+ },
+ {
+ "epoch": 0.618572927597062,
+ "grad_norm": 0.16231008092636912,
+ "learning_rate": 1.9311070582592327e-05,
+ "loss": 0.4561542570590973,
+ "step": 1179
+ },
+ {
+ "epoch": 0.6190975865687304,
+ "grad_norm": 0.16790382249577043,
+ "learning_rate": 1.9276539210751292e-05,
+ "loss": 0.4484325647354126,
+ "step": 1180
+ },
+ {
+ "epoch": 0.6196222455403987,
+ "grad_norm": 0.18794729932605855,
+ "learning_rate": 1.9242030180212323e-05,
+ "loss": 0.44169026613235474,
+ "step": 1181
+ },
+ {
+ "epoch": 0.6201469045120671,
+ "grad_norm": 0.16948051243889814,
+ "learning_rate": 1.920754358472872e-05,
+ "loss": 0.4592876434326172,
+ "step": 1182
+ },
+ {
+ "epoch": 0.6206715634837355,
+ "grad_norm": 0.16762558710029374,
+ "learning_rate": 1.917307951799286e-05,
+ "loss": 0.4402095079421997,
+ "step": 1183
+ },
+ {
+ "epoch": 0.621196222455404,
+ "grad_norm": 0.18170280116387003,
+ "learning_rate": 1.9138638073635885e-05,
+ "loss": 0.4633026123046875,
+ "step": 1184
+ },
+ {
+ "epoch": 0.6217208814270724,
+ "grad_norm": 0.1718051709136108,
+ "learning_rate": 1.9104219345227486e-05,
+ "loss": 0.4510066509246826,
+ "step": 1185
+ },
+ {
+ "epoch": 0.6222455403987408,
+ "grad_norm": 0.16829066326345576,
+ "learning_rate": 1.906982342627565e-05,
+ "loss": 0.4632919430732727,
+ "step": 1186
+ },
+ {
+ "epoch": 0.6227701993704092,
+ "grad_norm": 0.16669141808548274,
+ "learning_rate": 1.9035450410226395e-05,
+ "loss": 0.455544650554657,
+ "step": 1187
+ },
+ {
+ "epoch": 0.6232948583420777,
+ "grad_norm": 0.16679996559722232,
+ "learning_rate": 1.90011003904635e-05,
+ "loss": 0.459994375705719,
+ "step": 1188
+ },
+ {
+ "epoch": 0.6238195173137461,
+ "grad_norm": 0.17379754713278164,
+ "learning_rate": 1.8966773460308284e-05,
+ "loss": 0.4576749801635742,
+ "step": 1189
+ },
+ {
+ "epoch": 0.6243441762854145,
+ "grad_norm": 0.17088539733389246,
+ "learning_rate": 1.8932469713019335e-05,
+ "loss": 0.44853052496910095,
+ "step": 1190
+ },
+ {
+ "epoch": 0.6248688352570829,
+ "grad_norm": 0.15556883889201015,
+ "learning_rate": 1.8898189241792248e-05,
+ "loss": 0.44060447812080383,
+ "step": 1191
+ },
+ {
+ "epoch": 0.6253934942287513,
+ "grad_norm": 0.1640093917439215,
+ "learning_rate": 1.8863932139759388e-05,
+ "loss": 0.4531911611557007,
+ "step": 1192
+ },
+ {
+ "epoch": 0.6259181532004198,
+ "grad_norm": 0.16444284826591576,
+ "learning_rate": 1.882969849998964e-05,
+ "loss": 0.44176363945007324,
+ "step": 1193
+ },
+ {
+ "epoch": 0.6264428121720882,
+ "grad_norm": 0.1696462520995642,
+ "learning_rate": 1.8795488415488143e-05,
+ "loss": 0.44516652822494507,
+ "step": 1194
+ },
+ {
+ "epoch": 0.6269674711437565,
+ "grad_norm": 0.16810724705191796,
+ "learning_rate": 1.8761301979196032e-05,
+ "loss": 0.4542563855648041,
+ "step": 1195
+ },
+ {
+ "epoch": 0.6274921301154249,
+ "grad_norm": 0.17334813313855824,
+ "learning_rate": 1.8727139283990204e-05,
+ "loss": 0.47769248485565186,
+ "step": 1196
+ },
+ {
+ "epoch": 0.6280167890870934,
+ "grad_norm": 0.16308212035303235,
+ "learning_rate": 1.8693000422683043e-05,
+ "loss": 0.4407799243927002,
+ "step": 1197
+ },
+ {
+ "epoch": 0.6285414480587618,
+ "grad_norm": 0.16853319139928977,
+ "learning_rate": 1.8658885488022204e-05,
+ "loss": 0.4565497040748596,
+ "step": 1198
+ },
+ {
+ "epoch": 0.6290661070304302,
+ "grad_norm": 0.16694968781284703,
+ "learning_rate": 1.8624794572690328e-05,
+ "loss": 0.44522374868392944,
+ "step": 1199
+ },
+ {
+ "epoch": 0.6295907660020986,
+ "grad_norm": 0.16418579468339103,
+ "learning_rate": 1.8590727769304788e-05,
+ "loss": 0.4441048502922058,
+ "step": 1200
+ },
+ {
+ "epoch": 0.630115424973767,
+ "grad_norm": 0.1661930777124156,
+ "learning_rate": 1.8556685170417486e-05,
+ "loss": 0.457327663898468,
+ "step": 1201
+ },
+ {
+ "epoch": 0.6306400839454355,
+ "grad_norm": 0.16536630027039814,
+ "learning_rate": 1.852266686851452e-05,
+ "loss": 0.462494432926178,
+ "step": 1202
+ },
+ {
+ "epoch": 0.6311647429171039,
+ "grad_norm": 0.16611474088307454,
+ "learning_rate": 1.8488672956016013e-05,
+ "loss": 0.4557701647281647,
+ "step": 1203
+ },
+ {
+ "epoch": 0.6316894018887723,
+ "grad_norm": 0.16759343079749697,
+ "learning_rate": 1.8454703525275813e-05,
+ "loss": 0.45069995522499084,
+ "step": 1204
+ },
+ {
+ "epoch": 0.6322140608604407,
+ "grad_norm": 0.16186676294485336,
+ "learning_rate": 1.8420758668581255e-05,
+ "loss": 0.44954800605773926,
+ "step": 1205
+ },
+ {
+ "epoch": 0.6327387198321092,
+ "grad_norm": 0.15687159835368553,
+ "learning_rate": 1.8386838478152916e-05,
+ "loss": 0.4490429759025574,
+ "step": 1206
+ },
+ {
+ "epoch": 0.6332633788037776,
+ "grad_norm": 0.16242951976115577,
+ "learning_rate": 1.8352943046144372e-05,
+ "loss": 0.45819586515426636,
+ "step": 1207
+ },
+ {
+ "epoch": 0.633788037775446,
+ "grad_norm": 0.16516087266517232,
+ "learning_rate": 1.831907246464191e-05,
+ "loss": 0.44769883155822754,
+ "step": 1208
+ },
+ {
+ "epoch": 0.6343126967471143,
+ "grad_norm": 0.16709475871512644,
+ "learning_rate": 1.8285226825664332e-05,
+ "loss": 0.4699128568172455,
+ "step": 1209
+ },
+ {
+ "epoch": 0.6348373557187827,
+ "grad_norm": 0.16288623150595774,
+ "learning_rate": 1.8251406221162636e-05,
+ "loss": 0.45322558283805847,
+ "step": 1210
+ },
+ {
+ "epoch": 0.6353620146904512,
+ "grad_norm": 0.1675722848054237,
+ "learning_rate": 1.821761074301986e-05,
+ "loss": 0.4752347469329834,
+ "step": 1211
+ },
+ {
+ "epoch": 0.6358866736621196,
+ "grad_norm": 0.163285963514504,
+ "learning_rate": 1.818384048305074e-05,
+ "loss": 0.441958487033844,
+ "step": 1212
+ },
+ {
+ "epoch": 0.636411332633788,
+ "grad_norm": 0.17078102287557764,
+ "learning_rate": 1.8150095533001517e-05,
+ "loss": 0.4551333487033844,
+ "step": 1213
+ },
+ {
+ "epoch": 0.6369359916054564,
+ "grad_norm": 0.16188540746097138,
+ "learning_rate": 1.811637598454966e-05,
+ "loss": 0.4470779597759247,
+ "step": 1214
+ },
+ {
+ "epoch": 0.6374606505771249,
+ "grad_norm": 0.15235628218958286,
+ "learning_rate": 1.808268192930364e-05,
+ "loss": 0.43604153394699097,
+ "step": 1215
+ },
+ {
+ "epoch": 0.6379853095487933,
+ "grad_norm": 0.1604601949959104,
+ "learning_rate": 1.8049013458802665e-05,
+ "loss": 0.4445713758468628,
+ "step": 1216
+ },
+ {
+ "epoch": 0.6385099685204617,
+ "grad_norm": 0.15511797131215901,
+ "learning_rate": 1.8015370664516427e-05,
+ "loss": 0.4455301761627197,
+ "step": 1217
+ },
+ {
+ "epoch": 0.6390346274921301,
+ "grad_norm": 0.16815669546944095,
+ "learning_rate": 1.798175363784487e-05,
+ "loss": 0.46135014295578003,
+ "step": 1218
+ },
+ {
+ "epoch": 0.6395592864637986,
+ "grad_norm": 0.161558177688922,
+ "learning_rate": 1.7948162470117917e-05,
+ "loss": 0.4519304037094116,
+ "step": 1219
+ },
+ {
+ "epoch": 0.640083945435467,
+ "grad_norm": 0.16677842130048232,
+ "learning_rate": 1.7914597252595265e-05,
+ "loss": 0.4550550878047943,
+ "step": 1220
+ },
+ {
+ "epoch": 0.6406086044071354,
+ "grad_norm": 0.1597331480423138,
+ "learning_rate": 1.7881058076466095e-05,
+ "loss": 0.441120445728302,
+ "step": 1221
+ },
+ {
+ "epoch": 0.6411332633788038,
+ "grad_norm": 0.15951106545246824,
+ "learning_rate": 1.7847545032848844e-05,
+ "loss": 0.4471382200717926,
+ "step": 1222
+ },
+ {
+ "epoch": 0.6416579223504721,
+ "grad_norm": 0.16157787674667565,
+ "learning_rate": 1.7814058212790947e-05,
+ "loss": 0.4550648331642151,
+ "step": 1223
+ },
+ {
+ "epoch": 0.6421825813221406,
+ "grad_norm": 0.16989186763352415,
+ "learning_rate": 1.7780597707268583e-05,
+ "loss": 0.4515089988708496,
+ "step": 1224
+ },
+ {
+ "epoch": 0.642707240293809,
+ "grad_norm": 0.15952153223099508,
+ "learning_rate": 1.7747163607186468e-05,
+ "loss": 0.4658638834953308,
+ "step": 1225
+ },
+ {
+ "epoch": 0.6432318992654774,
+ "grad_norm": 0.15205926699647218,
+ "learning_rate": 1.7713756003377567e-05,
+ "loss": 0.4330868422985077,
+ "step": 1226
+ },
+ {
+ "epoch": 0.6437565582371458,
+ "grad_norm": 0.15784806480936794,
+ "learning_rate": 1.7680374986602865e-05,
+ "loss": 0.4506068527698517,
+ "step": 1227
+ },
+ {
+ "epoch": 0.6442812172088143,
+ "grad_norm": 0.16006709280686687,
+ "learning_rate": 1.7647020647551105e-05,
+ "loss": 0.446941077709198,
+ "step": 1228
+ },
+ {
+ "epoch": 0.6448058761804827,
+ "grad_norm": 0.16064631158222062,
+ "learning_rate": 1.761369307683856e-05,
+ "loss": 0.4418525695800781,
+ "step": 1229
+ },
+ {
+ "epoch": 0.6453305351521511,
+ "grad_norm": 0.16454136480444392,
+ "learning_rate": 1.7580392365008767e-05,
+ "loss": 0.46268391609191895,
+ "step": 1230
+ },
+ {
+ "epoch": 0.6458551941238195,
+ "grad_norm": 0.15401246535771462,
+ "learning_rate": 1.754711860253233e-05,
+ "loss": 0.4372989535331726,
+ "step": 1231
+ },
+ {
+ "epoch": 0.646379853095488,
+ "grad_norm": 0.1544567975028383,
+ "learning_rate": 1.7513871879806603e-05,
+ "loss": 0.44806575775146484,
+ "step": 1232
+ },
+ {
+ "epoch": 0.6469045120671564,
+ "grad_norm": 0.16411225112730965,
+ "learning_rate": 1.748065228715549e-05,
+ "loss": 0.43782058358192444,
+ "step": 1233
+ },
+ {
+ "epoch": 0.6474291710388248,
+ "grad_norm": 0.1552767832972093,
+ "learning_rate": 1.7447459914829183e-05,
+ "loss": 0.44531378149986267,
+ "step": 1234
+ },
+ {
+ "epoch": 0.6479538300104932,
+ "grad_norm": 0.16214695257123513,
+ "learning_rate": 1.741429485300392e-05,
+ "loss": 0.47399842739105225,
+ "step": 1235
+ },
+ {
+ "epoch": 0.6484784889821616,
+ "grad_norm": 0.16684861111484112,
+ "learning_rate": 1.7381157191781766e-05,
+ "loss": 0.4654282331466675,
+ "step": 1236
+ },
+ {
+ "epoch": 0.64900314795383,
+ "grad_norm": 0.15901864178630273,
+ "learning_rate": 1.7348047021190334e-05,
+ "loss": 0.45128414034843445,
+ "step": 1237
+ },
+ {
+ "epoch": 0.6495278069254984,
+ "grad_norm": 0.15360991961742707,
+ "learning_rate": 1.7314964431182523e-05,
+ "loss": 0.4303375482559204,
+ "step": 1238
+ },
+ {
+ "epoch": 0.6500524658971668,
+ "grad_norm": 0.1633151574385741,
+ "learning_rate": 1.728190951163633e-05,
+ "loss": 0.45568403601646423,
+ "step": 1239
+ },
+ {
+ "epoch": 0.6505771248688352,
+ "grad_norm": 0.16377853571058645,
+ "learning_rate": 1.724888235235457e-05,
+ "loss": 0.46503543853759766,
+ "step": 1240
+ },
+ {
+ "epoch": 0.6511017838405037,
+ "grad_norm": 0.1595023801275368,
+ "learning_rate": 1.7215883043064655e-05,
+ "loss": 0.46337294578552246,
+ "step": 1241
+ },
+ {
+ "epoch": 0.6516264428121721,
+ "grad_norm": 0.16324796182212292,
+ "learning_rate": 1.7182911673418313e-05,
+ "loss": 0.46197354793548584,
+ "step": 1242
+ },
+ {
+ "epoch": 0.6521511017838405,
+ "grad_norm": 0.16353734830879765,
+ "learning_rate": 1.714996833299138e-05,
+ "loss": 0.45741671323776245,
+ "step": 1243
+ },
+ {
+ "epoch": 0.6526757607555089,
+ "grad_norm": 0.15941587678263386,
+ "learning_rate": 1.7117053111283528e-05,
+ "loss": 0.4514503479003906,
+ "step": 1244
+ },
+ {
+ "epoch": 0.6532004197271774,
+ "grad_norm": 0.1642653230145008,
+ "learning_rate": 1.708416609771804e-05,
+ "loss": 0.4437655210494995,
+ "step": 1245
+ },
+ {
+ "epoch": 0.6537250786988458,
+ "grad_norm": 0.15814155519130152,
+ "learning_rate": 1.7051307381641587e-05,
+ "loss": 0.4314163327217102,
+ "step": 1246
+ },
+ {
+ "epoch": 0.6542497376705142,
+ "grad_norm": 0.16086078370768056,
+ "learning_rate": 1.701847705232394e-05,
+ "loss": 0.4526430666446686,
+ "step": 1247
+ },
+ {
+ "epoch": 0.6547743966421826,
+ "grad_norm": 0.16389360190715183,
+ "learning_rate": 1.698567519895775e-05,
+ "loss": 0.4278966188430786,
+ "step": 1248
+ },
+ {
+ "epoch": 0.655299055613851,
+ "grad_norm": 0.16303142287804262,
+ "learning_rate": 1.6952901910658316e-05,
+ "loss": 0.45778220891952515,
+ "step": 1249
+ },
+ {
+ "epoch": 0.6558237145855194,
+ "grad_norm": 0.1675209956904048,
+ "learning_rate": 1.6920157276463318e-05,
+ "loss": 0.4489403963088989,
+ "step": 1250
+ },
+ {
+ "epoch": 0.6563483735571878,
+ "grad_norm": 0.16228311551074323,
+ "learning_rate": 1.6887441385332604e-05,
+ "loss": 0.4534342885017395,
+ "step": 1251
+ },
+ {
+ "epoch": 0.6568730325288562,
+ "grad_norm": 0.16408957851046913,
+ "learning_rate": 1.6854754326147916e-05,
+ "loss": 0.4549208879470825,
+ "step": 1252
+ },
+ {
+ "epoch": 0.6573976915005246,
+ "grad_norm": 0.15967922111432845,
+ "learning_rate": 1.682209618771268e-05,
+ "loss": 0.4451141357421875,
+ "step": 1253
+ },
+ {
+ "epoch": 0.657922350472193,
+ "grad_norm": 0.16440698722204208,
+ "learning_rate": 1.678946705875176e-05,
+ "loss": 0.454200804233551,
+ "step": 1254
+ },
+ {
+ "epoch": 0.6584470094438615,
+ "grad_norm": 0.16251120184938497,
+ "learning_rate": 1.675686702791117e-05,
+ "loss": 0.4580027461051941,
+ "step": 1255
+ },
+ {
+ "epoch": 0.6589716684155299,
+ "grad_norm": 0.1643717386150457,
+ "learning_rate": 1.6724296183757924e-05,
+ "loss": 0.4422311782836914,
+ "step": 1256
+ },
+ {
+ "epoch": 0.6594963273871983,
+ "grad_norm": 0.1644586414378722,
+ "learning_rate": 1.66917546147797e-05,
+ "loss": 0.4593179225921631,
+ "step": 1257
+ },
+ {
+ "epoch": 0.6600209863588667,
+ "grad_norm": 0.16751074186967685,
+ "learning_rate": 1.6659242409384657e-05,
+ "loss": 0.457930326461792,
+ "step": 1258
+ },
+ {
+ "epoch": 0.6605456453305352,
+ "grad_norm": 0.15398418609155334,
+ "learning_rate": 1.6626759655901182e-05,
+ "loss": 0.44002899527549744,
+ "step": 1259
+ },
+ {
+ "epoch": 0.6610703043022036,
+ "grad_norm": 0.16328113175790226,
+ "learning_rate": 1.659430644257763e-05,
+ "loss": 0.4500438868999481,
+ "step": 1260
+ },
+ {
+ "epoch": 0.661594963273872,
+ "grad_norm": 0.1694244340615106,
+ "learning_rate": 1.656188285758214e-05,
+ "loss": 0.4818427264690399,
+ "step": 1261
+ },
+ {
+ "epoch": 0.6621196222455404,
+ "grad_norm": 0.1578320701754368,
+ "learning_rate": 1.652948898900232e-05,
+ "loss": 0.4492436647415161,
+ "step": 1262
+ },
+ {
+ "epoch": 0.6626442812172089,
+ "grad_norm": 0.16538865917828469,
+ "learning_rate": 1.6497124924845064e-05,
+ "loss": 0.45070338249206543,
+ "step": 1263
+ },
+ {
+ "epoch": 0.6631689401888772,
+ "grad_norm": 0.15670817848256852,
+ "learning_rate": 1.6464790753036296e-05,
+ "loss": 0.45439285039901733,
+ "step": 1264
+ },
+ {
+ "epoch": 0.6636935991605456,
+ "grad_norm": 0.16184786585965885,
+ "learning_rate": 1.64324865614207e-05,
+ "loss": 0.446522980928421,
+ "step": 1265
+ },
+ {
+ "epoch": 0.664218258132214,
+ "grad_norm": 0.15719726462251984,
+ "learning_rate": 1.6400212437761554e-05,
+ "loss": 0.4495280981063843,
+ "step": 1266
+ },
+ {
+ "epoch": 0.6647429171038824,
+ "grad_norm": 0.1680804414134405,
+ "learning_rate": 1.6367968469740416e-05,
+ "loss": 0.4652596712112427,
+ "step": 1267
+ },
+ {
+ "epoch": 0.6652675760755509,
+ "grad_norm": 0.15919541846683566,
+ "learning_rate": 1.6335754744956937e-05,
+ "loss": 0.43842798471450806,
+ "step": 1268
+ },
+ {
+ "epoch": 0.6657922350472193,
+ "grad_norm": 0.16322634187139418,
+ "learning_rate": 1.630357135092859e-05,
+ "loss": 0.45242664217948914,
+ "step": 1269
+ },
+ {
+ "epoch": 0.6663168940188877,
+ "grad_norm": 0.15644867734734783,
+ "learning_rate": 1.6271418375090445e-05,
+ "loss": 0.44801655411720276,
+ "step": 1270
+ },
+ {
+ "epoch": 0.6668415529905561,
+ "grad_norm": 0.1514717416124423,
+ "learning_rate": 1.623929590479496e-05,
+ "loss": 0.4506194293498993,
+ "step": 1271
+ },
+ {
+ "epoch": 0.6673662119622246,
+ "grad_norm": 0.1740345085495889,
+ "learning_rate": 1.6207204027311683e-05,
+ "loss": 0.4818177819252014,
+ "step": 1272
+ },
+ {
+ "epoch": 0.667890870933893,
+ "grad_norm": 0.15797701900765457,
+ "learning_rate": 1.6175142829827067e-05,
+ "loss": 0.4639382064342499,
+ "step": 1273
+ },
+ {
+ "epoch": 0.6684155299055614,
+ "grad_norm": 0.15663812316653875,
+ "learning_rate": 1.6143112399444198e-05,
+ "loss": 0.46217501163482666,
+ "step": 1274
+ },
+ {
+ "epoch": 0.6689401888772298,
+ "grad_norm": 0.15474018977043758,
+ "learning_rate": 1.6111112823182592e-05,
+ "loss": 0.43691688776016235,
+ "step": 1275
+ },
+ {
+ "epoch": 0.6694648478488983,
+ "grad_norm": 0.1615634942529431,
+ "learning_rate": 1.6079144187977923e-05,
+ "loss": 0.4412601590156555,
+ "step": 1276
+ },
+ {
+ "epoch": 0.6699895068205667,
+ "grad_norm": 0.15440022842754444,
+ "learning_rate": 1.604720658068184e-05,
+ "loss": 0.4523273706436157,
+ "step": 1277
+ },
+ {
+ "epoch": 0.670514165792235,
+ "grad_norm": 0.1634985665776242,
+ "learning_rate": 1.6015300088061664e-05,
+ "loss": 0.46264398097991943,
+ "step": 1278
+ },
+ {
+ "epoch": 0.6710388247639034,
+ "grad_norm": 0.15037886370689527,
+ "learning_rate": 1.598342479680018e-05,
+ "loss": 0.451465904712677,
+ "step": 1279
+ },
+ {
+ "epoch": 0.6715634837355718,
+ "grad_norm": 0.16895242437746613,
+ "learning_rate": 1.595158079349542e-05,
+ "loss": 0.4466145634651184,
+ "step": 1280
+ },
+ {
+ "epoch": 0.6720881427072403,
+ "grad_norm": 0.1629548326999642,
+ "learning_rate": 1.5919768164660425e-05,
+ "loss": 0.45415782928466797,
+ "step": 1281
+ },
+ {
+ "epoch": 0.6726128016789087,
+ "grad_norm": 0.1563872102268991,
+ "learning_rate": 1.588798699672298e-05,
+ "loss": 0.4386737048625946,
+ "step": 1282
+ },
+ {
+ "epoch": 0.6731374606505771,
+ "grad_norm": 0.16038172773344717,
+ "learning_rate": 1.5856237376025403e-05,
+ "loss": 0.446260929107666,
+ "step": 1283
+ },
+ {
+ "epoch": 0.6736621196222455,
+ "grad_norm": 0.16189780714407623,
+ "learning_rate": 1.582451938882431e-05,
+ "loss": 0.45458072423934937,
+ "step": 1284
+ },
+ {
+ "epoch": 0.674186778593914,
+ "grad_norm": 0.15204521785049285,
+ "learning_rate": 1.5792833121290367e-05,
+ "loss": 0.4425617754459381,
+ "step": 1285
+ },
+ {
+ "epoch": 0.6747114375655824,
+ "grad_norm": 0.16237376704792778,
+ "learning_rate": 1.5761178659508063e-05,
+ "loss": 0.44631800055503845,
+ "step": 1286
+ },
+ {
+ "epoch": 0.6752360965372508,
+ "grad_norm": 0.1582063618466069,
+ "learning_rate": 1.57295560894755e-05,
+ "loss": 0.44280534982681274,
+ "step": 1287
+ },
+ {
+ "epoch": 0.6757607555089192,
+ "grad_norm": 0.16587032349847394,
+ "learning_rate": 1.5697965497104116e-05,
+ "loss": 0.4495997130870819,
+ "step": 1288
+ },
+ {
+ "epoch": 0.6762854144805877,
+ "grad_norm": 0.1557269262755563,
+ "learning_rate": 1.566640696821848e-05,
+ "loss": 0.4493359625339508,
+ "step": 1289
+ },
+ {
+ "epoch": 0.6768100734522561,
+ "grad_norm": 0.16038490357012256,
+ "learning_rate": 1.5634880588556043e-05,
+ "loss": 0.4447544813156128,
+ "step": 1290
+ },
+ {
+ "epoch": 0.6773347324239245,
+ "grad_norm": 0.15355201908546406,
+ "learning_rate": 1.560338644376692e-05,
+ "loss": 0.4468432664871216,
+ "step": 1291
+ },
+ {
+ "epoch": 0.6778593913955928,
+ "grad_norm": 0.15077009515589776,
+ "learning_rate": 1.5571924619413668e-05,
+ "loss": 0.44959044456481934,
+ "step": 1292
+ },
+ {
+ "epoch": 0.6783840503672612,
+ "grad_norm": 0.1583038146613965,
+ "learning_rate": 1.5540495200971012e-05,
+ "loss": 0.42450425028800964,
+ "step": 1293
+ },
+ {
+ "epoch": 0.6789087093389297,
+ "grad_norm": 0.15950687190598087,
+ "learning_rate": 1.5509098273825635e-05,
+ "loss": 0.4527238607406616,
+ "step": 1294
+ },
+ {
+ "epoch": 0.6794333683105981,
+ "grad_norm": 0.1526262208847808,
+ "learning_rate": 1.5477733923275968e-05,
+ "loss": 0.4379724860191345,
+ "step": 1295
+ },
+ {
+ "epoch": 0.6799580272822665,
+ "grad_norm": 0.15541579650021778,
+ "learning_rate": 1.5446402234531934e-05,
+ "loss": 0.4462166130542755,
+ "step": 1296
+ },
+ {
+ "epoch": 0.6804826862539349,
+ "grad_norm": 0.15483930734313786,
+ "learning_rate": 1.5415103292714718e-05,
+ "loss": 0.4390607178211212,
+ "step": 1297
+ },
+ {
+ "epoch": 0.6810073452256034,
+ "grad_norm": 0.1578005731149154,
+ "learning_rate": 1.5383837182856532e-05,
+ "loss": 0.44462254643440247,
+ "step": 1298
+ },
+ {
+ "epoch": 0.6815320041972718,
+ "grad_norm": 0.15830183884958823,
+ "learning_rate": 1.5352603989900406e-05,
+ "loss": 0.4563537538051605,
+ "step": 1299
+ },
+ {
+ "epoch": 0.6820566631689402,
+ "grad_norm": 0.15605472163445522,
+ "learning_rate": 1.5321403798699923e-05,
+ "loss": 0.43852412700653076,
+ "step": 1300
+ },
+ {
+ "epoch": 0.6825813221406086,
+ "grad_norm": 0.15770470227141278,
+ "learning_rate": 1.5290236694019017e-05,
+ "loss": 0.4438081979751587,
+ "step": 1301
+ },
+ {
+ "epoch": 0.683105981112277,
+ "grad_norm": 0.16175339845012077,
+ "learning_rate": 1.5259102760531747e-05,
+ "loss": 0.46670734882354736,
+ "step": 1302
+ },
+ {
+ "epoch": 0.6836306400839455,
+ "grad_norm": 0.16359504716154477,
+ "learning_rate": 1.5228002082822036e-05,
+ "loss": 0.4590187668800354,
+ "step": 1303
+ },
+ {
+ "epoch": 0.6841552990556139,
+ "grad_norm": 0.15422879500254713,
+ "learning_rate": 1.5196934745383457e-05,
+ "loss": 0.45287445187568665,
+ "step": 1304
+ },
+ {
+ "epoch": 0.6846799580272823,
+ "grad_norm": 0.15929902143574254,
+ "learning_rate": 1.5165900832619015e-05,
+ "loss": 0.44286859035491943,
+ "step": 1305
+ },
+ {
+ "epoch": 0.6852046169989506,
+ "grad_norm": 0.14838868714016124,
+ "learning_rate": 1.5134900428840904e-05,
+ "loss": 0.45155346393585205,
+ "step": 1306
+ },
+ {
+ "epoch": 0.6857292759706191,
+ "grad_norm": 0.15494288640397672,
+ "learning_rate": 1.5103933618270283e-05,
+ "loss": 0.44010141491889954,
+ "step": 1307
+ },
+ {
+ "epoch": 0.6862539349422875,
+ "grad_norm": 0.1686262360598338,
+ "learning_rate": 1.5073000485037042e-05,
+ "loss": 0.446555495262146,
+ "step": 1308
+ },
+ {
+ "epoch": 0.6867785939139559,
+ "grad_norm": 0.16122110823982602,
+ "learning_rate": 1.5042101113179579e-05,
+ "loss": 0.44640374183654785,
+ "step": 1309
+ },
+ {
+ "epoch": 0.6873032528856243,
+ "grad_norm": 0.16849937797919867,
+ "learning_rate": 1.5011235586644572e-05,
+ "loss": 0.4465785026550293,
+ "step": 1310
+ },
+ {
+ "epoch": 0.6878279118572928,
+ "grad_norm": 0.16062971262800754,
+ "learning_rate": 1.498040398928674e-05,
+ "loss": 0.445321649312973,
+ "step": 1311
+ },
+ {
+ "epoch": 0.6883525708289612,
+ "grad_norm": 0.1608953185993034,
+ "learning_rate": 1.4949606404868638e-05,
+ "loss": 0.46518227458000183,
+ "step": 1312
+ },
+ {
+ "epoch": 0.6888772298006296,
+ "grad_norm": 0.15880255672830915,
+ "learning_rate": 1.491884291706041e-05,
+ "loss": 0.44582217931747437,
+ "step": 1313
+ },
+ {
+ "epoch": 0.689401888772298,
+ "grad_norm": 0.16715145200789822,
+ "learning_rate": 1.4888113609439563e-05,
+ "loss": 0.4435983896255493,
+ "step": 1314
+ },
+ {
+ "epoch": 0.6899265477439664,
+ "grad_norm": 0.15263740340694454,
+ "learning_rate": 1.485741856549075e-05,
+ "loss": 0.4405066967010498,
+ "step": 1315
+ },
+ {
+ "epoch": 0.6904512067156349,
+ "grad_norm": 0.16247034518813966,
+ "learning_rate": 1.4826757868605517e-05,
+ "loss": 0.4544215500354767,
+ "step": 1316
+ },
+ {
+ "epoch": 0.6909758656873033,
+ "grad_norm": 0.15649146705978162,
+ "learning_rate": 1.4796131602082142e-05,
+ "loss": 0.4454413056373596,
+ "step": 1317
+ },
+ {
+ "epoch": 0.6915005246589717,
+ "grad_norm": 0.16286718636058178,
+ "learning_rate": 1.4765539849125318e-05,
+ "loss": 0.4497183561325073,
+ "step": 1318
+ },
+ {
+ "epoch": 0.6920251836306401,
+ "grad_norm": 0.15812342085005582,
+ "learning_rate": 1.473498269284599e-05,
+ "loss": 0.445032000541687,
+ "step": 1319
+ },
+ {
+ "epoch": 0.6925498426023085,
+ "grad_norm": 0.1558101290511678,
+ "learning_rate": 1.4704460216261123e-05,
+ "loss": 0.45769619941711426,
+ "step": 1320
+ },
+ {
+ "epoch": 0.6930745015739769,
+ "grad_norm": 0.15642889059495207,
+ "learning_rate": 1.4673972502293428e-05,
+ "loss": 0.44788360595703125,
+ "step": 1321
+ },
+ {
+ "epoch": 0.6935991605456453,
+ "grad_norm": 0.15722513518113074,
+ "learning_rate": 1.4643519633771222e-05,
+ "loss": 0.449631929397583,
+ "step": 1322
+ },
+ {
+ "epoch": 0.6941238195173137,
+ "grad_norm": 0.15587633748081542,
+ "learning_rate": 1.4613101693428125e-05,
+ "loss": 0.4450187087059021,
+ "step": 1323
+ },
+ {
+ "epoch": 0.6946484784889821,
+ "grad_norm": 0.1577107074882479,
+ "learning_rate": 1.458271876390287e-05,
+ "loss": 0.45108890533447266,
+ "step": 1324
+ },
+ {
+ "epoch": 0.6951731374606506,
+ "grad_norm": 0.15649839486990696,
+ "learning_rate": 1.4552370927739073e-05,
+ "loss": 0.46430301666259766,
+ "step": 1325
+ },
+ {
+ "epoch": 0.695697796432319,
+ "grad_norm": 0.18573600420690636,
+ "learning_rate": 1.4522058267385008e-05,
+ "loss": 0.4701972007751465,
+ "step": 1326
+ },
+ {
+ "epoch": 0.6962224554039874,
+ "grad_norm": 0.16296497084037542,
+ "learning_rate": 1.4491780865193402e-05,
+ "loss": 0.4607710838317871,
+ "step": 1327
+ },
+ {
+ "epoch": 0.6967471143756558,
+ "grad_norm": 0.161124007217343,
+ "learning_rate": 1.446153880342117e-05,
+ "loss": 0.45095115900039673,
+ "step": 1328
+ },
+ {
+ "epoch": 0.6972717733473243,
+ "grad_norm": 0.1537884039166766,
+ "learning_rate": 1.4431332164229222e-05,
+ "loss": 0.424068808555603,
+ "step": 1329
+ },
+ {
+ "epoch": 0.6977964323189927,
+ "grad_norm": 0.16588834503456087,
+ "learning_rate": 1.4401161029682237e-05,
+ "loss": 0.4622839093208313,
+ "step": 1330
+ },
+ {
+ "epoch": 0.6983210912906611,
+ "grad_norm": 0.16100510604015544,
+ "learning_rate": 1.4371025481748435e-05,
+ "loss": 0.4660705626010895,
+ "step": 1331
+ },
+ {
+ "epoch": 0.6988457502623295,
+ "grad_norm": 0.15559888940590075,
+ "learning_rate": 1.4340925602299337e-05,
+ "loss": 0.4484976530075073,
+ "step": 1332
+ },
+ {
+ "epoch": 0.6993704092339978,
+ "grad_norm": 0.15503821966213488,
+ "learning_rate": 1.4310861473109616e-05,
+ "loss": 0.45636600255966187,
+ "step": 1333
+ },
+ {
+ "epoch": 0.6998950682056663,
+ "grad_norm": 0.15539074896109334,
+ "learning_rate": 1.4280833175856743e-05,
+ "loss": 0.45863068103790283,
+ "step": 1334
+ },
+ {
+ "epoch": 0.7004197271773347,
+ "grad_norm": 0.15664821009154872,
+ "learning_rate": 1.4250840792120896e-05,
+ "loss": 0.44043463468551636,
+ "step": 1335
+ },
+ {
+ "epoch": 0.7009443861490031,
+ "grad_norm": 0.15825478808520285,
+ "learning_rate": 1.4220884403384655e-05,
+ "loss": 0.46522051095962524,
+ "step": 1336
+ },
+ {
+ "epoch": 0.7014690451206715,
+ "grad_norm": 0.16004268568141544,
+ "learning_rate": 1.4190964091032843e-05,
+ "loss": 0.46114686131477356,
+ "step": 1337
+ },
+ {
+ "epoch": 0.70199370409234,
+ "grad_norm": 0.16413139949945635,
+ "learning_rate": 1.4161079936352234e-05,
+ "loss": 0.4457515478134155,
+ "step": 1338
+ },
+ {
+ "epoch": 0.7025183630640084,
+ "grad_norm": 0.1538890577972702,
+ "learning_rate": 1.4131232020531398e-05,
+ "loss": 0.4616093337535858,
+ "step": 1339
+ },
+ {
+ "epoch": 0.7030430220356768,
+ "grad_norm": 0.16240737854323284,
+ "learning_rate": 1.4101420424660434e-05,
+ "loss": 0.44631537795066833,
+ "step": 1340
+ },
+ {
+ "epoch": 0.7035676810073452,
+ "grad_norm": 0.15984708603406073,
+ "learning_rate": 1.4071645229730762e-05,
+ "loss": 0.4366717040538788,
+ "step": 1341
+ },
+ {
+ "epoch": 0.7040923399790137,
+ "grad_norm": 0.15149695695635904,
+ "learning_rate": 1.4041906516634946e-05,
+ "loss": 0.44129428267478943,
+ "step": 1342
+ },
+ {
+ "epoch": 0.7046169989506821,
+ "grad_norm": 0.15041167501768415,
+ "learning_rate": 1.4012204366166399e-05,
+ "loss": 0.43654775619506836,
+ "step": 1343
+ },
+ {
+ "epoch": 0.7051416579223505,
+ "grad_norm": 0.1596097627963756,
+ "learning_rate": 1.3982538859019217e-05,
+ "loss": 0.4652445614337921,
+ "step": 1344
+ },
+ {
+ "epoch": 0.7056663168940189,
+ "grad_norm": 0.15621704205027884,
+ "learning_rate": 1.3952910075787941e-05,
+ "loss": 0.44577041268348694,
+ "step": 1345
+ },
+ {
+ "epoch": 0.7061909758656874,
+ "grad_norm": 0.16073041098008783,
+ "learning_rate": 1.3923318096967341e-05,
+ "loss": 0.4391976594924927,
+ "step": 1346
+ },
+ {
+ "epoch": 0.7067156348373557,
+ "grad_norm": 0.14793853946372448,
+ "learning_rate": 1.3893763002952204e-05,
+ "loss": 0.4463799297809601,
+ "step": 1347
+ },
+ {
+ "epoch": 0.7072402938090241,
+ "grad_norm": 0.16176080358867354,
+ "learning_rate": 1.3864244874037097e-05,
+ "loss": 0.45183271169662476,
+ "step": 1348
+ },
+ {
+ "epoch": 0.7077649527806925,
+ "grad_norm": 0.16372771179026924,
+ "learning_rate": 1.3834763790416171e-05,
+ "loss": 0.44205576181411743,
+ "step": 1349
+ },
+ {
+ "epoch": 0.7082896117523609,
+ "grad_norm": 0.15135222516127125,
+ "learning_rate": 1.3805319832182937e-05,
+ "loss": 0.4390261173248291,
+ "step": 1350
+ },
+ {
+ "epoch": 0.7088142707240294,
+ "grad_norm": 0.15493183225792365,
+ "learning_rate": 1.377591307933002e-05,
+ "loss": 0.45837974548339844,
+ "step": 1351
+ },
+ {
+ "epoch": 0.7093389296956978,
+ "grad_norm": 0.16405906906516504,
+ "learning_rate": 1.3746543611749008e-05,
+ "loss": 0.45032668113708496,
+ "step": 1352
+ },
+ {
+ "epoch": 0.7098635886673662,
+ "grad_norm": 0.15725677976524505,
+ "learning_rate": 1.3717211509230155e-05,
+ "loss": 0.4486224949359894,
+ "step": 1353
+ },
+ {
+ "epoch": 0.7103882476390346,
+ "grad_norm": 0.15781469066983164,
+ "learning_rate": 1.3687916851462227e-05,
+ "loss": 0.4502638578414917,
+ "step": 1354
+ },
+ {
+ "epoch": 0.710912906610703,
+ "grad_norm": 0.16115007729054298,
+ "learning_rate": 1.3658659718032241e-05,
+ "loss": 0.4508914351463318,
+ "step": 1355
+ },
+ {
+ "epoch": 0.7114375655823715,
+ "grad_norm": 0.16755060475968744,
+ "learning_rate": 1.3629440188425286e-05,
+ "loss": 0.45653945207595825,
+ "step": 1356
+ },
+ {
+ "epoch": 0.7119622245540399,
+ "grad_norm": 0.1574994294756123,
+ "learning_rate": 1.3600258342024268e-05,
+ "loss": 0.4349849820137024,
+ "step": 1357
+ },
+ {
+ "epoch": 0.7124868835257083,
+ "grad_norm": 0.16104440922657715,
+ "learning_rate": 1.3571114258109758e-05,
+ "loss": 0.4501233696937561,
+ "step": 1358
+ },
+ {
+ "epoch": 0.7130115424973767,
+ "grad_norm": 0.15163184290939022,
+ "learning_rate": 1.3542008015859691e-05,
+ "loss": 0.44171035289764404,
+ "step": 1359
+ },
+ {
+ "epoch": 0.7135362014690452,
+ "grad_norm": 0.15955896674456582,
+ "learning_rate": 1.3512939694349211e-05,
+ "loss": 0.4558671712875366,
+ "step": 1360
+ },
+ {
+ "epoch": 0.7140608604407135,
+ "grad_norm": 0.1589376751536802,
+ "learning_rate": 1.3483909372550446e-05,
+ "loss": 0.4430835247039795,
+ "step": 1361
+ },
+ {
+ "epoch": 0.7145855194123819,
+ "grad_norm": 0.15979088397822155,
+ "learning_rate": 1.3454917129332279e-05,
+ "loss": 0.4426245391368866,
+ "step": 1362
+ },
+ {
+ "epoch": 0.7151101783840503,
+ "grad_norm": 0.15493179714072167,
+ "learning_rate": 1.3425963043460143e-05,
+ "loss": 0.4550474286079407,
+ "step": 1363
+ },
+ {
+ "epoch": 0.7156348373557188,
+ "grad_norm": 0.15396498968995032,
+ "learning_rate": 1.3397047193595808e-05,
+ "loss": 0.44220536947250366,
+ "step": 1364
+ },
+ {
+ "epoch": 0.7161594963273872,
+ "grad_norm": 0.1489346822667273,
+ "learning_rate": 1.336816965829717e-05,
+ "loss": 0.4437323212623596,
+ "step": 1365
+ },
+ {
+ "epoch": 0.7166841552990556,
+ "grad_norm": 0.16311964931734477,
+ "learning_rate": 1.333933051601802e-05,
+ "loss": 0.45308777689933777,
+ "step": 1366
+ },
+ {
+ "epoch": 0.717208814270724,
+ "grad_norm": 0.1534354619021179,
+ "learning_rate": 1.331052984510785e-05,
+ "loss": 0.4334886372089386,
+ "step": 1367
+ },
+ {
+ "epoch": 0.7177334732423925,
+ "grad_norm": 0.15541089758065404,
+ "learning_rate": 1.328176772381165e-05,
+ "loss": 0.45248323678970337,
+ "step": 1368
+ },
+ {
+ "epoch": 0.7182581322140609,
+ "grad_norm": 0.1617921305700596,
+ "learning_rate": 1.3253044230269651e-05,
+ "loss": 0.44092947244644165,
+ "step": 1369
+ },
+ {
+ "epoch": 0.7187827911857293,
+ "grad_norm": 0.15047353793760151,
+ "learning_rate": 1.3224359442517165e-05,
+ "loss": 0.4424414336681366,
+ "step": 1370
+ },
+ {
+ "epoch": 0.7193074501573977,
+ "grad_norm": 0.15958089102258743,
+ "learning_rate": 1.319571343848433e-05,
+ "loss": 0.4458580017089844,
+ "step": 1371
+ },
+ {
+ "epoch": 0.7198321091290661,
+ "grad_norm": 0.15349258306968733,
+ "learning_rate": 1.3167106295995924e-05,
+ "loss": 0.4285386800765991,
+ "step": 1372
+ },
+ {
+ "epoch": 0.7203567681007346,
+ "grad_norm": 0.16042529563759886,
+ "learning_rate": 1.3138538092771164e-05,
+ "loss": 0.4497690200805664,
+ "step": 1373
+ },
+ {
+ "epoch": 0.720881427072403,
+ "grad_norm": 0.1565113736588754,
+ "learning_rate": 1.3110008906423454e-05,
+ "loss": 0.43562617897987366,
+ "step": 1374
+ },
+ {
+ "epoch": 0.7214060860440713,
+ "grad_norm": 0.15177291086871506,
+ "learning_rate": 1.3081518814460215e-05,
+ "loss": 0.44456908106803894,
+ "step": 1375
+ },
+ {
+ "epoch": 0.7219307450157397,
+ "grad_norm": 0.15928490480439225,
+ "learning_rate": 1.3053067894282622e-05,
+ "loss": 0.46170780062675476,
+ "step": 1376
+ },
+ {
+ "epoch": 0.7224554039874082,
+ "grad_norm": 0.1554814469217892,
+ "learning_rate": 1.302465622318549e-05,
+ "loss": 0.4422169327735901,
+ "step": 1377
+ },
+ {
+ "epoch": 0.7229800629590766,
+ "grad_norm": 0.15375370214966386,
+ "learning_rate": 1.2996283878356947e-05,
+ "loss": 0.4564763903617859,
+ "step": 1378
+ },
+ {
+ "epoch": 0.723504721930745,
+ "grad_norm": 0.15942841969395802,
+ "learning_rate": 1.2967950936878311e-05,
+ "loss": 0.45121875405311584,
+ "step": 1379
+ },
+ {
+ "epoch": 0.7240293809024134,
+ "grad_norm": 0.15731308716774903,
+ "learning_rate": 1.2939657475723844e-05,
+ "loss": 0.4285027086734772,
+ "step": 1380
+ },
+ {
+ "epoch": 0.7245540398740818,
+ "grad_norm": 0.15997753613932147,
+ "learning_rate": 1.291140357176054e-05,
+ "loss": 0.4566386342048645,
+ "step": 1381
+ },
+ {
+ "epoch": 0.7250786988457503,
+ "grad_norm": 0.15333715267473727,
+ "learning_rate": 1.2883189301747927e-05,
+ "loss": 0.45093846321105957,
+ "step": 1382
+ },
+ {
+ "epoch": 0.7256033578174187,
+ "grad_norm": 0.1514510304709061,
+ "learning_rate": 1.2855014742337874e-05,
+ "loss": 0.44256967306137085,
+ "step": 1383
+ },
+ {
+ "epoch": 0.7261280167890871,
+ "grad_norm": 0.15624301298995746,
+ "learning_rate": 1.2826879970074346e-05,
+ "loss": 0.4361875653266907,
+ "step": 1384
+ },
+ {
+ "epoch": 0.7266526757607555,
+ "grad_norm": 0.15767850122006694,
+ "learning_rate": 1.279878506139322e-05,
+ "loss": 0.4477671980857849,
+ "step": 1385
+ },
+ {
+ "epoch": 0.727177334732424,
+ "grad_norm": 0.15537676969694064,
+ "learning_rate": 1.2770730092622072e-05,
+ "loss": 0.4339832663536072,
+ "step": 1386
+ },
+ {
+ "epoch": 0.7277019937040924,
+ "grad_norm": 0.15944113633281293,
+ "learning_rate": 1.2742715139979965e-05,
+ "loss": 0.45407935976982117,
+ "step": 1387
+ },
+ {
+ "epoch": 0.7282266526757608,
+ "grad_norm": 0.16763519518606218,
+ "learning_rate": 1.271474027957727e-05,
+ "loss": 0.46442392468452454,
+ "step": 1388
+ },
+ {
+ "epoch": 0.7287513116474291,
+ "grad_norm": 0.1564572774601902,
+ "learning_rate": 1.2686805587415423e-05,
+ "loss": 0.4463968276977539,
+ "step": 1389
+ },
+ {
+ "epoch": 0.7292759706190975,
+ "grad_norm": 0.16384626314986128,
+ "learning_rate": 1.2658911139386709e-05,
+ "loss": 0.45519885420799255,
+ "step": 1390
+ },
+ {
+ "epoch": 0.729800629590766,
+ "grad_norm": 0.1562023802820964,
+ "learning_rate": 1.2631057011274106e-05,
+ "loss": 0.4365222454071045,
+ "step": 1391
+ },
+ {
+ "epoch": 0.7303252885624344,
+ "grad_norm": 0.15597287232940926,
+ "learning_rate": 1.2603243278751042e-05,
+ "loss": 0.4328266680240631,
+ "step": 1392
+ },
+ {
+ "epoch": 0.7308499475341028,
+ "grad_norm": 0.14762833706132167,
+ "learning_rate": 1.2575470017381214e-05,
+ "loss": 0.43902426958084106,
+ "step": 1393
+ },
+ {
+ "epoch": 0.7313746065057712,
+ "grad_norm": 0.1688371515856315,
+ "learning_rate": 1.2547737302618351e-05,
+ "loss": 0.4649963676929474,
+ "step": 1394
+ },
+ {
+ "epoch": 0.7318992654774397,
+ "grad_norm": 0.15501514509586015,
+ "learning_rate": 1.2520045209806024e-05,
+ "loss": 0.4501558542251587,
+ "step": 1395
+ },
+ {
+ "epoch": 0.7324239244491081,
+ "grad_norm": 0.15516748917693537,
+ "learning_rate": 1.2492393814177457e-05,
+ "loss": 0.4405498206615448,
+ "step": 1396
+ },
+ {
+ "epoch": 0.7329485834207765,
+ "grad_norm": 0.17115400468199243,
+ "learning_rate": 1.2464783190855295e-05,
+ "loss": 0.4445379674434662,
+ "step": 1397
+ },
+ {
+ "epoch": 0.7334732423924449,
+ "grad_norm": 0.15966386341604186,
+ "learning_rate": 1.243721341485143e-05,
+ "loss": 0.4769889712333679,
+ "step": 1398
+ },
+ {
+ "epoch": 0.7339979013641134,
+ "grad_norm": 0.15614706359781472,
+ "learning_rate": 1.2409684561066766e-05,
+ "loss": 0.43643918633461,
+ "step": 1399
+ },
+ {
+ "epoch": 0.7345225603357818,
+ "grad_norm": 0.17467968960064226,
+ "learning_rate": 1.238219670429104e-05,
+ "loss": 0.45413005352020264,
+ "step": 1400
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 3.704774381797376e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1450/chat_template.jinja b/checkpoints/checkpoint-1450/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1450/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1450/config.json b/checkpoints/checkpoint-1450/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1450/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1450/expert_prune_metadata.json b/checkpoints/checkpoint-1450/expert_prune_metadata.json
new file mode 100644
index 0000000000000000000000000000000000000000..16e01c8a10636331e334732fa9629579c68df44e
--- /dev/null
+++ b/checkpoints/checkpoint-1450/expert_prune_metadata.json
@@ -0,0 +1,14 @@
+{
+ "source_model": "/mnt/data/wuwei/model_list/Qwen3.6-35B-A3B/models--Qwen--Qwen3.6-35B-A3B/snapshots/995ad96eacd98c81ed38be0c5b274b04031597b0",
+ "output_model": "/mnt/data/wuwei/model_list/qwen3-19b-a3b",
+ "total_parameters": 18990568816,
+ "parameter_billions_decimal": 18.990568816,
+ "directory_size_label": "19B",
+ "original_num_experts": 256,
+ "kept_experts_per_layer": 128,
+ "num_profiled_layers": 40,
+ "mtp_weights": "removed because Qwen3_5MoeForConditionalGeneration does not instantiate or use the source checkpoint's mtp.* training-only tensors",
+ "expert_order": "ascending original expert index; router rows and expert tensors use the same remapping",
+ "router_statistics": "router_stats.json and router_stats.pt",
+ "selection_score": "top-8 router hit frequency * mean(abs(actual expert output after down_proj and before routing-weight multiplication))"
+}
diff --git a/checkpoints/checkpoint-1450/generation_config.json b/checkpoints/checkpoint-1450/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1450/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1450/model.safetensors b/checkpoints/checkpoint-1450/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..7f5c4e30b5c6a714df5d6bea6fb05f88b97000c2
--- /dev/null
+++ b/checkpoints/checkpoint-1450/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:36465ae327419957f43f1c61cecfb2c3fbf10e13631143480a47cda70cee17da
+size 37983421000
diff --git a/checkpoints/checkpoint-1450/preprocessor_config.json b/checkpoints/checkpoint-1450/preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ec43d1f859ecadf800ebae92fda4a812d34975d3
--- /dev/null
+++ b/checkpoints/checkpoint-1450/preprocessor_config.json
@@ -0,0 +1,26 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "image_mean": [
+ 0.48145466,
+ 0.4578275,
+ 0.40821073
+ ],
+ "image_processor_type": "Qwen2VLImageProcessorFast",
+ "image_std": [
+ 0.26862954,
+ 0.26130258,
+ 0.27577711
+ ],
+ "merge_size": 2,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 1003520,
+ "shortest_edge": 3136
+ },
+ "temporal_patch_size": 2
+}
diff --git a/checkpoints/checkpoint-1450/tokenizer.json b/checkpoints/checkpoint-1450/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1450/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1450/tokenizer_config.json b/checkpoints/checkpoint-1450/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1450/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1450/trainer_state.json b/checkpoints/checkpoint-1450/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..70b3a2984d7e4432a7b95ed6c2a6994c9065299f
--- /dev/null
+++ b/checkpoints/checkpoint-1450/trainer_state.json
@@ -0,0 +1,10184 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.7607555089192025,
+ "eval_steps": 500,
+ "global_step": 1450,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ },
+ {
+ "epoch": 0.5776495278069255,
+ "grad_norm": 0.1862053032854098,
+ "learning_rate": 2.2065553964033126e-05,
+ "loss": 0.4816875457763672,
+ "step": 1101
+ },
+ {
+ "epoch": 0.5781741867785939,
+ "grad_norm": 0.16996015894363117,
+ "learning_rate": 2.20295734045532e-05,
+ "loss": 0.4471197724342346,
+ "step": 1102
+ },
+ {
+ "epoch": 0.5786988457502623,
+ "grad_norm": 0.16889227008390692,
+ "learning_rate": 2.1993607706997725e-05,
+ "loss": 0.468391478061676,
+ "step": 1103
+ },
+ {
+ "epoch": 0.5792235047219307,
+ "grad_norm": 0.16670931681562373,
+ "learning_rate": 2.195765696907745e-05,
+ "loss": 0.44934162497520447,
+ "step": 1104
+ },
+ {
+ "epoch": 0.5797481636935992,
+ "grad_norm": 0.16328418741223716,
+ "learning_rate": 2.1921721288462482e-05,
+ "loss": 0.4558270275592804,
+ "step": 1105
+ },
+ {
+ "epoch": 0.5802728226652676,
+ "grad_norm": 0.17364660408703936,
+ "learning_rate": 2.188580076278203e-05,
+ "loss": 0.4597751796245575,
+ "step": 1106
+ },
+ {
+ "epoch": 0.580797481636936,
+ "grad_norm": 0.1628618453800601,
+ "learning_rate": 2.18498954896241e-05,
+ "loss": 0.4521859288215637,
+ "step": 1107
+ },
+ {
+ "epoch": 0.5813221406086044,
+ "grad_norm": 0.16813817951871482,
+ "learning_rate": 2.1814005566535302e-05,
+ "loss": 0.4581005573272705,
+ "step": 1108
+ },
+ {
+ "epoch": 0.5818467995802729,
+ "grad_norm": 0.16825923629726025,
+ "learning_rate": 2.1778131091020516e-05,
+ "loss": 0.4433916211128235,
+ "step": 1109
+ },
+ {
+ "epoch": 0.5823714585519413,
+ "grad_norm": 0.17194743467883944,
+ "learning_rate": 2.1742272160542654e-05,
+ "loss": 0.47296425700187683,
+ "step": 1110
+ },
+ {
+ "epoch": 0.5828961175236097,
+ "grad_norm": 0.16947730546039969,
+ "learning_rate": 2.170642887252241e-05,
+ "loss": 0.4815937876701355,
+ "step": 1111
+ },
+ {
+ "epoch": 0.583420776495278,
+ "grad_norm": 0.1657449928668473,
+ "learning_rate": 2.167060132433798e-05,
+ "loss": 0.4479164779186249,
+ "step": 1112
+ },
+ {
+ "epoch": 0.5839454354669464,
+ "grad_norm": 0.1658952516607364,
+ "learning_rate": 2.1634789613324778e-05,
+ "loss": 0.4637463688850403,
+ "step": 1113
+ },
+ {
+ "epoch": 0.5844700944386149,
+ "grad_norm": 0.16777002188998236,
+ "learning_rate": 2.1598993836775218e-05,
+ "loss": 0.457524836063385,
+ "step": 1114
+ },
+ {
+ "epoch": 0.5849947534102833,
+ "grad_norm": 0.15837566447341322,
+ "learning_rate": 2.1563214091938405e-05,
+ "loss": 0.44938087463378906,
+ "step": 1115
+ },
+ {
+ "epoch": 0.5855194123819517,
+ "grad_norm": 0.17162206617035314,
+ "learning_rate": 2.1527450476019896e-05,
+ "loss": 0.4393450617790222,
+ "step": 1116
+ },
+ {
+ "epoch": 0.5860440713536201,
+ "grad_norm": 0.16224101921433814,
+ "learning_rate": 2.1491703086181426e-05,
+ "loss": 0.451913982629776,
+ "step": 1117
+ },
+ {
+ "epoch": 0.5865687303252886,
+ "grad_norm": 0.1736253504575144,
+ "learning_rate": 2.145597201954066e-05,
+ "loss": 0.4611290395259857,
+ "step": 1118
+ },
+ {
+ "epoch": 0.587093389296957,
+ "grad_norm": 0.16308408333737234,
+ "learning_rate": 2.1420257373170903e-05,
+ "loss": 0.45898568630218506,
+ "step": 1119
+ },
+ {
+ "epoch": 0.5876180482686254,
+ "grad_norm": 0.16948634794175524,
+ "learning_rate": 2.1384559244100848e-05,
+ "loss": 0.45982882380485535,
+ "step": 1120
+ },
+ {
+ "epoch": 0.5881427072402938,
+ "grad_norm": 0.16676366538565252,
+ "learning_rate": 2.1348877729314336e-05,
+ "loss": 0.44826576113700867,
+ "step": 1121
+ },
+ {
+ "epoch": 0.5886673662119623,
+ "grad_norm": 0.16431125630712295,
+ "learning_rate": 2.1313212925750037e-05,
+ "loss": 0.4616445302963257,
+ "step": 1122
+ },
+ {
+ "epoch": 0.5891920251836307,
+ "grad_norm": 0.1631659061616955,
+ "learning_rate": 2.1277564930301252e-05,
+ "loss": 0.45651188492774963,
+ "step": 1123
+ },
+ {
+ "epoch": 0.5897166841552991,
+ "grad_norm": 0.15805950502689237,
+ "learning_rate": 2.1241933839815604e-05,
+ "loss": 0.4486045837402344,
+ "step": 1124
+ },
+ {
+ "epoch": 0.5902413431269675,
+ "grad_norm": 0.16530288038603502,
+ "learning_rate": 2.1206319751094773e-05,
+ "loss": 0.4485066533088684,
+ "step": 1125
+ },
+ {
+ "epoch": 0.5907660020986358,
+ "grad_norm": 0.16061903194239047,
+ "learning_rate": 2.1170722760894295e-05,
+ "loss": 0.4565330743789673,
+ "step": 1126
+ },
+ {
+ "epoch": 0.5912906610703043,
+ "grad_norm": 0.1643346999353627,
+ "learning_rate": 2.1135142965923195e-05,
+ "loss": 0.45560717582702637,
+ "step": 1127
+ },
+ {
+ "epoch": 0.5918153200419727,
+ "grad_norm": 0.17484644315215753,
+ "learning_rate": 2.109958046284382e-05,
+ "loss": 0.4795891046524048,
+ "step": 1128
+ },
+ {
+ "epoch": 0.5923399790136411,
+ "grad_norm": 0.1623762678622651,
+ "learning_rate": 2.1064035348271528e-05,
+ "loss": 0.47395604848861694,
+ "step": 1129
+ },
+ {
+ "epoch": 0.5928646379853095,
+ "grad_norm": 0.16888824241669015,
+ "learning_rate": 2.1028507718774456e-05,
+ "loss": 0.4601455330848694,
+ "step": 1130
+ },
+ {
+ "epoch": 0.593389296956978,
+ "grad_norm": 0.16291744568806302,
+ "learning_rate": 2.0992997670873193e-05,
+ "loss": 0.46248236298561096,
+ "step": 1131
+ },
+ {
+ "epoch": 0.5939139559286464,
+ "grad_norm": 0.16185598691236988,
+ "learning_rate": 2.0957505301040603e-05,
+ "loss": 0.45085301995277405,
+ "step": 1132
+ },
+ {
+ "epoch": 0.5944386149003148,
+ "grad_norm": 0.1678000806117814,
+ "learning_rate": 2.0922030705701518e-05,
+ "loss": 0.44939014315605164,
+ "step": 1133
+ },
+ {
+ "epoch": 0.5949632738719832,
+ "grad_norm": 0.16104474282288994,
+ "learning_rate": 2.088657398123246e-05,
+ "loss": 0.4313976764678955,
+ "step": 1134
+ },
+ {
+ "epoch": 0.5954879328436516,
+ "grad_norm": 0.1631128968199669,
+ "learning_rate": 2.085113522396141e-05,
+ "loss": 0.4508278965950012,
+ "step": 1135
+ },
+ {
+ "epoch": 0.5960125918153201,
+ "grad_norm": 0.16307688195844533,
+ "learning_rate": 2.0815714530167557e-05,
+ "loss": 0.4467576742172241,
+ "step": 1136
+ },
+ {
+ "epoch": 0.5965372507869885,
+ "grad_norm": 0.16456832304532207,
+ "learning_rate": 2.0780311996080976e-05,
+ "loss": 0.4621703624725342,
+ "step": 1137
+ },
+ {
+ "epoch": 0.5970619097586569,
+ "grad_norm": 0.16346217748325367,
+ "learning_rate": 2.0744927717882435e-05,
+ "loss": 0.4589449465274811,
+ "step": 1138
+ },
+ {
+ "epoch": 0.5975865687303253,
+ "grad_norm": 0.16180691501076994,
+ "learning_rate": 2.0709561791703097e-05,
+ "loss": 0.43698400259017944,
+ "step": 1139
+ },
+ {
+ "epoch": 0.5981112277019937,
+ "grad_norm": 0.15695459976898835,
+ "learning_rate": 2.067421431362427e-05,
+ "loss": 0.452271044254303,
+ "step": 1140
+ },
+ {
+ "epoch": 0.5986358866736621,
+ "grad_norm": 0.15930893553721054,
+ "learning_rate": 2.063888537967713e-05,
+ "loss": 0.44868963956832886,
+ "step": 1141
+ },
+ {
+ "epoch": 0.5991605456453305,
+ "grad_norm": 0.15963802284017928,
+ "learning_rate": 2.0603575085842485e-05,
+ "loss": 0.4376864731311798,
+ "step": 1142
+ },
+ {
+ "epoch": 0.5996852046169989,
+ "grad_norm": 0.16610266900507178,
+ "learning_rate": 2.05682835280505e-05,
+ "loss": 0.44676727056503296,
+ "step": 1143
+ },
+ {
+ "epoch": 0.6002098635886673,
+ "grad_norm": 0.15978253236669612,
+ "learning_rate": 2.053301080218044e-05,
+ "loss": 0.434773713350296,
+ "step": 1144
+ },
+ {
+ "epoch": 0.6007345225603358,
+ "grad_norm": 0.16305141803526324,
+ "learning_rate": 2.0497757004060403e-05,
+ "loss": 0.4529780149459839,
+ "step": 1145
+ },
+ {
+ "epoch": 0.6012591815320042,
+ "grad_norm": 0.1649737864101215,
+ "learning_rate": 2.046252222946707e-05,
+ "loss": 0.4571434259414673,
+ "step": 1146
+ },
+ {
+ "epoch": 0.6017838405036726,
+ "grad_norm": 0.15738824968326598,
+ "learning_rate": 2.0427306574125427e-05,
+ "loss": 0.4446289539337158,
+ "step": 1147
+ },
+ {
+ "epoch": 0.602308499475341,
+ "grad_norm": 0.1731139535185187,
+ "learning_rate": 2.039211013370853e-05,
+ "loss": 0.4551090598106384,
+ "step": 1148
+ },
+ {
+ "epoch": 0.6028331584470095,
+ "grad_norm": 0.15857251900043037,
+ "learning_rate": 2.0356933003837237e-05,
+ "loss": 0.4644010663032532,
+ "step": 1149
+ },
+ {
+ "epoch": 0.6033578174186779,
+ "grad_norm": 0.1589582861134211,
+ "learning_rate": 2.032177528007992e-05,
+ "loss": 0.45741650462150574,
+ "step": 1150
+ },
+ {
+ "epoch": 0.6038824763903463,
+ "grad_norm": 0.15639092706575122,
+ "learning_rate": 2.0286637057952257e-05,
+ "loss": 0.4375547766685486,
+ "step": 1151
+ },
+ {
+ "epoch": 0.6044071353620147,
+ "grad_norm": 0.16335605581337187,
+ "learning_rate": 2.0251518432916933e-05,
+ "loss": 0.4636140465736389,
+ "step": 1152
+ },
+ {
+ "epoch": 0.6049317943336832,
+ "grad_norm": 0.15437437113359356,
+ "learning_rate": 2.0216419500383372e-05,
+ "loss": 0.44632142782211304,
+ "step": 1153
+ },
+ {
+ "epoch": 0.6054564533053515,
+ "grad_norm": 0.15522039198698664,
+ "learning_rate": 2.0181340355707545e-05,
+ "loss": 0.449049174785614,
+ "step": 1154
+ },
+ {
+ "epoch": 0.6059811122770199,
+ "grad_norm": 0.1599951917909323,
+ "learning_rate": 2.0146281094191604e-05,
+ "loss": 0.44395962357521057,
+ "step": 1155
+ },
+ {
+ "epoch": 0.6065057712486883,
+ "grad_norm": 0.16866621620527716,
+ "learning_rate": 2.011124181108373e-05,
+ "loss": 0.45693716406822205,
+ "step": 1156
+ },
+ {
+ "epoch": 0.6070304302203567,
+ "grad_norm": 0.15508935650135075,
+ "learning_rate": 2.0076222601577815e-05,
+ "loss": 0.4525151252746582,
+ "step": 1157
+ },
+ {
+ "epoch": 0.6075550891920252,
+ "grad_norm": 0.16508998118005971,
+ "learning_rate": 2.004122356081321e-05,
+ "loss": 0.466006875038147,
+ "step": 1158
+ },
+ {
+ "epoch": 0.6080797481636936,
+ "grad_norm": 0.15817418864022237,
+ "learning_rate": 2.0006244783874466e-05,
+ "loss": 0.4496110677719116,
+ "step": 1159
+ },
+ {
+ "epoch": 0.608604407135362,
+ "grad_norm": 0.16127270259749824,
+ "learning_rate": 1.997128636579109e-05,
+ "loss": 0.45918628573417664,
+ "step": 1160
+ },
+ {
+ "epoch": 0.6091290661070304,
+ "grad_norm": 0.1702599924159677,
+ "learning_rate": 1.9936348401537292e-05,
+ "loss": 0.47218555212020874,
+ "step": 1161
+ },
+ {
+ "epoch": 0.6096537250786989,
+ "grad_norm": 0.15690176248846127,
+ "learning_rate": 1.9901430986031695e-05,
+ "loss": 0.4516097903251648,
+ "step": 1162
+ },
+ {
+ "epoch": 0.6101783840503673,
+ "grad_norm": 0.1713614405614144,
+ "learning_rate": 1.9866534214137085e-05,
+ "loss": 0.45030856132507324,
+ "step": 1163
+ },
+ {
+ "epoch": 0.6107030430220357,
+ "grad_norm": 0.1670865563026229,
+ "learning_rate": 1.9831658180660195e-05,
+ "loss": 0.45603489875793457,
+ "step": 1164
+ },
+ {
+ "epoch": 0.6112277019937041,
+ "grad_norm": 0.16897539113401952,
+ "learning_rate": 1.9796802980351392e-05,
+ "loss": 0.46314537525177,
+ "step": 1165
+ },
+ {
+ "epoch": 0.6117523609653726,
+ "grad_norm": 0.15832868870697644,
+ "learning_rate": 1.976196870790446e-05,
+ "loss": 0.44784611463546753,
+ "step": 1166
+ },
+ {
+ "epoch": 0.612277019937041,
+ "grad_norm": 0.16595365551891114,
+ "learning_rate": 1.9727155457956315e-05,
+ "loss": 0.4409624934196472,
+ "step": 1167
+ },
+ {
+ "epoch": 0.6128016789087093,
+ "grad_norm": 0.1642630411212017,
+ "learning_rate": 1.969236332508677e-05,
+ "loss": 0.46120351552963257,
+ "step": 1168
+ },
+ {
+ "epoch": 0.6133263378803777,
+ "grad_norm": 0.16481796135868812,
+ "learning_rate": 1.9657592403818253e-05,
+ "loss": 0.45544904470443726,
+ "step": 1169
+ },
+ {
+ "epoch": 0.6138509968520461,
+ "grad_norm": 0.1629832779852364,
+ "learning_rate": 1.962284278861557e-05,
+ "loss": 0.4578462839126587,
+ "step": 1170
+ },
+ {
+ "epoch": 0.6143756558237146,
+ "grad_norm": 0.15916219613603688,
+ "learning_rate": 1.9588114573885664e-05,
+ "loss": 0.4546983540058136,
+ "step": 1171
+ },
+ {
+ "epoch": 0.614900314795383,
+ "grad_norm": 0.1660520165166075,
+ "learning_rate": 1.95534078539773e-05,
+ "loss": 0.4601942300796509,
+ "step": 1172
+ },
+ {
+ "epoch": 0.6154249737670514,
+ "grad_norm": 0.16924794578577768,
+ "learning_rate": 1.9518722723180877e-05,
+ "loss": 0.45997846126556396,
+ "step": 1173
+ },
+ {
+ "epoch": 0.6159496327387198,
+ "grad_norm": 0.16597646133045382,
+ "learning_rate": 1.9484059275728127e-05,
+ "loss": 0.4594605565071106,
+ "step": 1174
+ },
+ {
+ "epoch": 0.6164742917103883,
+ "grad_norm": 0.16964243792519748,
+ "learning_rate": 1.9449417605791876e-05,
+ "loss": 0.4638690650463104,
+ "step": 1175
+ },
+ {
+ "epoch": 0.6169989506820567,
+ "grad_norm": 0.1758591426791696,
+ "learning_rate": 1.9414797807485796e-05,
+ "loss": 0.4666447639465332,
+ "step": 1176
+ },
+ {
+ "epoch": 0.6175236096537251,
+ "grad_norm": 0.16217661227891753,
+ "learning_rate": 1.9380199974864126e-05,
+ "loss": 0.44252535700798035,
+ "step": 1177
+ },
+ {
+ "epoch": 0.6180482686253935,
+ "grad_norm": 0.1576446184499019,
+ "learning_rate": 1.934562420192142e-05,
+ "loss": 0.4499641954898834,
+ "step": 1178
+ },
+ {
+ "epoch": 0.618572927597062,
+ "grad_norm": 0.16231008092636912,
+ "learning_rate": 1.9311070582592327e-05,
+ "loss": 0.4561542570590973,
+ "step": 1179
+ },
+ {
+ "epoch": 0.6190975865687304,
+ "grad_norm": 0.16790382249577043,
+ "learning_rate": 1.9276539210751292e-05,
+ "loss": 0.4484325647354126,
+ "step": 1180
+ },
+ {
+ "epoch": 0.6196222455403987,
+ "grad_norm": 0.18794729932605855,
+ "learning_rate": 1.9242030180212323e-05,
+ "loss": 0.44169026613235474,
+ "step": 1181
+ },
+ {
+ "epoch": 0.6201469045120671,
+ "grad_norm": 0.16948051243889814,
+ "learning_rate": 1.920754358472872e-05,
+ "loss": 0.4592876434326172,
+ "step": 1182
+ },
+ {
+ "epoch": 0.6206715634837355,
+ "grad_norm": 0.16762558710029374,
+ "learning_rate": 1.917307951799286e-05,
+ "loss": 0.4402095079421997,
+ "step": 1183
+ },
+ {
+ "epoch": 0.621196222455404,
+ "grad_norm": 0.18170280116387003,
+ "learning_rate": 1.9138638073635885e-05,
+ "loss": 0.4633026123046875,
+ "step": 1184
+ },
+ {
+ "epoch": 0.6217208814270724,
+ "grad_norm": 0.1718051709136108,
+ "learning_rate": 1.9104219345227486e-05,
+ "loss": 0.4510066509246826,
+ "step": 1185
+ },
+ {
+ "epoch": 0.6222455403987408,
+ "grad_norm": 0.16829066326345576,
+ "learning_rate": 1.906982342627565e-05,
+ "loss": 0.4632919430732727,
+ "step": 1186
+ },
+ {
+ "epoch": 0.6227701993704092,
+ "grad_norm": 0.16669141808548274,
+ "learning_rate": 1.9035450410226395e-05,
+ "loss": 0.455544650554657,
+ "step": 1187
+ },
+ {
+ "epoch": 0.6232948583420777,
+ "grad_norm": 0.16679996559722232,
+ "learning_rate": 1.90011003904635e-05,
+ "loss": 0.459994375705719,
+ "step": 1188
+ },
+ {
+ "epoch": 0.6238195173137461,
+ "grad_norm": 0.17379754713278164,
+ "learning_rate": 1.8966773460308284e-05,
+ "loss": 0.4576749801635742,
+ "step": 1189
+ },
+ {
+ "epoch": 0.6243441762854145,
+ "grad_norm": 0.17088539733389246,
+ "learning_rate": 1.8932469713019335e-05,
+ "loss": 0.44853052496910095,
+ "step": 1190
+ },
+ {
+ "epoch": 0.6248688352570829,
+ "grad_norm": 0.15556883889201015,
+ "learning_rate": 1.8898189241792248e-05,
+ "loss": 0.44060447812080383,
+ "step": 1191
+ },
+ {
+ "epoch": 0.6253934942287513,
+ "grad_norm": 0.1640093917439215,
+ "learning_rate": 1.8863932139759388e-05,
+ "loss": 0.4531911611557007,
+ "step": 1192
+ },
+ {
+ "epoch": 0.6259181532004198,
+ "grad_norm": 0.16444284826591576,
+ "learning_rate": 1.882969849998964e-05,
+ "loss": 0.44176363945007324,
+ "step": 1193
+ },
+ {
+ "epoch": 0.6264428121720882,
+ "grad_norm": 0.1696462520995642,
+ "learning_rate": 1.8795488415488143e-05,
+ "loss": 0.44516652822494507,
+ "step": 1194
+ },
+ {
+ "epoch": 0.6269674711437565,
+ "grad_norm": 0.16810724705191796,
+ "learning_rate": 1.8761301979196032e-05,
+ "loss": 0.4542563855648041,
+ "step": 1195
+ },
+ {
+ "epoch": 0.6274921301154249,
+ "grad_norm": 0.17334813313855824,
+ "learning_rate": 1.8727139283990204e-05,
+ "loss": 0.47769248485565186,
+ "step": 1196
+ },
+ {
+ "epoch": 0.6280167890870934,
+ "grad_norm": 0.16308212035303235,
+ "learning_rate": 1.8693000422683043e-05,
+ "loss": 0.4407799243927002,
+ "step": 1197
+ },
+ {
+ "epoch": 0.6285414480587618,
+ "grad_norm": 0.16853319139928977,
+ "learning_rate": 1.8658885488022204e-05,
+ "loss": 0.4565497040748596,
+ "step": 1198
+ },
+ {
+ "epoch": 0.6290661070304302,
+ "grad_norm": 0.16694968781284703,
+ "learning_rate": 1.8624794572690328e-05,
+ "loss": 0.44522374868392944,
+ "step": 1199
+ },
+ {
+ "epoch": 0.6295907660020986,
+ "grad_norm": 0.16418579468339103,
+ "learning_rate": 1.8590727769304788e-05,
+ "loss": 0.4441048502922058,
+ "step": 1200
+ },
+ {
+ "epoch": 0.630115424973767,
+ "grad_norm": 0.1661930777124156,
+ "learning_rate": 1.8556685170417486e-05,
+ "loss": 0.457327663898468,
+ "step": 1201
+ },
+ {
+ "epoch": 0.6306400839454355,
+ "grad_norm": 0.16536630027039814,
+ "learning_rate": 1.852266686851452e-05,
+ "loss": 0.462494432926178,
+ "step": 1202
+ },
+ {
+ "epoch": 0.6311647429171039,
+ "grad_norm": 0.16611474088307454,
+ "learning_rate": 1.8488672956016013e-05,
+ "loss": 0.4557701647281647,
+ "step": 1203
+ },
+ {
+ "epoch": 0.6316894018887723,
+ "grad_norm": 0.16759343079749697,
+ "learning_rate": 1.8454703525275813e-05,
+ "loss": 0.45069995522499084,
+ "step": 1204
+ },
+ {
+ "epoch": 0.6322140608604407,
+ "grad_norm": 0.16186676294485336,
+ "learning_rate": 1.8420758668581255e-05,
+ "loss": 0.44954800605773926,
+ "step": 1205
+ },
+ {
+ "epoch": 0.6327387198321092,
+ "grad_norm": 0.15687159835368553,
+ "learning_rate": 1.8386838478152916e-05,
+ "loss": 0.4490429759025574,
+ "step": 1206
+ },
+ {
+ "epoch": 0.6332633788037776,
+ "grad_norm": 0.16242951976115577,
+ "learning_rate": 1.8352943046144372e-05,
+ "loss": 0.45819586515426636,
+ "step": 1207
+ },
+ {
+ "epoch": 0.633788037775446,
+ "grad_norm": 0.16516087266517232,
+ "learning_rate": 1.831907246464191e-05,
+ "loss": 0.44769883155822754,
+ "step": 1208
+ },
+ {
+ "epoch": 0.6343126967471143,
+ "grad_norm": 0.16709475871512644,
+ "learning_rate": 1.8285226825664332e-05,
+ "loss": 0.4699128568172455,
+ "step": 1209
+ },
+ {
+ "epoch": 0.6348373557187827,
+ "grad_norm": 0.16288623150595774,
+ "learning_rate": 1.8251406221162636e-05,
+ "loss": 0.45322558283805847,
+ "step": 1210
+ },
+ {
+ "epoch": 0.6353620146904512,
+ "grad_norm": 0.1675722848054237,
+ "learning_rate": 1.821761074301986e-05,
+ "loss": 0.4752347469329834,
+ "step": 1211
+ },
+ {
+ "epoch": 0.6358866736621196,
+ "grad_norm": 0.163285963514504,
+ "learning_rate": 1.818384048305074e-05,
+ "loss": 0.441958487033844,
+ "step": 1212
+ },
+ {
+ "epoch": 0.636411332633788,
+ "grad_norm": 0.17078102287557764,
+ "learning_rate": 1.8150095533001517e-05,
+ "loss": 0.4551333487033844,
+ "step": 1213
+ },
+ {
+ "epoch": 0.6369359916054564,
+ "grad_norm": 0.16188540746097138,
+ "learning_rate": 1.811637598454966e-05,
+ "loss": 0.4470779597759247,
+ "step": 1214
+ },
+ {
+ "epoch": 0.6374606505771249,
+ "grad_norm": 0.15235628218958286,
+ "learning_rate": 1.808268192930364e-05,
+ "loss": 0.43604153394699097,
+ "step": 1215
+ },
+ {
+ "epoch": 0.6379853095487933,
+ "grad_norm": 0.1604601949959104,
+ "learning_rate": 1.8049013458802665e-05,
+ "loss": 0.4445713758468628,
+ "step": 1216
+ },
+ {
+ "epoch": 0.6385099685204617,
+ "grad_norm": 0.15511797131215901,
+ "learning_rate": 1.8015370664516427e-05,
+ "loss": 0.4455301761627197,
+ "step": 1217
+ },
+ {
+ "epoch": 0.6390346274921301,
+ "grad_norm": 0.16815669546944095,
+ "learning_rate": 1.798175363784487e-05,
+ "loss": 0.46135014295578003,
+ "step": 1218
+ },
+ {
+ "epoch": 0.6395592864637986,
+ "grad_norm": 0.161558177688922,
+ "learning_rate": 1.7948162470117917e-05,
+ "loss": 0.4519304037094116,
+ "step": 1219
+ },
+ {
+ "epoch": 0.640083945435467,
+ "grad_norm": 0.16677842130048232,
+ "learning_rate": 1.7914597252595265e-05,
+ "loss": 0.4550550878047943,
+ "step": 1220
+ },
+ {
+ "epoch": 0.6406086044071354,
+ "grad_norm": 0.1597331480423138,
+ "learning_rate": 1.7881058076466095e-05,
+ "loss": 0.441120445728302,
+ "step": 1221
+ },
+ {
+ "epoch": 0.6411332633788038,
+ "grad_norm": 0.15951106545246824,
+ "learning_rate": 1.7847545032848844e-05,
+ "loss": 0.4471382200717926,
+ "step": 1222
+ },
+ {
+ "epoch": 0.6416579223504721,
+ "grad_norm": 0.16157787674667565,
+ "learning_rate": 1.7814058212790947e-05,
+ "loss": 0.4550648331642151,
+ "step": 1223
+ },
+ {
+ "epoch": 0.6421825813221406,
+ "grad_norm": 0.16989186763352415,
+ "learning_rate": 1.7780597707268583e-05,
+ "loss": 0.4515089988708496,
+ "step": 1224
+ },
+ {
+ "epoch": 0.642707240293809,
+ "grad_norm": 0.15952153223099508,
+ "learning_rate": 1.7747163607186468e-05,
+ "loss": 0.4658638834953308,
+ "step": 1225
+ },
+ {
+ "epoch": 0.6432318992654774,
+ "grad_norm": 0.15205926699647218,
+ "learning_rate": 1.7713756003377567e-05,
+ "loss": 0.4330868422985077,
+ "step": 1226
+ },
+ {
+ "epoch": 0.6437565582371458,
+ "grad_norm": 0.15784806480936794,
+ "learning_rate": 1.7680374986602865e-05,
+ "loss": 0.4506068527698517,
+ "step": 1227
+ },
+ {
+ "epoch": 0.6442812172088143,
+ "grad_norm": 0.16006709280686687,
+ "learning_rate": 1.7647020647551105e-05,
+ "loss": 0.446941077709198,
+ "step": 1228
+ },
+ {
+ "epoch": 0.6448058761804827,
+ "grad_norm": 0.16064631158222062,
+ "learning_rate": 1.761369307683856e-05,
+ "loss": 0.4418525695800781,
+ "step": 1229
+ },
+ {
+ "epoch": 0.6453305351521511,
+ "grad_norm": 0.16454136480444392,
+ "learning_rate": 1.7580392365008767e-05,
+ "loss": 0.46268391609191895,
+ "step": 1230
+ },
+ {
+ "epoch": 0.6458551941238195,
+ "grad_norm": 0.15401246535771462,
+ "learning_rate": 1.754711860253233e-05,
+ "loss": 0.4372989535331726,
+ "step": 1231
+ },
+ {
+ "epoch": 0.646379853095488,
+ "grad_norm": 0.1544567975028383,
+ "learning_rate": 1.7513871879806603e-05,
+ "loss": 0.44806575775146484,
+ "step": 1232
+ },
+ {
+ "epoch": 0.6469045120671564,
+ "grad_norm": 0.16411225112730965,
+ "learning_rate": 1.748065228715549e-05,
+ "loss": 0.43782058358192444,
+ "step": 1233
+ },
+ {
+ "epoch": 0.6474291710388248,
+ "grad_norm": 0.1552767832972093,
+ "learning_rate": 1.7447459914829183e-05,
+ "loss": 0.44531378149986267,
+ "step": 1234
+ },
+ {
+ "epoch": 0.6479538300104932,
+ "grad_norm": 0.16214695257123513,
+ "learning_rate": 1.741429485300392e-05,
+ "loss": 0.47399842739105225,
+ "step": 1235
+ },
+ {
+ "epoch": 0.6484784889821616,
+ "grad_norm": 0.16684861111484112,
+ "learning_rate": 1.7381157191781766e-05,
+ "loss": 0.4654282331466675,
+ "step": 1236
+ },
+ {
+ "epoch": 0.64900314795383,
+ "grad_norm": 0.15901864178630273,
+ "learning_rate": 1.7348047021190334e-05,
+ "loss": 0.45128414034843445,
+ "step": 1237
+ },
+ {
+ "epoch": 0.6495278069254984,
+ "grad_norm": 0.15360991961742707,
+ "learning_rate": 1.7314964431182523e-05,
+ "loss": 0.4303375482559204,
+ "step": 1238
+ },
+ {
+ "epoch": 0.6500524658971668,
+ "grad_norm": 0.1633151574385741,
+ "learning_rate": 1.728190951163633e-05,
+ "loss": 0.45568403601646423,
+ "step": 1239
+ },
+ {
+ "epoch": 0.6505771248688352,
+ "grad_norm": 0.16377853571058645,
+ "learning_rate": 1.724888235235457e-05,
+ "loss": 0.46503543853759766,
+ "step": 1240
+ },
+ {
+ "epoch": 0.6511017838405037,
+ "grad_norm": 0.1595023801275368,
+ "learning_rate": 1.7215883043064655e-05,
+ "loss": 0.46337294578552246,
+ "step": 1241
+ },
+ {
+ "epoch": 0.6516264428121721,
+ "grad_norm": 0.16324796182212292,
+ "learning_rate": 1.7182911673418313e-05,
+ "loss": 0.46197354793548584,
+ "step": 1242
+ },
+ {
+ "epoch": 0.6521511017838405,
+ "grad_norm": 0.16353734830879765,
+ "learning_rate": 1.714996833299138e-05,
+ "loss": 0.45741671323776245,
+ "step": 1243
+ },
+ {
+ "epoch": 0.6526757607555089,
+ "grad_norm": 0.15941587678263386,
+ "learning_rate": 1.7117053111283528e-05,
+ "loss": 0.4514503479003906,
+ "step": 1244
+ },
+ {
+ "epoch": 0.6532004197271774,
+ "grad_norm": 0.1642653230145008,
+ "learning_rate": 1.708416609771804e-05,
+ "loss": 0.4437655210494995,
+ "step": 1245
+ },
+ {
+ "epoch": 0.6537250786988458,
+ "grad_norm": 0.15814155519130152,
+ "learning_rate": 1.7051307381641587e-05,
+ "loss": 0.4314163327217102,
+ "step": 1246
+ },
+ {
+ "epoch": 0.6542497376705142,
+ "grad_norm": 0.16086078370768056,
+ "learning_rate": 1.701847705232394e-05,
+ "loss": 0.4526430666446686,
+ "step": 1247
+ },
+ {
+ "epoch": 0.6547743966421826,
+ "grad_norm": 0.16389360190715183,
+ "learning_rate": 1.698567519895775e-05,
+ "loss": 0.4278966188430786,
+ "step": 1248
+ },
+ {
+ "epoch": 0.655299055613851,
+ "grad_norm": 0.16303142287804262,
+ "learning_rate": 1.6952901910658316e-05,
+ "loss": 0.45778220891952515,
+ "step": 1249
+ },
+ {
+ "epoch": 0.6558237145855194,
+ "grad_norm": 0.1675209956904048,
+ "learning_rate": 1.6920157276463318e-05,
+ "loss": 0.4489403963088989,
+ "step": 1250
+ },
+ {
+ "epoch": 0.6563483735571878,
+ "grad_norm": 0.16228311551074323,
+ "learning_rate": 1.6887441385332604e-05,
+ "loss": 0.4534342885017395,
+ "step": 1251
+ },
+ {
+ "epoch": 0.6568730325288562,
+ "grad_norm": 0.16408957851046913,
+ "learning_rate": 1.6854754326147916e-05,
+ "loss": 0.4549208879470825,
+ "step": 1252
+ },
+ {
+ "epoch": 0.6573976915005246,
+ "grad_norm": 0.15967922111432845,
+ "learning_rate": 1.682209618771268e-05,
+ "loss": 0.4451141357421875,
+ "step": 1253
+ },
+ {
+ "epoch": 0.657922350472193,
+ "grad_norm": 0.16440698722204208,
+ "learning_rate": 1.678946705875176e-05,
+ "loss": 0.454200804233551,
+ "step": 1254
+ },
+ {
+ "epoch": 0.6584470094438615,
+ "grad_norm": 0.16251120184938497,
+ "learning_rate": 1.675686702791117e-05,
+ "loss": 0.4580027461051941,
+ "step": 1255
+ },
+ {
+ "epoch": 0.6589716684155299,
+ "grad_norm": 0.1643717386150457,
+ "learning_rate": 1.6724296183757924e-05,
+ "loss": 0.4422311782836914,
+ "step": 1256
+ },
+ {
+ "epoch": 0.6594963273871983,
+ "grad_norm": 0.1644586414378722,
+ "learning_rate": 1.66917546147797e-05,
+ "loss": 0.4593179225921631,
+ "step": 1257
+ },
+ {
+ "epoch": 0.6600209863588667,
+ "grad_norm": 0.16751074186967685,
+ "learning_rate": 1.6659242409384657e-05,
+ "loss": 0.457930326461792,
+ "step": 1258
+ },
+ {
+ "epoch": 0.6605456453305352,
+ "grad_norm": 0.15398418609155334,
+ "learning_rate": 1.6626759655901182e-05,
+ "loss": 0.44002899527549744,
+ "step": 1259
+ },
+ {
+ "epoch": 0.6610703043022036,
+ "grad_norm": 0.16328113175790226,
+ "learning_rate": 1.659430644257763e-05,
+ "loss": 0.4500438868999481,
+ "step": 1260
+ },
+ {
+ "epoch": 0.661594963273872,
+ "grad_norm": 0.1694244340615106,
+ "learning_rate": 1.656188285758214e-05,
+ "loss": 0.4818427264690399,
+ "step": 1261
+ },
+ {
+ "epoch": 0.6621196222455404,
+ "grad_norm": 0.1578320701754368,
+ "learning_rate": 1.652948898900232e-05,
+ "loss": 0.4492436647415161,
+ "step": 1262
+ },
+ {
+ "epoch": 0.6626442812172089,
+ "grad_norm": 0.16538865917828469,
+ "learning_rate": 1.6497124924845064e-05,
+ "loss": 0.45070338249206543,
+ "step": 1263
+ },
+ {
+ "epoch": 0.6631689401888772,
+ "grad_norm": 0.15670817848256852,
+ "learning_rate": 1.6464790753036296e-05,
+ "loss": 0.45439285039901733,
+ "step": 1264
+ },
+ {
+ "epoch": 0.6636935991605456,
+ "grad_norm": 0.16184786585965885,
+ "learning_rate": 1.64324865614207e-05,
+ "loss": 0.446522980928421,
+ "step": 1265
+ },
+ {
+ "epoch": 0.664218258132214,
+ "grad_norm": 0.15719726462251984,
+ "learning_rate": 1.6400212437761554e-05,
+ "loss": 0.4495280981063843,
+ "step": 1266
+ },
+ {
+ "epoch": 0.6647429171038824,
+ "grad_norm": 0.1680804414134405,
+ "learning_rate": 1.6367968469740416e-05,
+ "loss": 0.4652596712112427,
+ "step": 1267
+ },
+ {
+ "epoch": 0.6652675760755509,
+ "grad_norm": 0.15919541846683566,
+ "learning_rate": 1.6335754744956937e-05,
+ "loss": 0.43842798471450806,
+ "step": 1268
+ },
+ {
+ "epoch": 0.6657922350472193,
+ "grad_norm": 0.16322634187139418,
+ "learning_rate": 1.630357135092859e-05,
+ "loss": 0.45242664217948914,
+ "step": 1269
+ },
+ {
+ "epoch": 0.6663168940188877,
+ "grad_norm": 0.15644867734734783,
+ "learning_rate": 1.6271418375090445e-05,
+ "loss": 0.44801655411720276,
+ "step": 1270
+ },
+ {
+ "epoch": 0.6668415529905561,
+ "grad_norm": 0.1514717416124423,
+ "learning_rate": 1.623929590479496e-05,
+ "loss": 0.4506194293498993,
+ "step": 1271
+ },
+ {
+ "epoch": 0.6673662119622246,
+ "grad_norm": 0.1740345085495889,
+ "learning_rate": 1.6207204027311683e-05,
+ "loss": 0.4818177819252014,
+ "step": 1272
+ },
+ {
+ "epoch": 0.667890870933893,
+ "grad_norm": 0.15797701900765457,
+ "learning_rate": 1.6175142829827067e-05,
+ "loss": 0.4639382064342499,
+ "step": 1273
+ },
+ {
+ "epoch": 0.6684155299055614,
+ "grad_norm": 0.15663812316653875,
+ "learning_rate": 1.6143112399444198e-05,
+ "loss": 0.46217501163482666,
+ "step": 1274
+ },
+ {
+ "epoch": 0.6689401888772298,
+ "grad_norm": 0.15474018977043758,
+ "learning_rate": 1.6111112823182592e-05,
+ "loss": 0.43691688776016235,
+ "step": 1275
+ },
+ {
+ "epoch": 0.6694648478488983,
+ "grad_norm": 0.1615634942529431,
+ "learning_rate": 1.6079144187977923e-05,
+ "loss": 0.4412601590156555,
+ "step": 1276
+ },
+ {
+ "epoch": 0.6699895068205667,
+ "grad_norm": 0.15440022842754444,
+ "learning_rate": 1.604720658068184e-05,
+ "loss": 0.4523273706436157,
+ "step": 1277
+ },
+ {
+ "epoch": 0.670514165792235,
+ "grad_norm": 0.1634985665776242,
+ "learning_rate": 1.6015300088061664e-05,
+ "loss": 0.46264398097991943,
+ "step": 1278
+ },
+ {
+ "epoch": 0.6710388247639034,
+ "grad_norm": 0.15037886370689527,
+ "learning_rate": 1.598342479680018e-05,
+ "loss": 0.451465904712677,
+ "step": 1279
+ },
+ {
+ "epoch": 0.6715634837355718,
+ "grad_norm": 0.16895242437746613,
+ "learning_rate": 1.595158079349542e-05,
+ "loss": 0.4466145634651184,
+ "step": 1280
+ },
+ {
+ "epoch": 0.6720881427072403,
+ "grad_norm": 0.1629548326999642,
+ "learning_rate": 1.5919768164660425e-05,
+ "loss": 0.45415782928466797,
+ "step": 1281
+ },
+ {
+ "epoch": 0.6726128016789087,
+ "grad_norm": 0.1563872102268991,
+ "learning_rate": 1.588798699672298e-05,
+ "loss": 0.4386737048625946,
+ "step": 1282
+ },
+ {
+ "epoch": 0.6731374606505771,
+ "grad_norm": 0.16038172773344717,
+ "learning_rate": 1.5856237376025403e-05,
+ "loss": 0.446260929107666,
+ "step": 1283
+ },
+ {
+ "epoch": 0.6736621196222455,
+ "grad_norm": 0.16189780714407623,
+ "learning_rate": 1.582451938882431e-05,
+ "loss": 0.45458072423934937,
+ "step": 1284
+ },
+ {
+ "epoch": 0.674186778593914,
+ "grad_norm": 0.15204521785049285,
+ "learning_rate": 1.5792833121290367e-05,
+ "loss": 0.4425617754459381,
+ "step": 1285
+ },
+ {
+ "epoch": 0.6747114375655824,
+ "grad_norm": 0.16237376704792778,
+ "learning_rate": 1.5761178659508063e-05,
+ "loss": 0.44631800055503845,
+ "step": 1286
+ },
+ {
+ "epoch": 0.6752360965372508,
+ "grad_norm": 0.1582063618466069,
+ "learning_rate": 1.57295560894755e-05,
+ "loss": 0.44280534982681274,
+ "step": 1287
+ },
+ {
+ "epoch": 0.6757607555089192,
+ "grad_norm": 0.16587032349847394,
+ "learning_rate": 1.5697965497104116e-05,
+ "loss": 0.4495997130870819,
+ "step": 1288
+ },
+ {
+ "epoch": 0.6762854144805877,
+ "grad_norm": 0.1557269262755563,
+ "learning_rate": 1.566640696821848e-05,
+ "loss": 0.4493359625339508,
+ "step": 1289
+ },
+ {
+ "epoch": 0.6768100734522561,
+ "grad_norm": 0.16038490357012256,
+ "learning_rate": 1.5634880588556043e-05,
+ "loss": 0.4447544813156128,
+ "step": 1290
+ },
+ {
+ "epoch": 0.6773347324239245,
+ "grad_norm": 0.15355201908546406,
+ "learning_rate": 1.560338644376692e-05,
+ "loss": 0.4468432664871216,
+ "step": 1291
+ },
+ {
+ "epoch": 0.6778593913955928,
+ "grad_norm": 0.15077009515589776,
+ "learning_rate": 1.5571924619413668e-05,
+ "loss": 0.44959044456481934,
+ "step": 1292
+ },
+ {
+ "epoch": 0.6783840503672612,
+ "grad_norm": 0.1583038146613965,
+ "learning_rate": 1.5540495200971012e-05,
+ "loss": 0.42450425028800964,
+ "step": 1293
+ },
+ {
+ "epoch": 0.6789087093389297,
+ "grad_norm": 0.15950687190598087,
+ "learning_rate": 1.5509098273825635e-05,
+ "loss": 0.4527238607406616,
+ "step": 1294
+ },
+ {
+ "epoch": 0.6794333683105981,
+ "grad_norm": 0.1526262208847808,
+ "learning_rate": 1.5477733923275968e-05,
+ "loss": 0.4379724860191345,
+ "step": 1295
+ },
+ {
+ "epoch": 0.6799580272822665,
+ "grad_norm": 0.15541579650021778,
+ "learning_rate": 1.5446402234531934e-05,
+ "loss": 0.4462166130542755,
+ "step": 1296
+ },
+ {
+ "epoch": 0.6804826862539349,
+ "grad_norm": 0.15483930734313786,
+ "learning_rate": 1.5415103292714718e-05,
+ "loss": 0.4390607178211212,
+ "step": 1297
+ },
+ {
+ "epoch": 0.6810073452256034,
+ "grad_norm": 0.1578005731149154,
+ "learning_rate": 1.5383837182856532e-05,
+ "loss": 0.44462254643440247,
+ "step": 1298
+ },
+ {
+ "epoch": 0.6815320041972718,
+ "grad_norm": 0.15830183884958823,
+ "learning_rate": 1.5352603989900406e-05,
+ "loss": 0.4563537538051605,
+ "step": 1299
+ },
+ {
+ "epoch": 0.6820566631689402,
+ "grad_norm": 0.15605472163445522,
+ "learning_rate": 1.5321403798699923e-05,
+ "loss": 0.43852412700653076,
+ "step": 1300
+ },
+ {
+ "epoch": 0.6825813221406086,
+ "grad_norm": 0.15770470227141278,
+ "learning_rate": 1.5290236694019017e-05,
+ "loss": 0.4438081979751587,
+ "step": 1301
+ },
+ {
+ "epoch": 0.683105981112277,
+ "grad_norm": 0.16175339845012077,
+ "learning_rate": 1.5259102760531747e-05,
+ "loss": 0.46670734882354736,
+ "step": 1302
+ },
+ {
+ "epoch": 0.6836306400839455,
+ "grad_norm": 0.16359504716154477,
+ "learning_rate": 1.5228002082822036e-05,
+ "loss": 0.4590187668800354,
+ "step": 1303
+ },
+ {
+ "epoch": 0.6841552990556139,
+ "grad_norm": 0.15422879500254713,
+ "learning_rate": 1.5196934745383457e-05,
+ "loss": 0.45287445187568665,
+ "step": 1304
+ },
+ {
+ "epoch": 0.6846799580272823,
+ "grad_norm": 0.15929902143574254,
+ "learning_rate": 1.5165900832619015e-05,
+ "loss": 0.44286859035491943,
+ "step": 1305
+ },
+ {
+ "epoch": 0.6852046169989506,
+ "grad_norm": 0.14838868714016124,
+ "learning_rate": 1.5134900428840904e-05,
+ "loss": 0.45155346393585205,
+ "step": 1306
+ },
+ {
+ "epoch": 0.6857292759706191,
+ "grad_norm": 0.15494288640397672,
+ "learning_rate": 1.5103933618270283e-05,
+ "loss": 0.44010141491889954,
+ "step": 1307
+ },
+ {
+ "epoch": 0.6862539349422875,
+ "grad_norm": 0.1686262360598338,
+ "learning_rate": 1.5073000485037042e-05,
+ "loss": 0.446555495262146,
+ "step": 1308
+ },
+ {
+ "epoch": 0.6867785939139559,
+ "grad_norm": 0.16122110823982602,
+ "learning_rate": 1.5042101113179579e-05,
+ "loss": 0.44640374183654785,
+ "step": 1309
+ },
+ {
+ "epoch": 0.6873032528856243,
+ "grad_norm": 0.16849937797919867,
+ "learning_rate": 1.5011235586644572e-05,
+ "loss": 0.4465785026550293,
+ "step": 1310
+ },
+ {
+ "epoch": 0.6878279118572928,
+ "grad_norm": 0.16062971262800754,
+ "learning_rate": 1.498040398928674e-05,
+ "loss": 0.445321649312973,
+ "step": 1311
+ },
+ {
+ "epoch": 0.6883525708289612,
+ "grad_norm": 0.1608953185993034,
+ "learning_rate": 1.4949606404868638e-05,
+ "loss": 0.46518227458000183,
+ "step": 1312
+ },
+ {
+ "epoch": 0.6888772298006296,
+ "grad_norm": 0.15880255672830915,
+ "learning_rate": 1.491884291706041e-05,
+ "loss": 0.44582217931747437,
+ "step": 1313
+ },
+ {
+ "epoch": 0.689401888772298,
+ "grad_norm": 0.16715145200789822,
+ "learning_rate": 1.4888113609439563e-05,
+ "loss": 0.4435983896255493,
+ "step": 1314
+ },
+ {
+ "epoch": 0.6899265477439664,
+ "grad_norm": 0.15263740340694454,
+ "learning_rate": 1.485741856549075e-05,
+ "loss": 0.4405066967010498,
+ "step": 1315
+ },
+ {
+ "epoch": 0.6904512067156349,
+ "grad_norm": 0.16247034518813966,
+ "learning_rate": 1.4826757868605517e-05,
+ "loss": 0.4544215500354767,
+ "step": 1316
+ },
+ {
+ "epoch": 0.6909758656873033,
+ "grad_norm": 0.15649146705978162,
+ "learning_rate": 1.4796131602082142e-05,
+ "loss": 0.4454413056373596,
+ "step": 1317
+ },
+ {
+ "epoch": 0.6915005246589717,
+ "grad_norm": 0.16286718636058178,
+ "learning_rate": 1.4765539849125318e-05,
+ "loss": 0.4497183561325073,
+ "step": 1318
+ },
+ {
+ "epoch": 0.6920251836306401,
+ "grad_norm": 0.15812342085005582,
+ "learning_rate": 1.473498269284599e-05,
+ "loss": 0.445032000541687,
+ "step": 1319
+ },
+ {
+ "epoch": 0.6925498426023085,
+ "grad_norm": 0.1558101290511678,
+ "learning_rate": 1.4704460216261123e-05,
+ "loss": 0.45769619941711426,
+ "step": 1320
+ },
+ {
+ "epoch": 0.6930745015739769,
+ "grad_norm": 0.15642889059495207,
+ "learning_rate": 1.4673972502293428e-05,
+ "loss": 0.44788360595703125,
+ "step": 1321
+ },
+ {
+ "epoch": 0.6935991605456453,
+ "grad_norm": 0.15722513518113074,
+ "learning_rate": 1.4643519633771222e-05,
+ "loss": 0.449631929397583,
+ "step": 1322
+ },
+ {
+ "epoch": 0.6941238195173137,
+ "grad_norm": 0.15587633748081542,
+ "learning_rate": 1.4613101693428125e-05,
+ "loss": 0.4450187087059021,
+ "step": 1323
+ },
+ {
+ "epoch": 0.6946484784889821,
+ "grad_norm": 0.1577107074882479,
+ "learning_rate": 1.458271876390287e-05,
+ "loss": 0.45108890533447266,
+ "step": 1324
+ },
+ {
+ "epoch": 0.6951731374606506,
+ "grad_norm": 0.15649839486990696,
+ "learning_rate": 1.4552370927739073e-05,
+ "loss": 0.46430301666259766,
+ "step": 1325
+ },
+ {
+ "epoch": 0.695697796432319,
+ "grad_norm": 0.18573600420690636,
+ "learning_rate": 1.4522058267385008e-05,
+ "loss": 0.4701972007751465,
+ "step": 1326
+ },
+ {
+ "epoch": 0.6962224554039874,
+ "grad_norm": 0.16296497084037542,
+ "learning_rate": 1.4491780865193402e-05,
+ "loss": 0.4607710838317871,
+ "step": 1327
+ },
+ {
+ "epoch": 0.6967471143756558,
+ "grad_norm": 0.161124007217343,
+ "learning_rate": 1.446153880342117e-05,
+ "loss": 0.45095115900039673,
+ "step": 1328
+ },
+ {
+ "epoch": 0.6972717733473243,
+ "grad_norm": 0.1537884039166766,
+ "learning_rate": 1.4431332164229222e-05,
+ "loss": 0.424068808555603,
+ "step": 1329
+ },
+ {
+ "epoch": 0.6977964323189927,
+ "grad_norm": 0.16588834503456087,
+ "learning_rate": 1.4401161029682237e-05,
+ "loss": 0.4622839093208313,
+ "step": 1330
+ },
+ {
+ "epoch": 0.6983210912906611,
+ "grad_norm": 0.16100510604015544,
+ "learning_rate": 1.4371025481748435e-05,
+ "loss": 0.4660705626010895,
+ "step": 1331
+ },
+ {
+ "epoch": 0.6988457502623295,
+ "grad_norm": 0.15559888940590075,
+ "learning_rate": 1.4340925602299337e-05,
+ "loss": 0.4484976530075073,
+ "step": 1332
+ },
+ {
+ "epoch": 0.6993704092339978,
+ "grad_norm": 0.15503821966213488,
+ "learning_rate": 1.4310861473109616e-05,
+ "loss": 0.45636600255966187,
+ "step": 1333
+ },
+ {
+ "epoch": 0.6998950682056663,
+ "grad_norm": 0.15539074896109334,
+ "learning_rate": 1.4280833175856743e-05,
+ "loss": 0.45863068103790283,
+ "step": 1334
+ },
+ {
+ "epoch": 0.7004197271773347,
+ "grad_norm": 0.15664821009154872,
+ "learning_rate": 1.4250840792120896e-05,
+ "loss": 0.44043463468551636,
+ "step": 1335
+ },
+ {
+ "epoch": 0.7009443861490031,
+ "grad_norm": 0.15825478808520285,
+ "learning_rate": 1.4220884403384655e-05,
+ "loss": 0.46522051095962524,
+ "step": 1336
+ },
+ {
+ "epoch": 0.7014690451206715,
+ "grad_norm": 0.16004268568141544,
+ "learning_rate": 1.4190964091032843e-05,
+ "loss": 0.46114686131477356,
+ "step": 1337
+ },
+ {
+ "epoch": 0.70199370409234,
+ "grad_norm": 0.16413139949945635,
+ "learning_rate": 1.4161079936352234e-05,
+ "loss": 0.4457515478134155,
+ "step": 1338
+ },
+ {
+ "epoch": 0.7025183630640084,
+ "grad_norm": 0.1538890577972702,
+ "learning_rate": 1.4131232020531398e-05,
+ "loss": 0.4616093337535858,
+ "step": 1339
+ },
+ {
+ "epoch": 0.7030430220356768,
+ "grad_norm": 0.16240737854323284,
+ "learning_rate": 1.4101420424660434e-05,
+ "loss": 0.44631537795066833,
+ "step": 1340
+ },
+ {
+ "epoch": 0.7035676810073452,
+ "grad_norm": 0.15984708603406073,
+ "learning_rate": 1.4071645229730762e-05,
+ "loss": 0.4366717040538788,
+ "step": 1341
+ },
+ {
+ "epoch": 0.7040923399790137,
+ "grad_norm": 0.15149695695635904,
+ "learning_rate": 1.4041906516634946e-05,
+ "loss": 0.44129428267478943,
+ "step": 1342
+ },
+ {
+ "epoch": 0.7046169989506821,
+ "grad_norm": 0.15041167501768415,
+ "learning_rate": 1.4012204366166399e-05,
+ "loss": 0.43654775619506836,
+ "step": 1343
+ },
+ {
+ "epoch": 0.7051416579223505,
+ "grad_norm": 0.1596097627963756,
+ "learning_rate": 1.3982538859019217e-05,
+ "loss": 0.4652445614337921,
+ "step": 1344
+ },
+ {
+ "epoch": 0.7056663168940189,
+ "grad_norm": 0.15621704205027884,
+ "learning_rate": 1.3952910075787941e-05,
+ "loss": 0.44577041268348694,
+ "step": 1345
+ },
+ {
+ "epoch": 0.7061909758656874,
+ "grad_norm": 0.16073041098008783,
+ "learning_rate": 1.3923318096967341e-05,
+ "loss": 0.4391976594924927,
+ "step": 1346
+ },
+ {
+ "epoch": 0.7067156348373557,
+ "grad_norm": 0.14793853946372448,
+ "learning_rate": 1.3893763002952204e-05,
+ "loss": 0.4463799297809601,
+ "step": 1347
+ },
+ {
+ "epoch": 0.7072402938090241,
+ "grad_norm": 0.16176080358867354,
+ "learning_rate": 1.3864244874037097e-05,
+ "loss": 0.45183271169662476,
+ "step": 1348
+ },
+ {
+ "epoch": 0.7077649527806925,
+ "grad_norm": 0.16372771179026924,
+ "learning_rate": 1.3834763790416171e-05,
+ "loss": 0.44205576181411743,
+ "step": 1349
+ },
+ {
+ "epoch": 0.7082896117523609,
+ "grad_norm": 0.15135222516127125,
+ "learning_rate": 1.3805319832182937e-05,
+ "loss": 0.4390261173248291,
+ "step": 1350
+ },
+ {
+ "epoch": 0.7088142707240294,
+ "grad_norm": 0.15493183225792365,
+ "learning_rate": 1.377591307933002e-05,
+ "loss": 0.45837974548339844,
+ "step": 1351
+ },
+ {
+ "epoch": 0.7093389296956978,
+ "grad_norm": 0.16405906906516504,
+ "learning_rate": 1.3746543611749008e-05,
+ "loss": 0.45032668113708496,
+ "step": 1352
+ },
+ {
+ "epoch": 0.7098635886673662,
+ "grad_norm": 0.15725677976524505,
+ "learning_rate": 1.3717211509230155e-05,
+ "loss": 0.4486224949359894,
+ "step": 1353
+ },
+ {
+ "epoch": 0.7103882476390346,
+ "grad_norm": 0.15781469066983164,
+ "learning_rate": 1.3687916851462227e-05,
+ "loss": 0.4502638578414917,
+ "step": 1354
+ },
+ {
+ "epoch": 0.710912906610703,
+ "grad_norm": 0.16115007729054298,
+ "learning_rate": 1.3658659718032241e-05,
+ "loss": 0.4508914351463318,
+ "step": 1355
+ },
+ {
+ "epoch": 0.7114375655823715,
+ "grad_norm": 0.16755060475968744,
+ "learning_rate": 1.3629440188425286e-05,
+ "loss": 0.45653945207595825,
+ "step": 1356
+ },
+ {
+ "epoch": 0.7119622245540399,
+ "grad_norm": 0.1574994294756123,
+ "learning_rate": 1.3600258342024268e-05,
+ "loss": 0.4349849820137024,
+ "step": 1357
+ },
+ {
+ "epoch": 0.7124868835257083,
+ "grad_norm": 0.16104440922657715,
+ "learning_rate": 1.3571114258109758e-05,
+ "loss": 0.4501233696937561,
+ "step": 1358
+ },
+ {
+ "epoch": 0.7130115424973767,
+ "grad_norm": 0.15163184290939022,
+ "learning_rate": 1.3542008015859691e-05,
+ "loss": 0.44171035289764404,
+ "step": 1359
+ },
+ {
+ "epoch": 0.7135362014690452,
+ "grad_norm": 0.15955896674456582,
+ "learning_rate": 1.3512939694349211e-05,
+ "loss": 0.4558671712875366,
+ "step": 1360
+ },
+ {
+ "epoch": 0.7140608604407135,
+ "grad_norm": 0.1589376751536802,
+ "learning_rate": 1.3483909372550446e-05,
+ "loss": 0.4430835247039795,
+ "step": 1361
+ },
+ {
+ "epoch": 0.7145855194123819,
+ "grad_norm": 0.15979088397822155,
+ "learning_rate": 1.3454917129332279e-05,
+ "loss": 0.4426245391368866,
+ "step": 1362
+ },
+ {
+ "epoch": 0.7151101783840503,
+ "grad_norm": 0.15493179714072167,
+ "learning_rate": 1.3425963043460143e-05,
+ "loss": 0.4550474286079407,
+ "step": 1363
+ },
+ {
+ "epoch": 0.7156348373557188,
+ "grad_norm": 0.15396498968995032,
+ "learning_rate": 1.3397047193595808e-05,
+ "loss": 0.44220536947250366,
+ "step": 1364
+ },
+ {
+ "epoch": 0.7161594963273872,
+ "grad_norm": 0.1489346822667273,
+ "learning_rate": 1.336816965829717e-05,
+ "loss": 0.4437323212623596,
+ "step": 1365
+ },
+ {
+ "epoch": 0.7166841552990556,
+ "grad_norm": 0.16311964931734477,
+ "learning_rate": 1.333933051601802e-05,
+ "loss": 0.45308777689933777,
+ "step": 1366
+ },
+ {
+ "epoch": 0.717208814270724,
+ "grad_norm": 0.1534354619021179,
+ "learning_rate": 1.331052984510785e-05,
+ "loss": 0.4334886372089386,
+ "step": 1367
+ },
+ {
+ "epoch": 0.7177334732423925,
+ "grad_norm": 0.15541089758065404,
+ "learning_rate": 1.328176772381165e-05,
+ "loss": 0.45248323678970337,
+ "step": 1368
+ },
+ {
+ "epoch": 0.7182581322140609,
+ "grad_norm": 0.1617921305700596,
+ "learning_rate": 1.3253044230269651e-05,
+ "loss": 0.44092947244644165,
+ "step": 1369
+ },
+ {
+ "epoch": 0.7187827911857293,
+ "grad_norm": 0.15047353793760151,
+ "learning_rate": 1.3224359442517165e-05,
+ "loss": 0.4424414336681366,
+ "step": 1370
+ },
+ {
+ "epoch": 0.7193074501573977,
+ "grad_norm": 0.15958089102258743,
+ "learning_rate": 1.319571343848433e-05,
+ "loss": 0.4458580017089844,
+ "step": 1371
+ },
+ {
+ "epoch": 0.7198321091290661,
+ "grad_norm": 0.15349258306968733,
+ "learning_rate": 1.3167106295995924e-05,
+ "loss": 0.4285386800765991,
+ "step": 1372
+ },
+ {
+ "epoch": 0.7203567681007346,
+ "grad_norm": 0.16042529563759886,
+ "learning_rate": 1.3138538092771164e-05,
+ "loss": 0.4497690200805664,
+ "step": 1373
+ },
+ {
+ "epoch": 0.720881427072403,
+ "grad_norm": 0.1565113736588754,
+ "learning_rate": 1.3110008906423454e-05,
+ "loss": 0.43562617897987366,
+ "step": 1374
+ },
+ {
+ "epoch": 0.7214060860440713,
+ "grad_norm": 0.15177291086871506,
+ "learning_rate": 1.3081518814460215e-05,
+ "loss": 0.44456908106803894,
+ "step": 1375
+ },
+ {
+ "epoch": 0.7219307450157397,
+ "grad_norm": 0.15928490480439225,
+ "learning_rate": 1.3053067894282622e-05,
+ "loss": 0.46170780062675476,
+ "step": 1376
+ },
+ {
+ "epoch": 0.7224554039874082,
+ "grad_norm": 0.1554814469217892,
+ "learning_rate": 1.302465622318549e-05,
+ "loss": 0.4422169327735901,
+ "step": 1377
+ },
+ {
+ "epoch": 0.7229800629590766,
+ "grad_norm": 0.15375370214966386,
+ "learning_rate": 1.2996283878356947e-05,
+ "loss": 0.4564763903617859,
+ "step": 1378
+ },
+ {
+ "epoch": 0.723504721930745,
+ "grad_norm": 0.15942841969395802,
+ "learning_rate": 1.2967950936878311e-05,
+ "loss": 0.45121875405311584,
+ "step": 1379
+ },
+ {
+ "epoch": 0.7240293809024134,
+ "grad_norm": 0.15731308716774903,
+ "learning_rate": 1.2939657475723844e-05,
+ "loss": 0.4285027086734772,
+ "step": 1380
+ },
+ {
+ "epoch": 0.7245540398740818,
+ "grad_norm": 0.15997753613932147,
+ "learning_rate": 1.291140357176054e-05,
+ "loss": 0.4566386342048645,
+ "step": 1381
+ },
+ {
+ "epoch": 0.7250786988457503,
+ "grad_norm": 0.15333715267473727,
+ "learning_rate": 1.2883189301747927e-05,
+ "loss": 0.45093846321105957,
+ "step": 1382
+ },
+ {
+ "epoch": 0.7256033578174187,
+ "grad_norm": 0.1514510304709061,
+ "learning_rate": 1.2855014742337874e-05,
+ "loss": 0.44256967306137085,
+ "step": 1383
+ },
+ {
+ "epoch": 0.7261280167890871,
+ "grad_norm": 0.15624301298995746,
+ "learning_rate": 1.2826879970074346e-05,
+ "loss": 0.4361875653266907,
+ "step": 1384
+ },
+ {
+ "epoch": 0.7266526757607555,
+ "grad_norm": 0.15767850122006694,
+ "learning_rate": 1.279878506139322e-05,
+ "loss": 0.4477671980857849,
+ "step": 1385
+ },
+ {
+ "epoch": 0.727177334732424,
+ "grad_norm": 0.15537676969694064,
+ "learning_rate": 1.2770730092622072e-05,
+ "loss": 0.4339832663536072,
+ "step": 1386
+ },
+ {
+ "epoch": 0.7277019937040924,
+ "grad_norm": 0.15944113633281293,
+ "learning_rate": 1.2742715139979965e-05,
+ "loss": 0.45407935976982117,
+ "step": 1387
+ },
+ {
+ "epoch": 0.7282266526757608,
+ "grad_norm": 0.16763519518606218,
+ "learning_rate": 1.271474027957727e-05,
+ "loss": 0.46442392468452454,
+ "step": 1388
+ },
+ {
+ "epoch": 0.7287513116474291,
+ "grad_norm": 0.1564572774601902,
+ "learning_rate": 1.2686805587415423e-05,
+ "loss": 0.4463968276977539,
+ "step": 1389
+ },
+ {
+ "epoch": 0.7292759706190975,
+ "grad_norm": 0.16384626314986128,
+ "learning_rate": 1.2658911139386709e-05,
+ "loss": 0.45519885420799255,
+ "step": 1390
+ },
+ {
+ "epoch": 0.729800629590766,
+ "grad_norm": 0.1562023802820964,
+ "learning_rate": 1.2631057011274106e-05,
+ "loss": 0.4365222454071045,
+ "step": 1391
+ },
+ {
+ "epoch": 0.7303252885624344,
+ "grad_norm": 0.15597287232940926,
+ "learning_rate": 1.2603243278751042e-05,
+ "loss": 0.4328266680240631,
+ "step": 1392
+ },
+ {
+ "epoch": 0.7308499475341028,
+ "grad_norm": 0.14762833706132167,
+ "learning_rate": 1.2575470017381214e-05,
+ "loss": 0.43902426958084106,
+ "step": 1393
+ },
+ {
+ "epoch": 0.7313746065057712,
+ "grad_norm": 0.1688371515856315,
+ "learning_rate": 1.2547737302618351e-05,
+ "loss": 0.4649963676929474,
+ "step": 1394
+ },
+ {
+ "epoch": 0.7318992654774397,
+ "grad_norm": 0.15501514509586015,
+ "learning_rate": 1.2520045209806024e-05,
+ "loss": 0.4501558542251587,
+ "step": 1395
+ },
+ {
+ "epoch": 0.7324239244491081,
+ "grad_norm": 0.15516748917693537,
+ "learning_rate": 1.2492393814177457e-05,
+ "loss": 0.4405498206615448,
+ "step": 1396
+ },
+ {
+ "epoch": 0.7329485834207765,
+ "grad_norm": 0.17115400468199243,
+ "learning_rate": 1.2464783190855295e-05,
+ "loss": 0.4445379674434662,
+ "step": 1397
+ },
+ {
+ "epoch": 0.7334732423924449,
+ "grad_norm": 0.15966386341604186,
+ "learning_rate": 1.243721341485143e-05,
+ "loss": 0.4769889712333679,
+ "step": 1398
+ },
+ {
+ "epoch": 0.7339979013641134,
+ "grad_norm": 0.15614706359781472,
+ "learning_rate": 1.2409684561066766e-05,
+ "loss": 0.43643918633461,
+ "step": 1399
+ },
+ {
+ "epoch": 0.7345225603357818,
+ "grad_norm": 0.17467968960064226,
+ "learning_rate": 1.238219670429104e-05,
+ "loss": 0.45413005352020264,
+ "step": 1400
+ },
+ {
+ "epoch": 0.7350472193074502,
+ "grad_norm": 0.1544591919002847,
+ "learning_rate": 1.2354749919202598e-05,
+ "loss": 0.45218425989151,
+ "step": 1401
+ },
+ {
+ "epoch": 0.7355718782791185,
+ "grad_norm": 0.16946410088675617,
+ "learning_rate": 1.2327344280368217e-05,
+ "loss": 0.4459337592124939,
+ "step": 1402
+ },
+ {
+ "epoch": 0.7360965372507869,
+ "grad_norm": 0.17454129080094982,
+ "learning_rate": 1.2299979862242882e-05,
+ "loss": 0.440622478723526,
+ "step": 1403
+ },
+ {
+ "epoch": 0.7366211962224554,
+ "grad_norm": 0.1551459393434639,
+ "learning_rate": 1.2272656739169588e-05,
+ "loss": 0.43796059489250183,
+ "step": 1404
+ },
+ {
+ "epoch": 0.7371458551941238,
+ "grad_norm": 0.16505319132579585,
+ "learning_rate": 1.2245374985379147e-05,
+ "loss": 0.44425907731056213,
+ "step": 1405
+ },
+ {
+ "epoch": 0.7376705141657922,
+ "grad_norm": 0.1681701201995967,
+ "learning_rate": 1.2218134674989978e-05,
+ "loss": 0.45886728167533875,
+ "step": 1406
+ },
+ {
+ "epoch": 0.7381951731374606,
+ "grad_norm": 0.15787202261507643,
+ "learning_rate": 1.2190935882007894e-05,
+ "loss": 0.4489729404449463,
+ "step": 1407
+ },
+ {
+ "epoch": 0.7387198321091291,
+ "grad_norm": 0.1580152926319956,
+ "learning_rate": 1.2163778680325947e-05,
+ "loss": 0.4368497133255005,
+ "step": 1408
+ },
+ {
+ "epoch": 0.7392444910807975,
+ "grad_norm": 0.16798156319918758,
+ "learning_rate": 1.2136663143724163e-05,
+ "loss": 0.4503224492073059,
+ "step": 1409
+ },
+ {
+ "epoch": 0.7397691500524659,
+ "grad_norm": 0.164907453178928,
+ "learning_rate": 1.2109589345869387e-05,
+ "loss": 0.46600162982940674,
+ "step": 1410
+ },
+ {
+ "epoch": 0.7402938090241343,
+ "grad_norm": 0.16237894750513804,
+ "learning_rate": 1.2082557360315065e-05,
+ "loss": 0.45209893584251404,
+ "step": 1411
+ },
+ {
+ "epoch": 0.7408184679958028,
+ "grad_norm": 0.1600179850065201,
+ "learning_rate": 1.2055567260501054e-05,
+ "loss": 0.4404461979866028,
+ "step": 1412
+ },
+ {
+ "epoch": 0.7413431269674712,
+ "grad_norm": 0.15094395289646126,
+ "learning_rate": 1.2028619119753402e-05,
+ "loss": 0.42805302143096924,
+ "step": 1413
+ },
+ {
+ "epoch": 0.7418677859391396,
+ "grad_norm": 0.1560803755960921,
+ "learning_rate": 1.2001713011284193e-05,
+ "loss": 0.4421588182449341,
+ "step": 1414
+ },
+ {
+ "epoch": 0.742392444910808,
+ "grad_norm": 0.1570850315783027,
+ "learning_rate": 1.1974849008191288e-05,
+ "loss": 0.4540778696537018,
+ "step": 1415
+ },
+ {
+ "epoch": 0.7429171038824763,
+ "grad_norm": 0.15804301837393614,
+ "learning_rate": 1.1948027183458182e-05,
+ "loss": 0.4601803421974182,
+ "step": 1416
+ },
+ {
+ "epoch": 0.7434417628541448,
+ "grad_norm": 0.15329202123031233,
+ "learning_rate": 1.1921247609953747e-05,
+ "loss": 0.4551991820335388,
+ "step": 1417
+ },
+ {
+ "epoch": 0.7439664218258132,
+ "grad_norm": 0.15953306858721886,
+ "learning_rate": 1.1894510360432106e-05,
+ "loss": 0.4297157824039459,
+ "step": 1418
+ },
+ {
+ "epoch": 0.7444910807974816,
+ "grad_norm": 0.16006724351151672,
+ "learning_rate": 1.1867815507532377e-05,
+ "loss": 0.4468337297439575,
+ "step": 1419
+ },
+ {
+ "epoch": 0.74501573976915,
+ "grad_norm": 0.15300788223514664,
+ "learning_rate": 1.1841163123778502e-05,
+ "loss": 0.44412750005722046,
+ "step": 1420
+ },
+ {
+ "epoch": 0.7455403987408185,
+ "grad_norm": 0.1534395618553107,
+ "learning_rate": 1.1814553281579034e-05,
+ "loss": 0.4501240849494934,
+ "step": 1421
+ },
+ {
+ "epoch": 0.7460650577124869,
+ "grad_norm": 0.15384102301255653,
+ "learning_rate": 1.1787986053226954e-05,
+ "loss": 0.45214593410491943,
+ "step": 1422
+ },
+ {
+ "epoch": 0.7465897166841553,
+ "grad_norm": 0.15393815588029566,
+ "learning_rate": 1.176146151089949e-05,
+ "loss": 0.44750508666038513,
+ "step": 1423
+ },
+ {
+ "epoch": 0.7471143756558237,
+ "grad_norm": 0.15864106035518064,
+ "learning_rate": 1.1734979726657878e-05,
+ "loss": 0.4473385214805603,
+ "step": 1424
+ },
+ {
+ "epoch": 0.7476390346274921,
+ "grad_norm": 0.15035802419668112,
+ "learning_rate": 1.17085407724472e-05,
+ "loss": 0.4568535089492798,
+ "step": 1425
+ },
+ {
+ "epoch": 0.7481636935991606,
+ "grad_norm": 0.14513260265639483,
+ "learning_rate": 1.1682144720096168e-05,
+ "loss": 0.43895506858825684,
+ "step": 1426
+ },
+ {
+ "epoch": 0.748688352570829,
+ "grad_norm": 0.15405098264322387,
+ "learning_rate": 1.1655791641316954e-05,
+ "loss": 0.4408157467842102,
+ "step": 1427
+ },
+ {
+ "epoch": 0.7492130115424974,
+ "grad_norm": 0.1580598028770981,
+ "learning_rate": 1.1629481607704963e-05,
+ "loss": 0.4624619483947754,
+ "step": 1428
+ },
+ {
+ "epoch": 0.7497376705141658,
+ "grad_norm": 0.14844582674985438,
+ "learning_rate": 1.1603214690738683e-05,
+ "loss": 0.4332723021507263,
+ "step": 1429
+ },
+ {
+ "epoch": 0.7502623294858342,
+ "grad_norm": 0.15177050924813218,
+ "learning_rate": 1.1576990961779447e-05,
+ "loss": 0.4521399140357971,
+ "step": 1430
+ },
+ {
+ "epoch": 0.7507869884575026,
+ "grad_norm": 0.1529493955525006,
+ "learning_rate": 1.1550810492071243e-05,
+ "loss": 0.4454570412635803,
+ "step": 1431
+ },
+ {
+ "epoch": 0.751311647429171,
+ "grad_norm": 0.15120682752180228,
+ "learning_rate": 1.1524673352740543e-05,
+ "loss": 0.43875277042388916,
+ "step": 1432
+ },
+ {
+ "epoch": 0.7518363064008394,
+ "grad_norm": 0.1467903729247548,
+ "learning_rate": 1.1498579614796121e-05,
+ "loss": 0.4373723566532135,
+ "step": 1433
+ },
+ {
+ "epoch": 0.7523609653725079,
+ "grad_norm": 0.17163249169426106,
+ "learning_rate": 1.1472529349128812e-05,
+ "loss": 0.4545165002346039,
+ "step": 1434
+ },
+ {
+ "epoch": 0.7528856243441763,
+ "grad_norm": 0.14756845154182477,
+ "learning_rate": 1.144652262651136e-05,
+ "loss": 0.43648290634155273,
+ "step": 1435
+ },
+ {
+ "epoch": 0.7534102833158447,
+ "grad_norm": 0.16264595781092334,
+ "learning_rate": 1.1420559517598205e-05,
+ "loss": 0.46860024333000183,
+ "step": 1436
+ },
+ {
+ "epoch": 0.7539349422875131,
+ "grad_norm": 0.15536522876020278,
+ "learning_rate": 1.1394640092925302e-05,
+ "loss": 0.4501672685146332,
+ "step": 1437
+ },
+ {
+ "epoch": 0.7544596012591815,
+ "grad_norm": 0.16184288744523595,
+ "learning_rate": 1.1368764422909919e-05,
+ "loss": 0.4542751908302307,
+ "step": 1438
+ },
+ {
+ "epoch": 0.75498426023085,
+ "grad_norm": 0.15868996516705539,
+ "learning_rate": 1.1342932577850473e-05,
+ "loss": 0.46219199895858765,
+ "step": 1439
+ },
+ {
+ "epoch": 0.7555089192025184,
+ "grad_norm": 0.15896636432385472,
+ "learning_rate": 1.13171446279263e-05,
+ "loss": 0.4495980441570282,
+ "step": 1440
+ },
+ {
+ "epoch": 0.7560335781741868,
+ "grad_norm": 0.1535420584032188,
+ "learning_rate": 1.1291400643197484e-05,
+ "loss": 0.45162540674209595,
+ "step": 1441
+ },
+ {
+ "epoch": 0.7565582371458552,
+ "grad_norm": 0.15414897245083417,
+ "learning_rate": 1.1265700693604668e-05,
+ "loss": 0.44752851128578186,
+ "step": 1442
+ },
+ {
+ "epoch": 0.7570828961175237,
+ "grad_norm": 0.15700054057233076,
+ "learning_rate": 1.1240044848968856e-05,
+ "loss": 0.4516165256500244,
+ "step": 1443
+ },
+ {
+ "epoch": 0.757607555089192,
+ "grad_norm": 0.1564534139702123,
+ "learning_rate": 1.1214433178991258e-05,
+ "loss": 0.459886372089386,
+ "step": 1444
+ },
+ {
+ "epoch": 0.7581322140608604,
+ "grad_norm": 0.15412437202500157,
+ "learning_rate": 1.1188865753253022e-05,
+ "loss": 0.4439350366592407,
+ "step": 1445
+ },
+ {
+ "epoch": 0.7586568730325288,
+ "grad_norm": 0.15615271855095464,
+ "learning_rate": 1.116334264121513e-05,
+ "loss": 0.45808249711990356,
+ "step": 1446
+ },
+ {
+ "epoch": 0.7591815320041972,
+ "grad_norm": 0.1541136399166323,
+ "learning_rate": 1.1137863912218154e-05,
+ "loss": 0.4474959373474121,
+ "step": 1447
+ },
+ {
+ "epoch": 0.7597061909758657,
+ "grad_norm": 0.15443512615409669,
+ "learning_rate": 1.1112429635482118e-05,
+ "loss": 0.4358808398246765,
+ "step": 1448
+ },
+ {
+ "epoch": 0.7602308499475341,
+ "grad_norm": 0.15662992428712963,
+ "learning_rate": 1.1087039880106247e-05,
+ "loss": 0.46792712807655334,
+ "step": 1449
+ },
+ {
+ "epoch": 0.7607555089192025,
+ "grad_norm": 0.1515920350740043,
+ "learning_rate": 1.106169471506882e-05,
+ "loss": 0.4230170249938965,
+ "step": 1450
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 3.838217406749082e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1450/video_preprocessor_config.json b/checkpoints/checkpoint-1450/video_preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..a933f0ec6941222c33aeba7980bd5acf8dd98e91
--- /dev/null
+++ b/checkpoints/checkpoint-1450/video_preprocessor_config.json
@@ -0,0 +1,30 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "do_sample_frames": true,
+ "fps": 2,
+ "image_mean": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "image_std": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "max_frames": 768,
+ "merge_size": 2,
+ "min_frames": 4,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 786432,
+ "shortest_edge": 131072
+ },
+ "temporal_patch_size": 2,
+ "video_processor_type": "Qwen3VLVideoProcessor"
+}
diff --git a/checkpoints/checkpoint-1550/chat_template.jinja b/checkpoints/checkpoint-1550/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1550/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1550/config.json b/checkpoints/checkpoint-1550/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1550/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1550/expert_prune_metadata.json b/checkpoints/checkpoint-1550/expert_prune_metadata.json
new file mode 100644
index 0000000000000000000000000000000000000000..16e01c8a10636331e334732fa9629579c68df44e
--- /dev/null
+++ b/checkpoints/checkpoint-1550/expert_prune_metadata.json
@@ -0,0 +1,14 @@
+{
+ "source_model": "/mnt/data/wuwei/model_list/Qwen3.6-35B-A3B/models--Qwen--Qwen3.6-35B-A3B/snapshots/995ad96eacd98c81ed38be0c5b274b04031597b0",
+ "output_model": "/mnt/data/wuwei/model_list/qwen3-19b-a3b",
+ "total_parameters": 18990568816,
+ "parameter_billions_decimal": 18.990568816,
+ "directory_size_label": "19B",
+ "original_num_experts": 256,
+ "kept_experts_per_layer": 128,
+ "num_profiled_layers": 40,
+ "mtp_weights": "removed because Qwen3_5MoeForConditionalGeneration does not instantiate or use the source checkpoint's mtp.* training-only tensors",
+ "expert_order": "ascending original expert index; router rows and expert tensors use the same remapping",
+ "router_statistics": "router_stats.json and router_stats.pt",
+ "selection_score": "top-8 router hit frequency * mean(abs(actual expert output after down_proj and before routing-weight multiplication))"
+}
diff --git a/checkpoints/checkpoint-1550/generation_config.json b/checkpoints/checkpoint-1550/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1550/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1550/model.safetensors b/checkpoints/checkpoint-1550/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..783f32c63798756827b068a74f5deb78c86a7c8e
--- /dev/null
+++ b/checkpoints/checkpoint-1550/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3539effe6ad2943540812cdd09dc7f682348217f99f074d5d4091acb29585dd
+size 37983421000
diff --git a/checkpoints/checkpoint-1550/preprocessor_config.json b/checkpoints/checkpoint-1550/preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ec43d1f859ecadf800ebae92fda4a812d34975d3
--- /dev/null
+++ b/checkpoints/checkpoint-1550/preprocessor_config.json
@@ -0,0 +1,26 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "image_mean": [
+ 0.48145466,
+ 0.4578275,
+ 0.40821073
+ ],
+ "image_processor_type": "Qwen2VLImageProcessorFast",
+ "image_std": [
+ 0.26862954,
+ 0.26130258,
+ 0.27577711
+ ],
+ "merge_size": 2,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 1003520,
+ "shortest_edge": 3136
+ },
+ "temporal_patch_size": 2
+}
diff --git a/checkpoints/checkpoint-1550/tokenizer.json b/checkpoints/checkpoint-1550/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1550/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1550/tokenizer_config.json b/checkpoints/checkpoint-1550/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1550/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1550/trainer_state.json b/checkpoints/checkpoint-1550/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..b9327b21cd5a7e5d37228046885dc5d483748dff
--- /dev/null
+++ b/checkpoints/checkpoint-1550/trainer_state.json
@@ -0,0 +1,10884 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.8132214060860441,
+ "eval_steps": 500,
+ "global_step": 1550,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ },
+ {
+ "epoch": 0.5776495278069255,
+ "grad_norm": 0.1862053032854098,
+ "learning_rate": 2.2065553964033126e-05,
+ "loss": 0.4816875457763672,
+ "step": 1101
+ },
+ {
+ "epoch": 0.5781741867785939,
+ "grad_norm": 0.16996015894363117,
+ "learning_rate": 2.20295734045532e-05,
+ "loss": 0.4471197724342346,
+ "step": 1102
+ },
+ {
+ "epoch": 0.5786988457502623,
+ "grad_norm": 0.16889227008390692,
+ "learning_rate": 2.1993607706997725e-05,
+ "loss": 0.468391478061676,
+ "step": 1103
+ },
+ {
+ "epoch": 0.5792235047219307,
+ "grad_norm": 0.16670931681562373,
+ "learning_rate": 2.195765696907745e-05,
+ "loss": 0.44934162497520447,
+ "step": 1104
+ },
+ {
+ "epoch": 0.5797481636935992,
+ "grad_norm": 0.16328418741223716,
+ "learning_rate": 2.1921721288462482e-05,
+ "loss": 0.4558270275592804,
+ "step": 1105
+ },
+ {
+ "epoch": 0.5802728226652676,
+ "grad_norm": 0.17364660408703936,
+ "learning_rate": 2.188580076278203e-05,
+ "loss": 0.4597751796245575,
+ "step": 1106
+ },
+ {
+ "epoch": 0.580797481636936,
+ "grad_norm": 0.1628618453800601,
+ "learning_rate": 2.18498954896241e-05,
+ "loss": 0.4521859288215637,
+ "step": 1107
+ },
+ {
+ "epoch": 0.5813221406086044,
+ "grad_norm": 0.16813817951871482,
+ "learning_rate": 2.1814005566535302e-05,
+ "loss": 0.4581005573272705,
+ "step": 1108
+ },
+ {
+ "epoch": 0.5818467995802729,
+ "grad_norm": 0.16825923629726025,
+ "learning_rate": 2.1778131091020516e-05,
+ "loss": 0.4433916211128235,
+ "step": 1109
+ },
+ {
+ "epoch": 0.5823714585519413,
+ "grad_norm": 0.17194743467883944,
+ "learning_rate": 2.1742272160542654e-05,
+ "loss": 0.47296425700187683,
+ "step": 1110
+ },
+ {
+ "epoch": 0.5828961175236097,
+ "grad_norm": 0.16947730546039969,
+ "learning_rate": 2.170642887252241e-05,
+ "loss": 0.4815937876701355,
+ "step": 1111
+ },
+ {
+ "epoch": 0.583420776495278,
+ "grad_norm": 0.1657449928668473,
+ "learning_rate": 2.167060132433798e-05,
+ "loss": 0.4479164779186249,
+ "step": 1112
+ },
+ {
+ "epoch": 0.5839454354669464,
+ "grad_norm": 0.1658952516607364,
+ "learning_rate": 2.1634789613324778e-05,
+ "loss": 0.4637463688850403,
+ "step": 1113
+ },
+ {
+ "epoch": 0.5844700944386149,
+ "grad_norm": 0.16777002188998236,
+ "learning_rate": 2.1598993836775218e-05,
+ "loss": 0.457524836063385,
+ "step": 1114
+ },
+ {
+ "epoch": 0.5849947534102833,
+ "grad_norm": 0.15837566447341322,
+ "learning_rate": 2.1563214091938405e-05,
+ "loss": 0.44938087463378906,
+ "step": 1115
+ },
+ {
+ "epoch": 0.5855194123819517,
+ "grad_norm": 0.17162206617035314,
+ "learning_rate": 2.1527450476019896e-05,
+ "loss": 0.4393450617790222,
+ "step": 1116
+ },
+ {
+ "epoch": 0.5860440713536201,
+ "grad_norm": 0.16224101921433814,
+ "learning_rate": 2.1491703086181426e-05,
+ "loss": 0.451913982629776,
+ "step": 1117
+ },
+ {
+ "epoch": 0.5865687303252886,
+ "grad_norm": 0.1736253504575144,
+ "learning_rate": 2.145597201954066e-05,
+ "loss": 0.4611290395259857,
+ "step": 1118
+ },
+ {
+ "epoch": 0.587093389296957,
+ "grad_norm": 0.16308408333737234,
+ "learning_rate": 2.1420257373170903e-05,
+ "loss": 0.45898568630218506,
+ "step": 1119
+ },
+ {
+ "epoch": 0.5876180482686254,
+ "grad_norm": 0.16948634794175524,
+ "learning_rate": 2.1384559244100848e-05,
+ "loss": 0.45982882380485535,
+ "step": 1120
+ },
+ {
+ "epoch": 0.5881427072402938,
+ "grad_norm": 0.16676366538565252,
+ "learning_rate": 2.1348877729314336e-05,
+ "loss": 0.44826576113700867,
+ "step": 1121
+ },
+ {
+ "epoch": 0.5886673662119623,
+ "grad_norm": 0.16431125630712295,
+ "learning_rate": 2.1313212925750037e-05,
+ "loss": 0.4616445302963257,
+ "step": 1122
+ },
+ {
+ "epoch": 0.5891920251836307,
+ "grad_norm": 0.1631659061616955,
+ "learning_rate": 2.1277564930301252e-05,
+ "loss": 0.45651188492774963,
+ "step": 1123
+ },
+ {
+ "epoch": 0.5897166841552991,
+ "grad_norm": 0.15805950502689237,
+ "learning_rate": 2.1241933839815604e-05,
+ "loss": 0.4486045837402344,
+ "step": 1124
+ },
+ {
+ "epoch": 0.5902413431269675,
+ "grad_norm": 0.16530288038603502,
+ "learning_rate": 2.1206319751094773e-05,
+ "loss": 0.4485066533088684,
+ "step": 1125
+ },
+ {
+ "epoch": 0.5907660020986358,
+ "grad_norm": 0.16061903194239047,
+ "learning_rate": 2.1170722760894295e-05,
+ "loss": 0.4565330743789673,
+ "step": 1126
+ },
+ {
+ "epoch": 0.5912906610703043,
+ "grad_norm": 0.1643346999353627,
+ "learning_rate": 2.1135142965923195e-05,
+ "loss": 0.45560717582702637,
+ "step": 1127
+ },
+ {
+ "epoch": 0.5918153200419727,
+ "grad_norm": 0.17484644315215753,
+ "learning_rate": 2.109958046284382e-05,
+ "loss": 0.4795891046524048,
+ "step": 1128
+ },
+ {
+ "epoch": 0.5923399790136411,
+ "grad_norm": 0.1623762678622651,
+ "learning_rate": 2.1064035348271528e-05,
+ "loss": 0.47395604848861694,
+ "step": 1129
+ },
+ {
+ "epoch": 0.5928646379853095,
+ "grad_norm": 0.16888824241669015,
+ "learning_rate": 2.1028507718774456e-05,
+ "loss": 0.4601455330848694,
+ "step": 1130
+ },
+ {
+ "epoch": 0.593389296956978,
+ "grad_norm": 0.16291744568806302,
+ "learning_rate": 2.0992997670873193e-05,
+ "loss": 0.46248236298561096,
+ "step": 1131
+ },
+ {
+ "epoch": 0.5939139559286464,
+ "grad_norm": 0.16185598691236988,
+ "learning_rate": 2.0957505301040603e-05,
+ "loss": 0.45085301995277405,
+ "step": 1132
+ },
+ {
+ "epoch": 0.5944386149003148,
+ "grad_norm": 0.1678000806117814,
+ "learning_rate": 2.0922030705701518e-05,
+ "loss": 0.44939014315605164,
+ "step": 1133
+ },
+ {
+ "epoch": 0.5949632738719832,
+ "grad_norm": 0.16104474282288994,
+ "learning_rate": 2.088657398123246e-05,
+ "loss": 0.4313976764678955,
+ "step": 1134
+ },
+ {
+ "epoch": 0.5954879328436516,
+ "grad_norm": 0.1631128968199669,
+ "learning_rate": 2.085113522396141e-05,
+ "loss": 0.4508278965950012,
+ "step": 1135
+ },
+ {
+ "epoch": 0.5960125918153201,
+ "grad_norm": 0.16307688195844533,
+ "learning_rate": 2.0815714530167557e-05,
+ "loss": 0.4467576742172241,
+ "step": 1136
+ },
+ {
+ "epoch": 0.5965372507869885,
+ "grad_norm": 0.16456832304532207,
+ "learning_rate": 2.0780311996080976e-05,
+ "loss": 0.4621703624725342,
+ "step": 1137
+ },
+ {
+ "epoch": 0.5970619097586569,
+ "grad_norm": 0.16346217748325367,
+ "learning_rate": 2.0744927717882435e-05,
+ "loss": 0.4589449465274811,
+ "step": 1138
+ },
+ {
+ "epoch": 0.5975865687303253,
+ "grad_norm": 0.16180691501076994,
+ "learning_rate": 2.0709561791703097e-05,
+ "loss": 0.43698400259017944,
+ "step": 1139
+ },
+ {
+ "epoch": 0.5981112277019937,
+ "grad_norm": 0.15695459976898835,
+ "learning_rate": 2.067421431362427e-05,
+ "loss": 0.452271044254303,
+ "step": 1140
+ },
+ {
+ "epoch": 0.5986358866736621,
+ "grad_norm": 0.15930893553721054,
+ "learning_rate": 2.063888537967713e-05,
+ "loss": 0.44868963956832886,
+ "step": 1141
+ },
+ {
+ "epoch": 0.5991605456453305,
+ "grad_norm": 0.15963802284017928,
+ "learning_rate": 2.0603575085842485e-05,
+ "loss": 0.4376864731311798,
+ "step": 1142
+ },
+ {
+ "epoch": 0.5996852046169989,
+ "grad_norm": 0.16610266900507178,
+ "learning_rate": 2.05682835280505e-05,
+ "loss": 0.44676727056503296,
+ "step": 1143
+ },
+ {
+ "epoch": 0.6002098635886673,
+ "grad_norm": 0.15978253236669612,
+ "learning_rate": 2.053301080218044e-05,
+ "loss": 0.434773713350296,
+ "step": 1144
+ },
+ {
+ "epoch": 0.6007345225603358,
+ "grad_norm": 0.16305141803526324,
+ "learning_rate": 2.0497757004060403e-05,
+ "loss": 0.4529780149459839,
+ "step": 1145
+ },
+ {
+ "epoch": 0.6012591815320042,
+ "grad_norm": 0.1649737864101215,
+ "learning_rate": 2.046252222946707e-05,
+ "loss": 0.4571434259414673,
+ "step": 1146
+ },
+ {
+ "epoch": 0.6017838405036726,
+ "grad_norm": 0.15738824968326598,
+ "learning_rate": 2.0427306574125427e-05,
+ "loss": 0.4446289539337158,
+ "step": 1147
+ },
+ {
+ "epoch": 0.602308499475341,
+ "grad_norm": 0.1731139535185187,
+ "learning_rate": 2.039211013370853e-05,
+ "loss": 0.4551090598106384,
+ "step": 1148
+ },
+ {
+ "epoch": 0.6028331584470095,
+ "grad_norm": 0.15857251900043037,
+ "learning_rate": 2.0356933003837237e-05,
+ "loss": 0.4644010663032532,
+ "step": 1149
+ },
+ {
+ "epoch": 0.6033578174186779,
+ "grad_norm": 0.1589582861134211,
+ "learning_rate": 2.032177528007992e-05,
+ "loss": 0.45741650462150574,
+ "step": 1150
+ },
+ {
+ "epoch": 0.6038824763903463,
+ "grad_norm": 0.15639092706575122,
+ "learning_rate": 2.0286637057952257e-05,
+ "loss": 0.4375547766685486,
+ "step": 1151
+ },
+ {
+ "epoch": 0.6044071353620147,
+ "grad_norm": 0.16335605581337187,
+ "learning_rate": 2.0251518432916933e-05,
+ "loss": 0.4636140465736389,
+ "step": 1152
+ },
+ {
+ "epoch": 0.6049317943336832,
+ "grad_norm": 0.15437437113359356,
+ "learning_rate": 2.0216419500383372e-05,
+ "loss": 0.44632142782211304,
+ "step": 1153
+ },
+ {
+ "epoch": 0.6054564533053515,
+ "grad_norm": 0.15522039198698664,
+ "learning_rate": 2.0181340355707545e-05,
+ "loss": 0.449049174785614,
+ "step": 1154
+ },
+ {
+ "epoch": 0.6059811122770199,
+ "grad_norm": 0.1599951917909323,
+ "learning_rate": 2.0146281094191604e-05,
+ "loss": 0.44395962357521057,
+ "step": 1155
+ },
+ {
+ "epoch": 0.6065057712486883,
+ "grad_norm": 0.16866621620527716,
+ "learning_rate": 2.011124181108373e-05,
+ "loss": 0.45693716406822205,
+ "step": 1156
+ },
+ {
+ "epoch": 0.6070304302203567,
+ "grad_norm": 0.15508935650135075,
+ "learning_rate": 2.0076222601577815e-05,
+ "loss": 0.4525151252746582,
+ "step": 1157
+ },
+ {
+ "epoch": 0.6075550891920252,
+ "grad_norm": 0.16508998118005971,
+ "learning_rate": 2.004122356081321e-05,
+ "loss": 0.466006875038147,
+ "step": 1158
+ },
+ {
+ "epoch": 0.6080797481636936,
+ "grad_norm": 0.15817418864022237,
+ "learning_rate": 2.0006244783874466e-05,
+ "loss": 0.4496110677719116,
+ "step": 1159
+ },
+ {
+ "epoch": 0.608604407135362,
+ "grad_norm": 0.16127270259749824,
+ "learning_rate": 1.997128636579109e-05,
+ "loss": 0.45918628573417664,
+ "step": 1160
+ },
+ {
+ "epoch": 0.6091290661070304,
+ "grad_norm": 0.1702599924159677,
+ "learning_rate": 1.9936348401537292e-05,
+ "loss": 0.47218555212020874,
+ "step": 1161
+ },
+ {
+ "epoch": 0.6096537250786989,
+ "grad_norm": 0.15690176248846127,
+ "learning_rate": 1.9901430986031695e-05,
+ "loss": 0.4516097903251648,
+ "step": 1162
+ },
+ {
+ "epoch": 0.6101783840503673,
+ "grad_norm": 0.1713614405614144,
+ "learning_rate": 1.9866534214137085e-05,
+ "loss": 0.45030856132507324,
+ "step": 1163
+ },
+ {
+ "epoch": 0.6107030430220357,
+ "grad_norm": 0.1670865563026229,
+ "learning_rate": 1.9831658180660195e-05,
+ "loss": 0.45603489875793457,
+ "step": 1164
+ },
+ {
+ "epoch": 0.6112277019937041,
+ "grad_norm": 0.16897539113401952,
+ "learning_rate": 1.9796802980351392e-05,
+ "loss": 0.46314537525177,
+ "step": 1165
+ },
+ {
+ "epoch": 0.6117523609653726,
+ "grad_norm": 0.15832868870697644,
+ "learning_rate": 1.976196870790446e-05,
+ "loss": 0.44784611463546753,
+ "step": 1166
+ },
+ {
+ "epoch": 0.612277019937041,
+ "grad_norm": 0.16595365551891114,
+ "learning_rate": 1.9727155457956315e-05,
+ "loss": 0.4409624934196472,
+ "step": 1167
+ },
+ {
+ "epoch": 0.6128016789087093,
+ "grad_norm": 0.1642630411212017,
+ "learning_rate": 1.969236332508677e-05,
+ "loss": 0.46120351552963257,
+ "step": 1168
+ },
+ {
+ "epoch": 0.6133263378803777,
+ "grad_norm": 0.16481796135868812,
+ "learning_rate": 1.9657592403818253e-05,
+ "loss": 0.45544904470443726,
+ "step": 1169
+ },
+ {
+ "epoch": 0.6138509968520461,
+ "grad_norm": 0.1629832779852364,
+ "learning_rate": 1.962284278861557e-05,
+ "loss": 0.4578462839126587,
+ "step": 1170
+ },
+ {
+ "epoch": 0.6143756558237146,
+ "grad_norm": 0.15916219613603688,
+ "learning_rate": 1.9588114573885664e-05,
+ "loss": 0.4546983540058136,
+ "step": 1171
+ },
+ {
+ "epoch": 0.614900314795383,
+ "grad_norm": 0.1660520165166075,
+ "learning_rate": 1.95534078539773e-05,
+ "loss": 0.4601942300796509,
+ "step": 1172
+ },
+ {
+ "epoch": 0.6154249737670514,
+ "grad_norm": 0.16924794578577768,
+ "learning_rate": 1.9518722723180877e-05,
+ "loss": 0.45997846126556396,
+ "step": 1173
+ },
+ {
+ "epoch": 0.6159496327387198,
+ "grad_norm": 0.16597646133045382,
+ "learning_rate": 1.9484059275728127e-05,
+ "loss": 0.4594605565071106,
+ "step": 1174
+ },
+ {
+ "epoch": 0.6164742917103883,
+ "grad_norm": 0.16964243792519748,
+ "learning_rate": 1.9449417605791876e-05,
+ "loss": 0.4638690650463104,
+ "step": 1175
+ },
+ {
+ "epoch": 0.6169989506820567,
+ "grad_norm": 0.1758591426791696,
+ "learning_rate": 1.9414797807485796e-05,
+ "loss": 0.4666447639465332,
+ "step": 1176
+ },
+ {
+ "epoch": 0.6175236096537251,
+ "grad_norm": 0.16217661227891753,
+ "learning_rate": 1.9380199974864126e-05,
+ "loss": 0.44252535700798035,
+ "step": 1177
+ },
+ {
+ "epoch": 0.6180482686253935,
+ "grad_norm": 0.1576446184499019,
+ "learning_rate": 1.934562420192142e-05,
+ "loss": 0.4499641954898834,
+ "step": 1178
+ },
+ {
+ "epoch": 0.618572927597062,
+ "grad_norm": 0.16231008092636912,
+ "learning_rate": 1.9311070582592327e-05,
+ "loss": 0.4561542570590973,
+ "step": 1179
+ },
+ {
+ "epoch": 0.6190975865687304,
+ "grad_norm": 0.16790382249577043,
+ "learning_rate": 1.9276539210751292e-05,
+ "loss": 0.4484325647354126,
+ "step": 1180
+ },
+ {
+ "epoch": 0.6196222455403987,
+ "grad_norm": 0.18794729932605855,
+ "learning_rate": 1.9242030180212323e-05,
+ "loss": 0.44169026613235474,
+ "step": 1181
+ },
+ {
+ "epoch": 0.6201469045120671,
+ "grad_norm": 0.16948051243889814,
+ "learning_rate": 1.920754358472872e-05,
+ "loss": 0.4592876434326172,
+ "step": 1182
+ },
+ {
+ "epoch": 0.6206715634837355,
+ "grad_norm": 0.16762558710029374,
+ "learning_rate": 1.917307951799286e-05,
+ "loss": 0.4402095079421997,
+ "step": 1183
+ },
+ {
+ "epoch": 0.621196222455404,
+ "grad_norm": 0.18170280116387003,
+ "learning_rate": 1.9138638073635885e-05,
+ "loss": 0.4633026123046875,
+ "step": 1184
+ },
+ {
+ "epoch": 0.6217208814270724,
+ "grad_norm": 0.1718051709136108,
+ "learning_rate": 1.9104219345227486e-05,
+ "loss": 0.4510066509246826,
+ "step": 1185
+ },
+ {
+ "epoch": 0.6222455403987408,
+ "grad_norm": 0.16829066326345576,
+ "learning_rate": 1.906982342627565e-05,
+ "loss": 0.4632919430732727,
+ "step": 1186
+ },
+ {
+ "epoch": 0.6227701993704092,
+ "grad_norm": 0.16669141808548274,
+ "learning_rate": 1.9035450410226395e-05,
+ "loss": 0.455544650554657,
+ "step": 1187
+ },
+ {
+ "epoch": 0.6232948583420777,
+ "grad_norm": 0.16679996559722232,
+ "learning_rate": 1.90011003904635e-05,
+ "loss": 0.459994375705719,
+ "step": 1188
+ },
+ {
+ "epoch": 0.6238195173137461,
+ "grad_norm": 0.17379754713278164,
+ "learning_rate": 1.8966773460308284e-05,
+ "loss": 0.4576749801635742,
+ "step": 1189
+ },
+ {
+ "epoch": 0.6243441762854145,
+ "grad_norm": 0.17088539733389246,
+ "learning_rate": 1.8932469713019335e-05,
+ "loss": 0.44853052496910095,
+ "step": 1190
+ },
+ {
+ "epoch": 0.6248688352570829,
+ "grad_norm": 0.15556883889201015,
+ "learning_rate": 1.8898189241792248e-05,
+ "loss": 0.44060447812080383,
+ "step": 1191
+ },
+ {
+ "epoch": 0.6253934942287513,
+ "grad_norm": 0.1640093917439215,
+ "learning_rate": 1.8863932139759388e-05,
+ "loss": 0.4531911611557007,
+ "step": 1192
+ },
+ {
+ "epoch": 0.6259181532004198,
+ "grad_norm": 0.16444284826591576,
+ "learning_rate": 1.882969849998964e-05,
+ "loss": 0.44176363945007324,
+ "step": 1193
+ },
+ {
+ "epoch": 0.6264428121720882,
+ "grad_norm": 0.1696462520995642,
+ "learning_rate": 1.8795488415488143e-05,
+ "loss": 0.44516652822494507,
+ "step": 1194
+ },
+ {
+ "epoch": 0.6269674711437565,
+ "grad_norm": 0.16810724705191796,
+ "learning_rate": 1.8761301979196032e-05,
+ "loss": 0.4542563855648041,
+ "step": 1195
+ },
+ {
+ "epoch": 0.6274921301154249,
+ "grad_norm": 0.17334813313855824,
+ "learning_rate": 1.8727139283990204e-05,
+ "loss": 0.47769248485565186,
+ "step": 1196
+ },
+ {
+ "epoch": 0.6280167890870934,
+ "grad_norm": 0.16308212035303235,
+ "learning_rate": 1.8693000422683043e-05,
+ "loss": 0.4407799243927002,
+ "step": 1197
+ },
+ {
+ "epoch": 0.6285414480587618,
+ "grad_norm": 0.16853319139928977,
+ "learning_rate": 1.8658885488022204e-05,
+ "loss": 0.4565497040748596,
+ "step": 1198
+ },
+ {
+ "epoch": 0.6290661070304302,
+ "grad_norm": 0.16694968781284703,
+ "learning_rate": 1.8624794572690328e-05,
+ "loss": 0.44522374868392944,
+ "step": 1199
+ },
+ {
+ "epoch": 0.6295907660020986,
+ "grad_norm": 0.16418579468339103,
+ "learning_rate": 1.8590727769304788e-05,
+ "loss": 0.4441048502922058,
+ "step": 1200
+ },
+ {
+ "epoch": 0.630115424973767,
+ "grad_norm": 0.1661930777124156,
+ "learning_rate": 1.8556685170417486e-05,
+ "loss": 0.457327663898468,
+ "step": 1201
+ },
+ {
+ "epoch": 0.6306400839454355,
+ "grad_norm": 0.16536630027039814,
+ "learning_rate": 1.852266686851452e-05,
+ "loss": 0.462494432926178,
+ "step": 1202
+ },
+ {
+ "epoch": 0.6311647429171039,
+ "grad_norm": 0.16611474088307454,
+ "learning_rate": 1.8488672956016013e-05,
+ "loss": 0.4557701647281647,
+ "step": 1203
+ },
+ {
+ "epoch": 0.6316894018887723,
+ "grad_norm": 0.16759343079749697,
+ "learning_rate": 1.8454703525275813e-05,
+ "loss": 0.45069995522499084,
+ "step": 1204
+ },
+ {
+ "epoch": 0.6322140608604407,
+ "grad_norm": 0.16186676294485336,
+ "learning_rate": 1.8420758668581255e-05,
+ "loss": 0.44954800605773926,
+ "step": 1205
+ },
+ {
+ "epoch": 0.6327387198321092,
+ "grad_norm": 0.15687159835368553,
+ "learning_rate": 1.8386838478152916e-05,
+ "loss": 0.4490429759025574,
+ "step": 1206
+ },
+ {
+ "epoch": 0.6332633788037776,
+ "grad_norm": 0.16242951976115577,
+ "learning_rate": 1.8352943046144372e-05,
+ "loss": 0.45819586515426636,
+ "step": 1207
+ },
+ {
+ "epoch": 0.633788037775446,
+ "grad_norm": 0.16516087266517232,
+ "learning_rate": 1.831907246464191e-05,
+ "loss": 0.44769883155822754,
+ "step": 1208
+ },
+ {
+ "epoch": 0.6343126967471143,
+ "grad_norm": 0.16709475871512644,
+ "learning_rate": 1.8285226825664332e-05,
+ "loss": 0.4699128568172455,
+ "step": 1209
+ },
+ {
+ "epoch": 0.6348373557187827,
+ "grad_norm": 0.16288623150595774,
+ "learning_rate": 1.8251406221162636e-05,
+ "loss": 0.45322558283805847,
+ "step": 1210
+ },
+ {
+ "epoch": 0.6353620146904512,
+ "grad_norm": 0.1675722848054237,
+ "learning_rate": 1.821761074301986e-05,
+ "loss": 0.4752347469329834,
+ "step": 1211
+ },
+ {
+ "epoch": 0.6358866736621196,
+ "grad_norm": 0.163285963514504,
+ "learning_rate": 1.818384048305074e-05,
+ "loss": 0.441958487033844,
+ "step": 1212
+ },
+ {
+ "epoch": 0.636411332633788,
+ "grad_norm": 0.17078102287557764,
+ "learning_rate": 1.8150095533001517e-05,
+ "loss": 0.4551333487033844,
+ "step": 1213
+ },
+ {
+ "epoch": 0.6369359916054564,
+ "grad_norm": 0.16188540746097138,
+ "learning_rate": 1.811637598454966e-05,
+ "loss": 0.4470779597759247,
+ "step": 1214
+ },
+ {
+ "epoch": 0.6374606505771249,
+ "grad_norm": 0.15235628218958286,
+ "learning_rate": 1.808268192930364e-05,
+ "loss": 0.43604153394699097,
+ "step": 1215
+ },
+ {
+ "epoch": 0.6379853095487933,
+ "grad_norm": 0.1604601949959104,
+ "learning_rate": 1.8049013458802665e-05,
+ "loss": 0.4445713758468628,
+ "step": 1216
+ },
+ {
+ "epoch": 0.6385099685204617,
+ "grad_norm": 0.15511797131215901,
+ "learning_rate": 1.8015370664516427e-05,
+ "loss": 0.4455301761627197,
+ "step": 1217
+ },
+ {
+ "epoch": 0.6390346274921301,
+ "grad_norm": 0.16815669546944095,
+ "learning_rate": 1.798175363784487e-05,
+ "loss": 0.46135014295578003,
+ "step": 1218
+ },
+ {
+ "epoch": 0.6395592864637986,
+ "grad_norm": 0.161558177688922,
+ "learning_rate": 1.7948162470117917e-05,
+ "loss": 0.4519304037094116,
+ "step": 1219
+ },
+ {
+ "epoch": 0.640083945435467,
+ "grad_norm": 0.16677842130048232,
+ "learning_rate": 1.7914597252595265e-05,
+ "loss": 0.4550550878047943,
+ "step": 1220
+ },
+ {
+ "epoch": 0.6406086044071354,
+ "grad_norm": 0.1597331480423138,
+ "learning_rate": 1.7881058076466095e-05,
+ "loss": 0.441120445728302,
+ "step": 1221
+ },
+ {
+ "epoch": 0.6411332633788038,
+ "grad_norm": 0.15951106545246824,
+ "learning_rate": 1.7847545032848844e-05,
+ "loss": 0.4471382200717926,
+ "step": 1222
+ },
+ {
+ "epoch": 0.6416579223504721,
+ "grad_norm": 0.16157787674667565,
+ "learning_rate": 1.7814058212790947e-05,
+ "loss": 0.4550648331642151,
+ "step": 1223
+ },
+ {
+ "epoch": 0.6421825813221406,
+ "grad_norm": 0.16989186763352415,
+ "learning_rate": 1.7780597707268583e-05,
+ "loss": 0.4515089988708496,
+ "step": 1224
+ },
+ {
+ "epoch": 0.642707240293809,
+ "grad_norm": 0.15952153223099508,
+ "learning_rate": 1.7747163607186468e-05,
+ "loss": 0.4658638834953308,
+ "step": 1225
+ },
+ {
+ "epoch": 0.6432318992654774,
+ "grad_norm": 0.15205926699647218,
+ "learning_rate": 1.7713756003377567e-05,
+ "loss": 0.4330868422985077,
+ "step": 1226
+ },
+ {
+ "epoch": 0.6437565582371458,
+ "grad_norm": 0.15784806480936794,
+ "learning_rate": 1.7680374986602865e-05,
+ "loss": 0.4506068527698517,
+ "step": 1227
+ },
+ {
+ "epoch": 0.6442812172088143,
+ "grad_norm": 0.16006709280686687,
+ "learning_rate": 1.7647020647551105e-05,
+ "loss": 0.446941077709198,
+ "step": 1228
+ },
+ {
+ "epoch": 0.6448058761804827,
+ "grad_norm": 0.16064631158222062,
+ "learning_rate": 1.761369307683856e-05,
+ "loss": 0.4418525695800781,
+ "step": 1229
+ },
+ {
+ "epoch": 0.6453305351521511,
+ "grad_norm": 0.16454136480444392,
+ "learning_rate": 1.7580392365008767e-05,
+ "loss": 0.46268391609191895,
+ "step": 1230
+ },
+ {
+ "epoch": 0.6458551941238195,
+ "grad_norm": 0.15401246535771462,
+ "learning_rate": 1.754711860253233e-05,
+ "loss": 0.4372989535331726,
+ "step": 1231
+ },
+ {
+ "epoch": 0.646379853095488,
+ "grad_norm": 0.1544567975028383,
+ "learning_rate": 1.7513871879806603e-05,
+ "loss": 0.44806575775146484,
+ "step": 1232
+ },
+ {
+ "epoch": 0.6469045120671564,
+ "grad_norm": 0.16411225112730965,
+ "learning_rate": 1.748065228715549e-05,
+ "loss": 0.43782058358192444,
+ "step": 1233
+ },
+ {
+ "epoch": 0.6474291710388248,
+ "grad_norm": 0.1552767832972093,
+ "learning_rate": 1.7447459914829183e-05,
+ "loss": 0.44531378149986267,
+ "step": 1234
+ },
+ {
+ "epoch": 0.6479538300104932,
+ "grad_norm": 0.16214695257123513,
+ "learning_rate": 1.741429485300392e-05,
+ "loss": 0.47399842739105225,
+ "step": 1235
+ },
+ {
+ "epoch": 0.6484784889821616,
+ "grad_norm": 0.16684861111484112,
+ "learning_rate": 1.7381157191781766e-05,
+ "loss": 0.4654282331466675,
+ "step": 1236
+ },
+ {
+ "epoch": 0.64900314795383,
+ "grad_norm": 0.15901864178630273,
+ "learning_rate": 1.7348047021190334e-05,
+ "loss": 0.45128414034843445,
+ "step": 1237
+ },
+ {
+ "epoch": 0.6495278069254984,
+ "grad_norm": 0.15360991961742707,
+ "learning_rate": 1.7314964431182523e-05,
+ "loss": 0.4303375482559204,
+ "step": 1238
+ },
+ {
+ "epoch": 0.6500524658971668,
+ "grad_norm": 0.1633151574385741,
+ "learning_rate": 1.728190951163633e-05,
+ "loss": 0.45568403601646423,
+ "step": 1239
+ },
+ {
+ "epoch": 0.6505771248688352,
+ "grad_norm": 0.16377853571058645,
+ "learning_rate": 1.724888235235457e-05,
+ "loss": 0.46503543853759766,
+ "step": 1240
+ },
+ {
+ "epoch": 0.6511017838405037,
+ "grad_norm": 0.1595023801275368,
+ "learning_rate": 1.7215883043064655e-05,
+ "loss": 0.46337294578552246,
+ "step": 1241
+ },
+ {
+ "epoch": 0.6516264428121721,
+ "grad_norm": 0.16324796182212292,
+ "learning_rate": 1.7182911673418313e-05,
+ "loss": 0.46197354793548584,
+ "step": 1242
+ },
+ {
+ "epoch": 0.6521511017838405,
+ "grad_norm": 0.16353734830879765,
+ "learning_rate": 1.714996833299138e-05,
+ "loss": 0.45741671323776245,
+ "step": 1243
+ },
+ {
+ "epoch": 0.6526757607555089,
+ "grad_norm": 0.15941587678263386,
+ "learning_rate": 1.7117053111283528e-05,
+ "loss": 0.4514503479003906,
+ "step": 1244
+ },
+ {
+ "epoch": 0.6532004197271774,
+ "grad_norm": 0.1642653230145008,
+ "learning_rate": 1.708416609771804e-05,
+ "loss": 0.4437655210494995,
+ "step": 1245
+ },
+ {
+ "epoch": 0.6537250786988458,
+ "grad_norm": 0.15814155519130152,
+ "learning_rate": 1.7051307381641587e-05,
+ "loss": 0.4314163327217102,
+ "step": 1246
+ },
+ {
+ "epoch": 0.6542497376705142,
+ "grad_norm": 0.16086078370768056,
+ "learning_rate": 1.701847705232394e-05,
+ "loss": 0.4526430666446686,
+ "step": 1247
+ },
+ {
+ "epoch": 0.6547743966421826,
+ "grad_norm": 0.16389360190715183,
+ "learning_rate": 1.698567519895775e-05,
+ "loss": 0.4278966188430786,
+ "step": 1248
+ },
+ {
+ "epoch": 0.655299055613851,
+ "grad_norm": 0.16303142287804262,
+ "learning_rate": 1.6952901910658316e-05,
+ "loss": 0.45778220891952515,
+ "step": 1249
+ },
+ {
+ "epoch": 0.6558237145855194,
+ "grad_norm": 0.1675209956904048,
+ "learning_rate": 1.6920157276463318e-05,
+ "loss": 0.4489403963088989,
+ "step": 1250
+ },
+ {
+ "epoch": 0.6563483735571878,
+ "grad_norm": 0.16228311551074323,
+ "learning_rate": 1.6887441385332604e-05,
+ "loss": 0.4534342885017395,
+ "step": 1251
+ },
+ {
+ "epoch": 0.6568730325288562,
+ "grad_norm": 0.16408957851046913,
+ "learning_rate": 1.6854754326147916e-05,
+ "loss": 0.4549208879470825,
+ "step": 1252
+ },
+ {
+ "epoch": 0.6573976915005246,
+ "grad_norm": 0.15967922111432845,
+ "learning_rate": 1.682209618771268e-05,
+ "loss": 0.4451141357421875,
+ "step": 1253
+ },
+ {
+ "epoch": 0.657922350472193,
+ "grad_norm": 0.16440698722204208,
+ "learning_rate": 1.678946705875176e-05,
+ "loss": 0.454200804233551,
+ "step": 1254
+ },
+ {
+ "epoch": 0.6584470094438615,
+ "grad_norm": 0.16251120184938497,
+ "learning_rate": 1.675686702791117e-05,
+ "loss": 0.4580027461051941,
+ "step": 1255
+ },
+ {
+ "epoch": 0.6589716684155299,
+ "grad_norm": 0.1643717386150457,
+ "learning_rate": 1.6724296183757924e-05,
+ "loss": 0.4422311782836914,
+ "step": 1256
+ },
+ {
+ "epoch": 0.6594963273871983,
+ "grad_norm": 0.1644586414378722,
+ "learning_rate": 1.66917546147797e-05,
+ "loss": 0.4593179225921631,
+ "step": 1257
+ },
+ {
+ "epoch": 0.6600209863588667,
+ "grad_norm": 0.16751074186967685,
+ "learning_rate": 1.6659242409384657e-05,
+ "loss": 0.457930326461792,
+ "step": 1258
+ },
+ {
+ "epoch": 0.6605456453305352,
+ "grad_norm": 0.15398418609155334,
+ "learning_rate": 1.6626759655901182e-05,
+ "loss": 0.44002899527549744,
+ "step": 1259
+ },
+ {
+ "epoch": 0.6610703043022036,
+ "grad_norm": 0.16328113175790226,
+ "learning_rate": 1.659430644257763e-05,
+ "loss": 0.4500438868999481,
+ "step": 1260
+ },
+ {
+ "epoch": 0.661594963273872,
+ "grad_norm": 0.1694244340615106,
+ "learning_rate": 1.656188285758214e-05,
+ "loss": 0.4818427264690399,
+ "step": 1261
+ },
+ {
+ "epoch": 0.6621196222455404,
+ "grad_norm": 0.1578320701754368,
+ "learning_rate": 1.652948898900232e-05,
+ "loss": 0.4492436647415161,
+ "step": 1262
+ },
+ {
+ "epoch": 0.6626442812172089,
+ "grad_norm": 0.16538865917828469,
+ "learning_rate": 1.6497124924845064e-05,
+ "loss": 0.45070338249206543,
+ "step": 1263
+ },
+ {
+ "epoch": 0.6631689401888772,
+ "grad_norm": 0.15670817848256852,
+ "learning_rate": 1.6464790753036296e-05,
+ "loss": 0.45439285039901733,
+ "step": 1264
+ },
+ {
+ "epoch": 0.6636935991605456,
+ "grad_norm": 0.16184786585965885,
+ "learning_rate": 1.64324865614207e-05,
+ "loss": 0.446522980928421,
+ "step": 1265
+ },
+ {
+ "epoch": 0.664218258132214,
+ "grad_norm": 0.15719726462251984,
+ "learning_rate": 1.6400212437761554e-05,
+ "loss": 0.4495280981063843,
+ "step": 1266
+ },
+ {
+ "epoch": 0.6647429171038824,
+ "grad_norm": 0.1680804414134405,
+ "learning_rate": 1.6367968469740416e-05,
+ "loss": 0.4652596712112427,
+ "step": 1267
+ },
+ {
+ "epoch": 0.6652675760755509,
+ "grad_norm": 0.15919541846683566,
+ "learning_rate": 1.6335754744956937e-05,
+ "loss": 0.43842798471450806,
+ "step": 1268
+ },
+ {
+ "epoch": 0.6657922350472193,
+ "grad_norm": 0.16322634187139418,
+ "learning_rate": 1.630357135092859e-05,
+ "loss": 0.45242664217948914,
+ "step": 1269
+ },
+ {
+ "epoch": 0.6663168940188877,
+ "grad_norm": 0.15644867734734783,
+ "learning_rate": 1.6271418375090445e-05,
+ "loss": 0.44801655411720276,
+ "step": 1270
+ },
+ {
+ "epoch": 0.6668415529905561,
+ "grad_norm": 0.1514717416124423,
+ "learning_rate": 1.623929590479496e-05,
+ "loss": 0.4506194293498993,
+ "step": 1271
+ },
+ {
+ "epoch": 0.6673662119622246,
+ "grad_norm": 0.1740345085495889,
+ "learning_rate": 1.6207204027311683e-05,
+ "loss": 0.4818177819252014,
+ "step": 1272
+ },
+ {
+ "epoch": 0.667890870933893,
+ "grad_norm": 0.15797701900765457,
+ "learning_rate": 1.6175142829827067e-05,
+ "loss": 0.4639382064342499,
+ "step": 1273
+ },
+ {
+ "epoch": 0.6684155299055614,
+ "grad_norm": 0.15663812316653875,
+ "learning_rate": 1.6143112399444198e-05,
+ "loss": 0.46217501163482666,
+ "step": 1274
+ },
+ {
+ "epoch": 0.6689401888772298,
+ "grad_norm": 0.15474018977043758,
+ "learning_rate": 1.6111112823182592e-05,
+ "loss": 0.43691688776016235,
+ "step": 1275
+ },
+ {
+ "epoch": 0.6694648478488983,
+ "grad_norm": 0.1615634942529431,
+ "learning_rate": 1.6079144187977923e-05,
+ "loss": 0.4412601590156555,
+ "step": 1276
+ },
+ {
+ "epoch": 0.6699895068205667,
+ "grad_norm": 0.15440022842754444,
+ "learning_rate": 1.604720658068184e-05,
+ "loss": 0.4523273706436157,
+ "step": 1277
+ },
+ {
+ "epoch": 0.670514165792235,
+ "grad_norm": 0.1634985665776242,
+ "learning_rate": 1.6015300088061664e-05,
+ "loss": 0.46264398097991943,
+ "step": 1278
+ },
+ {
+ "epoch": 0.6710388247639034,
+ "grad_norm": 0.15037886370689527,
+ "learning_rate": 1.598342479680018e-05,
+ "loss": 0.451465904712677,
+ "step": 1279
+ },
+ {
+ "epoch": 0.6715634837355718,
+ "grad_norm": 0.16895242437746613,
+ "learning_rate": 1.595158079349542e-05,
+ "loss": 0.4466145634651184,
+ "step": 1280
+ },
+ {
+ "epoch": 0.6720881427072403,
+ "grad_norm": 0.1629548326999642,
+ "learning_rate": 1.5919768164660425e-05,
+ "loss": 0.45415782928466797,
+ "step": 1281
+ },
+ {
+ "epoch": 0.6726128016789087,
+ "grad_norm": 0.1563872102268991,
+ "learning_rate": 1.588798699672298e-05,
+ "loss": 0.4386737048625946,
+ "step": 1282
+ },
+ {
+ "epoch": 0.6731374606505771,
+ "grad_norm": 0.16038172773344717,
+ "learning_rate": 1.5856237376025403e-05,
+ "loss": 0.446260929107666,
+ "step": 1283
+ },
+ {
+ "epoch": 0.6736621196222455,
+ "grad_norm": 0.16189780714407623,
+ "learning_rate": 1.582451938882431e-05,
+ "loss": 0.45458072423934937,
+ "step": 1284
+ },
+ {
+ "epoch": 0.674186778593914,
+ "grad_norm": 0.15204521785049285,
+ "learning_rate": 1.5792833121290367e-05,
+ "loss": 0.4425617754459381,
+ "step": 1285
+ },
+ {
+ "epoch": 0.6747114375655824,
+ "grad_norm": 0.16237376704792778,
+ "learning_rate": 1.5761178659508063e-05,
+ "loss": 0.44631800055503845,
+ "step": 1286
+ },
+ {
+ "epoch": 0.6752360965372508,
+ "grad_norm": 0.1582063618466069,
+ "learning_rate": 1.57295560894755e-05,
+ "loss": 0.44280534982681274,
+ "step": 1287
+ },
+ {
+ "epoch": 0.6757607555089192,
+ "grad_norm": 0.16587032349847394,
+ "learning_rate": 1.5697965497104116e-05,
+ "loss": 0.4495997130870819,
+ "step": 1288
+ },
+ {
+ "epoch": 0.6762854144805877,
+ "grad_norm": 0.1557269262755563,
+ "learning_rate": 1.566640696821848e-05,
+ "loss": 0.4493359625339508,
+ "step": 1289
+ },
+ {
+ "epoch": 0.6768100734522561,
+ "grad_norm": 0.16038490357012256,
+ "learning_rate": 1.5634880588556043e-05,
+ "loss": 0.4447544813156128,
+ "step": 1290
+ },
+ {
+ "epoch": 0.6773347324239245,
+ "grad_norm": 0.15355201908546406,
+ "learning_rate": 1.560338644376692e-05,
+ "loss": 0.4468432664871216,
+ "step": 1291
+ },
+ {
+ "epoch": 0.6778593913955928,
+ "grad_norm": 0.15077009515589776,
+ "learning_rate": 1.5571924619413668e-05,
+ "loss": 0.44959044456481934,
+ "step": 1292
+ },
+ {
+ "epoch": 0.6783840503672612,
+ "grad_norm": 0.1583038146613965,
+ "learning_rate": 1.5540495200971012e-05,
+ "loss": 0.42450425028800964,
+ "step": 1293
+ },
+ {
+ "epoch": 0.6789087093389297,
+ "grad_norm": 0.15950687190598087,
+ "learning_rate": 1.5509098273825635e-05,
+ "loss": 0.4527238607406616,
+ "step": 1294
+ },
+ {
+ "epoch": 0.6794333683105981,
+ "grad_norm": 0.1526262208847808,
+ "learning_rate": 1.5477733923275968e-05,
+ "loss": 0.4379724860191345,
+ "step": 1295
+ },
+ {
+ "epoch": 0.6799580272822665,
+ "grad_norm": 0.15541579650021778,
+ "learning_rate": 1.5446402234531934e-05,
+ "loss": 0.4462166130542755,
+ "step": 1296
+ },
+ {
+ "epoch": 0.6804826862539349,
+ "grad_norm": 0.15483930734313786,
+ "learning_rate": 1.5415103292714718e-05,
+ "loss": 0.4390607178211212,
+ "step": 1297
+ },
+ {
+ "epoch": 0.6810073452256034,
+ "grad_norm": 0.1578005731149154,
+ "learning_rate": 1.5383837182856532e-05,
+ "loss": 0.44462254643440247,
+ "step": 1298
+ },
+ {
+ "epoch": 0.6815320041972718,
+ "grad_norm": 0.15830183884958823,
+ "learning_rate": 1.5352603989900406e-05,
+ "loss": 0.4563537538051605,
+ "step": 1299
+ },
+ {
+ "epoch": 0.6820566631689402,
+ "grad_norm": 0.15605472163445522,
+ "learning_rate": 1.5321403798699923e-05,
+ "loss": 0.43852412700653076,
+ "step": 1300
+ },
+ {
+ "epoch": 0.6825813221406086,
+ "grad_norm": 0.15770470227141278,
+ "learning_rate": 1.5290236694019017e-05,
+ "loss": 0.4438081979751587,
+ "step": 1301
+ },
+ {
+ "epoch": 0.683105981112277,
+ "grad_norm": 0.16175339845012077,
+ "learning_rate": 1.5259102760531747e-05,
+ "loss": 0.46670734882354736,
+ "step": 1302
+ },
+ {
+ "epoch": 0.6836306400839455,
+ "grad_norm": 0.16359504716154477,
+ "learning_rate": 1.5228002082822036e-05,
+ "loss": 0.4590187668800354,
+ "step": 1303
+ },
+ {
+ "epoch": 0.6841552990556139,
+ "grad_norm": 0.15422879500254713,
+ "learning_rate": 1.5196934745383457e-05,
+ "loss": 0.45287445187568665,
+ "step": 1304
+ },
+ {
+ "epoch": 0.6846799580272823,
+ "grad_norm": 0.15929902143574254,
+ "learning_rate": 1.5165900832619015e-05,
+ "loss": 0.44286859035491943,
+ "step": 1305
+ },
+ {
+ "epoch": 0.6852046169989506,
+ "grad_norm": 0.14838868714016124,
+ "learning_rate": 1.5134900428840904e-05,
+ "loss": 0.45155346393585205,
+ "step": 1306
+ },
+ {
+ "epoch": 0.6857292759706191,
+ "grad_norm": 0.15494288640397672,
+ "learning_rate": 1.5103933618270283e-05,
+ "loss": 0.44010141491889954,
+ "step": 1307
+ },
+ {
+ "epoch": 0.6862539349422875,
+ "grad_norm": 0.1686262360598338,
+ "learning_rate": 1.5073000485037042e-05,
+ "loss": 0.446555495262146,
+ "step": 1308
+ },
+ {
+ "epoch": 0.6867785939139559,
+ "grad_norm": 0.16122110823982602,
+ "learning_rate": 1.5042101113179579e-05,
+ "loss": 0.44640374183654785,
+ "step": 1309
+ },
+ {
+ "epoch": 0.6873032528856243,
+ "grad_norm": 0.16849937797919867,
+ "learning_rate": 1.5011235586644572e-05,
+ "loss": 0.4465785026550293,
+ "step": 1310
+ },
+ {
+ "epoch": 0.6878279118572928,
+ "grad_norm": 0.16062971262800754,
+ "learning_rate": 1.498040398928674e-05,
+ "loss": 0.445321649312973,
+ "step": 1311
+ },
+ {
+ "epoch": 0.6883525708289612,
+ "grad_norm": 0.1608953185993034,
+ "learning_rate": 1.4949606404868638e-05,
+ "loss": 0.46518227458000183,
+ "step": 1312
+ },
+ {
+ "epoch": 0.6888772298006296,
+ "grad_norm": 0.15880255672830915,
+ "learning_rate": 1.491884291706041e-05,
+ "loss": 0.44582217931747437,
+ "step": 1313
+ },
+ {
+ "epoch": 0.689401888772298,
+ "grad_norm": 0.16715145200789822,
+ "learning_rate": 1.4888113609439563e-05,
+ "loss": 0.4435983896255493,
+ "step": 1314
+ },
+ {
+ "epoch": 0.6899265477439664,
+ "grad_norm": 0.15263740340694454,
+ "learning_rate": 1.485741856549075e-05,
+ "loss": 0.4405066967010498,
+ "step": 1315
+ },
+ {
+ "epoch": 0.6904512067156349,
+ "grad_norm": 0.16247034518813966,
+ "learning_rate": 1.4826757868605517e-05,
+ "loss": 0.4544215500354767,
+ "step": 1316
+ },
+ {
+ "epoch": 0.6909758656873033,
+ "grad_norm": 0.15649146705978162,
+ "learning_rate": 1.4796131602082142e-05,
+ "loss": 0.4454413056373596,
+ "step": 1317
+ },
+ {
+ "epoch": 0.6915005246589717,
+ "grad_norm": 0.16286718636058178,
+ "learning_rate": 1.4765539849125318e-05,
+ "loss": 0.4497183561325073,
+ "step": 1318
+ },
+ {
+ "epoch": 0.6920251836306401,
+ "grad_norm": 0.15812342085005582,
+ "learning_rate": 1.473498269284599e-05,
+ "loss": 0.445032000541687,
+ "step": 1319
+ },
+ {
+ "epoch": 0.6925498426023085,
+ "grad_norm": 0.1558101290511678,
+ "learning_rate": 1.4704460216261123e-05,
+ "loss": 0.45769619941711426,
+ "step": 1320
+ },
+ {
+ "epoch": 0.6930745015739769,
+ "grad_norm": 0.15642889059495207,
+ "learning_rate": 1.4673972502293428e-05,
+ "loss": 0.44788360595703125,
+ "step": 1321
+ },
+ {
+ "epoch": 0.6935991605456453,
+ "grad_norm": 0.15722513518113074,
+ "learning_rate": 1.4643519633771222e-05,
+ "loss": 0.449631929397583,
+ "step": 1322
+ },
+ {
+ "epoch": 0.6941238195173137,
+ "grad_norm": 0.15587633748081542,
+ "learning_rate": 1.4613101693428125e-05,
+ "loss": 0.4450187087059021,
+ "step": 1323
+ },
+ {
+ "epoch": 0.6946484784889821,
+ "grad_norm": 0.1577107074882479,
+ "learning_rate": 1.458271876390287e-05,
+ "loss": 0.45108890533447266,
+ "step": 1324
+ },
+ {
+ "epoch": 0.6951731374606506,
+ "grad_norm": 0.15649839486990696,
+ "learning_rate": 1.4552370927739073e-05,
+ "loss": 0.46430301666259766,
+ "step": 1325
+ },
+ {
+ "epoch": 0.695697796432319,
+ "grad_norm": 0.18573600420690636,
+ "learning_rate": 1.4522058267385008e-05,
+ "loss": 0.4701972007751465,
+ "step": 1326
+ },
+ {
+ "epoch": 0.6962224554039874,
+ "grad_norm": 0.16296497084037542,
+ "learning_rate": 1.4491780865193402e-05,
+ "loss": 0.4607710838317871,
+ "step": 1327
+ },
+ {
+ "epoch": 0.6967471143756558,
+ "grad_norm": 0.161124007217343,
+ "learning_rate": 1.446153880342117e-05,
+ "loss": 0.45095115900039673,
+ "step": 1328
+ },
+ {
+ "epoch": 0.6972717733473243,
+ "grad_norm": 0.1537884039166766,
+ "learning_rate": 1.4431332164229222e-05,
+ "loss": 0.424068808555603,
+ "step": 1329
+ },
+ {
+ "epoch": 0.6977964323189927,
+ "grad_norm": 0.16588834503456087,
+ "learning_rate": 1.4401161029682237e-05,
+ "loss": 0.4622839093208313,
+ "step": 1330
+ },
+ {
+ "epoch": 0.6983210912906611,
+ "grad_norm": 0.16100510604015544,
+ "learning_rate": 1.4371025481748435e-05,
+ "loss": 0.4660705626010895,
+ "step": 1331
+ },
+ {
+ "epoch": 0.6988457502623295,
+ "grad_norm": 0.15559888940590075,
+ "learning_rate": 1.4340925602299337e-05,
+ "loss": 0.4484976530075073,
+ "step": 1332
+ },
+ {
+ "epoch": 0.6993704092339978,
+ "grad_norm": 0.15503821966213488,
+ "learning_rate": 1.4310861473109616e-05,
+ "loss": 0.45636600255966187,
+ "step": 1333
+ },
+ {
+ "epoch": 0.6998950682056663,
+ "grad_norm": 0.15539074896109334,
+ "learning_rate": 1.4280833175856743e-05,
+ "loss": 0.45863068103790283,
+ "step": 1334
+ },
+ {
+ "epoch": 0.7004197271773347,
+ "grad_norm": 0.15664821009154872,
+ "learning_rate": 1.4250840792120896e-05,
+ "loss": 0.44043463468551636,
+ "step": 1335
+ },
+ {
+ "epoch": 0.7009443861490031,
+ "grad_norm": 0.15825478808520285,
+ "learning_rate": 1.4220884403384655e-05,
+ "loss": 0.46522051095962524,
+ "step": 1336
+ },
+ {
+ "epoch": 0.7014690451206715,
+ "grad_norm": 0.16004268568141544,
+ "learning_rate": 1.4190964091032843e-05,
+ "loss": 0.46114686131477356,
+ "step": 1337
+ },
+ {
+ "epoch": 0.70199370409234,
+ "grad_norm": 0.16413139949945635,
+ "learning_rate": 1.4161079936352234e-05,
+ "loss": 0.4457515478134155,
+ "step": 1338
+ },
+ {
+ "epoch": 0.7025183630640084,
+ "grad_norm": 0.1538890577972702,
+ "learning_rate": 1.4131232020531398e-05,
+ "loss": 0.4616093337535858,
+ "step": 1339
+ },
+ {
+ "epoch": 0.7030430220356768,
+ "grad_norm": 0.16240737854323284,
+ "learning_rate": 1.4101420424660434e-05,
+ "loss": 0.44631537795066833,
+ "step": 1340
+ },
+ {
+ "epoch": 0.7035676810073452,
+ "grad_norm": 0.15984708603406073,
+ "learning_rate": 1.4071645229730762e-05,
+ "loss": 0.4366717040538788,
+ "step": 1341
+ },
+ {
+ "epoch": 0.7040923399790137,
+ "grad_norm": 0.15149695695635904,
+ "learning_rate": 1.4041906516634946e-05,
+ "loss": 0.44129428267478943,
+ "step": 1342
+ },
+ {
+ "epoch": 0.7046169989506821,
+ "grad_norm": 0.15041167501768415,
+ "learning_rate": 1.4012204366166399e-05,
+ "loss": 0.43654775619506836,
+ "step": 1343
+ },
+ {
+ "epoch": 0.7051416579223505,
+ "grad_norm": 0.1596097627963756,
+ "learning_rate": 1.3982538859019217e-05,
+ "loss": 0.4652445614337921,
+ "step": 1344
+ },
+ {
+ "epoch": 0.7056663168940189,
+ "grad_norm": 0.15621704205027884,
+ "learning_rate": 1.3952910075787941e-05,
+ "loss": 0.44577041268348694,
+ "step": 1345
+ },
+ {
+ "epoch": 0.7061909758656874,
+ "grad_norm": 0.16073041098008783,
+ "learning_rate": 1.3923318096967341e-05,
+ "loss": 0.4391976594924927,
+ "step": 1346
+ },
+ {
+ "epoch": 0.7067156348373557,
+ "grad_norm": 0.14793853946372448,
+ "learning_rate": 1.3893763002952204e-05,
+ "loss": 0.4463799297809601,
+ "step": 1347
+ },
+ {
+ "epoch": 0.7072402938090241,
+ "grad_norm": 0.16176080358867354,
+ "learning_rate": 1.3864244874037097e-05,
+ "loss": 0.45183271169662476,
+ "step": 1348
+ },
+ {
+ "epoch": 0.7077649527806925,
+ "grad_norm": 0.16372771179026924,
+ "learning_rate": 1.3834763790416171e-05,
+ "loss": 0.44205576181411743,
+ "step": 1349
+ },
+ {
+ "epoch": 0.7082896117523609,
+ "grad_norm": 0.15135222516127125,
+ "learning_rate": 1.3805319832182937e-05,
+ "loss": 0.4390261173248291,
+ "step": 1350
+ },
+ {
+ "epoch": 0.7088142707240294,
+ "grad_norm": 0.15493183225792365,
+ "learning_rate": 1.377591307933002e-05,
+ "loss": 0.45837974548339844,
+ "step": 1351
+ },
+ {
+ "epoch": 0.7093389296956978,
+ "grad_norm": 0.16405906906516504,
+ "learning_rate": 1.3746543611749008e-05,
+ "loss": 0.45032668113708496,
+ "step": 1352
+ },
+ {
+ "epoch": 0.7098635886673662,
+ "grad_norm": 0.15725677976524505,
+ "learning_rate": 1.3717211509230155e-05,
+ "loss": 0.4486224949359894,
+ "step": 1353
+ },
+ {
+ "epoch": 0.7103882476390346,
+ "grad_norm": 0.15781469066983164,
+ "learning_rate": 1.3687916851462227e-05,
+ "loss": 0.4502638578414917,
+ "step": 1354
+ },
+ {
+ "epoch": 0.710912906610703,
+ "grad_norm": 0.16115007729054298,
+ "learning_rate": 1.3658659718032241e-05,
+ "loss": 0.4508914351463318,
+ "step": 1355
+ },
+ {
+ "epoch": 0.7114375655823715,
+ "grad_norm": 0.16755060475968744,
+ "learning_rate": 1.3629440188425286e-05,
+ "loss": 0.45653945207595825,
+ "step": 1356
+ },
+ {
+ "epoch": 0.7119622245540399,
+ "grad_norm": 0.1574994294756123,
+ "learning_rate": 1.3600258342024268e-05,
+ "loss": 0.4349849820137024,
+ "step": 1357
+ },
+ {
+ "epoch": 0.7124868835257083,
+ "grad_norm": 0.16104440922657715,
+ "learning_rate": 1.3571114258109758e-05,
+ "loss": 0.4501233696937561,
+ "step": 1358
+ },
+ {
+ "epoch": 0.7130115424973767,
+ "grad_norm": 0.15163184290939022,
+ "learning_rate": 1.3542008015859691e-05,
+ "loss": 0.44171035289764404,
+ "step": 1359
+ },
+ {
+ "epoch": 0.7135362014690452,
+ "grad_norm": 0.15955896674456582,
+ "learning_rate": 1.3512939694349211e-05,
+ "loss": 0.4558671712875366,
+ "step": 1360
+ },
+ {
+ "epoch": 0.7140608604407135,
+ "grad_norm": 0.1589376751536802,
+ "learning_rate": 1.3483909372550446e-05,
+ "loss": 0.4430835247039795,
+ "step": 1361
+ },
+ {
+ "epoch": 0.7145855194123819,
+ "grad_norm": 0.15979088397822155,
+ "learning_rate": 1.3454917129332279e-05,
+ "loss": 0.4426245391368866,
+ "step": 1362
+ },
+ {
+ "epoch": 0.7151101783840503,
+ "grad_norm": 0.15493179714072167,
+ "learning_rate": 1.3425963043460143e-05,
+ "loss": 0.4550474286079407,
+ "step": 1363
+ },
+ {
+ "epoch": 0.7156348373557188,
+ "grad_norm": 0.15396498968995032,
+ "learning_rate": 1.3397047193595808e-05,
+ "loss": 0.44220536947250366,
+ "step": 1364
+ },
+ {
+ "epoch": 0.7161594963273872,
+ "grad_norm": 0.1489346822667273,
+ "learning_rate": 1.336816965829717e-05,
+ "loss": 0.4437323212623596,
+ "step": 1365
+ },
+ {
+ "epoch": 0.7166841552990556,
+ "grad_norm": 0.16311964931734477,
+ "learning_rate": 1.333933051601802e-05,
+ "loss": 0.45308777689933777,
+ "step": 1366
+ },
+ {
+ "epoch": 0.717208814270724,
+ "grad_norm": 0.1534354619021179,
+ "learning_rate": 1.331052984510785e-05,
+ "loss": 0.4334886372089386,
+ "step": 1367
+ },
+ {
+ "epoch": 0.7177334732423925,
+ "grad_norm": 0.15541089758065404,
+ "learning_rate": 1.328176772381165e-05,
+ "loss": 0.45248323678970337,
+ "step": 1368
+ },
+ {
+ "epoch": 0.7182581322140609,
+ "grad_norm": 0.1617921305700596,
+ "learning_rate": 1.3253044230269651e-05,
+ "loss": 0.44092947244644165,
+ "step": 1369
+ },
+ {
+ "epoch": 0.7187827911857293,
+ "grad_norm": 0.15047353793760151,
+ "learning_rate": 1.3224359442517165e-05,
+ "loss": 0.4424414336681366,
+ "step": 1370
+ },
+ {
+ "epoch": 0.7193074501573977,
+ "grad_norm": 0.15958089102258743,
+ "learning_rate": 1.319571343848433e-05,
+ "loss": 0.4458580017089844,
+ "step": 1371
+ },
+ {
+ "epoch": 0.7198321091290661,
+ "grad_norm": 0.15349258306968733,
+ "learning_rate": 1.3167106295995924e-05,
+ "loss": 0.4285386800765991,
+ "step": 1372
+ },
+ {
+ "epoch": 0.7203567681007346,
+ "grad_norm": 0.16042529563759886,
+ "learning_rate": 1.3138538092771164e-05,
+ "loss": 0.4497690200805664,
+ "step": 1373
+ },
+ {
+ "epoch": 0.720881427072403,
+ "grad_norm": 0.1565113736588754,
+ "learning_rate": 1.3110008906423454e-05,
+ "loss": 0.43562617897987366,
+ "step": 1374
+ },
+ {
+ "epoch": 0.7214060860440713,
+ "grad_norm": 0.15177291086871506,
+ "learning_rate": 1.3081518814460215e-05,
+ "loss": 0.44456908106803894,
+ "step": 1375
+ },
+ {
+ "epoch": 0.7219307450157397,
+ "grad_norm": 0.15928490480439225,
+ "learning_rate": 1.3053067894282622e-05,
+ "loss": 0.46170780062675476,
+ "step": 1376
+ },
+ {
+ "epoch": 0.7224554039874082,
+ "grad_norm": 0.1554814469217892,
+ "learning_rate": 1.302465622318549e-05,
+ "loss": 0.4422169327735901,
+ "step": 1377
+ },
+ {
+ "epoch": 0.7229800629590766,
+ "grad_norm": 0.15375370214966386,
+ "learning_rate": 1.2996283878356947e-05,
+ "loss": 0.4564763903617859,
+ "step": 1378
+ },
+ {
+ "epoch": 0.723504721930745,
+ "grad_norm": 0.15942841969395802,
+ "learning_rate": 1.2967950936878311e-05,
+ "loss": 0.45121875405311584,
+ "step": 1379
+ },
+ {
+ "epoch": 0.7240293809024134,
+ "grad_norm": 0.15731308716774903,
+ "learning_rate": 1.2939657475723844e-05,
+ "loss": 0.4285027086734772,
+ "step": 1380
+ },
+ {
+ "epoch": 0.7245540398740818,
+ "grad_norm": 0.15997753613932147,
+ "learning_rate": 1.291140357176054e-05,
+ "loss": 0.4566386342048645,
+ "step": 1381
+ },
+ {
+ "epoch": 0.7250786988457503,
+ "grad_norm": 0.15333715267473727,
+ "learning_rate": 1.2883189301747927e-05,
+ "loss": 0.45093846321105957,
+ "step": 1382
+ },
+ {
+ "epoch": 0.7256033578174187,
+ "grad_norm": 0.1514510304709061,
+ "learning_rate": 1.2855014742337874e-05,
+ "loss": 0.44256967306137085,
+ "step": 1383
+ },
+ {
+ "epoch": 0.7261280167890871,
+ "grad_norm": 0.15624301298995746,
+ "learning_rate": 1.2826879970074346e-05,
+ "loss": 0.4361875653266907,
+ "step": 1384
+ },
+ {
+ "epoch": 0.7266526757607555,
+ "grad_norm": 0.15767850122006694,
+ "learning_rate": 1.279878506139322e-05,
+ "loss": 0.4477671980857849,
+ "step": 1385
+ },
+ {
+ "epoch": 0.727177334732424,
+ "grad_norm": 0.15537676969694064,
+ "learning_rate": 1.2770730092622072e-05,
+ "loss": 0.4339832663536072,
+ "step": 1386
+ },
+ {
+ "epoch": 0.7277019937040924,
+ "grad_norm": 0.15944113633281293,
+ "learning_rate": 1.2742715139979965e-05,
+ "loss": 0.45407935976982117,
+ "step": 1387
+ },
+ {
+ "epoch": 0.7282266526757608,
+ "grad_norm": 0.16763519518606218,
+ "learning_rate": 1.271474027957727e-05,
+ "loss": 0.46442392468452454,
+ "step": 1388
+ },
+ {
+ "epoch": 0.7287513116474291,
+ "grad_norm": 0.1564572774601902,
+ "learning_rate": 1.2686805587415423e-05,
+ "loss": 0.4463968276977539,
+ "step": 1389
+ },
+ {
+ "epoch": 0.7292759706190975,
+ "grad_norm": 0.16384626314986128,
+ "learning_rate": 1.2658911139386709e-05,
+ "loss": 0.45519885420799255,
+ "step": 1390
+ },
+ {
+ "epoch": 0.729800629590766,
+ "grad_norm": 0.1562023802820964,
+ "learning_rate": 1.2631057011274106e-05,
+ "loss": 0.4365222454071045,
+ "step": 1391
+ },
+ {
+ "epoch": 0.7303252885624344,
+ "grad_norm": 0.15597287232940926,
+ "learning_rate": 1.2603243278751042e-05,
+ "loss": 0.4328266680240631,
+ "step": 1392
+ },
+ {
+ "epoch": 0.7308499475341028,
+ "grad_norm": 0.14762833706132167,
+ "learning_rate": 1.2575470017381214e-05,
+ "loss": 0.43902426958084106,
+ "step": 1393
+ },
+ {
+ "epoch": 0.7313746065057712,
+ "grad_norm": 0.1688371515856315,
+ "learning_rate": 1.2547737302618351e-05,
+ "loss": 0.4649963676929474,
+ "step": 1394
+ },
+ {
+ "epoch": 0.7318992654774397,
+ "grad_norm": 0.15501514509586015,
+ "learning_rate": 1.2520045209806024e-05,
+ "loss": 0.4501558542251587,
+ "step": 1395
+ },
+ {
+ "epoch": 0.7324239244491081,
+ "grad_norm": 0.15516748917693537,
+ "learning_rate": 1.2492393814177457e-05,
+ "loss": 0.4405498206615448,
+ "step": 1396
+ },
+ {
+ "epoch": 0.7329485834207765,
+ "grad_norm": 0.17115400468199243,
+ "learning_rate": 1.2464783190855295e-05,
+ "loss": 0.4445379674434662,
+ "step": 1397
+ },
+ {
+ "epoch": 0.7334732423924449,
+ "grad_norm": 0.15966386341604186,
+ "learning_rate": 1.243721341485143e-05,
+ "loss": 0.4769889712333679,
+ "step": 1398
+ },
+ {
+ "epoch": 0.7339979013641134,
+ "grad_norm": 0.15614706359781472,
+ "learning_rate": 1.2409684561066766e-05,
+ "loss": 0.43643918633461,
+ "step": 1399
+ },
+ {
+ "epoch": 0.7345225603357818,
+ "grad_norm": 0.17467968960064226,
+ "learning_rate": 1.238219670429104e-05,
+ "loss": 0.45413005352020264,
+ "step": 1400
+ },
+ {
+ "epoch": 0.7350472193074502,
+ "grad_norm": 0.1544517430810819,
+ "learning_rate": 1.2354749919202598e-05,
+ "loss": 0.45218425989151,
+ "step": 1401
+ },
+ {
+ "epoch": 0.7355718782791185,
+ "grad_norm": 0.16551083673374484,
+ "learning_rate": 1.2327344280368217e-05,
+ "loss": 0.4459628462791443,
+ "step": 1402
+ },
+ {
+ "epoch": 0.7360965372507869,
+ "grad_norm": 0.17614490459613263,
+ "learning_rate": 1.2299979862242882e-05,
+ "loss": 0.4406280815601349,
+ "step": 1403
+ },
+ {
+ "epoch": 0.7366211962224554,
+ "grad_norm": 0.15554925394435032,
+ "learning_rate": 1.2272656739169588e-05,
+ "loss": 0.4380159378051758,
+ "step": 1404
+ },
+ {
+ "epoch": 0.7371458551941238,
+ "grad_norm": 0.16475466799870916,
+ "learning_rate": 1.2245374985379147e-05,
+ "loss": 0.4442812204360962,
+ "step": 1405
+ },
+ {
+ "epoch": 0.7376705141657922,
+ "grad_norm": 0.16877301168875541,
+ "learning_rate": 1.2218134674989978e-05,
+ "loss": 0.4589429199695587,
+ "step": 1406
+ },
+ {
+ "epoch": 0.7381951731374606,
+ "grad_norm": 0.1583278589457438,
+ "learning_rate": 1.2190935882007894e-05,
+ "loss": 0.44896745681762695,
+ "step": 1407
+ },
+ {
+ "epoch": 0.7387198321091291,
+ "grad_norm": 0.1582022541656192,
+ "learning_rate": 1.2163778680325947e-05,
+ "loss": 0.4367997646331787,
+ "step": 1408
+ },
+ {
+ "epoch": 0.7392444910807975,
+ "grad_norm": 0.1669573982300068,
+ "learning_rate": 1.2136663143724163e-05,
+ "loss": 0.4503386616706848,
+ "step": 1409
+ },
+ {
+ "epoch": 0.7397691500524659,
+ "grad_norm": 0.1637877523924539,
+ "learning_rate": 1.2109589345869387e-05,
+ "loss": 0.4660584330558777,
+ "step": 1410
+ },
+ {
+ "epoch": 0.7402938090241343,
+ "grad_norm": 0.1640136159943701,
+ "learning_rate": 1.2082557360315065e-05,
+ "loss": 0.45206528902053833,
+ "step": 1411
+ },
+ {
+ "epoch": 0.7408184679958028,
+ "grad_norm": 0.1596580168682554,
+ "learning_rate": 1.2055567260501054e-05,
+ "loss": 0.4405035376548767,
+ "step": 1412
+ },
+ {
+ "epoch": 0.7413431269674712,
+ "grad_norm": 0.15224511741945293,
+ "learning_rate": 1.2028619119753402e-05,
+ "loss": 0.42810291051864624,
+ "step": 1413
+ },
+ {
+ "epoch": 0.7418677859391396,
+ "grad_norm": 0.15631381931817098,
+ "learning_rate": 1.2001713011284193e-05,
+ "loss": 0.4421626925468445,
+ "step": 1414
+ },
+ {
+ "epoch": 0.742392444910808,
+ "grad_norm": 0.15721819981265503,
+ "learning_rate": 1.1974849008191288e-05,
+ "loss": 0.4540696144104004,
+ "step": 1415
+ },
+ {
+ "epoch": 0.7429171038824763,
+ "grad_norm": 0.15904478578744463,
+ "learning_rate": 1.1948027183458182e-05,
+ "loss": 0.4601830244064331,
+ "step": 1416
+ },
+ {
+ "epoch": 0.7434417628541448,
+ "grad_norm": 0.1530755605235123,
+ "learning_rate": 1.1921247609953747e-05,
+ "loss": 0.4552589952945709,
+ "step": 1417
+ },
+ {
+ "epoch": 0.7439664218258132,
+ "grad_norm": 0.15375751072258462,
+ "learning_rate": 1.1894510360432106e-05,
+ "loss": 0.42965373396873474,
+ "step": 1418
+ },
+ {
+ "epoch": 0.7444910807974816,
+ "grad_norm": 0.15678686023937333,
+ "learning_rate": 1.1867815507532377e-05,
+ "loss": 0.44684016704559326,
+ "step": 1419
+ },
+ {
+ "epoch": 0.74501573976915,
+ "grad_norm": 0.15486952983783606,
+ "learning_rate": 1.1841163123778502e-05,
+ "loss": 0.44415023922920227,
+ "step": 1420
+ },
+ {
+ "epoch": 0.7455403987408185,
+ "grad_norm": 0.15342611490545593,
+ "learning_rate": 1.1814553281579034e-05,
+ "loss": 0.4501304626464844,
+ "step": 1421
+ },
+ {
+ "epoch": 0.7460650577124869,
+ "grad_norm": 0.15374260012788166,
+ "learning_rate": 1.1787986053226954e-05,
+ "loss": 0.4521749019622803,
+ "step": 1422
+ },
+ {
+ "epoch": 0.7465897166841553,
+ "grad_norm": 0.1543553258370361,
+ "learning_rate": 1.176146151089949e-05,
+ "loss": 0.44753748178482056,
+ "step": 1423
+ },
+ {
+ "epoch": 0.7471143756558237,
+ "grad_norm": 0.15921670193848306,
+ "learning_rate": 1.1734979726657878e-05,
+ "loss": 0.4473114311695099,
+ "step": 1424
+ },
+ {
+ "epoch": 0.7476390346274921,
+ "grad_norm": 0.15023498686186612,
+ "learning_rate": 1.17085407724472e-05,
+ "loss": 0.4568778872489929,
+ "step": 1425
+ },
+ {
+ "epoch": 0.7481636935991606,
+ "grad_norm": 0.14601156920274583,
+ "learning_rate": 1.1682144720096168e-05,
+ "loss": 0.4389464855194092,
+ "step": 1426
+ },
+ {
+ "epoch": 0.748688352570829,
+ "grad_norm": 0.1520548095149398,
+ "learning_rate": 1.1655791641316954e-05,
+ "loss": 0.440733402967453,
+ "step": 1427
+ },
+ {
+ "epoch": 0.7492130115424974,
+ "grad_norm": 0.1591915163097993,
+ "learning_rate": 1.1629481607704963e-05,
+ "loss": 0.46245887875556946,
+ "step": 1428
+ },
+ {
+ "epoch": 0.7497376705141658,
+ "grad_norm": 0.14867295658447263,
+ "learning_rate": 1.1603214690738683e-05,
+ "loss": 0.4332764744758606,
+ "step": 1429
+ },
+ {
+ "epoch": 0.7502623294858342,
+ "grad_norm": 0.1498957762903134,
+ "learning_rate": 1.1576990961779447e-05,
+ "loss": 0.45215466618537903,
+ "step": 1430
+ },
+ {
+ "epoch": 0.7507869884575026,
+ "grad_norm": 0.15430555803491824,
+ "learning_rate": 1.1550810492071243e-05,
+ "loss": 0.4454396963119507,
+ "step": 1431
+ },
+ {
+ "epoch": 0.751311647429171,
+ "grad_norm": 0.1517461021936533,
+ "learning_rate": 1.1524673352740543e-05,
+ "loss": 0.43869835138320923,
+ "step": 1432
+ },
+ {
+ "epoch": 0.7518363064008394,
+ "grad_norm": 0.14541657994692936,
+ "learning_rate": 1.1498579614796121e-05,
+ "loss": 0.4373054504394531,
+ "step": 1433
+ },
+ {
+ "epoch": 0.7523609653725079,
+ "grad_norm": 0.15539513763786977,
+ "learning_rate": 1.1472529349128812e-05,
+ "loss": 0.45451128482818604,
+ "step": 1434
+ },
+ {
+ "epoch": 0.7528856243441763,
+ "grad_norm": 0.1482043113690946,
+ "learning_rate": 1.144652262651136e-05,
+ "loss": 0.43644773960113525,
+ "step": 1435
+ },
+ {
+ "epoch": 0.7534102833158447,
+ "grad_norm": 0.16293628094190815,
+ "learning_rate": 1.1420559517598205e-05,
+ "loss": 0.4687269330024719,
+ "step": 1436
+ },
+ {
+ "epoch": 0.7539349422875131,
+ "grad_norm": 0.1553393974390287,
+ "learning_rate": 1.1394640092925302e-05,
+ "loss": 0.4501398205757141,
+ "step": 1437
+ },
+ {
+ "epoch": 0.7544596012591815,
+ "grad_norm": 0.16214837694535802,
+ "learning_rate": 1.1368764422909919e-05,
+ "loss": 0.4542577862739563,
+ "step": 1438
+ },
+ {
+ "epoch": 0.75498426023085,
+ "grad_norm": 0.15838878372428297,
+ "learning_rate": 1.1342932577850473e-05,
+ "loss": 0.46228161454200745,
+ "step": 1439
+ },
+ {
+ "epoch": 0.7555089192025184,
+ "grad_norm": 0.15090419497122687,
+ "learning_rate": 1.13171446279263e-05,
+ "loss": 0.4495932161808014,
+ "step": 1440
+ },
+ {
+ "epoch": 0.7560335781741868,
+ "grad_norm": 0.1543376786458394,
+ "learning_rate": 1.1291400643197484e-05,
+ "loss": 0.451630175113678,
+ "step": 1441
+ },
+ {
+ "epoch": 0.7565582371458552,
+ "grad_norm": 0.1539470751610696,
+ "learning_rate": 1.1265700693604668e-05,
+ "loss": 0.447547972202301,
+ "step": 1442
+ },
+ {
+ "epoch": 0.7570828961175237,
+ "grad_norm": 0.15761047703305445,
+ "learning_rate": 1.1240044848968856e-05,
+ "loss": 0.4516204595565796,
+ "step": 1443
+ },
+ {
+ "epoch": 0.757607555089192,
+ "grad_norm": 0.15675279683704735,
+ "learning_rate": 1.1214433178991258e-05,
+ "loss": 0.4597971737384796,
+ "step": 1444
+ },
+ {
+ "epoch": 0.7581322140608604,
+ "grad_norm": 0.15592110994904113,
+ "learning_rate": 1.1188865753253022e-05,
+ "loss": 0.443898469209671,
+ "step": 1445
+ },
+ {
+ "epoch": 0.7586568730325288,
+ "grad_norm": 0.15597501691030619,
+ "learning_rate": 1.116334264121513e-05,
+ "loss": 0.4581306278705597,
+ "step": 1446
+ },
+ {
+ "epoch": 0.7591815320041972,
+ "grad_norm": 0.15552304728589958,
+ "learning_rate": 1.1137863912218154e-05,
+ "loss": 0.4475254416465759,
+ "step": 1447
+ },
+ {
+ "epoch": 0.7597061909758657,
+ "grad_norm": 0.15759990123898035,
+ "learning_rate": 1.1112429635482118e-05,
+ "loss": 0.43585264682769775,
+ "step": 1448
+ },
+ {
+ "epoch": 0.7602308499475341,
+ "grad_norm": 0.15964657264676635,
+ "learning_rate": 1.1087039880106247e-05,
+ "loss": 0.46786290407180786,
+ "step": 1449
+ },
+ {
+ "epoch": 0.7607555089192025,
+ "grad_norm": 0.15155804051218852,
+ "learning_rate": 1.106169471506882e-05,
+ "loss": 0.42292726039886475,
+ "step": 1450
+ },
+ {
+ "epoch": 0.7612801678908709,
+ "grad_norm": 0.15519270966930884,
+ "learning_rate": 1.103639420922698e-05,
+ "loss": 0.4248972237110138,
+ "step": 1451
+ },
+ {
+ "epoch": 0.7618048268625394,
+ "grad_norm": 0.1526330511128821,
+ "learning_rate": 1.1011138431316537e-05,
+ "loss": 0.4317784309387207,
+ "step": 1452
+ },
+ {
+ "epoch": 0.7623294858342078,
+ "grad_norm": 0.15797682979272226,
+ "learning_rate": 1.0985927449951779e-05,
+ "loss": 0.44506698846817017,
+ "step": 1453
+ },
+ {
+ "epoch": 0.7628541448058762,
+ "grad_norm": 0.15479979838505462,
+ "learning_rate": 1.096076133362531e-05,
+ "loss": 0.4471631348133087,
+ "step": 1454
+ },
+ {
+ "epoch": 0.7633788037775446,
+ "grad_norm": 0.15712267274447342,
+ "learning_rate": 1.093564015070783e-05,
+ "loss": 0.4321915805339813,
+ "step": 1455
+ },
+ {
+ "epoch": 0.7639034627492131,
+ "grad_norm": 0.15857431494180016,
+ "learning_rate": 1.0910563969447969e-05,
+ "loss": 0.44455766677856445,
+ "step": 1456
+ },
+ {
+ "epoch": 0.7644281217208815,
+ "grad_norm": 0.15377057080997789,
+ "learning_rate": 1.0885532857972097e-05,
+ "loss": 0.4567924439907074,
+ "step": 1457
+ },
+ {
+ "epoch": 0.7649527806925498,
+ "grad_norm": 0.15544844005009104,
+ "learning_rate": 1.0860546884284143e-05,
+ "loss": 0.4491572976112366,
+ "step": 1458
+ },
+ {
+ "epoch": 0.7654774396642182,
+ "grad_norm": 0.16210735566479478,
+ "learning_rate": 1.08356061162654e-05,
+ "loss": 0.43823331594467163,
+ "step": 1459
+ },
+ {
+ "epoch": 0.7660020986358866,
+ "grad_norm": 0.15213559486379452,
+ "learning_rate": 1.0810710621674355e-05,
+ "loss": 0.45375770330429077,
+ "step": 1460
+ },
+ {
+ "epoch": 0.7665267576075551,
+ "grad_norm": 0.1727832306402992,
+ "learning_rate": 1.0785860468146494e-05,
+ "loss": 0.4639572501182556,
+ "step": 1461
+ },
+ {
+ "epoch": 0.7670514165792235,
+ "grad_norm": 0.1637555902013663,
+ "learning_rate": 1.076105572319412e-05,
+ "loss": 0.4355923533439636,
+ "step": 1462
+ },
+ {
+ "epoch": 0.7675760755508919,
+ "grad_norm": 0.16646936030470866,
+ "learning_rate": 1.0736296454206166e-05,
+ "loss": 0.4650861322879791,
+ "step": 1463
+ },
+ {
+ "epoch": 0.7681007345225603,
+ "grad_norm": 0.14987620615431385,
+ "learning_rate": 1.0711582728448033e-05,
+ "loss": 0.45445430278778076,
+ "step": 1464
+ },
+ {
+ "epoch": 0.7686253934942288,
+ "grad_norm": 0.16428618403547932,
+ "learning_rate": 1.0686914613061377e-05,
+ "loss": 0.46050935983657837,
+ "step": 1465
+ },
+ {
+ "epoch": 0.7691500524658972,
+ "grad_norm": 0.15052999053128632,
+ "learning_rate": 1.0662292175063949e-05,
+ "loss": 0.43385741114616394,
+ "step": 1466
+ },
+ {
+ "epoch": 0.7696747114375656,
+ "grad_norm": 0.1654949402740152,
+ "learning_rate": 1.0637715481349397e-05,
+ "loss": 0.4434410333633423,
+ "step": 1467
+ },
+ {
+ "epoch": 0.770199370409234,
+ "grad_norm": 0.15923836508490036,
+ "learning_rate": 1.0613184598687088e-05,
+ "loss": 0.4410185217857361,
+ "step": 1468
+ },
+ {
+ "epoch": 0.7707240293809025,
+ "grad_norm": 0.15813007111170696,
+ "learning_rate": 1.0588699593721957e-05,
+ "loss": 0.43588417768478394,
+ "step": 1469
+ },
+ {
+ "epoch": 0.7712486883525709,
+ "grad_norm": 0.16978817898431764,
+ "learning_rate": 1.0564260532974274e-05,
+ "loss": 0.4514843225479126,
+ "step": 1470
+ },
+ {
+ "epoch": 0.7717733473242392,
+ "grad_norm": 0.1613844852697828,
+ "learning_rate": 1.0539867482839499e-05,
+ "loss": 0.4554520845413208,
+ "step": 1471
+ },
+ {
+ "epoch": 0.7722980062959076,
+ "grad_norm": 0.1587788181984633,
+ "learning_rate": 1.0515520509588075e-05,
+ "loss": 0.43200671672821045,
+ "step": 1472
+ },
+ {
+ "epoch": 0.772822665267576,
+ "grad_norm": 0.1539930824827353,
+ "learning_rate": 1.0491219679365283e-05,
+ "loss": 0.4343116283416748,
+ "step": 1473
+ },
+ {
+ "epoch": 0.7733473242392445,
+ "grad_norm": 0.1561612537773709,
+ "learning_rate": 1.0466965058191049e-05,
+ "loss": 0.4477815330028534,
+ "step": 1474
+ },
+ {
+ "epoch": 0.7738719832109129,
+ "grad_norm": 0.169990777691657,
+ "learning_rate": 1.0442756711959745e-05,
+ "loss": 0.4476611614227295,
+ "step": 1475
+ },
+ {
+ "epoch": 0.7743966421825813,
+ "grad_norm": 0.1624313049193793,
+ "learning_rate": 1.041859470644002e-05,
+ "loss": 0.47136399149894714,
+ "step": 1476
+ },
+ {
+ "epoch": 0.7749213011542497,
+ "grad_norm": 0.15993315458994328,
+ "learning_rate": 1.0394479107274645e-05,
+ "loss": 0.4672670364379883,
+ "step": 1477
+ },
+ {
+ "epoch": 0.7754459601259182,
+ "grad_norm": 0.16755349272347217,
+ "learning_rate": 1.0370409979980294e-05,
+ "loss": 0.45135706663131714,
+ "step": 1478
+ },
+ {
+ "epoch": 0.7759706190975866,
+ "grad_norm": 0.15665059815830362,
+ "learning_rate": 1.0346387389947407e-05,
+ "loss": 0.45178359746932983,
+ "step": 1479
+ },
+ {
+ "epoch": 0.776495278069255,
+ "grad_norm": 0.1554612731452571,
+ "learning_rate": 1.0322411402439983e-05,
+ "loss": 0.4412590265274048,
+ "step": 1480
+ },
+ {
+ "epoch": 0.7770199370409234,
+ "grad_norm": 0.1532496869312065,
+ "learning_rate": 1.0298482082595408e-05,
+ "loss": 0.44104987382888794,
+ "step": 1481
+ },
+ {
+ "epoch": 0.7775445960125918,
+ "grad_norm": 0.15065868316960426,
+ "learning_rate": 1.0274599495424286e-05,
+ "loss": 0.4336603283882141,
+ "step": 1482
+ },
+ {
+ "epoch": 0.7780692549842603,
+ "grad_norm": 0.1548373981760097,
+ "learning_rate": 1.0250763705810267e-05,
+ "loss": 0.4658375680446625,
+ "step": 1483
+ },
+ {
+ "epoch": 0.7785939139559287,
+ "grad_norm": 0.15470181175866582,
+ "learning_rate": 1.0226974778509846e-05,
+ "loss": 0.45232075452804565,
+ "step": 1484
+ },
+ {
+ "epoch": 0.779118572927597,
+ "grad_norm": 0.15507371187727106,
+ "learning_rate": 1.0203232778152228e-05,
+ "loss": 0.43640458583831787,
+ "step": 1485
+ },
+ {
+ "epoch": 0.7796432318992654,
+ "grad_norm": 0.15269440736544085,
+ "learning_rate": 1.01795377692391e-05,
+ "loss": 0.4400939345359802,
+ "step": 1486
+ },
+ {
+ "epoch": 0.7801678908709339,
+ "grad_norm": 0.15610899550395085,
+ "learning_rate": 1.0155889816144495e-05,
+ "loss": 0.4546602666378021,
+ "step": 1487
+ },
+ {
+ "epoch": 0.7806925498426023,
+ "grad_norm": 0.1559328242601668,
+ "learning_rate": 1.0132288983114611e-05,
+ "loss": 0.45510709285736084,
+ "step": 1488
+ },
+ {
+ "epoch": 0.7812172088142707,
+ "grad_norm": 0.1537255194600787,
+ "learning_rate": 1.0108735334267639e-05,
+ "loss": 0.4475342631340027,
+ "step": 1489
+ },
+ {
+ "epoch": 0.7817418677859391,
+ "grad_norm": 0.14616575061664402,
+ "learning_rate": 1.0085228933593558e-05,
+ "loss": 0.4378190040588379,
+ "step": 1490
+ },
+ {
+ "epoch": 0.7822665267576075,
+ "grad_norm": 0.1624360967746598,
+ "learning_rate": 1.0061769844953999e-05,
+ "loss": 0.4439539909362793,
+ "step": 1491
+ },
+ {
+ "epoch": 0.782791185729276,
+ "grad_norm": 0.15407336125298413,
+ "learning_rate": 1.003835813208206e-05,
+ "loss": 0.44074907898902893,
+ "step": 1492
+ },
+ {
+ "epoch": 0.7833158447009444,
+ "grad_norm": 0.15786980710136125,
+ "learning_rate": 1.001499385858211e-05,
+ "loss": 0.4491667151451111,
+ "step": 1493
+ },
+ {
+ "epoch": 0.7838405036726128,
+ "grad_norm": 0.15338836738460088,
+ "learning_rate": 9.991677087929665e-06,
+ "loss": 0.446003794670105,
+ "step": 1494
+ },
+ {
+ "epoch": 0.7843651626442812,
+ "grad_norm": 0.156710894892,
+ "learning_rate": 9.968407883471165e-06,
+ "loss": 0.43604689836502075,
+ "step": 1495
+ },
+ {
+ "epoch": 0.7848898216159497,
+ "grad_norm": 0.16100539063191666,
+ "learning_rate": 9.945186308423832e-06,
+ "loss": 0.4431147873401642,
+ "step": 1496
+ },
+ {
+ "epoch": 0.7854144805876181,
+ "grad_norm": 0.15776018851122997,
+ "learning_rate": 9.92201242587548e-06,
+ "loss": 0.4532504975795746,
+ "step": 1497
+ },
+ {
+ "epoch": 0.7859391395592865,
+ "grad_norm": 0.1540333423540934,
+ "learning_rate": 9.898886298784368e-06,
+ "loss": 0.43605154752731323,
+ "step": 1498
+ },
+ {
+ "epoch": 0.7864637985309548,
+ "grad_norm": 0.16009602938972295,
+ "learning_rate": 9.875807989979005e-06,
+ "loss": 0.4605238437652588,
+ "step": 1499
+ },
+ {
+ "epoch": 0.7869884575026233,
+ "grad_norm": 0.15639368804171708,
+ "learning_rate": 9.852777562157983e-06,
+ "loss": 0.4509708881378174,
+ "step": 1500
+ },
+ {
+ "epoch": 0.7875131164742917,
+ "grad_norm": 0.15074011188741518,
+ "learning_rate": 9.829795077889828e-06,
+ "loss": 0.4527304172515869,
+ "step": 1501
+ },
+ {
+ "epoch": 0.7880377754459601,
+ "grad_norm": 0.15155083996490956,
+ "learning_rate": 9.806860599612799e-06,
+ "loss": 0.4306413531303406,
+ "step": 1502
+ },
+ {
+ "epoch": 0.7885624344176285,
+ "grad_norm": 0.1478956630795445,
+ "learning_rate": 9.783974189634734e-06,
+ "loss": 0.4434714913368225,
+ "step": 1503
+ },
+ {
+ "epoch": 0.789087093389297,
+ "grad_norm": 0.15327668434806432,
+ "learning_rate": 9.761135910132902e-06,
+ "loss": 0.44431227445602417,
+ "step": 1504
+ },
+ {
+ "epoch": 0.7896117523609654,
+ "grad_norm": 0.2243177585497212,
+ "learning_rate": 9.738345823153783e-06,
+ "loss": 0.4541007876396179,
+ "step": 1505
+ },
+ {
+ "epoch": 0.7901364113326338,
+ "grad_norm": 0.1600239865812282,
+ "learning_rate": 9.715603990612945e-06,
+ "loss": 0.4456910490989685,
+ "step": 1506
+ },
+ {
+ "epoch": 0.7906610703043022,
+ "grad_norm": 0.15304634075332066,
+ "learning_rate": 9.692910474294858e-06,
+ "loss": 0.43965545296669006,
+ "step": 1507
+ },
+ {
+ "epoch": 0.7911857292759706,
+ "grad_norm": 0.15657752062053032,
+ "learning_rate": 9.670265335852725e-06,
+ "loss": 0.43358081579208374,
+ "step": 1508
+ },
+ {
+ "epoch": 0.7917103882476391,
+ "grad_norm": 0.16110057783374115,
+ "learning_rate": 9.647668636808304e-06,
+ "loss": 0.45711368322372437,
+ "step": 1509
+ },
+ {
+ "epoch": 0.7922350472193075,
+ "grad_norm": 0.1551133690299623,
+ "learning_rate": 9.625120438551791e-06,
+ "loss": 0.4613519608974457,
+ "step": 1510
+ },
+ {
+ "epoch": 0.7927597061909759,
+ "grad_norm": 0.15690897380725946,
+ "learning_rate": 9.602620802341581e-06,
+ "loss": 0.4472261071205139,
+ "step": 1511
+ },
+ {
+ "epoch": 0.7932843651626443,
+ "grad_norm": 0.147796599322911,
+ "learning_rate": 9.580169789304152e-06,
+ "loss": 0.4447203576564789,
+ "step": 1512
+ },
+ {
+ "epoch": 0.7938090241343126,
+ "grad_norm": 0.152019559199633,
+ "learning_rate": 9.557767460433878e-06,
+ "loss": 0.445905864238739,
+ "step": 1513
+ },
+ {
+ "epoch": 0.7943336831059811,
+ "grad_norm": 0.1574763924872037,
+ "learning_rate": 9.535413876592872e-06,
+ "loss": 0.4502032995223999,
+ "step": 1514
+ },
+ {
+ "epoch": 0.7948583420776495,
+ "grad_norm": 0.1524218315499056,
+ "learning_rate": 9.513109098510816e-06,
+ "loss": 0.43753111362457275,
+ "step": 1515
+ },
+ {
+ "epoch": 0.7953830010493179,
+ "grad_norm": 0.15589147437937434,
+ "learning_rate": 9.490853186784801e-06,
+ "loss": 0.44524312019348145,
+ "step": 1516
+ },
+ {
+ "epoch": 0.7959076600209863,
+ "grad_norm": 0.15604009714081407,
+ "learning_rate": 9.468646201879156e-06,
+ "loss": 0.44141292572021484,
+ "step": 1517
+ },
+ {
+ "epoch": 0.7964323189926548,
+ "grad_norm": 0.1562432554004435,
+ "learning_rate": 9.44648820412529e-06,
+ "loss": 0.461681067943573,
+ "step": 1518
+ },
+ {
+ "epoch": 0.7969569779643232,
+ "grad_norm": 0.15233588531495665,
+ "learning_rate": 9.424379253721513e-06,
+ "loss": 0.4529497027397156,
+ "step": 1519
+ },
+ {
+ "epoch": 0.7974816369359916,
+ "grad_norm": 0.16856991685290923,
+ "learning_rate": 9.40231941073291e-06,
+ "loss": 0.4483596086502075,
+ "step": 1520
+ },
+ {
+ "epoch": 0.79800629590766,
+ "grad_norm": 0.1546876712981077,
+ "learning_rate": 9.380308735091125e-06,
+ "loss": 0.43846625089645386,
+ "step": 1521
+ },
+ {
+ "epoch": 0.7985309548793285,
+ "grad_norm": 0.15198637556890449,
+ "learning_rate": 9.35834728659424e-06,
+ "loss": 0.4625553488731384,
+ "step": 1522
+ },
+ {
+ "epoch": 0.7990556138509969,
+ "grad_norm": 0.1519777007198701,
+ "learning_rate": 9.336435124906592e-06,
+ "loss": 0.44183820486068726,
+ "step": 1523
+ },
+ {
+ "epoch": 0.7995802728226653,
+ "grad_norm": 0.16125823074979284,
+ "learning_rate": 9.314572309558616e-06,
+ "loss": 0.46057307720184326,
+ "step": 1524
+ },
+ {
+ "epoch": 0.8001049317943337,
+ "grad_norm": 0.16283323624848034,
+ "learning_rate": 9.2927588999467e-06,
+ "loss": 0.4530155062675476,
+ "step": 1525
+ },
+ {
+ "epoch": 0.8006295907660022,
+ "grad_norm": 0.14998654489027738,
+ "learning_rate": 9.270994955332984e-06,
+ "loss": 0.43155500292778015,
+ "step": 1526
+ },
+ {
+ "epoch": 0.8011542497376705,
+ "grad_norm": 0.15683095030848093,
+ "learning_rate": 9.249280534845241e-06,
+ "loss": 0.45097050070762634,
+ "step": 1527
+ },
+ {
+ "epoch": 0.8016789087093389,
+ "grad_norm": 0.1561578254492378,
+ "learning_rate": 9.227615697476674e-06,
+ "loss": 0.42797762155532837,
+ "step": 1528
+ },
+ {
+ "epoch": 0.8022035676810073,
+ "grad_norm": 0.15642385261469652,
+ "learning_rate": 9.206000502085813e-06,
+ "loss": 0.4560159742832184,
+ "step": 1529
+ },
+ {
+ "epoch": 0.8027282266526757,
+ "grad_norm": 0.1623126279097499,
+ "learning_rate": 9.1844350073963e-06,
+ "loss": 0.4704165458679199,
+ "step": 1530
+ },
+ {
+ "epoch": 0.8032528856243442,
+ "grad_norm": 0.15090934786253773,
+ "learning_rate": 9.162919271996755e-06,
+ "loss": 0.44642555713653564,
+ "step": 1531
+ },
+ {
+ "epoch": 0.8037775445960126,
+ "grad_norm": 0.15305216978534084,
+ "learning_rate": 9.14145335434062e-06,
+ "loss": 0.4634283781051636,
+ "step": 1532
+ },
+ {
+ "epoch": 0.804302203567681,
+ "grad_norm": 0.14653219861345698,
+ "learning_rate": 9.120037312745979e-06,
+ "loss": 0.44544070959091187,
+ "step": 1533
+ },
+ {
+ "epoch": 0.8048268625393494,
+ "grad_norm": 0.16247397938502464,
+ "learning_rate": 9.098671205395422e-06,
+ "loss": 0.4552367925643921,
+ "step": 1534
+ },
+ {
+ "epoch": 0.8053515215110179,
+ "grad_norm": 0.14963174165902007,
+ "learning_rate": 9.077355090335892e-06,
+ "loss": 0.43211787939071655,
+ "step": 1535
+ },
+ {
+ "epoch": 0.8058761804826863,
+ "grad_norm": 0.1554853846772283,
+ "learning_rate": 9.056089025478495e-06,
+ "loss": 0.4319286644458771,
+ "step": 1536
+ },
+ {
+ "epoch": 0.8064008394543547,
+ "grad_norm": 0.15840895821310472,
+ "learning_rate": 9.03487306859837e-06,
+ "loss": 0.4494994878768921,
+ "step": 1537
+ },
+ {
+ "epoch": 0.8069254984260231,
+ "grad_norm": 0.1543496716800982,
+ "learning_rate": 9.013707277334515e-06,
+ "loss": 0.45500510931015015,
+ "step": 1538
+ },
+ {
+ "epoch": 0.8074501573976915,
+ "grad_norm": 0.14797282318550045,
+ "learning_rate": 8.992591709189651e-06,
+ "loss": 0.4385896325111389,
+ "step": 1539
+ },
+ {
+ "epoch": 0.8079748163693599,
+ "grad_norm": 0.17045570682787758,
+ "learning_rate": 8.971526421530066e-06,
+ "loss": 0.4585988521575928,
+ "step": 1540
+ },
+ {
+ "epoch": 0.8084994753410283,
+ "grad_norm": 0.14758220713044884,
+ "learning_rate": 8.950511471585407e-06,
+ "loss": 0.4295082092285156,
+ "step": 1541
+ },
+ {
+ "epoch": 0.8090241343126967,
+ "grad_norm": 0.16005752068175036,
+ "learning_rate": 8.929546916448597e-06,
+ "loss": 0.4507679343223572,
+ "step": 1542
+ },
+ {
+ "epoch": 0.8095487932843651,
+ "grad_norm": 0.14884684083313074,
+ "learning_rate": 8.908632813075641e-06,
+ "loss": 0.43547600507736206,
+ "step": 1543
+ },
+ {
+ "epoch": 0.8100734522560336,
+ "grad_norm": 0.14860107595788136,
+ "learning_rate": 8.88776921828547e-06,
+ "loss": 0.42700642347335815,
+ "step": 1544
+ },
+ {
+ "epoch": 0.810598111227702,
+ "grad_norm": 0.1472630611843838,
+ "learning_rate": 8.86695618875981e-06,
+ "loss": 0.4473339319229126,
+ "step": 1545
+ },
+ {
+ "epoch": 0.8111227701993704,
+ "grad_norm": 0.15105523788317893,
+ "learning_rate": 8.846193781043e-06,
+ "loss": 0.44923150539398193,
+ "step": 1546
+ },
+ {
+ "epoch": 0.8116474291710388,
+ "grad_norm": 0.16121409015263036,
+ "learning_rate": 8.82548205154185e-06,
+ "loss": 0.4653576612472534,
+ "step": 1547
+ },
+ {
+ "epoch": 0.8121720881427072,
+ "grad_norm": 0.1595195150959421,
+ "learning_rate": 8.804821056525497e-06,
+ "loss": 0.4522424638271332,
+ "step": 1548
+ },
+ {
+ "epoch": 0.8126967471143757,
+ "grad_norm": 0.15163510106316194,
+ "learning_rate": 8.784210852125232e-06,
+ "loss": 0.43657630681991577,
+ "step": 1549
+ },
+ {
+ "epoch": 0.8132214060860441,
+ "grad_norm": 0.15155109347013793,
+ "learning_rate": 8.763651494334377e-06,
+ "loss": 0.43183624744415283,
+ "step": 1550
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 4.102341666222899e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1550/video_preprocessor_config.json b/checkpoints/checkpoint-1550/video_preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..a933f0ec6941222c33aeba7980bd5acf8dd98e91
--- /dev/null
+++ b/checkpoints/checkpoint-1550/video_preprocessor_config.json
@@ -0,0 +1,30 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "do_sample_frames": true,
+ "fps": 2,
+ "image_mean": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "image_std": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "max_frames": 768,
+ "merge_size": 2,
+ "min_frames": 4,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 786432,
+ "shortest_edge": 131072
+ },
+ "temporal_patch_size": 2,
+ "video_processor_type": "Qwen3VLVideoProcessor"
+}
diff --git a/checkpoints/checkpoint-1600/chat_template.jinja b/checkpoints/checkpoint-1600/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1600/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1600/config.json b/checkpoints/checkpoint-1600/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1600/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1600/expert_prune_metadata.json b/checkpoints/checkpoint-1600/expert_prune_metadata.json
new file mode 100644
index 0000000000000000000000000000000000000000..16e01c8a10636331e334732fa9629579c68df44e
--- /dev/null
+++ b/checkpoints/checkpoint-1600/expert_prune_metadata.json
@@ -0,0 +1,14 @@
+{
+ "source_model": "/mnt/data/wuwei/model_list/Qwen3.6-35B-A3B/models--Qwen--Qwen3.6-35B-A3B/snapshots/995ad96eacd98c81ed38be0c5b274b04031597b0",
+ "output_model": "/mnt/data/wuwei/model_list/qwen3-19b-a3b",
+ "total_parameters": 18990568816,
+ "parameter_billions_decimal": 18.990568816,
+ "directory_size_label": "19B",
+ "original_num_experts": 256,
+ "kept_experts_per_layer": 128,
+ "num_profiled_layers": 40,
+ "mtp_weights": "removed because Qwen3_5MoeForConditionalGeneration does not instantiate or use the source checkpoint's mtp.* training-only tensors",
+ "expert_order": "ascending original expert index; router rows and expert tensors use the same remapping",
+ "router_statistics": "router_stats.json and router_stats.pt",
+ "selection_score": "top-8 router hit frequency * mean(abs(actual expert output after down_proj and before routing-weight multiplication))"
+}
diff --git a/checkpoints/checkpoint-1600/generation_config.json b/checkpoints/checkpoint-1600/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1600/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1600/model.safetensors b/checkpoints/checkpoint-1600/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..421f714d41f99cab1c77f5bb785de6d6833a4657
--- /dev/null
+++ b/checkpoints/checkpoint-1600/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:434437476a77769e3fff73d8a8fa2c070a1456653e472057ff674789b9e64efe
+size 37983421000
diff --git a/checkpoints/checkpoint-1600/preprocessor_config.json b/checkpoints/checkpoint-1600/preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ec43d1f859ecadf800ebae92fda4a812d34975d3
--- /dev/null
+++ b/checkpoints/checkpoint-1600/preprocessor_config.json
@@ -0,0 +1,26 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "image_mean": [
+ 0.48145466,
+ 0.4578275,
+ 0.40821073
+ ],
+ "image_processor_type": "Qwen2VLImageProcessorFast",
+ "image_std": [
+ 0.26862954,
+ 0.26130258,
+ 0.27577711
+ ],
+ "merge_size": 2,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 1003520,
+ "shortest_edge": 3136
+ },
+ "temporal_patch_size": 2
+}
diff --git a/checkpoints/checkpoint-1600/tokenizer.json b/checkpoints/checkpoint-1600/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1600/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1600/tokenizer_config.json b/checkpoints/checkpoint-1600/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1600/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1600/trainer_state.json b/checkpoints/checkpoint-1600/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..96d7cec7a8538d20e2f5f32cadf25795f5ddaf51
--- /dev/null
+++ b/checkpoints/checkpoint-1600/trainer_state.json
@@ -0,0 +1,11234 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.8394543546694648,
+ "eval_steps": 500,
+ "global_step": 1600,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ },
+ {
+ "epoch": 0.5776495278069255,
+ "grad_norm": 0.1862053032854098,
+ "learning_rate": 2.2065553964033126e-05,
+ "loss": 0.4816875457763672,
+ "step": 1101
+ },
+ {
+ "epoch": 0.5781741867785939,
+ "grad_norm": 0.16996015894363117,
+ "learning_rate": 2.20295734045532e-05,
+ "loss": 0.4471197724342346,
+ "step": 1102
+ },
+ {
+ "epoch": 0.5786988457502623,
+ "grad_norm": 0.16889227008390692,
+ "learning_rate": 2.1993607706997725e-05,
+ "loss": 0.468391478061676,
+ "step": 1103
+ },
+ {
+ "epoch": 0.5792235047219307,
+ "grad_norm": 0.16670931681562373,
+ "learning_rate": 2.195765696907745e-05,
+ "loss": 0.44934162497520447,
+ "step": 1104
+ },
+ {
+ "epoch": 0.5797481636935992,
+ "grad_norm": 0.16328418741223716,
+ "learning_rate": 2.1921721288462482e-05,
+ "loss": 0.4558270275592804,
+ "step": 1105
+ },
+ {
+ "epoch": 0.5802728226652676,
+ "grad_norm": 0.17364660408703936,
+ "learning_rate": 2.188580076278203e-05,
+ "loss": 0.4597751796245575,
+ "step": 1106
+ },
+ {
+ "epoch": 0.580797481636936,
+ "grad_norm": 0.1628618453800601,
+ "learning_rate": 2.18498954896241e-05,
+ "loss": 0.4521859288215637,
+ "step": 1107
+ },
+ {
+ "epoch": 0.5813221406086044,
+ "grad_norm": 0.16813817951871482,
+ "learning_rate": 2.1814005566535302e-05,
+ "loss": 0.4581005573272705,
+ "step": 1108
+ },
+ {
+ "epoch": 0.5818467995802729,
+ "grad_norm": 0.16825923629726025,
+ "learning_rate": 2.1778131091020516e-05,
+ "loss": 0.4433916211128235,
+ "step": 1109
+ },
+ {
+ "epoch": 0.5823714585519413,
+ "grad_norm": 0.17194743467883944,
+ "learning_rate": 2.1742272160542654e-05,
+ "loss": 0.47296425700187683,
+ "step": 1110
+ },
+ {
+ "epoch": 0.5828961175236097,
+ "grad_norm": 0.16947730546039969,
+ "learning_rate": 2.170642887252241e-05,
+ "loss": 0.4815937876701355,
+ "step": 1111
+ },
+ {
+ "epoch": 0.583420776495278,
+ "grad_norm": 0.1657449928668473,
+ "learning_rate": 2.167060132433798e-05,
+ "loss": 0.4479164779186249,
+ "step": 1112
+ },
+ {
+ "epoch": 0.5839454354669464,
+ "grad_norm": 0.1658952516607364,
+ "learning_rate": 2.1634789613324778e-05,
+ "loss": 0.4637463688850403,
+ "step": 1113
+ },
+ {
+ "epoch": 0.5844700944386149,
+ "grad_norm": 0.16777002188998236,
+ "learning_rate": 2.1598993836775218e-05,
+ "loss": 0.457524836063385,
+ "step": 1114
+ },
+ {
+ "epoch": 0.5849947534102833,
+ "grad_norm": 0.15837566447341322,
+ "learning_rate": 2.1563214091938405e-05,
+ "loss": 0.44938087463378906,
+ "step": 1115
+ },
+ {
+ "epoch": 0.5855194123819517,
+ "grad_norm": 0.17162206617035314,
+ "learning_rate": 2.1527450476019896e-05,
+ "loss": 0.4393450617790222,
+ "step": 1116
+ },
+ {
+ "epoch": 0.5860440713536201,
+ "grad_norm": 0.16224101921433814,
+ "learning_rate": 2.1491703086181426e-05,
+ "loss": 0.451913982629776,
+ "step": 1117
+ },
+ {
+ "epoch": 0.5865687303252886,
+ "grad_norm": 0.1736253504575144,
+ "learning_rate": 2.145597201954066e-05,
+ "loss": 0.4611290395259857,
+ "step": 1118
+ },
+ {
+ "epoch": 0.587093389296957,
+ "grad_norm": 0.16308408333737234,
+ "learning_rate": 2.1420257373170903e-05,
+ "loss": 0.45898568630218506,
+ "step": 1119
+ },
+ {
+ "epoch": 0.5876180482686254,
+ "grad_norm": 0.16948634794175524,
+ "learning_rate": 2.1384559244100848e-05,
+ "loss": 0.45982882380485535,
+ "step": 1120
+ },
+ {
+ "epoch": 0.5881427072402938,
+ "grad_norm": 0.16676366538565252,
+ "learning_rate": 2.1348877729314336e-05,
+ "loss": 0.44826576113700867,
+ "step": 1121
+ },
+ {
+ "epoch": 0.5886673662119623,
+ "grad_norm": 0.16431125630712295,
+ "learning_rate": 2.1313212925750037e-05,
+ "loss": 0.4616445302963257,
+ "step": 1122
+ },
+ {
+ "epoch": 0.5891920251836307,
+ "grad_norm": 0.1631659061616955,
+ "learning_rate": 2.1277564930301252e-05,
+ "loss": 0.45651188492774963,
+ "step": 1123
+ },
+ {
+ "epoch": 0.5897166841552991,
+ "grad_norm": 0.15805950502689237,
+ "learning_rate": 2.1241933839815604e-05,
+ "loss": 0.4486045837402344,
+ "step": 1124
+ },
+ {
+ "epoch": 0.5902413431269675,
+ "grad_norm": 0.16530288038603502,
+ "learning_rate": 2.1206319751094773e-05,
+ "loss": 0.4485066533088684,
+ "step": 1125
+ },
+ {
+ "epoch": 0.5907660020986358,
+ "grad_norm": 0.16061903194239047,
+ "learning_rate": 2.1170722760894295e-05,
+ "loss": 0.4565330743789673,
+ "step": 1126
+ },
+ {
+ "epoch": 0.5912906610703043,
+ "grad_norm": 0.1643346999353627,
+ "learning_rate": 2.1135142965923195e-05,
+ "loss": 0.45560717582702637,
+ "step": 1127
+ },
+ {
+ "epoch": 0.5918153200419727,
+ "grad_norm": 0.17484644315215753,
+ "learning_rate": 2.109958046284382e-05,
+ "loss": 0.4795891046524048,
+ "step": 1128
+ },
+ {
+ "epoch": 0.5923399790136411,
+ "grad_norm": 0.1623762678622651,
+ "learning_rate": 2.1064035348271528e-05,
+ "loss": 0.47395604848861694,
+ "step": 1129
+ },
+ {
+ "epoch": 0.5928646379853095,
+ "grad_norm": 0.16888824241669015,
+ "learning_rate": 2.1028507718774456e-05,
+ "loss": 0.4601455330848694,
+ "step": 1130
+ },
+ {
+ "epoch": 0.593389296956978,
+ "grad_norm": 0.16291744568806302,
+ "learning_rate": 2.0992997670873193e-05,
+ "loss": 0.46248236298561096,
+ "step": 1131
+ },
+ {
+ "epoch": 0.5939139559286464,
+ "grad_norm": 0.16185598691236988,
+ "learning_rate": 2.0957505301040603e-05,
+ "loss": 0.45085301995277405,
+ "step": 1132
+ },
+ {
+ "epoch": 0.5944386149003148,
+ "grad_norm": 0.1678000806117814,
+ "learning_rate": 2.0922030705701518e-05,
+ "loss": 0.44939014315605164,
+ "step": 1133
+ },
+ {
+ "epoch": 0.5949632738719832,
+ "grad_norm": 0.16104474282288994,
+ "learning_rate": 2.088657398123246e-05,
+ "loss": 0.4313976764678955,
+ "step": 1134
+ },
+ {
+ "epoch": 0.5954879328436516,
+ "grad_norm": 0.1631128968199669,
+ "learning_rate": 2.085113522396141e-05,
+ "loss": 0.4508278965950012,
+ "step": 1135
+ },
+ {
+ "epoch": 0.5960125918153201,
+ "grad_norm": 0.16307688195844533,
+ "learning_rate": 2.0815714530167557e-05,
+ "loss": 0.4467576742172241,
+ "step": 1136
+ },
+ {
+ "epoch": 0.5965372507869885,
+ "grad_norm": 0.16456832304532207,
+ "learning_rate": 2.0780311996080976e-05,
+ "loss": 0.4621703624725342,
+ "step": 1137
+ },
+ {
+ "epoch": 0.5970619097586569,
+ "grad_norm": 0.16346217748325367,
+ "learning_rate": 2.0744927717882435e-05,
+ "loss": 0.4589449465274811,
+ "step": 1138
+ },
+ {
+ "epoch": 0.5975865687303253,
+ "grad_norm": 0.16180691501076994,
+ "learning_rate": 2.0709561791703097e-05,
+ "loss": 0.43698400259017944,
+ "step": 1139
+ },
+ {
+ "epoch": 0.5981112277019937,
+ "grad_norm": 0.15695459976898835,
+ "learning_rate": 2.067421431362427e-05,
+ "loss": 0.452271044254303,
+ "step": 1140
+ },
+ {
+ "epoch": 0.5986358866736621,
+ "grad_norm": 0.15930893553721054,
+ "learning_rate": 2.063888537967713e-05,
+ "loss": 0.44868963956832886,
+ "step": 1141
+ },
+ {
+ "epoch": 0.5991605456453305,
+ "grad_norm": 0.15963802284017928,
+ "learning_rate": 2.0603575085842485e-05,
+ "loss": 0.4376864731311798,
+ "step": 1142
+ },
+ {
+ "epoch": 0.5996852046169989,
+ "grad_norm": 0.16610266900507178,
+ "learning_rate": 2.05682835280505e-05,
+ "loss": 0.44676727056503296,
+ "step": 1143
+ },
+ {
+ "epoch": 0.6002098635886673,
+ "grad_norm": 0.15978253236669612,
+ "learning_rate": 2.053301080218044e-05,
+ "loss": 0.434773713350296,
+ "step": 1144
+ },
+ {
+ "epoch": 0.6007345225603358,
+ "grad_norm": 0.16305141803526324,
+ "learning_rate": 2.0497757004060403e-05,
+ "loss": 0.4529780149459839,
+ "step": 1145
+ },
+ {
+ "epoch": 0.6012591815320042,
+ "grad_norm": 0.1649737864101215,
+ "learning_rate": 2.046252222946707e-05,
+ "loss": 0.4571434259414673,
+ "step": 1146
+ },
+ {
+ "epoch": 0.6017838405036726,
+ "grad_norm": 0.15738824968326598,
+ "learning_rate": 2.0427306574125427e-05,
+ "loss": 0.4446289539337158,
+ "step": 1147
+ },
+ {
+ "epoch": 0.602308499475341,
+ "grad_norm": 0.1731139535185187,
+ "learning_rate": 2.039211013370853e-05,
+ "loss": 0.4551090598106384,
+ "step": 1148
+ },
+ {
+ "epoch": 0.6028331584470095,
+ "grad_norm": 0.15857251900043037,
+ "learning_rate": 2.0356933003837237e-05,
+ "loss": 0.4644010663032532,
+ "step": 1149
+ },
+ {
+ "epoch": 0.6033578174186779,
+ "grad_norm": 0.1589582861134211,
+ "learning_rate": 2.032177528007992e-05,
+ "loss": 0.45741650462150574,
+ "step": 1150
+ },
+ {
+ "epoch": 0.6038824763903463,
+ "grad_norm": 0.15639092706575122,
+ "learning_rate": 2.0286637057952257e-05,
+ "loss": 0.4375547766685486,
+ "step": 1151
+ },
+ {
+ "epoch": 0.6044071353620147,
+ "grad_norm": 0.16335605581337187,
+ "learning_rate": 2.0251518432916933e-05,
+ "loss": 0.4636140465736389,
+ "step": 1152
+ },
+ {
+ "epoch": 0.6049317943336832,
+ "grad_norm": 0.15437437113359356,
+ "learning_rate": 2.0216419500383372e-05,
+ "loss": 0.44632142782211304,
+ "step": 1153
+ },
+ {
+ "epoch": 0.6054564533053515,
+ "grad_norm": 0.15522039198698664,
+ "learning_rate": 2.0181340355707545e-05,
+ "loss": 0.449049174785614,
+ "step": 1154
+ },
+ {
+ "epoch": 0.6059811122770199,
+ "grad_norm": 0.1599951917909323,
+ "learning_rate": 2.0146281094191604e-05,
+ "loss": 0.44395962357521057,
+ "step": 1155
+ },
+ {
+ "epoch": 0.6065057712486883,
+ "grad_norm": 0.16866621620527716,
+ "learning_rate": 2.011124181108373e-05,
+ "loss": 0.45693716406822205,
+ "step": 1156
+ },
+ {
+ "epoch": 0.6070304302203567,
+ "grad_norm": 0.15508935650135075,
+ "learning_rate": 2.0076222601577815e-05,
+ "loss": 0.4525151252746582,
+ "step": 1157
+ },
+ {
+ "epoch": 0.6075550891920252,
+ "grad_norm": 0.16508998118005971,
+ "learning_rate": 2.004122356081321e-05,
+ "loss": 0.466006875038147,
+ "step": 1158
+ },
+ {
+ "epoch": 0.6080797481636936,
+ "grad_norm": 0.15817418864022237,
+ "learning_rate": 2.0006244783874466e-05,
+ "loss": 0.4496110677719116,
+ "step": 1159
+ },
+ {
+ "epoch": 0.608604407135362,
+ "grad_norm": 0.16127270259749824,
+ "learning_rate": 1.997128636579109e-05,
+ "loss": 0.45918628573417664,
+ "step": 1160
+ },
+ {
+ "epoch": 0.6091290661070304,
+ "grad_norm": 0.1702599924159677,
+ "learning_rate": 1.9936348401537292e-05,
+ "loss": 0.47218555212020874,
+ "step": 1161
+ },
+ {
+ "epoch": 0.6096537250786989,
+ "grad_norm": 0.15690176248846127,
+ "learning_rate": 1.9901430986031695e-05,
+ "loss": 0.4516097903251648,
+ "step": 1162
+ },
+ {
+ "epoch": 0.6101783840503673,
+ "grad_norm": 0.1713614405614144,
+ "learning_rate": 1.9866534214137085e-05,
+ "loss": 0.45030856132507324,
+ "step": 1163
+ },
+ {
+ "epoch": 0.6107030430220357,
+ "grad_norm": 0.1670865563026229,
+ "learning_rate": 1.9831658180660195e-05,
+ "loss": 0.45603489875793457,
+ "step": 1164
+ },
+ {
+ "epoch": 0.6112277019937041,
+ "grad_norm": 0.16897539113401952,
+ "learning_rate": 1.9796802980351392e-05,
+ "loss": 0.46314537525177,
+ "step": 1165
+ },
+ {
+ "epoch": 0.6117523609653726,
+ "grad_norm": 0.15832868870697644,
+ "learning_rate": 1.976196870790446e-05,
+ "loss": 0.44784611463546753,
+ "step": 1166
+ },
+ {
+ "epoch": 0.612277019937041,
+ "grad_norm": 0.16595365551891114,
+ "learning_rate": 1.9727155457956315e-05,
+ "loss": 0.4409624934196472,
+ "step": 1167
+ },
+ {
+ "epoch": 0.6128016789087093,
+ "grad_norm": 0.1642630411212017,
+ "learning_rate": 1.969236332508677e-05,
+ "loss": 0.46120351552963257,
+ "step": 1168
+ },
+ {
+ "epoch": 0.6133263378803777,
+ "grad_norm": 0.16481796135868812,
+ "learning_rate": 1.9657592403818253e-05,
+ "loss": 0.45544904470443726,
+ "step": 1169
+ },
+ {
+ "epoch": 0.6138509968520461,
+ "grad_norm": 0.1629832779852364,
+ "learning_rate": 1.962284278861557e-05,
+ "loss": 0.4578462839126587,
+ "step": 1170
+ },
+ {
+ "epoch": 0.6143756558237146,
+ "grad_norm": 0.15916219613603688,
+ "learning_rate": 1.9588114573885664e-05,
+ "loss": 0.4546983540058136,
+ "step": 1171
+ },
+ {
+ "epoch": 0.614900314795383,
+ "grad_norm": 0.1660520165166075,
+ "learning_rate": 1.95534078539773e-05,
+ "loss": 0.4601942300796509,
+ "step": 1172
+ },
+ {
+ "epoch": 0.6154249737670514,
+ "grad_norm": 0.16924794578577768,
+ "learning_rate": 1.9518722723180877e-05,
+ "loss": 0.45997846126556396,
+ "step": 1173
+ },
+ {
+ "epoch": 0.6159496327387198,
+ "grad_norm": 0.16597646133045382,
+ "learning_rate": 1.9484059275728127e-05,
+ "loss": 0.4594605565071106,
+ "step": 1174
+ },
+ {
+ "epoch": 0.6164742917103883,
+ "grad_norm": 0.16964243792519748,
+ "learning_rate": 1.9449417605791876e-05,
+ "loss": 0.4638690650463104,
+ "step": 1175
+ },
+ {
+ "epoch": 0.6169989506820567,
+ "grad_norm": 0.1758591426791696,
+ "learning_rate": 1.9414797807485796e-05,
+ "loss": 0.4666447639465332,
+ "step": 1176
+ },
+ {
+ "epoch": 0.6175236096537251,
+ "grad_norm": 0.16217661227891753,
+ "learning_rate": 1.9380199974864126e-05,
+ "loss": 0.44252535700798035,
+ "step": 1177
+ },
+ {
+ "epoch": 0.6180482686253935,
+ "grad_norm": 0.1576446184499019,
+ "learning_rate": 1.934562420192142e-05,
+ "loss": 0.4499641954898834,
+ "step": 1178
+ },
+ {
+ "epoch": 0.618572927597062,
+ "grad_norm": 0.16231008092636912,
+ "learning_rate": 1.9311070582592327e-05,
+ "loss": 0.4561542570590973,
+ "step": 1179
+ },
+ {
+ "epoch": 0.6190975865687304,
+ "grad_norm": 0.16790382249577043,
+ "learning_rate": 1.9276539210751292e-05,
+ "loss": 0.4484325647354126,
+ "step": 1180
+ },
+ {
+ "epoch": 0.6196222455403987,
+ "grad_norm": 0.18794729932605855,
+ "learning_rate": 1.9242030180212323e-05,
+ "loss": 0.44169026613235474,
+ "step": 1181
+ },
+ {
+ "epoch": 0.6201469045120671,
+ "grad_norm": 0.16948051243889814,
+ "learning_rate": 1.920754358472872e-05,
+ "loss": 0.4592876434326172,
+ "step": 1182
+ },
+ {
+ "epoch": 0.6206715634837355,
+ "grad_norm": 0.16762558710029374,
+ "learning_rate": 1.917307951799286e-05,
+ "loss": 0.4402095079421997,
+ "step": 1183
+ },
+ {
+ "epoch": 0.621196222455404,
+ "grad_norm": 0.18170280116387003,
+ "learning_rate": 1.9138638073635885e-05,
+ "loss": 0.4633026123046875,
+ "step": 1184
+ },
+ {
+ "epoch": 0.6217208814270724,
+ "grad_norm": 0.1718051709136108,
+ "learning_rate": 1.9104219345227486e-05,
+ "loss": 0.4510066509246826,
+ "step": 1185
+ },
+ {
+ "epoch": 0.6222455403987408,
+ "grad_norm": 0.16829066326345576,
+ "learning_rate": 1.906982342627565e-05,
+ "loss": 0.4632919430732727,
+ "step": 1186
+ },
+ {
+ "epoch": 0.6227701993704092,
+ "grad_norm": 0.16669141808548274,
+ "learning_rate": 1.9035450410226395e-05,
+ "loss": 0.455544650554657,
+ "step": 1187
+ },
+ {
+ "epoch": 0.6232948583420777,
+ "grad_norm": 0.16679996559722232,
+ "learning_rate": 1.90011003904635e-05,
+ "loss": 0.459994375705719,
+ "step": 1188
+ },
+ {
+ "epoch": 0.6238195173137461,
+ "grad_norm": 0.17379754713278164,
+ "learning_rate": 1.8966773460308284e-05,
+ "loss": 0.4576749801635742,
+ "step": 1189
+ },
+ {
+ "epoch": 0.6243441762854145,
+ "grad_norm": 0.17088539733389246,
+ "learning_rate": 1.8932469713019335e-05,
+ "loss": 0.44853052496910095,
+ "step": 1190
+ },
+ {
+ "epoch": 0.6248688352570829,
+ "grad_norm": 0.15556883889201015,
+ "learning_rate": 1.8898189241792248e-05,
+ "loss": 0.44060447812080383,
+ "step": 1191
+ },
+ {
+ "epoch": 0.6253934942287513,
+ "grad_norm": 0.1640093917439215,
+ "learning_rate": 1.8863932139759388e-05,
+ "loss": 0.4531911611557007,
+ "step": 1192
+ },
+ {
+ "epoch": 0.6259181532004198,
+ "grad_norm": 0.16444284826591576,
+ "learning_rate": 1.882969849998964e-05,
+ "loss": 0.44176363945007324,
+ "step": 1193
+ },
+ {
+ "epoch": 0.6264428121720882,
+ "grad_norm": 0.1696462520995642,
+ "learning_rate": 1.8795488415488143e-05,
+ "loss": 0.44516652822494507,
+ "step": 1194
+ },
+ {
+ "epoch": 0.6269674711437565,
+ "grad_norm": 0.16810724705191796,
+ "learning_rate": 1.8761301979196032e-05,
+ "loss": 0.4542563855648041,
+ "step": 1195
+ },
+ {
+ "epoch": 0.6274921301154249,
+ "grad_norm": 0.17334813313855824,
+ "learning_rate": 1.8727139283990204e-05,
+ "loss": 0.47769248485565186,
+ "step": 1196
+ },
+ {
+ "epoch": 0.6280167890870934,
+ "grad_norm": 0.16308212035303235,
+ "learning_rate": 1.8693000422683043e-05,
+ "loss": 0.4407799243927002,
+ "step": 1197
+ },
+ {
+ "epoch": 0.6285414480587618,
+ "grad_norm": 0.16853319139928977,
+ "learning_rate": 1.8658885488022204e-05,
+ "loss": 0.4565497040748596,
+ "step": 1198
+ },
+ {
+ "epoch": 0.6290661070304302,
+ "grad_norm": 0.16694968781284703,
+ "learning_rate": 1.8624794572690328e-05,
+ "loss": 0.44522374868392944,
+ "step": 1199
+ },
+ {
+ "epoch": 0.6295907660020986,
+ "grad_norm": 0.16418579468339103,
+ "learning_rate": 1.8590727769304788e-05,
+ "loss": 0.4441048502922058,
+ "step": 1200
+ },
+ {
+ "epoch": 0.630115424973767,
+ "grad_norm": 0.1661930777124156,
+ "learning_rate": 1.8556685170417486e-05,
+ "loss": 0.457327663898468,
+ "step": 1201
+ },
+ {
+ "epoch": 0.6306400839454355,
+ "grad_norm": 0.16536630027039814,
+ "learning_rate": 1.852266686851452e-05,
+ "loss": 0.462494432926178,
+ "step": 1202
+ },
+ {
+ "epoch": 0.6311647429171039,
+ "grad_norm": 0.16611474088307454,
+ "learning_rate": 1.8488672956016013e-05,
+ "loss": 0.4557701647281647,
+ "step": 1203
+ },
+ {
+ "epoch": 0.6316894018887723,
+ "grad_norm": 0.16759343079749697,
+ "learning_rate": 1.8454703525275813e-05,
+ "loss": 0.45069995522499084,
+ "step": 1204
+ },
+ {
+ "epoch": 0.6322140608604407,
+ "grad_norm": 0.16186676294485336,
+ "learning_rate": 1.8420758668581255e-05,
+ "loss": 0.44954800605773926,
+ "step": 1205
+ },
+ {
+ "epoch": 0.6327387198321092,
+ "grad_norm": 0.15687159835368553,
+ "learning_rate": 1.8386838478152916e-05,
+ "loss": 0.4490429759025574,
+ "step": 1206
+ },
+ {
+ "epoch": 0.6332633788037776,
+ "grad_norm": 0.16242951976115577,
+ "learning_rate": 1.8352943046144372e-05,
+ "loss": 0.45819586515426636,
+ "step": 1207
+ },
+ {
+ "epoch": 0.633788037775446,
+ "grad_norm": 0.16516087266517232,
+ "learning_rate": 1.831907246464191e-05,
+ "loss": 0.44769883155822754,
+ "step": 1208
+ },
+ {
+ "epoch": 0.6343126967471143,
+ "grad_norm": 0.16709475871512644,
+ "learning_rate": 1.8285226825664332e-05,
+ "loss": 0.4699128568172455,
+ "step": 1209
+ },
+ {
+ "epoch": 0.6348373557187827,
+ "grad_norm": 0.16288623150595774,
+ "learning_rate": 1.8251406221162636e-05,
+ "loss": 0.45322558283805847,
+ "step": 1210
+ },
+ {
+ "epoch": 0.6353620146904512,
+ "grad_norm": 0.1675722848054237,
+ "learning_rate": 1.821761074301986e-05,
+ "loss": 0.4752347469329834,
+ "step": 1211
+ },
+ {
+ "epoch": 0.6358866736621196,
+ "grad_norm": 0.163285963514504,
+ "learning_rate": 1.818384048305074e-05,
+ "loss": 0.441958487033844,
+ "step": 1212
+ },
+ {
+ "epoch": 0.636411332633788,
+ "grad_norm": 0.17078102287557764,
+ "learning_rate": 1.8150095533001517e-05,
+ "loss": 0.4551333487033844,
+ "step": 1213
+ },
+ {
+ "epoch": 0.6369359916054564,
+ "grad_norm": 0.16188540746097138,
+ "learning_rate": 1.811637598454966e-05,
+ "loss": 0.4470779597759247,
+ "step": 1214
+ },
+ {
+ "epoch": 0.6374606505771249,
+ "grad_norm": 0.15235628218958286,
+ "learning_rate": 1.808268192930364e-05,
+ "loss": 0.43604153394699097,
+ "step": 1215
+ },
+ {
+ "epoch": 0.6379853095487933,
+ "grad_norm": 0.1604601949959104,
+ "learning_rate": 1.8049013458802665e-05,
+ "loss": 0.4445713758468628,
+ "step": 1216
+ },
+ {
+ "epoch": 0.6385099685204617,
+ "grad_norm": 0.15511797131215901,
+ "learning_rate": 1.8015370664516427e-05,
+ "loss": 0.4455301761627197,
+ "step": 1217
+ },
+ {
+ "epoch": 0.6390346274921301,
+ "grad_norm": 0.16815669546944095,
+ "learning_rate": 1.798175363784487e-05,
+ "loss": 0.46135014295578003,
+ "step": 1218
+ },
+ {
+ "epoch": 0.6395592864637986,
+ "grad_norm": 0.161558177688922,
+ "learning_rate": 1.7948162470117917e-05,
+ "loss": 0.4519304037094116,
+ "step": 1219
+ },
+ {
+ "epoch": 0.640083945435467,
+ "grad_norm": 0.16677842130048232,
+ "learning_rate": 1.7914597252595265e-05,
+ "loss": 0.4550550878047943,
+ "step": 1220
+ },
+ {
+ "epoch": 0.6406086044071354,
+ "grad_norm": 0.1597331480423138,
+ "learning_rate": 1.7881058076466095e-05,
+ "loss": 0.441120445728302,
+ "step": 1221
+ },
+ {
+ "epoch": 0.6411332633788038,
+ "grad_norm": 0.15951106545246824,
+ "learning_rate": 1.7847545032848844e-05,
+ "loss": 0.4471382200717926,
+ "step": 1222
+ },
+ {
+ "epoch": 0.6416579223504721,
+ "grad_norm": 0.16157787674667565,
+ "learning_rate": 1.7814058212790947e-05,
+ "loss": 0.4550648331642151,
+ "step": 1223
+ },
+ {
+ "epoch": 0.6421825813221406,
+ "grad_norm": 0.16989186763352415,
+ "learning_rate": 1.7780597707268583e-05,
+ "loss": 0.4515089988708496,
+ "step": 1224
+ },
+ {
+ "epoch": 0.642707240293809,
+ "grad_norm": 0.15952153223099508,
+ "learning_rate": 1.7747163607186468e-05,
+ "loss": 0.4658638834953308,
+ "step": 1225
+ },
+ {
+ "epoch": 0.6432318992654774,
+ "grad_norm": 0.15205926699647218,
+ "learning_rate": 1.7713756003377567e-05,
+ "loss": 0.4330868422985077,
+ "step": 1226
+ },
+ {
+ "epoch": 0.6437565582371458,
+ "grad_norm": 0.15784806480936794,
+ "learning_rate": 1.7680374986602865e-05,
+ "loss": 0.4506068527698517,
+ "step": 1227
+ },
+ {
+ "epoch": 0.6442812172088143,
+ "grad_norm": 0.16006709280686687,
+ "learning_rate": 1.7647020647551105e-05,
+ "loss": 0.446941077709198,
+ "step": 1228
+ },
+ {
+ "epoch": 0.6448058761804827,
+ "grad_norm": 0.16064631158222062,
+ "learning_rate": 1.761369307683856e-05,
+ "loss": 0.4418525695800781,
+ "step": 1229
+ },
+ {
+ "epoch": 0.6453305351521511,
+ "grad_norm": 0.16454136480444392,
+ "learning_rate": 1.7580392365008767e-05,
+ "loss": 0.46268391609191895,
+ "step": 1230
+ },
+ {
+ "epoch": 0.6458551941238195,
+ "grad_norm": 0.15401246535771462,
+ "learning_rate": 1.754711860253233e-05,
+ "loss": 0.4372989535331726,
+ "step": 1231
+ },
+ {
+ "epoch": 0.646379853095488,
+ "grad_norm": 0.1544567975028383,
+ "learning_rate": 1.7513871879806603e-05,
+ "loss": 0.44806575775146484,
+ "step": 1232
+ },
+ {
+ "epoch": 0.6469045120671564,
+ "grad_norm": 0.16411225112730965,
+ "learning_rate": 1.748065228715549e-05,
+ "loss": 0.43782058358192444,
+ "step": 1233
+ },
+ {
+ "epoch": 0.6474291710388248,
+ "grad_norm": 0.1552767832972093,
+ "learning_rate": 1.7447459914829183e-05,
+ "loss": 0.44531378149986267,
+ "step": 1234
+ },
+ {
+ "epoch": 0.6479538300104932,
+ "grad_norm": 0.16214695257123513,
+ "learning_rate": 1.741429485300392e-05,
+ "loss": 0.47399842739105225,
+ "step": 1235
+ },
+ {
+ "epoch": 0.6484784889821616,
+ "grad_norm": 0.16684861111484112,
+ "learning_rate": 1.7381157191781766e-05,
+ "loss": 0.4654282331466675,
+ "step": 1236
+ },
+ {
+ "epoch": 0.64900314795383,
+ "grad_norm": 0.15901864178630273,
+ "learning_rate": 1.7348047021190334e-05,
+ "loss": 0.45128414034843445,
+ "step": 1237
+ },
+ {
+ "epoch": 0.6495278069254984,
+ "grad_norm": 0.15360991961742707,
+ "learning_rate": 1.7314964431182523e-05,
+ "loss": 0.4303375482559204,
+ "step": 1238
+ },
+ {
+ "epoch": 0.6500524658971668,
+ "grad_norm": 0.1633151574385741,
+ "learning_rate": 1.728190951163633e-05,
+ "loss": 0.45568403601646423,
+ "step": 1239
+ },
+ {
+ "epoch": 0.6505771248688352,
+ "grad_norm": 0.16377853571058645,
+ "learning_rate": 1.724888235235457e-05,
+ "loss": 0.46503543853759766,
+ "step": 1240
+ },
+ {
+ "epoch": 0.6511017838405037,
+ "grad_norm": 0.1595023801275368,
+ "learning_rate": 1.7215883043064655e-05,
+ "loss": 0.46337294578552246,
+ "step": 1241
+ },
+ {
+ "epoch": 0.6516264428121721,
+ "grad_norm": 0.16324796182212292,
+ "learning_rate": 1.7182911673418313e-05,
+ "loss": 0.46197354793548584,
+ "step": 1242
+ },
+ {
+ "epoch": 0.6521511017838405,
+ "grad_norm": 0.16353734830879765,
+ "learning_rate": 1.714996833299138e-05,
+ "loss": 0.45741671323776245,
+ "step": 1243
+ },
+ {
+ "epoch": 0.6526757607555089,
+ "grad_norm": 0.15941587678263386,
+ "learning_rate": 1.7117053111283528e-05,
+ "loss": 0.4514503479003906,
+ "step": 1244
+ },
+ {
+ "epoch": 0.6532004197271774,
+ "grad_norm": 0.1642653230145008,
+ "learning_rate": 1.708416609771804e-05,
+ "loss": 0.4437655210494995,
+ "step": 1245
+ },
+ {
+ "epoch": 0.6537250786988458,
+ "grad_norm": 0.15814155519130152,
+ "learning_rate": 1.7051307381641587e-05,
+ "loss": 0.4314163327217102,
+ "step": 1246
+ },
+ {
+ "epoch": 0.6542497376705142,
+ "grad_norm": 0.16086078370768056,
+ "learning_rate": 1.701847705232394e-05,
+ "loss": 0.4526430666446686,
+ "step": 1247
+ },
+ {
+ "epoch": 0.6547743966421826,
+ "grad_norm": 0.16389360190715183,
+ "learning_rate": 1.698567519895775e-05,
+ "loss": 0.4278966188430786,
+ "step": 1248
+ },
+ {
+ "epoch": 0.655299055613851,
+ "grad_norm": 0.16303142287804262,
+ "learning_rate": 1.6952901910658316e-05,
+ "loss": 0.45778220891952515,
+ "step": 1249
+ },
+ {
+ "epoch": 0.6558237145855194,
+ "grad_norm": 0.1675209956904048,
+ "learning_rate": 1.6920157276463318e-05,
+ "loss": 0.4489403963088989,
+ "step": 1250
+ },
+ {
+ "epoch": 0.6563483735571878,
+ "grad_norm": 0.16228311551074323,
+ "learning_rate": 1.6887441385332604e-05,
+ "loss": 0.4534342885017395,
+ "step": 1251
+ },
+ {
+ "epoch": 0.6568730325288562,
+ "grad_norm": 0.16408957851046913,
+ "learning_rate": 1.6854754326147916e-05,
+ "loss": 0.4549208879470825,
+ "step": 1252
+ },
+ {
+ "epoch": 0.6573976915005246,
+ "grad_norm": 0.15967922111432845,
+ "learning_rate": 1.682209618771268e-05,
+ "loss": 0.4451141357421875,
+ "step": 1253
+ },
+ {
+ "epoch": 0.657922350472193,
+ "grad_norm": 0.16440698722204208,
+ "learning_rate": 1.678946705875176e-05,
+ "loss": 0.454200804233551,
+ "step": 1254
+ },
+ {
+ "epoch": 0.6584470094438615,
+ "grad_norm": 0.16251120184938497,
+ "learning_rate": 1.675686702791117e-05,
+ "loss": 0.4580027461051941,
+ "step": 1255
+ },
+ {
+ "epoch": 0.6589716684155299,
+ "grad_norm": 0.1643717386150457,
+ "learning_rate": 1.6724296183757924e-05,
+ "loss": 0.4422311782836914,
+ "step": 1256
+ },
+ {
+ "epoch": 0.6594963273871983,
+ "grad_norm": 0.1644586414378722,
+ "learning_rate": 1.66917546147797e-05,
+ "loss": 0.4593179225921631,
+ "step": 1257
+ },
+ {
+ "epoch": 0.6600209863588667,
+ "grad_norm": 0.16751074186967685,
+ "learning_rate": 1.6659242409384657e-05,
+ "loss": 0.457930326461792,
+ "step": 1258
+ },
+ {
+ "epoch": 0.6605456453305352,
+ "grad_norm": 0.15398418609155334,
+ "learning_rate": 1.6626759655901182e-05,
+ "loss": 0.44002899527549744,
+ "step": 1259
+ },
+ {
+ "epoch": 0.6610703043022036,
+ "grad_norm": 0.16328113175790226,
+ "learning_rate": 1.659430644257763e-05,
+ "loss": 0.4500438868999481,
+ "step": 1260
+ },
+ {
+ "epoch": 0.661594963273872,
+ "grad_norm": 0.1694244340615106,
+ "learning_rate": 1.656188285758214e-05,
+ "loss": 0.4818427264690399,
+ "step": 1261
+ },
+ {
+ "epoch": 0.6621196222455404,
+ "grad_norm": 0.1578320701754368,
+ "learning_rate": 1.652948898900232e-05,
+ "loss": 0.4492436647415161,
+ "step": 1262
+ },
+ {
+ "epoch": 0.6626442812172089,
+ "grad_norm": 0.16538865917828469,
+ "learning_rate": 1.6497124924845064e-05,
+ "loss": 0.45070338249206543,
+ "step": 1263
+ },
+ {
+ "epoch": 0.6631689401888772,
+ "grad_norm": 0.15670817848256852,
+ "learning_rate": 1.6464790753036296e-05,
+ "loss": 0.45439285039901733,
+ "step": 1264
+ },
+ {
+ "epoch": 0.6636935991605456,
+ "grad_norm": 0.16184786585965885,
+ "learning_rate": 1.64324865614207e-05,
+ "loss": 0.446522980928421,
+ "step": 1265
+ },
+ {
+ "epoch": 0.664218258132214,
+ "grad_norm": 0.15719726462251984,
+ "learning_rate": 1.6400212437761554e-05,
+ "loss": 0.4495280981063843,
+ "step": 1266
+ },
+ {
+ "epoch": 0.6647429171038824,
+ "grad_norm": 0.1680804414134405,
+ "learning_rate": 1.6367968469740416e-05,
+ "loss": 0.4652596712112427,
+ "step": 1267
+ },
+ {
+ "epoch": 0.6652675760755509,
+ "grad_norm": 0.15919541846683566,
+ "learning_rate": 1.6335754744956937e-05,
+ "loss": 0.43842798471450806,
+ "step": 1268
+ },
+ {
+ "epoch": 0.6657922350472193,
+ "grad_norm": 0.16322634187139418,
+ "learning_rate": 1.630357135092859e-05,
+ "loss": 0.45242664217948914,
+ "step": 1269
+ },
+ {
+ "epoch": 0.6663168940188877,
+ "grad_norm": 0.15644867734734783,
+ "learning_rate": 1.6271418375090445e-05,
+ "loss": 0.44801655411720276,
+ "step": 1270
+ },
+ {
+ "epoch": 0.6668415529905561,
+ "grad_norm": 0.1514717416124423,
+ "learning_rate": 1.623929590479496e-05,
+ "loss": 0.4506194293498993,
+ "step": 1271
+ },
+ {
+ "epoch": 0.6673662119622246,
+ "grad_norm": 0.1740345085495889,
+ "learning_rate": 1.6207204027311683e-05,
+ "loss": 0.4818177819252014,
+ "step": 1272
+ },
+ {
+ "epoch": 0.667890870933893,
+ "grad_norm": 0.15797701900765457,
+ "learning_rate": 1.6175142829827067e-05,
+ "loss": 0.4639382064342499,
+ "step": 1273
+ },
+ {
+ "epoch": 0.6684155299055614,
+ "grad_norm": 0.15663812316653875,
+ "learning_rate": 1.6143112399444198e-05,
+ "loss": 0.46217501163482666,
+ "step": 1274
+ },
+ {
+ "epoch": 0.6689401888772298,
+ "grad_norm": 0.15474018977043758,
+ "learning_rate": 1.6111112823182592e-05,
+ "loss": 0.43691688776016235,
+ "step": 1275
+ },
+ {
+ "epoch": 0.6694648478488983,
+ "grad_norm": 0.1615634942529431,
+ "learning_rate": 1.6079144187977923e-05,
+ "loss": 0.4412601590156555,
+ "step": 1276
+ },
+ {
+ "epoch": 0.6699895068205667,
+ "grad_norm": 0.15440022842754444,
+ "learning_rate": 1.604720658068184e-05,
+ "loss": 0.4523273706436157,
+ "step": 1277
+ },
+ {
+ "epoch": 0.670514165792235,
+ "grad_norm": 0.1634985665776242,
+ "learning_rate": 1.6015300088061664e-05,
+ "loss": 0.46264398097991943,
+ "step": 1278
+ },
+ {
+ "epoch": 0.6710388247639034,
+ "grad_norm": 0.15037886370689527,
+ "learning_rate": 1.598342479680018e-05,
+ "loss": 0.451465904712677,
+ "step": 1279
+ },
+ {
+ "epoch": 0.6715634837355718,
+ "grad_norm": 0.16895242437746613,
+ "learning_rate": 1.595158079349542e-05,
+ "loss": 0.4466145634651184,
+ "step": 1280
+ },
+ {
+ "epoch": 0.6720881427072403,
+ "grad_norm": 0.1629548326999642,
+ "learning_rate": 1.5919768164660425e-05,
+ "loss": 0.45415782928466797,
+ "step": 1281
+ },
+ {
+ "epoch": 0.6726128016789087,
+ "grad_norm": 0.1563872102268991,
+ "learning_rate": 1.588798699672298e-05,
+ "loss": 0.4386737048625946,
+ "step": 1282
+ },
+ {
+ "epoch": 0.6731374606505771,
+ "grad_norm": 0.16038172773344717,
+ "learning_rate": 1.5856237376025403e-05,
+ "loss": 0.446260929107666,
+ "step": 1283
+ },
+ {
+ "epoch": 0.6736621196222455,
+ "grad_norm": 0.16189780714407623,
+ "learning_rate": 1.582451938882431e-05,
+ "loss": 0.45458072423934937,
+ "step": 1284
+ },
+ {
+ "epoch": 0.674186778593914,
+ "grad_norm": 0.15204521785049285,
+ "learning_rate": 1.5792833121290367e-05,
+ "loss": 0.4425617754459381,
+ "step": 1285
+ },
+ {
+ "epoch": 0.6747114375655824,
+ "grad_norm": 0.16237376704792778,
+ "learning_rate": 1.5761178659508063e-05,
+ "loss": 0.44631800055503845,
+ "step": 1286
+ },
+ {
+ "epoch": 0.6752360965372508,
+ "grad_norm": 0.1582063618466069,
+ "learning_rate": 1.57295560894755e-05,
+ "loss": 0.44280534982681274,
+ "step": 1287
+ },
+ {
+ "epoch": 0.6757607555089192,
+ "grad_norm": 0.16587032349847394,
+ "learning_rate": 1.5697965497104116e-05,
+ "loss": 0.4495997130870819,
+ "step": 1288
+ },
+ {
+ "epoch": 0.6762854144805877,
+ "grad_norm": 0.1557269262755563,
+ "learning_rate": 1.566640696821848e-05,
+ "loss": 0.4493359625339508,
+ "step": 1289
+ },
+ {
+ "epoch": 0.6768100734522561,
+ "grad_norm": 0.16038490357012256,
+ "learning_rate": 1.5634880588556043e-05,
+ "loss": 0.4447544813156128,
+ "step": 1290
+ },
+ {
+ "epoch": 0.6773347324239245,
+ "grad_norm": 0.15355201908546406,
+ "learning_rate": 1.560338644376692e-05,
+ "loss": 0.4468432664871216,
+ "step": 1291
+ },
+ {
+ "epoch": 0.6778593913955928,
+ "grad_norm": 0.15077009515589776,
+ "learning_rate": 1.5571924619413668e-05,
+ "loss": 0.44959044456481934,
+ "step": 1292
+ },
+ {
+ "epoch": 0.6783840503672612,
+ "grad_norm": 0.1583038146613965,
+ "learning_rate": 1.5540495200971012e-05,
+ "loss": 0.42450425028800964,
+ "step": 1293
+ },
+ {
+ "epoch": 0.6789087093389297,
+ "grad_norm": 0.15950687190598087,
+ "learning_rate": 1.5509098273825635e-05,
+ "loss": 0.4527238607406616,
+ "step": 1294
+ },
+ {
+ "epoch": 0.6794333683105981,
+ "grad_norm": 0.1526262208847808,
+ "learning_rate": 1.5477733923275968e-05,
+ "loss": 0.4379724860191345,
+ "step": 1295
+ },
+ {
+ "epoch": 0.6799580272822665,
+ "grad_norm": 0.15541579650021778,
+ "learning_rate": 1.5446402234531934e-05,
+ "loss": 0.4462166130542755,
+ "step": 1296
+ },
+ {
+ "epoch": 0.6804826862539349,
+ "grad_norm": 0.15483930734313786,
+ "learning_rate": 1.5415103292714718e-05,
+ "loss": 0.4390607178211212,
+ "step": 1297
+ },
+ {
+ "epoch": 0.6810073452256034,
+ "grad_norm": 0.1578005731149154,
+ "learning_rate": 1.5383837182856532e-05,
+ "loss": 0.44462254643440247,
+ "step": 1298
+ },
+ {
+ "epoch": 0.6815320041972718,
+ "grad_norm": 0.15830183884958823,
+ "learning_rate": 1.5352603989900406e-05,
+ "loss": 0.4563537538051605,
+ "step": 1299
+ },
+ {
+ "epoch": 0.6820566631689402,
+ "grad_norm": 0.15605472163445522,
+ "learning_rate": 1.5321403798699923e-05,
+ "loss": 0.43852412700653076,
+ "step": 1300
+ },
+ {
+ "epoch": 0.6825813221406086,
+ "grad_norm": 0.15770470227141278,
+ "learning_rate": 1.5290236694019017e-05,
+ "loss": 0.4438081979751587,
+ "step": 1301
+ },
+ {
+ "epoch": 0.683105981112277,
+ "grad_norm": 0.16175339845012077,
+ "learning_rate": 1.5259102760531747e-05,
+ "loss": 0.46670734882354736,
+ "step": 1302
+ },
+ {
+ "epoch": 0.6836306400839455,
+ "grad_norm": 0.16359504716154477,
+ "learning_rate": 1.5228002082822036e-05,
+ "loss": 0.4590187668800354,
+ "step": 1303
+ },
+ {
+ "epoch": 0.6841552990556139,
+ "grad_norm": 0.15422879500254713,
+ "learning_rate": 1.5196934745383457e-05,
+ "loss": 0.45287445187568665,
+ "step": 1304
+ },
+ {
+ "epoch": 0.6846799580272823,
+ "grad_norm": 0.15929902143574254,
+ "learning_rate": 1.5165900832619015e-05,
+ "loss": 0.44286859035491943,
+ "step": 1305
+ },
+ {
+ "epoch": 0.6852046169989506,
+ "grad_norm": 0.14838868714016124,
+ "learning_rate": 1.5134900428840904e-05,
+ "loss": 0.45155346393585205,
+ "step": 1306
+ },
+ {
+ "epoch": 0.6857292759706191,
+ "grad_norm": 0.15494288640397672,
+ "learning_rate": 1.5103933618270283e-05,
+ "loss": 0.44010141491889954,
+ "step": 1307
+ },
+ {
+ "epoch": 0.6862539349422875,
+ "grad_norm": 0.1686262360598338,
+ "learning_rate": 1.5073000485037042e-05,
+ "loss": 0.446555495262146,
+ "step": 1308
+ },
+ {
+ "epoch": 0.6867785939139559,
+ "grad_norm": 0.16122110823982602,
+ "learning_rate": 1.5042101113179579e-05,
+ "loss": 0.44640374183654785,
+ "step": 1309
+ },
+ {
+ "epoch": 0.6873032528856243,
+ "grad_norm": 0.16849937797919867,
+ "learning_rate": 1.5011235586644572e-05,
+ "loss": 0.4465785026550293,
+ "step": 1310
+ },
+ {
+ "epoch": 0.6878279118572928,
+ "grad_norm": 0.16062971262800754,
+ "learning_rate": 1.498040398928674e-05,
+ "loss": 0.445321649312973,
+ "step": 1311
+ },
+ {
+ "epoch": 0.6883525708289612,
+ "grad_norm": 0.1608953185993034,
+ "learning_rate": 1.4949606404868638e-05,
+ "loss": 0.46518227458000183,
+ "step": 1312
+ },
+ {
+ "epoch": 0.6888772298006296,
+ "grad_norm": 0.15880255672830915,
+ "learning_rate": 1.491884291706041e-05,
+ "loss": 0.44582217931747437,
+ "step": 1313
+ },
+ {
+ "epoch": 0.689401888772298,
+ "grad_norm": 0.16715145200789822,
+ "learning_rate": 1.4888113609439563e-05,
+ "loss": 0.4435983896255493,
+ "step": 1314
+ },
+ {
+ "epoch": 0.6899265477439664,
+ "grad_norm": 0.15263740340694454,
+ "learning_rate": 1.485741856549075e-05,
+ "loss": 0.4405066967010498,
+ "step": 1315
+ },
+ {
+ "epoch": 0.6904512067156349,
+ "grad_norm": 0.16247034518813966,
+ "learning_rate": 1.4826757868605517e-05,
+ "loss": 0.4544215500354767,
+ "step": 1316
+ },
+ {
+ "epoch": 0.6909758656873033,
+ "grad_norm": 0.15649146705978162,
+ "learning_rate": 1.4796131602082142e-05,
+ "loss": 0.4454413056373596,
+ "step": 1317
+ },
+ {
+ "epoch": 0.6915005246589717,
+ "grad_norm": 0.16286718636058178,
+ "learning_rate": 1.4765539849125318e-05,
+ "loss": 0.4497183561325073,
+ "step": 1318
+ },
+ {
+ "epoch": 0.6920251836306401,
+ "grad_norm": 0.15812342085005582,
+ "learning_rate": 1.473498269284599e-05,
+ "loss": 0.445032000541687,
+ "step": 1319
+ },
+ {
+ "epoch": 0.6925498426023085,
+ "grad_norm": 0.1558101290511678,
+ "learning_rate": 1.4704460216261123e-05,
+ "loss": 0.45769619941711426,
+ "step": 1320
+ },
+ {
+ "epoch": 0.6930745015739769,
+ "grad_norm": 0.15642889059495207,
+ "learning_rate": 1.4673972502293428e-05,
+ "loss": 0.44788360595703125,
+ "step": 1321
+ },
+ {
+ "epoch": 0.6935991605456453,
+ "grad_norm": 0.15722513518113074,
+ "learning_rate": 1.4643519633771222e-05,
+ "loss": 0.449631929397583,
+ "step": 1322
+ },
+ {
+ "epoch": 0.6941238195173137,
+ "grad_norm": 0.15587633748081542,
+ "learning_rate": 1.4613101693428125e-05,
+ "loss": 0.4450187087059021,
+ "step": 1323
+ },
+ {
+ "epoch": 0.6946484784889821,
+ "grad_norm": 0.1577107074882479,
+ "learning_rate": 1.458271876390287e-05,
+ "loss": 0.45108890533447266,
+ "step": 1324
+ },
+ {
+ "epoch": 0.6951731374606506,
+ "grad_norm": 0.15649839486990696,
+ "learning_rate": 1.4552370927739073e-05,
+ "loss": 0.46430301666259766,
+ "step": 1325
+ },
+ {
+ "epoch": 0.695697796432319,
+ "grad_norm": 0.18573600420690636,
+ "learning_rate": 1.4522058267385008e-05,
+ "loss": 0.4701972007751465,
+ "step": 1326
+ },
+ {
+ "epoch": 0.6962224554039874,
+ "grad_norm": 0.16296497084037542,
+ "learning_rate": 1.4491780865193402e-05,
+ "loss": 0.4607710838317871,
+ "step": 1327
+ },
+ {
+ "epoch": 0.6967471143756558,
+ "grad_norm": 0.161124007217343,
+ "learning_rate": 1.446153880342117e-05,
+ "loss": 0.45095115900039673,
+ "step": 1328
+ },
+ {
+ "epoch": 0.6972717733473243,
+ "grad_norm": 0.1537884039166766,
+ "learning_rate": 1.4431332164229222e-05,
+ "loss": 0.424068808555603,
+ "step": 1329
+ },
+ {
+ "epoch": 0.6977964323189927,
+ "grad_norm": 0.16588834503456087,
+ "learning_rate": 1.4401161029682237e-05,
+ "loss": 0.4622839093208313,
+ "step": 1330
+ },
+ {
+ "epoch": 0.6983210912906611,
+ "grad_norm": 0.16100510604015544,
+ "learning_rate": 1.4371025481748435e-05,
+ "loss": 0.4660705626010895,
+ "step": 1331
+ },
+ {
+ "epoch": 0.6988457502623295,
+ "grad_norm": 0.15559888940590075,
+ "learning_rate": 1.4340925602299337e-05,
+ "loss": 0.4484976530075073,
+ "step": 1332
+ },
+ {
+ "epoch": 0.6993704092339978,
+ "grad_norm": 0.15503821966213488,
+ "learning_rate": 1.4310861473109616e-05,
+ "loss": 0.45636600255966187,
+ "step": 1333
+ },
+ {
+ "epoch": 0.6998950682056663,
+ "grad_norm": 0.15539074896109334,
+ "learning_rate": 1.4280833175856743e-05,
+ "loss": 0.45863068103790283,
+ "step": 1334
+ },
+ {
+ "epoch": 0.7004197271773347,
+ "grad_norm": 0.15664821009154872,
+ "learning_rate": 1.4250840792120896e-05,
+ "loss": 0.44043463468551636,
+ "step": 1335
+ },
+ {
+ "epoch": 0.7009443861490031,
+ "grad_norm": 0.15825478808520285,
+ "learning_rate": 1.4220884403384655e-05,
+ "loss": 0.46522051095962524,
+ "step": 1336
+ },
+ {
+ "epoch": 0.7014690451206715,
+ "grad_norm": 0.16004268568141544,
+ "learning_rate": 1.4190964091032843e-05,
+ "loss": 0.46114686131477356,
+ "step": 1337
+ },
+ {
+ "epoch": 0.70199370409234,
+ "grad_norm": 0.16413139949945635,
+ "learning_rate": 1.4161079936352234e-05,
+ "loss": 0.4457515478134155,
+ "step": 1338
+ },
+ {
+ "epoch": 0.7025183630640084,
+ "grad_norm": 0.1538890577972702,
+ "learning_rate": 1.4131232020531398e-05,
+ "loss": 0.4616093337535858,
+ "step": 1339
+ },
+ {
+ "epoch": 0.7030430220356768,
+ "grad_norm": 0.16240737854323284,
+ "learning_rate": 1.4101420424660434e-05,
+ "loss": 0.44631537795066833,
+ "step": 1340
+ },
+ {
+ "epoch": 0.7035676810073452,
+ "grad_norm": 0.15984708603406073,
+ "learning_rate": 1.4071645229730762e-05,
+ "loss": 0.4366717040538788,
+ "step": 1341
+ },
+ {
+ "epoch": 0.7040923399790137,
+ "grad_norm": 0.15149695695635904,
+ "learning_rate": 1.4041906516634946e-05,
+ "loss": 0.44129428267478943,
+ "step": 1342
+ },
+ {
+ "epoch": 0.7046169989506821,
+ "grad_norm": 0.15041167501768415,
+ "learning_rate": 1.4012204366166399e-05,
+ "loss": 0.43654775619506836,
+ "step": 1343
+ },
+ {
+ "epoch": 0.7051416579223505,
+ "grad_norm": 0.1596097627963756,
+ "learning_rate": 1.3982538859019217e-05,
+ "loss": 0.4652445614337921,
+ "step": 1344
+ },
+ {
+ "epoch": 0.7056663168940189,
+ "grad_norm": 0.15621704205027884,
+ "learning_rate": 1.3952910075787941e-05,
+ "loss": 0.44577041268348694,
+ "step": 1345
+ },
+ {
+ "epoch": 0.7061909758656874,
+ "grad_norm": 0.16073041098008783,
+ "learning_rate": 1.3923318096967341e-05,
+ "loss": 0.4391976594924927,
+ "step": 1346
+ },
+ {
+ "epoch": 0.7067156348373557,
+ "grad_norm": 0.14793853946372448,
+ "learning_rate": 1.3893763002952204e-05,
+ "loss": 0.4463799297809601,
+ "step": 1347
+ },
+ {
+ "epoch": 0.7072402938090241,
+ "grad_norm": 0.16176080358867354,
+ "learning_rate": 1.3864244874037097e-05,
+ "loss": 0.45183271169662476,
+ "step": 1348
+ },
+ {
+ "epoch": 0.7077649527806925,
+ "grad_norm": 0.16372771179026924,
+ "learning_rate": 1.3834763790416171e-05,
+ "loss": 0.44205576181411743,
+ "step": 1349
+ },
+ {
+ "epoch": 0.7082896117523609,
+ "grad_norm": 0.15135222516127125,
+ "learning_rate": 1.3805319832182937e-05,
+ "loss": 0.4390261173248291,
+ "step": 1350
+ },
+ {
+ "epoch": 0.7088142707240294,
+ "grad_norm": 0.15493183225792365,
+ "learning_rate": 1.377591307933002e-05,
+ "loss": 0.45837974548339844,
+ "step": 1351
+ },
+ {
+ "epoch": 0.7093389296956978,
+ "grad_norm": 0.16405906906516504,
+ "learning_rate": 1.3746543611749008e-05,
+ "loss": 0.45032668113708496,
+ "step": 1352
+ },
+ {
+ "epoch": 0.7098635886673662,
+ "grad_norm": 0.15725677976524505,
+ "learning_rate": 1.3717211509230155e-05,
+ "loss": 0.4486224949359894,
+ "step": 1353
+ },
+ {
+ "epoch": 0.7103882476390346,
+ "grad_norm": 0.15781469066983164,
+ "learning_rate": 1.3687916851462227e-05,
+ "loss": 0.4502638578414917,
+ "step": 1354
+ },
+ {
+ "epoch": 0.710912906610703,
+ "grad_norm": 0.16115007729054298,
+ "learning_rate": 1.3658659718032241e-05,
+ "loss": 0.4508914351463318,
+ "step": 1355
+ },
+ {
+ "epoch": 0.7114375655823715,
+ "grad_norm": 0.16755060475968744,
+ "learning_rate": 1.3629440188425286e-05,
+ "loss": 0.45653945207595825,
+ "step": 1356
+ },
+ {
+ "epoch": 0.7119622245540399,
+ "grad_norm": 0.1574994294756123,
+ "learning_rate": 1.3600258342024268e-05,
+ "loss": 0.4349849820137024,
+ "step": 1357
+ },
+ {
+ "epoch": 0.7124868835257083,
+ "grad_norm": 0.16104440922657715,
+ "learning_rate": 1.3571114258109758e-05,
+ "loss": 0.4501233696937561,
+ "step": 1358
+ },
+ {
+ "epoch": 0.7130115424973767,
+ "grad_norm": 0.15163184290939022,
+ "learning_rate": 1.3542008015859691e-05,
+ "loss": 0.44171035289764404,
+ "step": 1359
+ },
+ {
+ "epoch": 0.7135362014690452,
+ "grad_norm": 0.15955896674456582,
+ "learning_rate": 1.3512939694349211e-05,
+ "loss": 0.4558671712875366,
+ "step": 1360
+ },
+ {
+ "epoch": 0.7140608604407135,
+ "grad_norm": 0.1589376751536802,
+ "learning_rate": 1.3483909372550446e-05,
+ "loss": 0.4430835247039795,
+ "step": 1361
+ },
+ {
+ "epoch": 0.7145855194123819,
+ "grad_norm": 0.15979088397822155,
+ "learning_rate": 1.3454917129332279e-05,
+ "loss": 0.4426245391368866,
+ "step": 1362
+ },
+ {
+ "epoch": 0.7151101783840503,
+ "grad_norm": 0.15493179714072167,
+ "learning_rate": 1.3425963043460143e-05,
+ "loss": 0.4550474286079407,
+ "step": 1363
+ },
+ {
+ "epoch": 0.7156348373557188,
+ "grad_norm": 0.15396498968995032,
+ "learning_rate": 1.3397047193595808e-05,
+ "loss": 0.44220536947250366,
+ "step": 1364
+ },
+ {
+ "epoch": 0.7161594963273872,
+ "grad_norm": 0.1489346822667273,
+ "learning_rate": 1.336816965829717e-05,
+ "loss": 0.4437323212623596,
+ "step": 1365
+ },
+ {
+ "epoch": 0.7166841552990556,
+ "grad_norm": 0.16311964931734477,
+ "learning_rate": 1.333933051601802e-05,
+ "loss": 0.45308777689933777,
+ "step": 1366
+ },
+ {
+ "epoch": 0.717208814270724,
+ "grad_norm": 0.1534354619021179,
+ "learning_rate": 1.331052984510785e-05,
+ "loss": 0.4334886372089386,
+ "step": 1367
+ },
+ {
+ "epoch": 0.7177334732423925,
+ "grad_norm": 0.15541089758065404,
+ "learning_rate": 1.328176772381165e-05,
+ "loss": 0.45248323678970337,
+ "step": 1368
+ },
+ {
+ "epoch": 0.7182581322140609,
+ "grad_norm": 0.1617921305700596,
+ "learning_rate": 1.3253044230269651e-05,
+ "loss": 0.44092947244644165,
+ "step": 1369
+ },
+ {
+ "epoch": 0.7187827911857293,
+ "grad_norm": 0.15047353793760151,
+ "learning_rate": 1.3224359442517165e-05,
+ "loss": 0.4424414336681366,
+ "step": 1370
+ },
+ {
+ "epoch": 0.7193074501573977,
+ "grad_norm": 0.15958089102258743,
+ "learning_rate": 1.319571343848433e-05,
+ "loss": 0.4458580017089844,
+ "step": 1371
+ },
+ {
+ "epoch": 0.7198321091290661,
+ "grad_norm": 0.15349258306968733,
+ "learning_rate": 1.3167106295995924e-05,
+ "loss": 0.4285386800765991,
+ "step": 1372
+ },
+ {
+ "epoch": 0.7203567681007346,
+ "grad_norm": 0.16042529563759886,
+ "learning_rate": 1.3138538092771164e-05,
+ "loss": 0.4497690200805664,
+ "step": 1373
+ },
+ {
+ "epoch": 0.720881427072403,
+ "grad_norm": 0.1565113736588754,
+ "learning_rate": 1.3110008906423454e-05,
+ "loss": 0.43562617897987366,
+ "step": 1374
+ },
+ {
+ "epoch": 0.7214060860440713,
+ "grad_norm": 0.15177291086871506,
+ "learning_rate": 1.3081518814460215e-05,
+ "loss": 0.44456908106803894,
+ "step": 1375
+ },
+ {
+ "epoch": 0.7219307450157397,
+ "grad_norm": 0.15928490480439225,
+ "learning_rate": 1.3053067894282622e-05,
+ "loss": 0.46170780062675476,
+ "step": 1376
+ },
+ {
+ "epoch": 0.7224554039874082,
+ "grad_norm": 0.1554814469217892,
+ "learning_rate": 1.302465622318549e-05,
+ "loss": 0.4422169327735901,
+ "step": 1377
+ },
+ {
+ "epoch": 0.7229800629590766,
+ "grad_norm": 0.15375370214966386,
+ "learning_rate": 1.2996283878356947e-05,
+ "loss": 0.4564763903617859,
+ "step": 1378
+ },
+ {
+ "epoch": 0.723504721930745,
+ "grad_norm": 0.15942841969395802,
+ "learning_rate": 1.2967950936878311e-05,
+ "loss": 0.45121875405311584,
+ "step": 1379
+ },
+ {
+ "epoch": 0.7240293809024134,
+ "grad_norm": 0.15731308716774903,
+ "learning_rate": 1.2939657475723844e-05,
+ "loss": 0.4285027086734772,
+ "step": 1380
+ },
+ {
+ "epoch": 0.7245540398740818,
+ "grad_norm": 0.15997753613932147,
+ "learning_rate": 1.291140357176054e-05,
+ "loss": 0.4566386342048645,
+ "step": 1381
+ },
+ {
+ "epoch": 0.7250786988457503,
+ "grad_norm": 0.15333715267473727,
+ "learning_rate": 1.2883189301747927e-05,
+ "loss": 0.45093846321105957,
+ "step": 1382
+ },
+ {
+ "epoch": 0.7256033578174187,
+ "grad_norm": 0.1514510304709061,
+ "learning_rate": 1.2855014742337874e-05,
+ "loss": 0.44256967306137085,
+ "step": 1383
+ },
+ {
+ "epoch": 0.7261280167890871,
+ "grad_norm": 0.15624301298995746,
+ "learning_rate": 1.2826879970074346e-05,
+ "loss": 0.4361875653266907,
+ "step": 1384
+ },
+ {
+ "epoch": 0.7266526757607555,
+ "grad_norm": 0.15767850122006694,
+ "learning_rate": 1.279878506139322e-05,
+ "loss": 0.4477671980857849,
+ "step": 1385
+ },
+ {
+ "epoch": 0.727177334732424,
+ "grad_norm": 0.15537676969694064,
+ "learning_rate": 1.2770730092622072e-05,
+ "loss": 0.4339832663536072,
+ "step": 1386
+ },
+ {
+ "epoch": 0.7277019937040924,
+ "grad_norm": 0.15944113633281293,
+ "learning_rate": 1.2742715139979965e-05,
+ "loss": 0.45407935976982117,
+ "step": 1387
+ },
+ {
+ "epoch": 0.7282266526757608,
+ "grad_norm": 0.16763519518606218,
+ "learning_rate": 1.271474027957727e-05,
+ "loss": 0.46442392468452454,
+ "step": 1388
+ },
+ {
+ "epoch": 0.7287513116474291,
+ "grad_norm": 0.1564572774601902,
+ "learning_rate": 1.2686805587415423e-05,
+ "loss": 0.4463968276977539,
+ "step": 1389
+ },
+ {
+ "epoch": 0.7292759706190975,
+ "grad_norm": 0.16384626314986128,
+ "learning_rate": 1.2658911139386709e-05,
+ "loss": 0.45519885420799255,
+ "step": 1390
+ },
+ {
+ "epoch": 0.729800629590766,
+ "grad_norm": 0.1562023802820964,
+ "learning_rate": 1.2631057011274106e-05,
+ "loss": 0.4365222454071045,
+ "step": 1391
+ },
+ {
+ "epoch": 0.7303252885624344,
+ "grad_norm": 0.15597287232940926,
+ "learning_rate": 1.2603243278751042e-05,
+ "loss": 0.4328266680240631,
+ "step": 1392
+ },
+ {
+ "epoch": 0.7308499475341028,
+ "grad_norm": 0.14762833706132167,
+ "learning_rate": 1.2575470017381214e-05,
+ "loss": 0.43902426958084106,
+ "step": 1393
+ },
+ {
+ "epoch": 0.7313746065057712,
+ "grad_norm": 0.1688371515856315,
+ "learning_rate": 1.2547737302618351e-05,
+ "loss": 0.4649963676929474,
+ "step": 1394
+ },
+ {
+ "epoch": 0.7318992654774397,
+ "grad_norm": 0.15501514509586015,
+ "learning_rate": 1.2520045209806024e-05,
+ "loss": 0.4501558542251587,
+ "step": 1395
+ },
+ {
+ "epoch": 0.7324239244491081,
+ "grad_norm": 0.15516748917693537,
+ "learning_rate": 1.2492393814177457e-05,
+ "loss": 0.4405498206615448,
+ "step": 1396
+ },
+ {
+ "epoch": 0.7329485834207765,
+ "grad_norm": 0.17115400468199243,
+ "learning_rate": 1.2464783190855295e-05,
+ "loss": 0.4445379674434662,
+ "step": 1397
+ },
+ {
+ "epoch": 0.7334732423924449,
+ "grad_norm": 0.15966386341604186,
+ "learning_rate": 1.243721341485143e-05,
+ "loss": 0.4769889712333679,
+ "step": 1398
+ },
+ {
+ "epoch": 0.7339979013641134,
+ "grad_norm": 0.15614706359781472,
+ "learning_rate": 1.2409684561066766e-05,
+ "loss": 0.43643918633461,
+ "step": 1399
+ },
+ {
+ "epoch": 0.7345225603357818,
+ "grad_norm": 0.17467968960064226,
+ "learning_rate": 1.238219670429104e-05,
+ "loss": 0.45413005352020264,
+ "step": 1400
+ },
+ {
+ "epoch": 0.7350472193074502,
+ "grad_norm": 0.1544517430810819,
+ "learning_rate": 1.2354749919202598e-05,
+ "loss": 0.45218425989151,
+ "step": 1401
+ },
+ {
+ "epoch": 0.7355718782791185,
+ "grad_norm": 0.16551083673374484,
+ "learning_rate": 1.2327344280368217e-05,
+ "loss": 0.4459628462791443,
+ "step": 1402
+ },
+ {
+ "epoch": 0.7360965372507869,
+ "grad_norm": 0.17614490459613263,
+ "learning_rate": 1.2299979862242882e-05,
+ "loss": 0.4406280815601349,
+ "step": 1403
+ },
+ {
+ "epoch": 0.7366211962224554,
+ "grad_norm": 0.15554925394435032,
+ "learning_rate": 1.2272656739169588e-05,
+ "loss": 0.4380159378051758,
+ "step": 1404
+ },
+ {
+ "epoch": 0.7371458551941238,
+ "grad_norm": 0.16475466799870916,
+ "learning_rate": 1.2245374985379147e-05,
+ "loss": 0.4442812204360962,
+ "step": 1405
+ },
+ {
+ "epoch": 0.7376705141657922,
+ "grad_norm": 0.16877301168875541,
+ "learning_rate": 1.2218134674989978e-05,
+ "loss": 0.4589429199695587,
+ "step": 1406
+ },
+ {
+ "epoch": 0.7381951731374606,
+ "grad_norm": 0.1583278589457438,
+ "learning_rate": 1.2190935882007894e-05,
+ "loss": 0.44896745681762695,
+ "step": 1407
+ },
+ {
+ "epoch": 0.7387198321091291,
+ "grad_norm": 0.1582022541656192,
+ "learning_rate": 1.2163778680325947e-05,
+ "loss": 0.4367997646331787,
+ "step": 1408
+ },
+ {
+ "epoch": 0.7392444910807975,
+ "grad_norm": 0.1669573982300068,
+ "learning_rate": 1.2136663143724163e-05,
+ "loss": 0.4503386616706848,
+ "step": 1409
+ },
+ {
+ "epoch": 0.7397691500524659,
+ "grad_norm": 0.1637877523924539,
+ "learning_rate": 1.2109589345869387e-05,
+ "loss": 0.4660584330558777,
+ "step": 1410
+ },
+ {
+ "epoch": 0.7402938090241343,
+ "grad_norm": 0.1640136159943701,
+ "learning_rate": 1.2082557360315065e-05,
+ "loss": 0.45206528902053833,
+ "step": 1411
+ },
+ {
+ "epoch": 0.7408184679958028,
+ "grad_norm": 0.1596580168682554,
+ "learning_rate": 1.2055567260501054e-05,
+ "loss": 0.4405035376548767,
+ "step": 1412
+ },
+ {
+ "epoch": 0.7413431269674712,
+ "grad_norm": 0.15224511741945293,
+ "learning_rate": 1.2028619119753402e-05,
+ "loss": 0.42810291051864624,
+ "step": 1413
+ },
+ {
+ "epoch": 0.7418677859391396,
+ "grad_norm": 0.15631381931817098,
+ "learning_rate": 1.2001713011284193e-05,
+ "loss": 0.4421626925468445,
+ "step": 1414
+ },
+ {
+ "epoch": 0.742392444910808,
+ "grad_norm": 0.15721819981265503,
+ "learning_rate": 1.1974849008191288e-05,
+ "loss": 0.4540696144104004,
+ "step": 1415
+ },
+ {
+ "epoch": 0.7429171038824763,
+ "grad_norm": 0.15904478578744463,
+ "learning_rate": 1.1948027183458182e-05,
+ "loss": 0.4601830244064331,
+ "step": 1416
+ },
+ {
+ "epoch": 0.7434417628541448,
+ "grad_norm": 0.1530755605235123,
+ "learning_rate": 1.1921247609953747e-05,
+ "loss": 0.4552589952945709,
+ "step": 1417
+ },
+ {
+ "epoch": 0.7439664218258132,
+ "grad_norm": 0.15375751072258462,
+ "learning_rate": 1.1894510360432106e-05,
+ "loss": 0.42965373396873474,
+ "step": 1418
+ },
+ {
+ "epoch": 0.7444910807974816,
+ "grad_norm": 0.15678686023937333,
+ "learning_rate": 1.1867815507532377e-05,
+ "loss": 0.44684016704559326,
+ "step": 1419
+ },
+ {
+ "epoch": 0.74501573976915,
+ "grad_norm": 0.15486952983783606,
+ "learning_rate": 1.1841163123778502e-05,
+ "loss": 0.44415023922920227,
+ "step": 1420
+ },
+ {
+ "epoch": 0.7455403987408185,
+ "grad_norm": 0.15342611490545593,
+ "learning_rate": 1.1814553281579034e-05,
+ "loss": 0.4501304626464844,
+ "step": 1421
+ },
+ {
+ "epoch": 0.7460650577124869,
+ "grad_norm": 0.15374260012788166,
+ "learning_rate": 1.1787986053226954e-05,
+ "loss": 0.4521749019622803,
+ "step": 1422
+ },
+ {
+ "epoch": 0.7465897166841553,
+ "grad_norm": 0.1543553258370361,
+ "learning_rate": 1.176146151089949e-05,
+ "loss": 0.44753748178482056,
+ "step": 1423
+ },
+ {
+ "epoch": 0.7471143756558237,
+ "grad_norm": 0.15921670193848306,
+ "learning_rate": 1.1734979726657878e-05,
+ "loss": 0.4473114311695099,
+ "step": 1424
+ },
+ {
+ "epoch": 0.7476390346274921,
+ "grad_norm": 0.15023498686186612,
+ "learning_rate": 1.17085407724472e-05,
+ "loss": 0.4568778872489929,
+ "step": 1425
+ },
+ {
+ "epoch": 0.7481636935991606,
+ "grad_norm": 0.14601156920274583,
+ "learning_rate": 1.1682144720096168e-05,
+ "loss": 0.4389464855194092,
+ "step": 1426
+ },
+ {
+ "epoch": 0.748688352570829,
+ "grad_norm": 0.1520548095149398,
+ "learning_rate": 1.1655791641316954e-05,
+ "loss": 0.440733402967453,
+ "step": 1427
+ },
+ {
+ "epoch": 0.7492130115424974,
+ "grad_norm": 0.1591915163097993,
+ "learning_rate": 1.1629481607704963e-05,
+ "loss": 0.46245887875556946,
+ "step": 1428
+ },
+ {
+ "epoch": 0.7497376705141658,
+ "grad_norm": 0.14867295658447263,
+ "learning_rate": 1.1603214690738683e-05,
+ "loss": 0.4332764744758606,
+ "step": 1429
+ },
+ {
+ "epoch": 0.7502623294858342,
+ "grad_norm": 0.1498957762903134,
+ "learning_rate": 1.1576990961779447e-05,
+ "loss": 0.45215466618537903,
+ "step": 1430
+ },
+ {
+ "epoch": 0.7507869884575026,
+ "grad_norm": 0.15430555803491824,
+ "learning_rate": 1.1550810492071243e-05,
+ "loss": 0.4454396963119507,
+ "step": 1431
+ },
+ {
+ "epoch": 0.751311647429171,
+ "grad_norm": 0.1517461021936533,
+ "learning_rate": 1.1524673352740543e-05,
+ "loss": 0.43869835138320923,
+ "step": 1432
+ },
+ {
+ "epoch": 0.7518363064008394,
+ "grad_norm": 0.14541657994692936,
+ "learning_rate": 1.1498579614796121e-05,
+ "loss": 0.4373054504394531,
+ "step": 1433
+ },
+ {
+ "epoch": 0.7523609653725079,
+ "grad_norm": 0.15539513763786977,
+ "learning_rate": 1.1472529349128812e-05,
+ "loss": 0.45451128482818604,
+ "step": 1434
+ },
+ {
+ "epoch": 0.7528856243441763,
+ "grad_norm": 0.1482043113690946,
+ "learning_rate": 1.144652262651136e-05,
+ "loss": 0.43644773960113525,
+ "step": 1435
+ },
+ {
+ "epoch": 0.7534102833158447,
+ "grad_norm": 0.16293628094190815,
+ "learning_rate": 1.1420559517598205e-05,
+ "loss": 0.4687269330024719,
+ "step": 1436
+ },
+ {
+ "epoch": 0.7539349422875131,
+ "grad_norm": 0.1553393974390287,
+ "learning_rate": 1.1394640092925302e-05,
+ "loss": 0.4501398205757141,
+ "step": 1437
+ },
+ {
+ "epoch": 0.7544596012591815,
+ "grad_norm": 0.16214837694535802,
+ "learning_rate": 1.1368764422909919e-05,
+ "loss": 0.4542577862739563,
+ "step": 1438
+ },
+ {
+ "epoch": 0.75498426023085,
+ "grad_norm": 0.15838878372428297,
+ "learning_rate": 1.1342932577850473e-05,
+ "loss": 0.46228161454200745,
+ "step": 1439
+ },
+ {
+ "epoch": 0.7555089192025184,
+ "grad_norm": 0.15090419497122687,
+ "learning_rate": 1.13171446279263e-05,
+ "loss": 0.4495932161808014,
+ "step": 1440
+ },
+ {
+ "epoch": 0.7560335781741868,
+ "grad_norm": 0.1543376786458394,
+ "learning_rate": 1.1291400643197484e-05,
+ "loss": 0.451630175113678,
+ "step": 1441
+ },
+ {
+ "epoch": 0.7565582371458552,
+ "grad_norm": 0.1539470751610696,
+ "learning_rate": 1.1265700693604668e-05,
+ "loss": 0.447547972202301,
+ "step": 1442
+ },
+ {
+ "epoch": 0.7570828961175237,
+ "grad_norm": 0.15761047703305445,
+ "learning_rate": 1.1240044848968856e-05,
+ "loss": 0.4516204595565796,
+ "step": 1443
+ },
+ {
+ "epoch": 0.757607555089192,
+ "grad_norm": 0.15675279683704735,
+ "learning_rate": 1.1214433178991258e-05,
+ "loss": 0.4597971737384796,
+ "step": 1444
+ },
+ {
+ "epoch": 0.7581322140608604,
+ "grad_norm": 0.15592110994904113,
+ "learning_rate": 1.1188865753253022e-05,
+ "loss": 0.443898469209671,
+ "step": 1445
+ },
+ {
+ "epoch": 0.7586568730325288,
+ "grad_norm": 0.15597501691030619,
+ "learning_rate": 1.116334264121513e-05,
+ "loss": 0.4581306278705597,
+ "step": 1446
+ },
+ {
+ "epoch": 0.7591815320041972,
+ "grad_norm": 0.15552304728589958,
+ "learning_rate": 1.1137863912218154e-05,
+ "loss": 0.4475254416465759,
+ "step": 1447
+ },
+ {
+ "epoch": 0.7597061909758657,
+ "grad_norm": 0.15759990123898035,
+ "learning_rate": 1.1112429635482118e-05,
+ "loss": 0.43585264682769775,
+ "step": 1448
+ },
+ {
+ "epoch": 0.7602308499475341,
+ "grad_norm": 0.15964657264676635,
+ "learning_rate": 1.1087039880106247e-05,
+ "loss": 0.46786290407180786,
+ "step": 1449
+ },
+ {
+ "epoch": 0.7607555089192025,
+ "grad_norm": 0.15155804051218852,
+ "learning_rate": 1.106169471506882e-05,
+ "loss": 0.42292726039886475,
+ "step": 1450
+ },
+ {
+ "epoch": 0.7612801678908709,
+ "grad_norm": 0.15519270966930884,
+ "learning_rate": 1.103639420922698e-05,
+ "loss": 0.4248972237110138,
+ "step": 1451
+ },
+ {
+ "epoch": 0.7618048268625394,
+ "grad_norm": 0.1526330511128821,
+ "learning_rate": 1.1011138431316537e-05,
+ "loss": 0.4317784309387207,
+ "step": 1452
+ },
+ {
+ "epoch": 0.7623294858342078,
+ "grad_norm": 0.15797682979272226,
+ "learning_rate": 1.0985927449951779e-05,
+ "loss": 0.44506698846817017,
+ "step": 1453
+ },
+ {
+ "epoch": 0.7628541448058762,
+ "grad_norm": 0.15479979838505462,
+ "learning_rate": 1.096076133362531e-05,
+ "loss": 0.4471631348133087,
+ "step": 1454
+ },
+ {
+ "epoch": 0.7633788037775446,
+ "grad_norm": 0.15712267274447342,
+ "learning_rate": 1.093564015070783e-05,
+ "loss": 0.4321915805339813,
+ "step": 1455
+ },
+ {
+ "epoch": 0.7639034627492131,
+ "grad_norm": 0.15857431494180016,
+ "learning_rate": 1.0910563969447969e-05,
+ "loss": 0.44455766677856445,
+ "step": 1456
+ },
+ {
+ "epoch": 0.7644281217208815,
+ "grad_norm": 0.15377057080997789,
+ "learning_rate": 1.0885532857972097e-05,
+ "loss": 0.4567924439907074,
+ "step": 1457
+ },
+ {
+ "epoch": 0.7649527806925498,
+ "grad_norm": 0.15544844005009104,
+ "learning_rate": 1.0860546884284143e-05,
+ "loss": 0.4491572976112366,
+ "step": 1458
+ },
+ {
+ "epoch": 0.7654774396642182,
+ "grad_norm": 0.16210735566479478,
+ "learning_rate": 1.08356061162654e-05,
+ "loss": 0.43823331594467163,
+ "step": 1459
+ },
+ {
+ "epoch": 0.7660020986358866,
+ "grad_norm": 0.15213559486379452,
+ "learning_rate": 1.0810710621674355e-05,
+ "loss": 0.45375770330429077,
+ "step": 1460
+ },
+ {
+ "epoch": 0.7665267576075551,
+ "grad_norm": 0.1727832306402992,
+ "learning_rate": 1.0785860468146494e-05,
+ "loss": 0.4639572501182556,
+ "step": 1461
+ },
+ {
+ "epoch": 0.7670514165792235,
+ "grad_norm": 0.1637555902013663,
+ "learning_rate": 1.076105572319412e-05,
+ "loss": 0.4355923533439636,
+ "step": 1462
+ },
+ {
+ "epoch": 0.7675760755508919,
+ "grad_norm": 0.16646936030470866,
+ "learning_rate": 1.0736296454206166e-05,
+ "loss": 0.4650861322879791,
+ "step": 1463
+ },
+ {
+ "epoch": 0.7681007345225603,
+ "grad_norm": 0.14987620615431385,
+ "learning_rate": 1.0711582728448033e-05,
+ "loss": 0.45445430278778076,
+ "step": 1464
+ },
+ {
+ "epoch": 0.7686253934942288,
+ "grad_norm": 0.16428618403547932,
+ "learning_rate": 1.0686914613061377e-05,
+ "loss": 0.46050935983657837,
+ "step": 1465
+ },
+ {
+ "epoch": 0.7691500524658972,
+ "grad_norm": 0.15052999053128632,
+ "learning_rate": 1.0662292175063949e-05,
+ "loss": 0.43385741114616394,
+ "step": 1466
+ },
+ {
+ "epoch": 0.7696747114375656,
+ "grad_norm": 0.1654949402740152,
+ "learning_rate": 1.0637715481349397e-05,
+ "loss": 0.4434410333633423,
+ "step": 1467
+ },
+ {
+ "epoch": 0.770199370409234,
+ "grad_norm": 0.15923836508490036,
+ "learning_rate": 1.0613184598687088e-05,
+ "loss": 0.4410185217857361,
+ "step": 1468
+ },
+ {
+ "epoch": 0.7707240293809025,
+ "grad_norm": 0.15813007111170696,
+ "learning_rate": 1.0588699593721957e-05,
+ "loss": 0.43588417768478394,
+ "step": 1469
+ },
+ {
+ "epoch": 0.7712486883525709,
+ "grad_norm": 0.16978817898431764,
+ "learning_rate": 1.0564260532974274e-05,
+ "loss": 0.4514843225479126,
+ "step": 1470
+ },
+ {
+ "epoch": 0.7717733473242392,
+ "grad_norm": 0.1613844852697828,
+ "learning_rate": 1.0539867482839499e-05,
+ "loss": 0.4554520845413208,
+ "step": 1471
+ },
+ {
+ "epoch": 0.7722980062959076,
+ "grad_norm": 0.1587788181984633,
+ "learning_rate": 1.0515520509588075e-05,
+ "loss": 0.43200671672821045,
+ "step": 1472
+ },
+ {
+ "epoch": 0.772822665267576,
+ "grad_norm": 0.1539930824827353,
+ "learning_rate": 1.0491219679365283e-05,
+ "loss": 0.4343116283416748,
+ "step": 1473
+ },
+ {
+ "epoch": 0.7733473242392445,
+ "grad_norm": 0.1561612537773709,
+ "learning_rate": 1.0466965058191049e-05,
+ "loss": 0.4477815330028534,
+ "step": 1474
+ },
+ {
+ "epoch": 0.7738719832109129,
+ "grad_norm": 0.169990777691657,
+ "learning_rate": 1.0442756711959745e-05,
+ "loss": 0.4476611614227295,
+ "step": 1475
+ },
+ {
+ "epoch": 0.7743966421825813,
+ "grad_norm": 0.1624313049193793,
+ "learning_rate": 1.041859470644002e-05,
+ "loss": 0.47136399149894714,
+ "step": 1476
+ },
+ {
+ "epoch": 0.7749213011542497,
+ "grad_norm": 0.15993315458994328,
+ "learning_rate": 1.0394479107274645e-05,
+ "loss": 0.4672670364379883,
+ "step": 1477
+ },
+ {
+ "epoch": 0.7754459601259182,
+ "grad_norm": 0.16755349272347217,
+ "learning_rate": 1.0370409979980294e-05,
+ "loss": 0.45135706663131714,
+ "step": 1478
+ },
+ {
+ "epoch": 0.7759706190975866,
+ "grad_norm": 0.15665059815830362,
+ "learning_rate": 1.0346387389947407e-05,
+ "loss": 0.45178359746932983,
+ "step": 1479
+ },
+ {
+ "epoch": 0.776495278069255,
+ "grad_norm": 0.1554612731452571,
+ "learning_rate": 1.0322411402439983e-05,
+ "loss": 0.4412590265274048,
+ "step": 1480
+ },
+ {
+ "epoch": 0.7770199370409234,
+ "grad_norm": 0.1532496869312065,
+ "learning_rate": 1.0298482082595408e-05,
+ "loss": 0.44104987382888794,
+ "step": 1481
+ },
+ {
+ "epoch": 0.7775445960125918,
+ "grad_norm": 0.15065868316960426,
+ "learning_rate": 1.0274599495424286e-05,
+ "loss": 0.4336603283882141,
+ "step": 1482
+ },
+ {
+ "epoch": 0.7780692549842603,
+ "grad_norm": 0.1548373981760097,
+ "learning_rate": 1.0250763705810267e-05,
+ "loss": 0.4658375680446625,
+ "step": 1483
+ },
+ {
+ "epoch": 0.7785939139559287,
+ "grad_norm": 0.15470181175866582,
+ "learning_rate": 1.0226974778509846e-05,
+ "loss": 0.45232075452804565,
+ "step": 1484
+ },
+ {
+ "epoch": 0.779118572927597,
+ "grad_norm": 0.15507371187727106,
+ "learning_rate": 1.0203232778152228e-05,
+ "loss": 0.43640458583831787,
+ "step": 1485
+ },
+ {
+ "epoch": 0.7796432318992654,
+ "grad_norm": 0.15269440736544085,
+ "learning_rate": 1.01795377692391e-05,
+ "loss": 0.4400939345359802,
+ "step": 1486
+ },
+ {
+ "epoch": 0.7801678908709339,
+ "grad_norm": 0.15610899550395085,
+ "learning_rate": 1.0155889816144495e-05,
+ "loss": 0.4546602666378021,
+ "step": 1487
+ },
+ {
+ "epoch": 0.7806925498426023,
+ "grad_norm": 0.1559328242601668,
+ "learning_rate": 1.0132288983114611e-05,
+ "loss": 0.45510709285736084,
+ "step": 1488
+ },
+ {
+ "epoch": 0.7812172088142707,
+ "grad_norm": 0.1537255194600787,
+ "learning_rate": 1.0108735334267639e-05,
+ "loss": 0.4475342631340027,
+ "step": 1489
+ },
+ {
+ "epoch": 0.7817418677859391,
+ "grad_norm": 0.14616575061664402,
+ "learning_rate": 1.0085228933593558e-05,
+ "loss": 0.4378190040588379,
+ "step": 1490
+ },
+ {
+ "epoch": 0.7822665267576075,
+ "grad_norm": 0.1624360967746598,
+ "learning_rate": 1.0061769844953999e-05,
+ "loss": 0.4439539909362793,
+ "step": 1491
+ },
+ {
+ "epoch": 0.782791185729276,
+ "grad_norm": 0.15407336125298413,
+ "learning_rate": 1.003835813208206e-05,
+ "loss": 0.44074907898902893,
+ "step": 1492
+ },
+ {
+ "epoch": 0.7833158447009444,
+ "grad_norm": 0.15786980710136125,
+ "learning_rate": 1.001499385858211e-05,
+ "loss": 0.4491667151451111,
+ "step": 1493
+ },
+ {
+ "epoch": 0.7838405036726128,
+ "grad_norm": 0.15338836738460088,
+ "learning_rate": 9.991677087929665e-06,
+ "loss": 0.446003794670105,
+ "step": 1494
+ },
+ {
+ "epoch": 0.7843651626442812,
+ "grad_norm": 0.156710894892,
+ "learning_rate": 9.968407883471165e-06,
+ "loss": 0.43604689836502075,
+ "step": 1495
+ },
+ {
+ "epoch": 0.7848898216159497,
+ "grad_norm": 0.16100539063191666,
+ "learning_rate": 9.945186308423832e-06,
+ "loss": 0.4431147873401642,
+ "step": 1496
+ },
+ {
+ "epoch": 0.7854144805876181,
+ "grad_norm": 0.15776018851122997,
+ "learning_rate": 9.92201242587548e-06,
+ "loss": 0.4532504975795746,
+ "step": 1497
+ },
+ {
+ "epoch": 0.7859391395592865,
+ "grad_norm": 0.1540333423540934,
+ "learning_rate": 9.898886298784368e-06,
+ "loss": 0.43605154752731323,
+ "step": 1498
+ },
+ {
+ "epoch": 0.7864637985309548,
+ "grad_norm": 0.16009602938972295,
+ "learning_rate": 9.875807989979005e-06,
+ "loss": 0.4605238437652588,
+ "step": 1499
+ },
+ {
+ "epoch": 0.7869884575026233,
+ "grad_norm": 0.15639368804171708,
+ "learning_rate": 9.852777562157983e-06,
+ "loss": 0.4509708881378174,
+ "step": 1500
+ },
+ {
+ "epoch": 0.7875131164742917,
+ "grad_norm": 0.15074011188741518,
+ "learning_rate": 9.829795077889828e-06,
+ "loss": 0.4527304172515869,
+ "step": 1501
+ },
+ {
+ "epoch": 0.7880377754459601,
+ "grad_norm": 0.15155083996490956,
+ "learning_rate": 9.806860599612799e-06,
+ "loss": 0.4306413531303406,
+ "step": 1502
+ },
+ {
+ "epoch": 0.7885624344176285,
+ "grad_norm": 0.1478956630795445,
+ "learning_rate": 9.783974189634734e-06,
+ "loss": 0.4434714913368225,
+ "step": 1503
+ },
+ {
+ "epoch": 0.789087093389297,
+ "grad_norm": 0.15327668434806432,
+ "learning_rate": 9.761135910132902e-06,
+ "loss": 0.44431227445602417,
+ "step": 1504
+ },
+ {
+ "epoch": 0.7896117523609654,
+ "grad_norm": 0.2243177585497212,
+ "learning_rate": 9.738345823153783e-06,
+ "loss": 0.4541007876396179,
+ "step": 1505
+ },
+ {
+ "epoch": 0.7901364113326338,
+ "grad_norm": 0.1600239865812282,
+ "learning_rate": 9.715603990612945e-06,
+ "loss": 0.4456910490989685,
+ "step": 1506
+ },
+ {
+ "epoch": 0.7906610703043022,
+ "grad_norm": 0.15304634075332066,
+ "learning_rate": 9.692910474294858e-06,
+ "loss": 0.43965545296669006,
+ "step": 1507
+ },
+ {
+ "epoch": 0.7911857292759706,
+ "grad_norm": 0.15657752062053032,
+ "learning_rate": 9.670265335852725e-06,
+ "loss": 0.43358081579208374,
+ "step": 1508
+ },
+ {
+ "epoch": 0.7917103882476391,
+ "grad_norm": 0.16110057783374115,
+ "learning_rate": 9.647668636808304e-06,
+ "loss": 0.45711368322372437,
+ "step": 1509
+ },
+ {
+ "epoch": 0.7922350472193075,
+ "grad_norm": 0.1551133690299623,
+ "learning_rate": 9.625120438551791e-06,
+ "loss": 0.4613519608974457,
+ "step": 1510
+ },
+ {
+ "epoch": 0.7927597061909759,
+ "grad_norm": 0.15690897380725946,
+ "learning_rate": 9.602620802341581e-06,
+ "loss": 0.4472261071205139,
+ "step": 1511
+ },
+ {
+ "epoch": 0.7932843651626443,
+ "grad_norm": 0.147796599322911,
+ "learning_rate": 9.580169789304152e-06,
+ "loss": 0.4447203576564789,
+ "step": 1512
+ },
+ {
+ "epoch": 0.7938090241343126,
+ "grad_norm": 0.152019559199633,
+ "learning_rate": 9.557767460433878e-06,
+ "loss": 0.445905864238739,
+ "step": 1513
+ },
+ {
+ "epoch": 0.7943336831059811,
+ "grad_norm": 0.1574763924872037,
+ "learning_rate": 9.535413876592872e-06,
+ "loss": 0.4502032995223999,
+ "step": 1514
+ },
+ {
+ "epoch": 0.7948583420776495,
+ "grad_norm": 0.1524218315499056,
+ "learning_rate": 9.513109098510816e-06,
+ "loss": 0.43753111362457275,
+ "step": 1515
+ },
+ {
+ "epoch": 0.7953830010493179,
+ "grad_norm": 0.15589147437937434,
+ "learning_rate": 9.490853186784801e-06,
+ "loss": 0.44524312019348145,
+ "step": 1516
+ },
+ {
+ "epoch": 0.7959076600209863,
+ "grad_norm": 0.15604009714081407,
+ "learning_rate": 9.468646201879156e-06,
+ "loss": 0.44141292572021484,
+ "step": 1517
+ },
+ {
+ "epoch": 0.7964323189926548,
+ "grad_norm": 0.1562432554004435,
+ "learning_rate": 9.44648820412529e-06,
+ "loss": 0.461681067943573,
+ "step": 1518
+ },
+ {
+ "epoch": 0.7969569779643232,
+ "grad_norm": 0.15233588531495665,
+ "learning_rate": 9.424379253721513e-06,
+ "loss": 0.4529497027397156,
+ "step": 1519
+ },
+ {
+ "epoch": 0.7974816369359916,
+ "grad_norm": 0.16856991685290923,
+ "learning_rate": 9.40231941073291e-06,
+ "loss": 0.4483596086502075,
+ "step": 1520
+ },
+ {
+ "epoch": 0.79800629590766,
+ "grad_norm": 0.1546876712981077,
+ "learning_rate": 9.380308735091125e-06,
+ "loss": 0.43846625089645386,
+ "step": 1521
+ },
+ {
+ "epoch": 0.7985309548793285,
+ "grad_norm": 0.15198637556890449,
+ "learning_rate": 9.35834728659424e-06,
+ "loss": 0.4625553488731384,
+ "step": 1522
+ },
+ {
+ "epoch": 0.7990556138509969,
+ "grad_norm": 0.1519777007198701,
+ "learning_rate": 9.336435124906592e-06,
+ "loss": 0.44183820486068726,
+ "step": 1523
+ },
+ {
+ "epoch": 0.7995802728226653,
+ "grad_norm": 0.16125823074979284,
+ "learning_rate": 9.314572309558616e-06,
+ "loss": 0.46057307720184326,
+ "step": 1524
+ },
+ {
+ "epoch": 0.8001049317943337,
+ "grad_norm": 0.16283323624848034,
+ "learning_rate": 9.2927588999467e-06,
+ "loss": 0.4530155062675476,
+ "step": 1525
+ },
+ {
+ "epoch": 0.8006295907660022,
+ "grad_norm": 0.14998654489027738,
+ "learning_rate": 9.270994955332984e-06,
+ "loss": 0.43155500292778015,
+ "step": 1526
+ },
+ {
+ "epoch": 0.8011542497376705,
+ "grad_norm": 0.15683095030848093,
+ "learning_rate": 9.249280534845241e-06,
+ "loss": 0.45097050070762634,
+ "step": 1527
+ },
+ {
+ "epoch": 0.8016789087093389,
+ "grad_norm": 0.1561578254492378,
+ "learning_rate": 9.227615697476674e-06,
+ "loss": 0.42797762155532837,
+ "step": 1528
+ },
+ {
+ "epoch": 0.8022035676810073,
+ "grad_norm": 0.15642385261469652,
+ "learning_rate": 9.206000502085813e-06,
+ "loss": 0.4560159742832184,
+ "step": 1529
+ },
+ {
+ "epoch": 0.8027282266526757,
+ "grad_norm": 0.1623126279097499,
+ "learning_rate": 9.1844350073963e-06,
+ "loss": 0.4704165458679199,
+ "step": 1530
+ },
+ {
+ "epoch": 0.8032528856243442,
+ "grad_norm": 0.15090934786253773,
+ "learning_rate": 9.162919271996755e-06,
+ "loss": 0.44642555713653564,
+ "step": 1531
+ },
+ {
+ "epoch": 0.8037775445960126,
+ "grad_norm": 0.15305216978534084,
+ "learning_rate": 9.14145335434062e-06,
+ "loss": 0.4634283781051636,
+ "step": 1532
+ },
+ {
+ "epoch": 0.804302203567681,
+ "grad_norm": 0.14653219861345698,
+ "learning_rate": 9.120037312745979e-06,
+ "loss": 0.44544070959091187,
+ "step": 1533
+ },
+ {
+ "epoch": 0.8048268625393494,
+ "grad_norm": 0.16247397938502464,
+ "learning_rate": 9.098671205395422e-06,
+ "loss": 0.4552367925643921,
+ "step": 1534
+ },
+ {
+ "epoch": 0.8053515215110179,
+ "grad_norm": 0.14963174165902007,
+ "learning_rate": 9.077355090335892e-06,
+ "loss": 0.43211787939071655,
+ "step": 1535
+ },
+ {
+ "epoch": 0.8058761804826863,
+ "grad_norm": 0.1554853846772283,
+ "learning_rate": 9.056089025478495e-06,
+ "loss": 0.4319286644458771,
+ "step": 1536
+ },
+ {
+ "epoch": 0.8064008394543547,
+ "grad_norm": 0.15840895821310472,
+ "learning_rate": 9.03487306859837e-06,
+ "loss": 0.4494994878768921,
+ "step": 1537
+ },
+ {
+ "epoch": 0.8069254984260231,
+ "grad_norm": 0.1543496716800982,
+ "learning_rate": 9.013707277334515e-06,
+ "loss": 0.45500510931015015,
+ "step": 1538
+ },
+ {
+ "epoch": 0.8074501573976915,
+ "grad_norm": 0.14797282318550045,
+ "learning_rate": 8.992591709189651e-06,
+ "loss": 0.4385896325111389,
+ "step": 1539
+ },
+ {
+ "epoch": 0.8079748163693599,
+ "grad_norm": 0.17045570682787758,
+ "learning_rate": 8.971526421530066e-06,
+ "loss": 0.4585988521575928,
+ "step": 1540
+ },
+ {
+ "epoch": 0.8084994753410283,
+ "grad_norm": 0.14758220713044884,
+ "learning_rate": 8.950511471585407e-06,
+ "loss": 0.4295082092285156,
+ "step": 1541
+ },
+ {
+ "epoch": 0.8090241343126967,
+ "grad_norm": 0.16005752068175036,
+ "learning_rate": 8.929546916448597e-06,
+ "loss": 0.4507679343223572,
+ "step": 1542
+ },
+ {
+ "epoch": 0.8095487932843651,
+ "grad_norm": 0.14884684083313074,
+ "learning_rate": 8.908632813075641e-06,
+ "loss": 0.43547600507736206,
+ "step": 1543
+ },
+ {
+ "epoch": 0.8100734522560336,
+ "grad_norm": 0.14860107595788136,
+ "learning_rate": 8.88776921828547e-06,
+ "loss": 0.42700642347335815,
+ "step": 1544
+ },
+ {
+ "epoch": 0.810598111227702,
+ "grad_norm": 0.1472630611843838,
+ "learning_rate": 8.86695618875981e-06,
+ "loss": 0.4473339319229126,
+ "step": 1545
+ },
+ {
+ "epoch": 0.8111227701993704,
+ "grad_norm": 0.15105523788317893,
+ "learning_rate": 8.846193781043e-06,
+ "loss": 0.44923150539398193,
+ "step": 1546
+ },
+ {
+ "epoch": 0.8116474291710388,
+ "grad_norm": 0.16121409015263036,
+ "learning_rate": 8.82548205154185e-06,
+ "loss": 0.4653576612472534,
+ "step": 1547
+ },
+ {
+ "epoch": 0.8121720881427072,
+ "grad_norm": 0.1595195150959421,
+ "learning_rate": 8.804821056525497e-06,
+ "loss": 0.4522424638271332,
+ "step": 1548
+ },
+ {
+ "epoch": 0.8126967471143757,
+ "grad_norm": 0.15163510106316194,
+ "learning_rate": 8.784210852125232e-06,
+ "loss": 0.43657630681991577,
+ "step": 1549
+ },
+ {
+ "epoch": 0.8132214060860441,
+ "grad_norm": 0.15155109347013793,
+ "learning_rate": 8.763651494334377e-06,
+ "loss": 0.43183624744415283,
+ "step": 1550
+ },
+ {
+ "epoch": 0.8137460650577125,
+ "grad_norm": 0.14930024958763324,
+ "learning_rate": 8.7431430390081e-06,
+ "loss": 0.43563947081565857,
+ "step": 1551
+ },
+ {
+ "epoch": 0.8142707240293809,
+ "grad_norm": 0.15361200153430712,
+ "learning_rate": 8.722685541863283e-06,
+ "loss": 0.4387405514717102,
+ "step": 1552
+ },
+ {
+ "epoch": 0.8147953830010494,
+ "grad_norm": 0.15549528796193493,
+ "learning_rate": 8.702279058478366e-06,
+ "loss": 0.44347938895225525,
+ "step": 1553
+ },
+ {
+ "epoch": 0.8153200419727177,
+ "grad_norm": 0.16243042867861712,
+ "learning_rate": 8.681923644293194e-06,
+ "loss": 0.44852834939956665,
+ "step": 1554
+ },
+ {
+ "epoch": 0.8158447009443861,
+ "grad_norm": 0.15706129652944026,
+ "learning_rate": 8.661619354608873e-06,
+ "loss": 0.45419150590896606,
+ "step": 1555
+ },
+ {
+ "epoch": 0.8163693599160545,
+ "grad_norm": 0.14854258573134127,
+ "learning_rate": 8.641366244587611e-06,
+ "loss": 0.44866859912872314,
+ "step": 1556
+ },
+ {
+ "epoch": 0.816894018887723,
+ "grad_norm": 0.14896937678017533,
+ "learning_rate": 8.62116436925258e-06,
+ "loss": 0.44596076011657715,
+ "step": 1557
+ },
+ {
+ "epoch": 0.8174186778593914,
+ "grad_norm": 0.15614508495620508,
+ "learning_rate": 8.601013783487746e-06,
+ "loss": 0.45645207166671753,
+ "step": 1558
+ },
+ {
+ "epoch": 0.8179433368310598,
+ "grad_norm": 0.1486538543050606,
+ "learning_rate": 8.580914542037734e-06,
+ "loss": 0.4592300355434418,
+ "step": 1559
+ },
+ {
+ "epoch": 0.8184679958027282,
+ "grad_norm": 0.15475320756830518,
+ "learning_rate": 8.560866699507702e-06,
+ "loss": 0.4605134129524231,
+ "step": 1560
+ },
+ {
+ "epoch": 0.8189926547743966,
+ "grad_norm": 0.14754406728057914,
+ "learning_rate": 8.540870310363137e-06,
+ "loss": 0.43651673197746277,
+ "step": 1561
+ },
+ {
+ "epoch": 0.8195173137460651,
+ "grad_norm": 0.14871990701077825,
+ "learning_rate": 8.520925428929755e-06,
+ "loss": 0.44830837845802307,
+ "step": 1562
+ },
+ {
+ "epoch": 0.8200419727177335,
+ "grad_norm": 0.15358241317712973,
+ "learning_rate": 8.501032109393334e-06,
+ "loss": 0.4546644687652588,
+ "step": 1563
+ },
+ {
+ "epoch": 0.8205666316894019,
+ "grad_norm": 0.15650856948126138,
+ "learning_rate": 8.481190405799569e-06,
+ "loss": 0.4423891305923462,
+ "step": 1564
+ },
+ {
+ "epoch": 0.8210912906610703,
+ "grad_norm": 0.15474996746242287,
+ "learning_rate": 8.461400372053923e-06,
+ "loss": 0.44242313504219055,
+ "step": 1565
+ },
+ {
+ "epoch": 0.8216159496327388,
+ "grad_norm": 0.16544926591962897,
+ "learning_rate": 8.441662061921497e-06,
+ "loss": 0.4487020969390869,
+ "step": 1566
+ },
+ {
+ "epoch": 0.8221406086044072,
+ "grad_norm": 0.1489733586650647,
+ "learning_rate": 8.421975529026855e-06,
+ "loss": 0.4371356964111328,
+ "step": 1567
+ },
+ {
+ "epoch": 0.8226652675760755,
+ "grad_norm": 0.15995455652906124,
+ "learning_rate": 8.402340826853907e-06,
+ "loss": 0.4566021263599396,
+ "step": 1568
+ },
+ {
+ "epoch": 0.8231899265477439,
+ "grad_norm": 0.1500716152622102,
+ "learning_rate": 8.382758008745726e-06,
+ "loss": 0.43888187408447266,
+ "step": 1569
+ },
+ {
+ "epoch": 0.8237145855194123,
+ "grad_norm": 0.15554904485712012,
+ "learning_rate": 8.363227127904464e-06,
+ "loss": 0.4584001898765564,
+ "step": 1570
+ },
+ {
+ "epoch": 0.8242392444910808,
+ "grad_norm": 0.14820917815884277,
+ "learning_rate": 8.343748237391144e-06,
+ "loss": 0.43289071321487427,
+ "step": 1571
+ },
+ {
+ "epoch": 0.8247639034627492,
+ "grad_norm": 0.15801200747859714,
+ "learning_rate": 8.324321390125547e-06,
+ "loss": 0.44196009635925293,
+ "step": 1572
+ },
+ {
+ "epoch": 0.8252885624344176,
+ "grad_norm": 0.15323277975380797,
+ "learning_rate": 8.304946638886078e-06,
+ "loss": 0.43244463205337524,
+ "step": 1573
+ },
+ {
+ "epoch": 0.825813221406086,
+ "grad_norm": 0.15076940176344772,
+ "learning_rate": 8.285624036309583e-06,
+ "loss": 0.44401073455810547,
+ "step": 1574
+ },
+ {
+ "epoch": 0.8263378803777545,
+ "grad_norm": 0.1544958300475393,
+ "learning_rate": 8.266353634891266e-06,
+ "loss": 0.45198434591293335,
+ "step": 1575
+ },
+ {
+ "epoch": 0.8268625393494229,
+ "grad_norm": 0.15093528292068448,
+ "learning_rate": 8.247135486984486e-06,
+ "loss": 0.4403747618198395,
+ "step": 1576
+ },
+ {
+ "epoch": 0.8273871983210913,
+ "grad_norm": 0.14680662784017148,
+ "learning_rate": 8.227969644800648e-06,
+ "loss": 0.4375043213367462,
+ "step": 1577
+ },
+ {
+ "epoch": 0.8279118572927597,
+ "grad_norm": 0.14791704009291914,
+ "learning_rate": 8.208856160409055e-06,
+ "loss": 0.43811577558517456,
+ "step": 1578
+ },
+ {
+ "epoch": 0.8284365162644282,
+ "grad_norm": 0.14924185710518617,
+ "learning_rate": 8.189795085736765e-06,
+ "loss": 0.43853428959846497,
+ "step": 1579
+ },
+ {
+ "epoch": 0.8289611752360966,
+ "grad_norm": 0.15420576936040198,
+ "learning_rate": 8.170786472568451e-06,
+ "loss": 0.445467084646225,
+ "step": 1580
+ },
+ {
+ "epoch": 0.829485834207765,
+ "grad_norm": 0.15577901772605834,
+ "learning_rate": 8.151830372546258e-06,
+ "loss": 0.4529452621936798,
+ "step": 1581
+ },
+ {
+ "epoch": 0.8300104931794333,
+ "grad_norm": 0.15387840488797552,
+ "learning_rate": 8.132926837169676e-06,
+ "loss": 0.46009817719459534,
+ "step": 1582
+ },
+ {
+ "epoch": 0.8305351521511017,
+ "grad_norm": 0.1527913309957562,
+ "learning_rate": 8.11407591779537e-06,
+ "loss": 0.4513983726501465,
+ "step": 1583
+ },
+ {
+ "epoch": 0.8310598111227702,
+ "grad_norm": 0.1537090286922398,
+ "learning_rate": 8.095277665637061e-06,
+ "loss": 0.4537739157676697,
+ "step": 1584
+ },
+ {
+ "epoch": 0.8315844700944386,
+ "grad_norm": 0.15722962107654076,
+ "learning_rate": 8.076532131765408e-06,
+ "loss": 0.45232969522476196,
+ "step": 1585
+ },
+ {
+ "epoch": 0.832109129066107,
+ "grad_norm": 0.1450142396611787,
+ "learning_rate": 8.057839367107817e-06,
+ "loss": 0.4270634651184082,
+ "step": 1586
+ },
+ {
+ "epoch": 0.8326337880377754,
+ "grad_norm": 0.150264001570592,
+ "learning_rate": 8.039199422448355e-06,
+ "loss": 0.43563878536224365,
+ "step": 1587
+ },
+ {
+ "epoch": 0.8331584470094439,
+ "grad_norm": 0.14904943307590535,
+ "learning_rate": 8.020612348427572e-06,
+ "loss": 0.4401509165763855,
+ "step": 1588
+ },
+ {
+ "epoch": 0.8336831059811123,
+ "grad_norm": 0.15453821662860034,
+ "learning_rate": 8.002078195542393e-06,
+ "loss": 0.44545161724090576,
+ "step": 1589
+ },
+ {
+ "epoch": 0.8342077649527807,
+ "grad_norm": 0.1570809678830383,
+ "learning_rate": 7.983597014145951e-06,
+ "loss": 0.45228466391563416,
+ "step": 1590
+ },
+ {
+ "epoch": 0.8347324239244491,
+ "grad_norm": 0.14741240769629818,
+ "learning_rate": 7.965168854447493e-06,
+ "loss": 0.4485381245613098,
+ "step": 1591
+ },
+ {
+ "epoch": 0.8352570828961176,
+ "grad_norm": 0.15362267095533516,
+ "learning_rate": 7.946793766512195e-06,
+ "loss": 0.4358059763908386,
+ "step": 1592
+ },
+ {
+ "epoch": 0.835781741867786,
+ "grad_norm": 0.1544754586495605,
+ "learning_rate": 7.92847180026106e-06,
+ "loss": 0.43327778577804565,
+ "step": 1593
+ },
+ {
+ "epoch": 0.8363064008394544,
+ "grad_norm": 0.15023225026164963,
+ "learning_rate": 7.91020300547077e-06,
+ "loss": 0.44359731674194336,
+ "step": 1594
+ },
+ {
+ "epoch": 0.8368310598111228,
+ "grad_norm": 0.15665530392095345,
+ "learning_rate": 7.891987431773545e-06,
+ "loss": 0.44503387808799744,
+ "step": 1595
+ },
+ {
+ "epoch": 0.8373557187827911,
+ "grad_norm": 0.14717614926392292,
+ "learning_rate": 7.873825128657033e-06,
+ "loss": 0.45314061641693115,
+ "step": 1596
+ },
+ {
+ "epoch": 0.8378803777544596,
+ "grad_norm": 0.15207464934030612,
+ "learning_rate": 7.855716145464131e-06,
+ "loss": 0.4526442885398865,
+ "step": 1597
+ },
+ {
+ "epoch": 0.838405036726128,
+ "grad_norm": 0.15129881541794873,
+ "learning_rate": 7.837660531392904e-06,
+ "loss": 0.45127853751182556,
+ "step": 1598
+ },
+ {
+ "epoch": 0.8389296956977964,
+ "grad_norm": 0.15623362715417285,
+ "learning_rate": 7.819658335496405e-06,
+ "loss": 0.46432793140411377,
+ "step": 1599
+ },
+ {
+ "epoch": 0.8394543546694648,
+ "grad_norm": 0.1529219879642125,
+ "learning_rate": 7.801709606682573e-06,
+ "loss": 0.4475358724594116,
+ "step": 1600
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 4.23430446865449e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1600/video_preprocessor_config.json b/checkpoints/checkpoint-1600/video_preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..a933f0ec6941222c33aeba7980bd5acf8dd98e91
--- /dev/null
+++ b/checkpoints/checkpoint-1600/video_preprocessor_config.json
@@ -0,0 +1,30 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "do_sample_frames": true,
+ "fps": 2,
+ "image_mean": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "image_std": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "max_frames": 768,
+ "merge_size": 2,
+ "min_frames": 4,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 786432,
+ "shortest_edge": 131072
+ },
+ "temporal_patch_size": 2,
+ "video_processor_type": "Qwen3VLVideoProcessor"
+}
diff --git a/checkpoints/checkpoint-1650/chat_template.jinja b/checkpoints/checkpoint-1650/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1650/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1650/config.json b/checkpoints/checkpoint-1650/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1650/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1650/expert_prune_metadata.json b/checkpoints/checkpoint-1650/expert_prune_metadata.json
new file mode 100644
index 0000000000000000000000000000000000000000..16e01c8a10636331e334732fa9629579c68df44e
--- /dev/null
+++ b/checkpoints/checkpoint-1650/expert_prune_metadata.json
@@ -0,0 +1,14 @@
+{
+ "source_model": "/mnt/data/wuwei/model_list/Qwen3.6-35B-A3B/models--Qwen--Qwen3.6-35B-A3B/snapshots/995ad96eacd98c81ed38be0c5b274b04031597b0",
+ "output_model": "/mnt/data/wuwei/model_list/qwen3-19b-a3b",
+ "total_parameters": 18990568816,
+ "parameter_billions_decimal": 18.990568816,
+ "directory_size_label": "19B",
+ "original_num_experts": 256,
+ "kept_experts_per_layer": 128,
+ "num_profiled_layers": 40,
+ "mtp_weights": "removed because Qwen3_5MoeForConditionalGeneration does not instantiate or use the source checkpoint's mtp.* training-only tensors",
+ "expert_order": "ascending original expert index; router rows and expert tensors use the same remapping",
+ "router_statistics": "router_stats.json and router_stats.pt",
+ "selection_score": "top-8 router hit frequency * mean(abs(actual expert output after down_proj and before routing-weight multiplication))"
+}
diff --git a/checkpoints/checkpoint-1650/generation_config.json b/checkpoints/checkpoint-1650/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1650/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1650/model.safetensors b/checkpoints/checkpoint-1650/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..e5da4ef72753a6134608dec52547ccbbfc53cd2e
--- /dev/null
+++ b/checkpoints/checkpoint-1650/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:39c16d1d2c176d009437615b69eb74ee27e0fe5eb45931154b19dfd6d17771af
+size 37983421000
diff --git a/checkpoints/checkpoint-1650/preprocessor_config.json b/checkpoints/checkpoint-1650/preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ec43d1f859ecadf800ebae92fda4a812d34975d3
--- /dev/null
+++ b/checkpoints/checkpoint-1650/preprocessor_config.json
@@ -0,0 +1,26 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "image_mean": [
+ 0.48145466,
+ 0.4578275,
+ 0.40821073
+ ],
+ "image_processor_type": "Qwen2VLImageProcessorFast",
+ "image_std": [
+ 0.26862954,
+ 0.26130258,
+ 0.27577711
+ ],
+ "merge_size": 2,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 1003520,
+ "shortest_edge": 3136
+ },
+ "temporal_patch_size": 2
+}
diff --git a/checkpoints/checkpoint-1650/tokenizer.json b/checkpoints/checkpoint-1650/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1650/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1650/tokenizer_config.json b/checkpoints/checkpoint-1650/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1650/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1650/trainer_state.json b/checkpoints/checkpoint-1650/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..0df90af0cd1ac6f72d1f7fa8cc96c6f2d4240892
--- /dev/null
+++ b/checkpoints/checkpoint-1650/trainer_state.json
@@ -0,0 +1,11584 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.8656873032528857,
+ "eval_steps": 500,
+ "global_step": 1650,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ },
+ {
+ "epoch": 0.5776495278069255,
+ "grad_norm": 0.1862053032854098,
+ "learning_rate": 2.2065553964033126e-05,
+ "loss": 0.4816875457763672,
+ "step": 1101
+ },
+ {
+ "epoch": 0.5781741867785939,
+ "grad_norm": 0.16996015894363117,
+ "learning_rate": 2.20295734045532e-05,
+ "loss": 0.4471197724342346,
+ "step": 1102
+ },
+ {
+ "epoch": 0.5786988457502623,
+ "grad_norm": 0.16889227008390692,
+ "learning_rate": 2.1993607706997725e-05,
+ "loss": 0.468391478061676,
+ "step": 1103
+ },
+ {
+ "epoch": 0.5792235047219307,
+ "grad_norm": 0.16670931681562373,
+ "learning_rate": 2.195765696907745e-05,
+ "loss": 0.44934162497520447,
+ "step": 1104
+ },
+ {
+ "epoch": 0.5797481636935992,
+ "grad_norm": 0.16328418741223716,
+ "learning_rate": 2.1921721288462482e-05,
+ "loss": 0.4558270275592804,
+ "step": 1105
+ },
+ {
+ "epoch": 0.5802728226652676,
+ "grad_norm": 0.17364660408703936,
+ "learning_rate": 2.188580076278203e-05,
+ "loss": 0.4597751796245575,
+ "step": 1106
+ },
+ {
+ "epoch": 0.580797481636936,
+ "grad_norm": 0.1628618453800601,
+ "learning_rate": 2.18498954896241e-05,
+ "loss": 0.4521859288215637,
+ "step": 1107
+ },
+ {
+ "epoch": 0.5813221406086044,
+ "grad_norm": 0.16813817951871482,
+ "learning_rate": 2.1814005566535302e-05,
+ "loss": 0.4581005573272705,
+ "step": 1108
+ },
+ {
+ "epoch": 0.5818467995802729,
+ "grad_norm": 0.16825923629726025,
+ "learning_rate": 2.1778131091020516e-05,
+ "loss": 0.4433916211128235,
+ "step": 1109
+ },
+ {
+ "epoch": 0.5823714585519413,
+ "grad_norm": 0.17194743467883944,
+ "learning_rate": 2.1742272160542654e-05,
+ "loss": 0.47296425700187683,
+ "step": 1110
+ },
+ {
+ "epoch": 0.5828961175236097,
+ "grad_norm": 0.16947730546039969,
+ "learning_rate": 2.170642887252241e-05,
+ "loss": 0.4815937876701355,
+ "step": 1111
+ },
+ {
+ "epoch": 0.583420776495278,
+ "grad_norm": 0.1657449928668473,
+ "learning_rate": 2.167060132433798e-05,
+ "loss": 0.4479164779186249,
+ "step": 1112
+ },
+ {
+ "epoch": 0.5839454354669464,
+ "grad_norm": 0.1658952516607364,
+ "learning_rate": 2.1634789613324778e-05,
+ "loss": 0.4637463688850403,
+ "step": 1113
+ },
+ {
+ "epoch": 0.5844700944386149,
+ "grad_norm": 0.16777002188998236,
+ "learning_rate": 2.1598993836775218e-05,
+ "loss": 0.457524836063385,
+ "step": 1114
+ },
+ {
+ "epoch": 0.5849947534102833,
+ "grad_norm": 0.15837566447341322,
+ "learning_rate": 2.1563214091938405e-05,
+ "loss": 0.44938087463378906,
+ "step": 1115
+ },
+ {
+ "epoch": 0.5855194123819517,
+ "grad_norm": 0.17162206617035314,
+ "learning_rate": 2.1527450476019896e-05,
+ "loss": 0.4393450617790222,
+ "step": 1116
+ },
+ {
+ "epoch": 0.5860440713536201,
+ "grad_norm": 0.16224101921433814,
+ "learning_rate": 2.1491703086181426e-05,
+ "loss": 0.451913982629776,
+ "step": 1117
+ },
+ {
+ "epoch": 0.5865687303252886,
+ "grad_norm": 0.1736253504575144,
+ "learning_rate": 2.145597201954066e-05,
+ "loss": 0.4611290395259857,
+ "step": 1118
+ },
+ {
+ "epoch": 0.587093389296957,
+ "grad_norm": 0.16308408333737234,
+ "learning_rate": 2.1420257373170903e-05,
+ "loss": 0.45898568630218506,
+ "step": 1119
+ },
+ {
+ "epoch": 0.5876180482686254,
+ "grad_norm": 0.16948634794175524,
+ "learning_rate": 2.1384559244100848e-05,
+ "loss": 0.45982882380485535,
+ "step": 1120
+ },
+ {
+ "epoch": 0.5881427072402938,
+ "grad_norm": 0.16676366538565252,
+ "learning_rate": 2.1348877729314336e-05,
+ "loss": 0.44826576113700867,
+ "step": 1121
+ },
+ {
+ "epoch": 0.5886673662119623,
+ "grad_norm": 0.16431125630712295,
+ "learning_rate": 2.1313212925750037e-05,
+ "loss": 0.4616445302963257,
+ "step": 1122
+ },
+ {
+ "epoch": 0.5891920251836307,
+ "grad_norm": 0.1631659061616955,
+ "learning_rate": 2.1277564930301252e-05,
+ "loss": 0.45651188492774963,
+ "step": 1123
+ },
+ {
+ "epoch": 0.5897166841552991,
+ "grad_norm": 0.15805950502689237,
+ "learning_rate": 2.1241933839815604e-05,
+ "loss": 0.4486045837402344,
+ "step": 1124
+ },
+ {
+ "epoch": 0.5902413431269675,
+ "grad_norm": 0.16530288038603502,
+ "learning_rate": 2.1206319751094773e-05,
+ "loss": 0.4485066533088684,
+ "step": 1125
+ },
+ {
+ "epoch": 0.5907660020986358,
+ "grad_norm": 0.16061903194239047,
+ "learning_rate": 2.1170722760894295e-05,
+ "loss": 0.4565330743789673,
+ "step": 1126
+ },
+ {
+ "epoch": 0.5912906610703043,
+ "grad_norm": 0.1643346999353627,
+ "learning_rate": 2.1135142965923195e-05,
+ "loss": 0.45560717582702637,
+ "step": 1127
+ },
+ {
+ "epoch": 0.5918153200419727,
+ "grad_norm": 0.17484644315215753,
+ "learning_rate": 2.109958046284382e-05,
+ "loss": 0.4795891046524048,
+ "step": 1128
+ },
+ {
+ "epoch": 0.5923399790136411,
+ "grad_norm": 0.1623762678622651,
+ "learning_rate": 2.1064035348271528e-05,
+ "loss": 0.47395604848861694,
+ "step": 1129
+ },
+ {
+ "epoch": 0.5928646379853095,
+ "grad_norm": 0.16888824241669015,
+ "learning_rate": 2.1028507718774456e-05,
+ "loss": 0.4601455330848694,
+ "step": 1130
+ },
+ {
+ "epoch": 0.593389296956978,
+ "grad_norm": 0.16291744568806302,
+ "learning_rate": 2.0992997670873193e-05,
+ "loss": 0.46248236298561096,
+ "step": 1131
+ },
+ {
+ "epoch": 0.5939139559286464,
+ "grad_norm": 0.16185598691236988,
+ "learning_rate": 2.0957505301040603e-05,
+ "loss": 0.45085301995277405,
+ "step": 1132
+ },
+ {
+ "epoch": 0.5944386149003148,
+ "grad_norm": 0.1678000806117814,
+ "learning_rate": 2.0922030705701518e-05,
+ "loss": 0.44939014315605164,
+ "step": 1133
+ },
+ {
+ "epoch": 0.5949632738719832,
+ "grad_norm": 0.16104474282288994,
+ "learning_rate": 2.088657398123246e-05,
+ "loss": 0.4313976764678955,
+ "step": 1134
+ },
+ {
+ "epoch": 0.5954879328436516,
+ "grad_norm": 0.1631128968199669,
+ "learning_rate": 2.085113522396141e-05,
+ "loss": 0.4508278965950012,
+ "step": 1135
+ },
+ {
+ "epoch": 0.5960125918153201,
+ "grad_norm": 0.16307688195844533,
+ "learning_rate": 2.0815714530167557e-05,
+ "loss": 0.4467576742172241,
+ "step": 1136
+ },
+ {
+ "epoch": 0.5965372507869885,
+ "grad_norm": 0.16456832304532207,
+ "learning_rate": 2.0780311996080976e-05,
+ "loss": 0.4621703624725342,
+ "step": 1137
+ },
+ {
+ "epoch": 0.5970619097586569,
+ "grad_norm": 0.16346217748325367,
+ "learning_rate": 2.0744927717882435e-05,
+ "loss": 0.4589449465274811,
+ "step": 1138
+ },
+ {
+ "epoch": 0.5975865687303253,
+ "grad_norm": 0.16180691501076994,
+ "learning_rate": 2.0709561791703097e-05,
+ "loss": 0.43698400259017944,
+ "step": 1139
+ },
+ {
+ "epoch": 0.5981112277019937,
+ "grad_norm": 0.15695459976898835,
+ "learning_rate": 2.067421431362427e-05,
+ "loss": 0.452271044254303,
+ "step": 1140
+ },
+ {
+ "epoch": 0.5986358866736621,
+ "grad_norm": 0.15930893553721054,
+ "learning_rate": 2.063888537967713e-05,
+ "loss": 0.44868963956832886,
+ "step": 1141
+ },
+ {
+ "epoch": 0.5991605456453305,
+ "grad_norm": 0.15963802284017928,
+ "learning_rate": 2.0603575085842485e-05,
+ "loss": 0.4376864731311798,
+ "step": 1142
+ },
+ {
+ "epoch": 0.5996852046169989,
+ "grad_norm": 0.16610266900507178,
+ "learning_rate": 2.05682835280505e-05,
+ "loss": 0.44676727056503296,
+ "step": 1143
+ },
+ {
+ "epoch": 0.6002098635886673,
+ "grad_norm": 0.15978253236669612,
+ "learning_rate": 2.053301080218044e-05,
+ "loss": 0.434773713350296,
+ "step": 1144
+ },
+ {
+ "epoch": 0.6007345225603358,
+ "grad_norm": 0.16305141803526324,
+ "learning_rate": 2.0497757004060403e-05,
+ "loss": 0.4529780149459839,
+ "step": 1145
+ },
+ {
+ "epoch": 0.6012591815320042,
+ "grad_norm": 0.1649737864101215,
+ "learning_rate": 2.046252222946707e-05,
+ "loss": 0.4571434259414673,
+ "step": 1146
+ },
+ {
+ "epoch": 0.6017838405036726,
+ "grad_norm": 0.15738824968326598,
+ "learning_rate": 2.0427306574125427e-05,
+ "loss": 0.4446289539337158,
+ "step": 1147
+ },
+ {
+ "epoch": 0.602308499475341,
+ "grad_norm": 0.1731139535185187,
+ "learning_rate": 2.039211013370853e-05,
+ "loss": 0.4551090598106384,
+ "step": 1148
+ },
+ {
+ "epoch": 0.6028331584470095,
+ "grad_norm": 0.15857251900043037,
+ "learning_rate": 2.0356933003837237e-05,
+ "loss": 0.4644010663032532,
+ "step": 1149
+ },
+ {
+ "epoch": 0.6033578174186779,
+ "grad_norm": 0.1589582861134211,
+ "learning_rate": 2.032177528007992e-05,
+ "loss": 0.45741650462150574,
+ "step": 1150
+ },
+ {
+ "epoch": 0.6038824763903463,
+ "grad_norm": 0.15639092706575122,
+ "learning_rate": 2.0286637057952257e-05,
+ "loss": 0.4375547766685486,
+ "step": 1151
+ },
+ {
+ "epoch": 0.6044071353620147,
+ "grad_norm": 0.16335605581337187,
+ "learning_rate": 2.0251518432916933e-05,
+ "loss": 0.4636140465736389,
+ "step": 1152
+ },
+ {
+ "epoch": 0.6049317943336832,
+ "grad_norm": 0.15437437113359356,
+ "learning_rate": 2.0216419500383372e-05,
+ "loss": 0.44632142782211304,
+ "step": 1153
+ },
+ {
+ "epoch": 0.6054564533053515,
+ "grad_norm": 0.15522039198698664,
+ "learning_rate": 2.0181340355707545e-05,
+ "loss": 0.449049174785614,
+ "step": 1154
+ },
+ {
+ "epoch": 0.6059811122770199,
+ "grad_norm": 0.1599951917909323,
+ "learning_rate": 2.0146281094191604e-05,
+ "loss": 0.44395962357521057,
+ "step": 1155
+ },
+ {
+ "epoch": 0.6065057712486883,
+ "grad_norm": 0.16866621620527716,
+ "learning_rate": 2.011124181108373e-05,
+ "loss": 0.45693716406822205,
+ "step": 1156
+ },
+ {
+ "epoch": 0.6070304302203567,
+ "grad_norm": 0.15508935650135075,
+ "learning_rate": 2.0076222601577815e-05,
+ "loss": 0.4525151252746582,
+ "step": 1157
+ },
+ {
+ "epoch": 0.6075550891920252,
+ "grad_norm": 0.16508998118005971,
+ "learning_rate": 2.004122356081321e-05,
+ "loss": 0.466006875038147,
+ "step": 1158
+ },
+ {
+ "epoch": 0.6080797481636936,
+ "grad_norm": 0.15817418864022237,
+ "learning_rate": 2.0006244783874466e-05,
+ "loss": 0.4496110677719116,
+ "step": 1159
+ },
+ {
+ "epoch": 0.608604407135362,
+ "grad_norm": 0.16127270259749824,
+ "learning_rate": 1.997128636579109e-05,
+ "loss": 0.45918628573417664,
+ "step": 1160
+ },
+ {
+ "epoch": 0.6091290661070304,
+ "grad_norm": 0.1702599924159677,
+ "learning_rate": 1.9936348401537292e-05,
+ "loss": 0.47218555212020874,
+ "step": 1161
+ },
+ {
+ "epoch": 0.6096537250786989,
+ "grad_norm": 0.15690176248846127,
+ "learning_rate": 1.9901430986031695e-05,
+ "loss": 0.4516097903251648,
+ "step": 1162
+ },
+ {
+ "epoch": 0.6101783840503673,
+ "grad_norm": 0.1713614405614144,
+ "learning_rate": 1.9866534214137085e-05,
+ "loss": 0.45030856132507324,
+ "step": 1163
+ },
+ {
+ "epoch": 0.6107030430220357,
+ "grad_norm": 0.1670865563026229,
+ "learning_rate": 1.9831658180660195e-05,
+ "loss": 0.45603489875793457,
+ "step": 1164
+ },
+ {
+ "epoch": 0.6112277019937041,
+ "grad_norm": 0.16897539113401952,
+ "learning_rate": 1.9796802980351392e-05,
+ "loss": 0.46314537525177,
+ "step": 1165
+ },
+ {
+ "epoch": 0.6117523609653726,
+ "grad_norm": 0.15832868870697644,
+ "learning_rate": 1.976196870790446e-05,
+ "loss": 0.44784611463546753,
+ "step": 1166
+ },
+ {
+ "epoch": 0.612277019937041,
+ "grad_norm": 0.16595365551891114,
+ "learning_rate": 1.9727155457956315e-05,
+ "loss": 0.4409624934196472,
+ "step": 1167
+ },
+ {
+ "epoch": 0.6128016789087093,
+ "grad_norm": 0.1642630411212017,
+ "learning_rate": 1.969236332508677e-05,
+ "loss": 0.46120351552963257,
+ "step": 1168
+ },
+ {
+ "epoch": 0.6133263378803777,
+ "grad_norm": 0.16481796135868812,
+ "learning_rate": 1.9657592403818253e-05,
+ "loss": 0.45544904470443726,
+ "step": 1169
+ },
+ {
+ "epoch": 0.6138509968520461,
+ "grad_norm": 0.1629832779852364,
+ "learning_rate": 1.962284278861557e-05,
+ "loss": 0.4578462839126587,
+ "step": 1170
+ },
+ {
+ "epoch": 0.6143756558237146,
+ "grad_norm": 0.15916219613603688,
+ "learning_rate": 1.9588114573885664e-05,
+ "loss": 0.4546983540058136,
+ "step": 1171
+ },
+ {
+ "epoch": 0.614900314795383,
+ "grad_norm": 0.1660520165166075,
+ "learning_rate": 1.95534078539773e-05,
+ "loss": 0.4601942300796509,
+ "step": 1172
+ },
+ {
+ "epoch": 0.6154249737670514,
+ "grad_norm": 0.16924794578577768,
+ "learning_rate": 1.9518722723180877e-05,
+ "loss": 0.45997846126556396,
+ "step": 1173
+ },
+ {
+ "epoch": 0.6159496327387198,
+ "grad_norm": 0.16597646133045382,
+ "learning_rate": 1.9484059275728127e-05,
+ "loss": 0.4594605565071106,
+ "step": 1174
+ },
+ {
+ "epoch": 0.6164742917103883,
+ "grad_norm": 0.16964243792519748,
+ "learning_rate": 1.9449417605791876e-05,
+ "loss": 0.4638690650463104,
+ "step": 1175
+ },
+ {
+ "epoch": 0.6169989506820567,
+ "grad_norm": 0.1758591426791696,
+ "learning_rate": 1.9414797807485796e-05,
+ "loss": 0.4666447639465332,
+ "step": 1176
+ },
+ {
+ "epoch": 0.6175236096537251,
+ "grad_norm": 0.16217661227891753,
+ "learning_rate": 1.9380199974864126e-05,
+ "loss": 0.44252535700798035,
+ "step": 1177
+ },
+ {
+ "epoch": 0.6180482686253935,
+ "grad_norm": 0.1576446184499019,
+ "learning_rate": 1.934562420192142e-05,
+ "loss": 0.4499641954898834,
+ "step": 1178
+ },
+ {
+ "epoch": 0.618572927597062,
+ "grad_norm": 0.16231008092636912,
+ "learning_rate": 1.9311070582592327e-05,
+ "loss": 0.4561542570590973,
+ "step": 1179
+ },
+ {
+ "epoch": 0.6190975865687304,
+ "grad_norm": 0.16790382249577043,
+ "learning_rate": 1.9276539210751292e-05,
+ "loss": 0.4484325647354126,
+ "step": 1180
+ },
+ {
+ "epoch": 0.6196222455403987,
+ "grad_norm": 0.18794729932605855,
+ "learning_rate": 1.9242030180212323e-05,
+ "loss": 0.44169026613235474,
+ "step": 1181
+ },
+ {
+ "epoch": 0.6201469045120671,
+ "grad_norm": 0.16948051243889814,
+ "learning_rate": 1.920754358472872e-05,
+ "loss": 0.4592876434326172,
+ "step": 1182
+ },
+ {
+ "epoch": 0.6206715634837355,
+ "grad_norm": 0.16762558710029374,
+ "learning_rate": 1.917307951799286e-05,
+ "loss": 0.4402095079421997,
+ "step": 1183
+ },
+ {
+ "epoch": 0.621196222455404,
+ "grad_norm": 0.18170280116387003,
+ "learning_rate": 1.9138638073635885e-05,
+ "loss": 0.4633026123046875,
+ "step": 1184
+ },
+ {
+ "epoch": 0.6217208814270724,
+ "grad_norm": 0.1718051709136108,
+ "learning_rate": 1.9104219345227486e-05,
+ "loss": 0.4510066509246826,
+ "step": 1185
+ },
+ {
+ "epoch": 0.6222455403987408,
+ "grad_norm": 0.16829066326345576,
+ "learning_rate": 1.906982342627565e-05,
+ "loss": 0.4632919430732727,
+ "step": 1186
+ },
+ {
+ "epoch": 0.6227701993704092,
+ "grad_norm": 0.16669141808548274,
+ "learning_rate": 1.9035450410226395e-05,
+ "loss": 0.455544650554657,
+ "step": 1187
+ },
+ {
+ "epoch": 0.6232948583420777,
+ "grad_norm": 0.16679996559722232,
+ "learning_rate": 1.90011003904635e-05,
+ "loss": 0.459994375705719,
+ "step": 1188
+ },
+ {
+ "epoch": 0.6238195173137461,
+ "grad_norm": 0.17379754713278164,
+ "learning_rate": 1.8966773460308284e-05,
+ "loss": 0.4576749801635742,
+ "step": 1189
+ },
+ {
+ "epoch": 0.6243441762854145,
+ "grad_norm": 0.17088539733389246,
+ "learning_rate": 1.8932469713019335e-05,
+ "loss": 0.44853052496910095,
+ "step": 1190
+ },
+ {
+ "epoch": 0.6248688352570829,
+ "grad_norm": 0.15556883889201015,
+ "learning_rate": 1.8898189241792248e-05,
+ "loss": 0.44060447812080383,
+ "step": 1191
+ },
+ {
+ "epoch": 0.6253934942287513,
+ "grad_norm": 0.1640093917439215,
+ "learning_rate": 1.8863932139759388e-05,
+ "loss": 0.4531911611557007,
+ "step": 1192
+ },
+ {
+ "epoch": 0.6259181532004198,
+ "grad_norm": 0.16444284826591576,
+ "learning_rate": 1.882969849998964e-05,
+ "loss": 0.44176363945007324,
+ "step": 1193
+ },
+ {
+ "epoch": 0.6264428121720882,
+ "grad_norm": 0.1696462520995642,
+ "learning_rate": 1.8795488415488143e-05,
+ "loss": 0.44516652822494507,
+ "step": 1194
+ },
+ {
+ "epoch": 0.6269674711437565,
+ "grad_norm": 0.16810724705191796,
+ "learning_rate": 1.8761301979196032e-05,
+ "loss": 0.4542563855648041,
+ "step": 1195
+ },
+ {
+ "epoch": 0.6274921301154249,
+ "grad_norm": 0.17334813313855824,
+ "learning_rate": 1.8727139283990204e-05,
+ "loss": 0.47769248485565186,
+ "step": 1196
+ },
+ {
+ "epoch": 0.6280167890870934,
+ "grad_norm": 0.16308212035303235,
+ "learning_rate": 1.8693000422683043e-05,
+ "loss": 0.4407799243927002,
+ "step": 1197
+ },
+ {
+ "epoch": 0.6285414480587618,
+ "grad_norm": 0.16853319139928977,
+ "learning_rate": 1.8658885488022204e-05,
+ "loss": 0.4565497040748596,
+ "step": 1198
+ },
+ {
+ "epoch": 0.6290661070304302,
+ "grad_norm": 0.16694968781284703,
+ "learning_rate": 1.8624794572690328e-05,
+ "loss": 0.44522374868392944,
+ "step": 1199
+ },
+ {
+ "epoch": 0.6295907660020986,
+ "grad_norm": 0.16418579468339103,
+ "learning_rate": 1.8590727769304788e-05,
+ "loss": 0.4441048502922058,
+ "step": 1200
+ },
+ {
+ "epoch": 0.630115424973767,
+ "grad_norm": 0.1661930777124156,
+ "learning_rate": 1.8556685170417486e-05,
+ "loss": 0.457327663898468,
+ "step": 1201
+ },
+ {
+ "epoch": 0.6306400839454355,
+ "grad_norm": 0.16536630027039814,
+ "learning_rate": 1.852266686851452e-05,
+ "loss": 0.462494432926178,
+ "step": 1202
+ },
+ {
+ "epoch": 0.6311647429171039,
+ "grad_norm": 0.16611474088307454,
+ "learning_rate": 1.8488672956016013e-05,
+ "loss": 0.4557701647281647,
+ "step": 1203
+ },
+ {
+ "epoch": 0.6316894018887723,
+ "grad_norm": 0.16759343079749697,
+ "learning_rate": 1.8454703525275813e-05,
+ "loss": 0.45069995522499084,
+ "step": 1204
+ },
+ {
+ "epoch": 0.6322140608604407,
+ "grad_norm": 0.16186676294485336,
+ "learning_rate": 1.8420758668581255e-05,
+ "loss": 0.44954800605773926,
+ "step": 1205
+ },
+ {
+ "epoch": 0.6327387198321092,
+ "grad_norm": 0.15687159835368553,
+ "learning_rate": 1.8386838478152916e-05,
+ "loss": 0.4490429759025574,
+ "step": 1206
+ },
+ {
+ "epoch": 0.6332633788037776,
+ "grad_norm": 0.16242951976115577,
+ "learning_rate": 1.8352943046144372e-05,
+ "loss": 0.45819586515426636,
+ "step": 1207
+ },
+ {
+ "epoch": 0.633788037775446,
+ "grad_norm": 0.16516087266517232,
+ "learning_rate": 1.831907246464191e-05,
+ "loss": 0.44769883155822754,
+ "step": 1208
+ },
+ {
+ "epoch": 0.6343126967471143,
+ "grad_norm": 0.16709475871512644,
+ "learning_rate": 1.8285226825664332e-05,
+ "loss": 0.4699128568172455,
+ "step": 1209
+ },
+ {
+ "epoch": 0.6348373557187827,
+ "grad_norm": 0.16288623150595774,
+ "learning_rate": 1.8251406221162636e-05,
+ "loss": 0.45322558283805847,
+ "step": 1210
+ },
+ {
+ "epoch": 0.6353620146904512,
+ "grad_norm": 0.1675722848054237,
+ "learning_rate": 1.821761074301986e-05,
+ "loss": 0.4752347469329834,
+ "step": 1211
+ },
+ {
+ "epoch": 0.6358866736621196,
+ "grad_norm": 0.163285963514504,
+ "learning_rate": 1.818384048305074e-05,
+ "loss": 0.441958487033844,
+ "step": 1212
+ },
+ {
+ "epoch": 0.636411332633788,
+ "grad_norm": 0.17078102287557764,
+ "learning_rate": 1.8150095533001517e-05,
+ "loss": 0.4551333487033844,
+ "step": 1213
+ },
+ {
+ "epoch": 0.6369359916054564,
+ "grad_norm": 0.16188540746097138,
+ "learning_rate": 1.811637598454966e-05,
+ "loss": 0.4470779597759247,
+ "step": 1214
+ },
+ {
+ "epoch": 0.6374606505771249,
+ "grad_norm": 0.15235628218958286,
+ "learning_rate": 1.808268192930364e-05,
+ "loss": 0.43604153394699097,
+ "step": 1215
+ },
+ {
+ "epoch": 0.6379853095487933,
+ "grad_norm": 0.1604601949959104,
+ "learning_rate": 1.8049013458802665e-05,
+ "loss": 0.4445713758468628,
+ "step": 1216
+ },
+ {
+ "epoch": 0.6385099685204617,
+ "grad_norm": 0.15511797131215901,
+ "learning_rate": 1.8015370664516427e-05,
+ "loss": 0.4455301761627197,
+ "step": 1217
+ },
+ {
+ "epoch": 0.6390346274921301,
+ "grad_norm": 0.16815669546944095,
+ "learning_rate": 1.798175363784487e-05,
+ "loss": 0.46135014295578003,
+ "step": 1218
+ },
+ {
+ "epoch": 0.6395592864637986,
+ "grad_norm": 0.161558177688922,
+ "learning_rate": 1.7948162470117917e-05,
+ "loss": 0.4519304037094116,
+ "step": 1219
+ },
+ {
+ "epoch": 0.640083945435467,
+ "grad_norm": 0.16677842130048232,
+ "learning_rate": 1.7914597252595265e-05,
+ "loss": 0.4550550878047943,
+ "step": 1220
+ },
+ {
+ "epoch": 0.6406086044071354,
+ "grad_norm": 0.1597331480423138,
+ "learning_rate": 1.7881058076466095e-05,
+ "loss": 0.441120445728302,
+ "step": 1221
+ },
+ {
+ "epoch": 0.6411332633788038,
+ "grad_norm": 0.15951106545246824,
+ "learning_rate": 1.7847545032848844e-05,
+ "loss": 0.4471382200717926,
+ "step": 1222
+ },
+ {
+ "epoch": 0.6416579223504721,
+ "grad_norm": 0.16157787674667565,
+ "learning_rate": 1.7814058212790947e-05,
+ "loss": 0.4550648331642151,
+ "step": 1223
+ },
+ {
+ "epoch": 0.6421825813221406,
+ "grad_norm": 0.16989186763352415,
+ "learning_rate": 1.7780597707268583e-05,
+ "loss": 0.4515089988708496,
+ "step": 1224
+ },
+ {
+ "epoch": 0.642707240293809,
+ "grad_norm": 0.15952153223099508,
+ "learning_rate": 1.7747163607186468e-05,
+ "loss": 0.4658638834953308,
+ "step": 1225
+ },
+ {
+ "epoch": 0.6432318992654774,
+ "grad_norm": 0.15205926699647218,
+ "learning_rate": 1.7713756003377567e-05,
+ "loss": 0.4330868422985077,
+ "step": 1226
+ },
+ {
+ "epoch": 0.6437565582371458,
+ "grad_norm": 0.15784806480936794,
+ "learning_rate": 1.7680374986602865e-05,
+ "loss": 0.4506068527698517,
+ "step": 1227
+ },
+ {
+ "epoch": 0.6442812172088143,
+ "grad_norm": 0.16006709280686687,
+ "learning_rate": 1.7647020647551105e-05,
+ "loss": 0.446941077709198,
+ "step": 1228
+ },
+ {
+ "epoch": 0.6448058761804827,
+ "grad_norm": 0.16064631158222062,
+ "learning_rate": 1.761369307683856e-05,
+ "loss": 0.4418525695800781,
+ "step": 1229
+ },
+ {
+ "epoch": 0.6453305351521511,
+ "grad_norm": 0.16454136480444392,
+ "learning_rate": 1.7580392365008767e-05,
+ "loss": 0.46268391609191895,
+ "step": 1230
+ },
+ {
+ "epoch": 0.6458551941238195,
+ "grad_norm": 0.15401246535771462,
+ "learning_rate": 1.754711860253233e-05,
+ "loss": 0.4372989535331726,
+ "step": 1231
+ },
+ {
+ "epoch": 0.646379853095488,
+ "grad_norm": 0.1544567975028383,
+ "learning_rate": 1.7513871879806603e-05,
+ "loss": 0.44806575775146484,
+ "step": 1232
+ },
+ {
+ "epoch": 0.6469045120671564,
+ "grad_norm": 0.16411225112730965,
+ "learning_rate": 1.748065228715549e-05,
+ "loss": 0.43782058358192444,
+ "step": 1233
+ },
+ {
+ "epoch": 0.6474291710388248,
+ "grad_norm": 0.1552767832972093,
+ "learning_rate": 1.7447459914829183e-05,
+ "loss": 0.44531378149986267,
+ "step": 1234
+ },
+ {
+ "epoch": 0.6479538300104932,
+ "grad_norm": 0.16214695257123513,
+ "learning_rate": 1.741429485300392e-05,
+ "loss": 0.47399842739105225,
+ "step": 1235
+ },
+ {
+ "epoch": 0.6484784889821616,
+ "grad_norm": 0.16684861111484112,
+ "learning_rate": 1.7381157191781766e-05,
+ "loss": 0.4654282331466675,
+ "step": 1236
+ },
+ {
+ "epoch": 0.64900314795383,
+ "grad_norm": 0.15901864178630273,
+ "learning_rate": 1.7348047021190334e-05,
+ "loss": 0.45128414034843445,
+ "step": 1237
+ },
+ {
+ "epoch": 0.6495278069254984,
+ "grad_norm": 0.15360991961742707,
+ "learning_rate": 1.7314964431182523e-05,
+ "loss": 0.4303375482559204,
+ "step": 1238
+ },
+ {
+ "epoch": 0.6500524658971668,
+ "grad_norm": 0.1633151574385741,
+ "learning_rate": 1.728190951163633e-05,
+ "loss": 0.45568403601646423,
+ "step": 1239
+ },
+ {
+ "epoch": 0.6505771248688352,
+ "grad_norm": 0.16377853571058645,
+ "learning_rate": 1.724888235235457e-05,
+ "loss": 0.46503543853759766,
+ "step": 1240
+ },
+ {
+ "epoch": 0.6511017838405037,
+ "grad_norm": 0.1595023801275368,
+ "learning_rate": 1.7215883043064655e-05,
+ "loss": 0.46337294578552246,
+ "step": 1241
+ },
+ {
+ "epoch": 0.6516264428121721,
+ "grad_norm": 0.16324796182212292,
+ "learning_rate": 1.7182911673418313e-05,
+ "loss": 0.46197354793548584,
+ "step": 1242
+ },
+ {
+ "epoch": 0.6521511017838405,
+ "grad_norm": 0.16353734830879765,
+ "learning_rate": 1.714996833299138e-05,
+ "loss": 0.45741671323776245,
+ "step": 1243
+ },
+ {
+ "epoch": 0.6526757607555089,
+ "grad_norm": 0.15941587678263386,
+ "learning_rate": 1.7117053111283528e-05,
+ "loss": 0.4514503479003906,
+ "step": 1244
+ },
+ {
+ "epoch": 0.6532004197271774,
+ "grad_norm": 0.1642653230145008,
+ "learning_rate": 1.708416609771804e-05,
+ "loss": 0.4437655210494995,
+ "step": 1245
+ },
+ {
+ "epoch": 0.6537250786988458,
+ "grad_norm": 0.15814155519130152,
+ "learning_rate": 1.7051307381641587e-05,
+ "loss": 0.4314163327217102,
+ "step": 1246
+ },
+ {
+ "epoch": 0.6542497376705142,
+ "grad_norm": 0.16086078370768056,
+ "learning_rate": 1.701847705232394e-05,
+ "loss": 0.4526430666446686,
+ "step": 1247
+ },
+ {
+ "epoch": 0.6547743966421826,
+ "grad_norm": 0.16389360190715183,
+ "learning_rate": 1.698567519895775e-05,
+ "loss": 0.4278966188430786,
+ "step": 1248
+ },
+ {
+ "epoch": 0.655299055613851,
+ "grad_norm": 0.16303142287804262,
+ "learning_rate": 1.6952901910658316e-05,
+ "loss": 0.45778220891952515,
+ "step": 1249
+ },
+ {
+ "epoch": 0.6558237145855194,
+ "grad_norm": 0.1675209956904048,
+ "learning_rate": 1.6920157276463318e-05,
+ "loss": 0.4489403963088989,
+ "step": 1250
+ },
+ {
+ "epoch": 0.6563483735571878,
+ "grad_norm": 0.16228311551074323,
+ "learning_rate": 1.6887441385332604e-05,
+ "loss": 0.4534342885017395,
+ "step": 1251
+ },
+ {
+ "epoch": 0.6568730325288562,
+ "grad_norm": 0.16408957851046913,
+ "learning_rate": 1.6854754326147916e-05,
+ "loss": 0.4549208879470825,
+ "step": 1252
+ },
+ {
+ "epoch": 0.6573976915005246,
+ "grad_norm": 0.15967922111432845,
+ "learning_rate": 1.682209618771268e-05,
+ "loss": 0.4451141357421875,
+ "step": 1253
+ },
+ {
+ "epoch": 0.657922350472193,
+ "grad_norm": 0.16440698722204208,
+ "learning_rate": 1.678946705875176e-05,
+ "loss": 0.454200804233551,
+ "step": 1254
+ },
+ {
+ "epoch": 0.6584470094438615,
+ "grad_norm": 0.16251120184938497,
+ "learning_rate": 1.675686702791117e-05,
+ "loss": 0.4580027461051941,
+ "step": 1255
+ },
+ {
+ "epoch": 0.6589716684155299,
+ "grad_norm": 0.1643717386150457,
+ "learning_rate": 1.6724296183757924e-05,
+ "loss": 0.4422311782836914,
+ "step": 1256
+ },
+ {
+ "epoch": 0.6594963273871983,
+ "grad_norm": 0.1644586414378722,
+ "learning_rate": 1.66917546147797e-05,
+ "loss": 0.4593179225921631,
+ "step": 1257
+ },
+ {
+ "epoch": 0.6600209863588667,
+ "grad_norm": 0.16751074186967685,
+ "learning_rate": 1.6659242409384657e-05,
+ "loss": 0.457930326461792,
+ "step": 1258
+ },
+ {
+ "epoch": 0.6605456453305352,
+ "grad_norm": 0.15398418609155334,
+ "learning_rate": 1.6626759655901182e-05,
+ "loss": 0.44002899527549744,
+ "step": 1259
+ },
+ {
+ "epoch": 0.6610703043022036,
+ "grad_norm": 0.16328113175790226,
+ "learning_rate": 1.659430644257763e-05,
+ "loss": 0.4500438868999481,
+ "step": 1260
+ },
+ {
+ "epoch": 0.661594963273872,
+ "grad_norm": 0.1694244340615106,
+ "learning_rate": 1.656188285758214e-05,
+ "loss": 0.4818427264690399,
+ "step": 1261
+ },
+ {
+ "epoch": 0.6621196222455404,
+ "grad_norm": 0.1578320701754368,
+ "learning_rate": 1.652948898900232e-05,
+ "loss": 0.4492436647415161,
+ "step": 1262
+ },
+ {
+ "epoch": 0.6626442812172089,
+ "grad_norm": 0.16538865917828469,
+ "learning_rate": 1.6497124924845064e-05,
+ "loss": 0.45070338249206543,
+ "step": 1263
+ },
+ {
+ "epoch": 0.6631689401888772,
+ "grad_norm": 0.15670817848256852,
+ "learning_rate": 1.6464790753036296e-05,
+ "loss": 0.45439285039901733,
+ "step": 1264
+ },
+ {
+ "epoch": 0.6636935991605456,
+ "grad_norm": 0.16184786585965885,
+ "learning_rate": 1.64324865614207e-05,
+ "loss": 0.446522980928421,
+ "step": 1265
+ },
+ {
+ "epoch": 0.664218258132214,
+ "grad_norm": 0.15719726462251984,
+ "learning_rate": 1.6400212437761554e-05,
+ "loss": 0.4495280981063843,
+ "step": 1266
+ },
+ {
+ "epoch": 0.6647429171038824,
+ "grad_norm": 0.1680804414134405,
+ "learning_rate": 1.6367968469740416e-05,
+ "loss": 0.4652596712112427,
+ "step": 1267
+ },
+ {
+ "epoch": 0.6652675760755509,
+ "grad_norm": 0.15919541846683566,
+ "learning_rate": 1.6335754744956937e-05,
+ "loss": 0.43842798471450806,
+ "step": 1268
+ },
+ {
+ "epoch": 0.6657922350472193,
+ "grad_norm": 0.16322634187139418,
+ "learning_rate": 1.630357135092859e-05,
+ "loss": 0.45242664217948914,
+ "step": 1269
+ },
+ {
+ "epoch": 0.6663168940188877,
+ "grad_norm": 0.15644867734734783,
+ "learning_rate": 1.6271418375090445e-05,
+ "loss": 0.44801655411720276,
+ "step": 1270
+ },
+ {
+ "epoch": 0.6668415529905561,
+ "grad_norm": 0.1514717416124423,
+ "learning_rate": 1.623929590479496e-05,
+ "loss": 0.4506194293498993,
+ "step": 1271
+ },
+ {
+ "epoch": 0.6673662119622246,
+ "grad_norm": 0.1740345085495889,
+ "learning_rate": 1.6207204027311683e-05,
+ "loss": 0.4818177819252014,
+ "step": 1272
+ },
+ {
+ "epoch": 0.667890870933893,
+ "grad_norm": 0.15797701900765457,
+ "learning_rate": 1.6175142829827067e-05,
+ "loss": 0.4639382064342499,
+ "step": 1273
+ },
+ {
+ "epoch": 0.6684155299055614,
+ "grad_norm": 0.15663812316653875,
+ "learning_rate": 1.6143112399444198e-05,
+ "loss": 0.46217501163482666,
+ "step": 1274
+ },
+ {
+ "epoch": 0.6689401888772298,
+ "grad_norm": 0.15474018977043758,
+ "learning_rate": 1.6111112823182592e-05,
+ "loss": 0.43691688776016235,
+ "step": 1275
+ },
+ {
+ "epoch": 0.6694648478488983,
+ "grad_norm": 0.1615634942529431,
+ "learning_rate": 1.6079144187977923e-05,
+ "loss": 0.4412601590156555,
+ "step": 1276
+ },
+ {
+ "epoch": 0.6699895068205667,
+ "grad_norm": 0.15440022842754444,
+ "learning_rate": 1.604720658068184e-05,
+ "loss": 0.4523273706436157,
+ "step": 1277
+ },
+ {
+ "epoch": 0.670514165792235,
+ "grad_norm": 0.1634985665776242,
+ "learning_rate": 1.6015300088061664e-05,
+ "loss": 0.46264398097991943,
+ "step": 1278
+ },
+ {
+ "epoch": 0.6710388247639034,
+ "grad_norm": 0.15037886370689527,
+ "learning_rate": 1.598342479680018e-05,
+ "loss": 0.451465904712677,
+ "step": 1279
+ },
+ {
+ "epoch": 0.6715634837355718,
+ "grad_norm": 0.16895242437746613,
+ "learning_rate": 1.595158079349542e-05,
+ "loss": 0.4466145634651184,
+ "step": 1280
+ },
+ {
+ "epoch": 0.6720881427072403,
+ "grad_norm": 0.1629548326999642,
+ "learning_rate": 1.5919768164660425e-05,
+ "loss": 0.45415782928466797,
+ "step": 1281
+ },
+ {
+ "epoch": 0.6726128016789087,
+ "grad_norm": 0.1563872102268991,
+ "learning_rate": 1.588798699672298e-05,
+ "loss": 0.4386737048625946,
+ "step": 1282
+ },
+ {
+ "epoch": 0.6731374606505771,
+ "grad_norm": 0.16038172773344717,
+ "learning_rate": 1.5856237376025403e-05,
+ "loss": 0.446260929107666,
+ "step": 1283
+ },
+ {
+ "epoch": 0.6736621196222455,
+ "grad_norm": 0.16189780714407623,
+ "learning_rate": 1.582451938882431e-05,
+ "loss": 0.45458072423934937,
+ "step": 1284
+ },
+ {
+ "epoch": 0.674186778593914,
+ "grad_norm": 0.15204521785049285,
+ "learning_rate": 1.5792833121290367e-05,
+ "loss": 0.4425617754459381,
+ "step": 1285
+ },
+ {
+ "epoch": 0.6747114375655824,
+ "grad_norm": 0.16237376704792778,
+ "learning_rate": 1.5761178659508063e-05,
+ "loss": 0.44631800055503845,
+ "step": 1286
+ },
+ {
+ "epoch": 0.6752360965372508,
+ "grad_norm": 0.1582063618466069,
+ "learning_rate": 1.57295560894755e-05,
+ "loss": 0.44280534982681274,
+ "step": 1287
+ },
+ {
+ "epoch": 0.6757607555089192,
+ "grad_norm": 0.16587032349847394,
+ "learning_rate": 1.5697965497104116e-05,
+ "loss": 0.4495997130870819,
+ "step": 1288
+ },
+ {
+ "epoch": 0.6762854144805877,
+ "grad_norm": 0.1557269262755563,
+ "learning_rate": 1.566640696821848e-05,
+ "loss": 0.4493359625339508,
+ "step": 1289
+ },
+ {
+ "epoch": 0.6768100734522561,
+ "grad_norm": 0.16038490357012256,
+ "learning_rate": 1.5634880588556043e-05,
+ "loss": 0.4447544813156128,
+ "step": 1290
+ },
+ {
+ "epoch": 0.6773347324239245,
+ "grad_norm": 0.15355201908546406,
+ "learning_rate": 1.560338644376692e-05,
+ "loss": 0.4468432664871216,
+ "step": 1291
+ },
+ {
+ "epoch": 0.6778593913955928,
+ "grad_norm": 0.15077009515589776,
+ "learning_rate": 1.5571924619413668e-05,
+ "loss": 0.44959044456481934,
+ "step": 1292
+ },
+ {
+ "epoch": 0.6783840503672612,
+ "grad_norm": 0.1583038146613965,
+ "learning_rate": 1.5540495200971012e-05,
+ "loss": 0.42450425028800964,
+ "step": 1293
+ },
+ {
+ "epoch": 0.6789087093389297,
+ "grad_norm": 0.15950687190598087,
+ "learning_rate": 1.5509098273825635e-05,
+ "loss": 0.4527238607406616,
+ "step": 1294
+ },
+ {
+ "epoch": 0.6794333683105981,
+ "grad_norm": 0.1526262208847808,
+ "learning_rate": 1.5477733923275968e-05,
+ "loss": 0.4379724860191345,
+ "step": 1295
+ },
+ {
+ "epoch": 0.6799580272822665,
+ "grad_norm": 0.15541579650021778,
+ "learning_rate": 1.5446402234531934e-05,
+ "loss": 0.4462166130542755,
+ "step": 1296
+ },
+ {
+ "epoch": 0.6804826862539349,
+ "grad_norm": 0.15483930734313786,
+ "learning_rate": 1.5415103292714718e-05,
+ "loss": 0.4390607178211212,
+ "step": 1297
+ },
+ {
+ "epoch": 0.6810073452256034,
+ "grad_norm": 0.1578005731149154,
+ "learning_rate": 1.5383837182856532e-05,
+ "loss": 0.44462254643440247,
+ "step": 1298
+ },
+ {
+ "epoch": 0.6815320041972718,
+ "grad_norm": 0.15830183884958823,
+ "learning_rate": 1.5352603989900406e-05,
+ "loss": 0.4563537538051605,
+ "step": 1299
+ },
+ {
+ "epoch": 0.6820566631689402,
+ "grad_norm": 0.15605472163445522,
+ "learning_rate": 1.5321403798699923e-05,
+ "loss": 0.43852412700653076,
+ "step": 1300
+ },
+ {
+ "epoch": 0.6825813221406086,
+ "grad_norm": 0.15770470227141278,
+ "learning_rate": 1.5290236694019017e-05,
+ "loss": 0.4438081979751587,
+ "step": 1301
+ },
+ {
+ "epoch": 0.683105981112277,
+ "grad_norm": 0.16175339845012077,
+ "learning_rate": 1.5259102760531747e-05,
+ "loss": 0.46670734882354736,
+ "step": 1302
+ },
+ {
+ "epoch": 0.6836306400839455,
+ "grad_norm": 0.16359504716154477,
+ "learning_rate": 1.5228002082822036e-05,
+ "loss": 0.4590187668800354,
+ "step": 1303
+ },
+ {
+ "epoch": 0.6841552990556139,
+ "grad_norm": 0.15422879500254713,
+ "learning_rate": 1.5196934745383457e-05,
+ "loss": 0.45287445187568665,
+ "step": 1304
+ },
+ {
+ "epoch": 0.6846799580272823,
+ "grad_norm": 0.15929902143574254,
+ "learning_rate": 1.5165900832619015e-05,
+ "loss": 0.44286859035491943,
+ "step": 1305
+ },
+ {
+ "epoch": 0.6852046169989506,
+ "grad_norm": 0.14838868714016124,
+ "learning_rate": 1.5134900428840904e-05,
+ "loss": 0.45155346393585205,
+ "step": 1306
+ },
+ {
+ "epoch": 0.6857292759706191,
+ "grad_norm": 0.15494288640397672,
+ "learning_rate": 1.5103933618270283e-05,
+ "loss": 0.44010141491889954,
+ "step": 1307
+ },
+ {
+ "epoch": 0.6862539349422875,
+ "grad_norm": 0.1686262360598338,
+ "learning_rate": 1.5073000485037042e-05,
+ "loss": 0.446555495262146,
+ "step": 1308
+ },
+ {
+ "epoch": 0.6867785939139559,
+ "grad_norm": 0.16122110823982602,
+ "learning_rate": 1.5042101113179579e-05,
+ "loss": 0.44640374183654785,
+ "step": 1309
+ },
+ {
+ "epoch": 0.6873032528856243,
+ "grad_norm": 0.16849937797919867,
+ "learning_rate": 1.5011235586644572e-05,
+ "loss": 0.4465785026550293,
+ "step": 1310
+ },
+ {
+ "epoch": 0.6878279118572928,
+ "grad_norm": 0.16062971262800754,
+ "learning_rate": 1.498040398928674e-05,
+ "loss": 0.445321649312973,
+ "step": 1311
+ },
+ {
+ "epoch": 0.6883525708289612,
+ "grad_norm": 0.1608953185993034,
+ "learning_rate": 1.4949606404868638e-05,
+ "loss": 0.46518227458000183,
+ "step": 1312
+ },
+ {
+ "epoch": 0.6888772298006296,
+ "grad_norm": 0.15880255672830915,
+ "learning_rate": 1.491884291706041e-05,
+ "loss": 0.44582217931747437,
+ "step": 1313
+ },
+ {
+ "epoch": 0.689401888772298,
+ "grad_norm": 0.16715145200789822,
+ "learning_rate": 1.4888113609439563e-05,
+ "loss": 0.4435983896255493,
+ "step": 1314
+ },
+ {
+ "epoch": 0.6899265477439664,
+ "grad_norm": 0.15263740340694454,
+ "learning_rate": 1.485741856549075e-05,
+ "loss": 0.4405066967010498,
+ "step": 1315
+ },
+ {
+ "epoch": 0.6904512067156349,
+ "grad_norm": 0.16247034518813966,
+ "learning_rate": 1.4826757868605517e-05,
+ "loss": 0.4544215500354767,
+ "step": 1316
+ },
+ {
+ "epoch": 0.6909758656873033,
+ "grad_norm": 0.15649146705978162,
+ "learning_rate": 1.4796131602082142e-05,
+ "loss": 0.4454413056373596,
+ "step": 1317
+ },
+ {
+ "epoch": 0.6915005246589717,
+ "grad_norm": 0.16286718636058178,
+ "learning_rate": 1.4765539849125318e-05,
+ "loss": 0.4497183561325073,
+ "step": 1318
+ },
+ {
+ "epoch": 0.6920251836306401,
+ "grad_norm": 0.15812342085005582,
+ "learning_rate": 1.473498269284599e-05,
+ "loss": 0.445032000541687,
+ "step": 1319
+ },
+ {
+ "epoch": 0.6925498426023085,
+ "grad_norm": 0.1558101290511678,
+ "learning_rate": 1.4704460216261123e-05,
+ "loss": 0.45769619941711426,
+ "step": 1320
+ },
+ {
+ "epoch": 0.6930745015739769,
+ "grad_norm": 0.15642889059495207,
+ "learning_rate": 1.4673972502293428e-05,
+ "loss": 0.44788360595703125,
+ "step": 1321
+ },
+ {
+ "epoch": 0.6935991605456453,
+ "grad_norm": 0.15722513518113074,
+ "learning_rate": 1.4643519633771222e-05,
+ "loss": 0.449631929397583,
+ "step": 1322
+ },
+ {
+ "epoch": 0.6941238195173137,
+ "grad_norm": 0.15587633748081542,
+ "learning_rate": 1.4613101693428125e-05,
+ "loss": 0.4450187087059021,
+ "step": 1323
+ },
+ {
+ "epoch": 0.6946484784889821,
+ "grad_norm": 0.1577107074882479,
+ "learning_rate": 1.458271876390287e-05,
+ "loss": 0.45108890533447266,
+ "step": 1324
+ },
+ {
+ "epoch": 0.6951731374606506,
+ "grad_norm": 0.15649839486990696,
+ "learning_rate": 1.4552370927739073e-05,
+ "loss": 0.46430301666259766,
+ "step": 1325
+ },
+ {
+ "epoch": 0.695697796432319,
+ "grad_norm": 0.18573600420690636,
+ "learning_rate": 1.4522058267385008e-05,
+ "loss": 0.4701972007751465,
+ "step": 1326
+ },
+ {
+ "epoch": 0.6962224554039874,
+ "grad_norm": 0.16296497084037542,
+ "learning_rate": 1.4491780865193402e-05,
+ "loss": 0.4607710838317871,
+ "step": 1327
+ },
+ {
+ "epoch": 0.6967471143756558,
+ "grad_norm": 0.161124007217343,
+ "learning_rate": 1.446153880342117e-05,
+ "loss": 0.45095115900039673,
+ "step": 1328
+ },
+ {
+ "epoch": 0.6972717733473243,
+ "grad_norm": 0.1537884039166766,
+ "learning_rate": 1.4431332164229222e-05,
+ "loss": 0.424068808555603,
+ "step": 1329
+ },
+ {
+ "epoch": 0.6977964323189927,
+ "grad_norm": 0.16588834503456087,
+ "learning_rate": 1.4401161029682237e-05,
+ "loss": 0.4622839093208313,
+ "step": 1330
+ },
+ {
+ "epoch": 0.6983210912906611,
+ "grad_norm": 0.16100510604015544,
+ "learning_rate": 1.4371025481748435e-05,
+ "loss": 0.4660705626010895,
+ "step": 1331
+ },
+ {
+ "epoch": 0.6988457502623295,
+ "grad_norm": 0.15559888940590075,
+ "learning_rate": 1.4340925602299337e-05,
+ "loss": 0.4484976530075073,
+ "step": 1332
+ },
+ {
+ "epoch": 0.6993704092339978,
+ "grad_norm": 0.15503821966213488,
+ "learning_rate": 1.4310861473109616e-05,
+ "loss": 0.45636600255966187,
+ "step": 1333
+ },
+ {
+ "epoch": 0.6998950682056663,
+ "grad_norm": 0.15539074896109334,
+ "learning_rate": 1.4280833175856743e-05,
+ "loss": 0.45863068103790283,
+ "step": 1334
+ },
+ {
+ "epoch": 0.7004197271773347,
+ "grad_norm": 0.15664821009154872,
+ "learning_rate": 1.4250840792120896e-05,
+ "loss": 0.44043463468551636,
+ "step": 1335
+ },
+ {
+ "epoch": 0.7009443861490031,
+ "grad_norm": 0.15825478808520285,
+ "learning_rate": 1.4220884403384655e-05,
+ "loss": 0.46522051095962524,
+ "step": 1336
+ },
+ {
+ "epoch": 0.7014690451206715,
+ "grad_norm": 0.16004268568141544,
+ "learning_rate": 1.4190964091032843e-05,
+ "loss": 0.46114686131477356,
+ "step": 1337
+ },
+ {
+ "epoch": 0.70199370409234,
+ "grad_norm": 0.16413139949945635,
+ "learning_rate": 1.4161079936352234e-05,
+ "loss": 0.4457515478134155,
+ "step": 1338
+ },
+ {
+ "epoch": 0.7025183630640084,
+ "grad_norm": 0.1538890577972702,
+ "learning_rate": 1.4131232020531398e-05,
+ "loss": 0.4616093337535858,
+ "step": 1339
+ },
+ {
+ "epoch": 0.7030430220356768,
+ "grad_norm": 0.16240737854323284,
+ "learning_rate": 1.4101420424660434e-05,
+ "loss": 0.44631537795066833,
+ "step": 1340
+ },
+ {
+ "epoch": 0.7035676810073452,
+ "grad_norm": 0.15984708603406073,
+ "learning_rate": 1.4071645229730762e-05,
+ "loss": 0.4366717040538788,
+ "step": 1341
+ },
+ {
+ "epoch": 0.7040923399790137,
+ "grad_norm": 0.15149695695635904,
+ "learning_rate": 1.4041906516634946e-05,
+ "loss": 0.44129428267478943,
+ "step": 1342
+ },
+ {
+ "epoch": 0.7046169989506821,
+ "grad_norm": 0.15041167501768415,
+ "learning_rate": 1.4012204366166399e-05,
+ "loss": 0.43654775619506836,
+ "step": 1343
+ },
+ {
+ "epoch": 0.7051416579223505,
+ "grad_norm": 0.1596097627963756,
+ "learning_rate": 1.3982538859019217e-05,
+ "loss": 0.4652445614337921,
+ "step": 1344
+ },
+ {
+ "epoch": 0.7056663168940189,
+ "grad_norm": 0.15621704205027884,
+ "learning_rate": 1.3952910075787941e-05,
+ "loss": 0.44577041268348694,
+ "step": 1345
+ },
+ {
+ "epoch": 0.7061909758656874,
+ "grad_norm": 0.16073041098008783,
+ "learning_rate": 1.3923318096967341e-05,
+ "loss": 0.4391976594924927,
+ "step": 1346
+ },
+ {
+ "epoch": 0.7067156348373557,
+ "grad_norm": 0.14793853946372448,
+ "learning_rate": 1.3893763002952204e-05,
+ "loss": 0.4463799297809601,
+ "step": 1347
+ },
+ {
+ "epoch": 0.7072402938090241,
+ "grad_norm": 0.16176080358867354,
+ "learning_rate": 1.3864244874037097e-05,
+ "loss": 0.45183271169662476,
+ "step": 1348
+ },
+ {
+ "epoch": 0.7077649527806925,
+ "grad_norm": 0.16372771179026924,
+ "learning_rate": 1.3834763790416171e-05,
+ "loss": 0.44205576181411743,
+ "step": 1349
+ },
+ {
+ "epoch": 0.7082896117523609,
+ "grad_norm": 0.15135222516127125,
+ "learning_rate": 1.3805319832182937e-05,
+ "loss": 0.4390261173248291,
+ "step": 1350
+ },
+ {
+ "epoch": 0.7088142707240294,
+ "grad_norm": 0.15493183225792365,
+ "learning_rate": 1.377591307933002e-05,
+ "loss": 0.45837974548339844,
+ "step": 1351
+ },
+ {
+ "epoch": 0.7093389296956978,
+ "grad_norm": 0.16405906906516504,
+ "learning_rate": 1.3746543611749008e-05,
+ "loss": 0.45032668113708496,
+ "step": 1352
+ },
+ {
+ "epoch": 0.7098635886673662,
+ "grad_norm": 0.15725677976524505,
+ "learning_rate": 1.3717211509230155e-05,
+ "loss": 0.4486224949359894,
+ "step": 1353
+ },
+ {
+ "epoch": 0.7103882476390346,
+ "grad_norm": 0.15781469066983164,
+ "learning_rate": 1.3687916851462227e-05,
+ "loss": 0.4502638578414917,
+ "step": 1354
+ },
+ {
+ "epoch": 0.710912906610703,
+ "grad_norm": 0.16115007729054298,
+ "learning_rate": 1.3658659718032241e-05,
+ "loss": 0.4508914351463318,
+ "step": 1355
+ },
+ {
+ "epoch": 0.7114375655823715,
+ "grad_norm": 0.16755060475968744,
+ "learning_rate": 1.3629440188425286e-05,
+ "loss": 0.45653945207595825,
+ "step": 1356
+ },
+ {
+ "epoch": 0.7119622245540399,
+ "grad_norm": 0.1574994294756123,
+ "learning_rate": 1.3600258342024268e-05,
+ "loss": 0.4349849820137024,
+ "step": 1357
+ },
+ {
+ "epoch": 0.7124868835257083,
+ "grad_norm": 0.16104440922657715,
+ "learning_rate": 1.3571114258109758e-05,
+ "loss": 0.4501233696937561,
+ "step": 1358
+ },
+ {
+ "epoch": 0.7130115424973767,
+ "grad_norm": 0.15163184290939022,
+ "learning_rate": 1.3542008015859691e-05,
+ "loss": 0.44171035289764404,
+ "step": 1359
+ },
+ {
+ "epoch": 0.7135362014690452,
+ "grad_norm": 0.15955896674456582,
+ "learning_rate": 1.3512939694349211e-05,
+ "loss": 0.4558671712875366,
+ "step": 1360
+ },
+ {
+ "epoch": 0.7140608604407135,
+ "grad_norm": 0.1589376751536802,
+ "learning_rate": 1.3483909372550446e-05,
+ "loss": 0.4430835247039795,
+ "step": 1361
+ },
+ {
+ "epoch": 0.7145855194123819,
+ "grad_norm": 0.15979088397822155,
+ "learning_rate": 1.3454917129332279e-05,
+ "loss": 0.4426245391368866,
+ "step": 1362
+ },
+ {
+ "epoch": 0.7151101783840503,
+ "grad_norm": 0.15493179714072167,
+ "learning_rate": 1.3425963043460143e-05,
+ "loss": 0.4550474286079407,
+ "step": 1363
+ },
+ {
+ "epoch": 0.7156348373557188,
+ "grad_norm": 0.15396498968995032,
+ "learning_rate": 1.3397047193595808e-05,
+ "loss": 0.44220536947250366,
+ "step": 1364
+ },
+ {
+ "epoch": 0.7161594963273872,
+ "grad_norm": 0.1489346822667273,
+ "learning_rate": 1.336816965829717e-05,
+ "loss": 0.4437323212623596,
+ "step": 1365
+ },
+ {
+ "epoch": 0.7166841552990556,
+ "grad_norm": 0.16311964931734477,
+ "learning_rate": 1.333933051601802e-05,
+ "loss": 0.45308777689933777,
+ "step": 1366
+ },
+ {
+ "epoch": 0.717208814270724,
+ "grad_norm": 0.1534354619021179,
+ "learning_rate": 1.331052984510785e-05,
+ "loss": 0.4334886372089386,
+ "step": 1367
+ },
+ {
+ "epoch": 0.7177334732423925,
+ "grad_norm": 0.15541089758065404,
+ "learning_rate": 1.328176772381165e-05,
+ "loss": 0.45248323678970337,
+ "step": 1368
+ },
+ {
+ "epoch": 0.7182581322140609,
+ "grad_norm": 0.1617921305700596,
+ "learning_rate": 1.3253044230269651e-05,
+ "loss": 0.44092947244644165,
+ "step": 1369
+ },
+ {
+ "epoch": 0.7187827911857293,
+ "grad_norm": 0.15047353793760151,
+ "learning_rate": 1.3224359442517165e-05,
+ "loss": 0.4424414336681366,
+ "step": 1370
+ },
+ {
+ "epoch": 0.7193074501573977,
+ "grad_norm": 0.15958089102258743,
+ "learning_rate": 1.319571343848433e-05,
+ "loss": 0.4458580017089844,
+ "step": 1371
+ },
+ {
+ "epoch": 0.7198321091290661,
+ "grad_norm": 0.15349258306968733,
+ "learning_rate": 1.3167106295995924e-05,
+ "loss": 0.4285386800765991,
+ "step": 1372
+ },
+ {
+ "epoch": 0.7203567681007346,
+ "grad_norm": 0.16042529563759886,
+ "learning_rate": 1.3138538092771164e-05,
+ "loss": 0.4497690200805664,
+ "step": 1373
+ },
+ {
+ "epoch": 0.720881427072403,
+ "grad_norm": 0.1565113736588754,
+ "learning_rate": 1.3110008906423454e-05,
+ "loss": 0.43562617897987366,
+ "step": 1374
+ },
+ {
+ "epoch": 0.7214060860440713,
+ "grad_norm": 0.15177291086871506,
+ "learning_rate": 1.3081518814460215e-05,
+ "loss": 0.44456908106803894,
+ "step": 1375
+ },
+ {
+ "epoch": 0.7219307450157397,
+ "grad_norm": 0.15928490480439225,
+ "learning_rate": 1.3053067894282622e-05,
+ "loss": 0.46170780062675476,
+ "step": 1376
+ },
+ {
+ "epoch": 0.7224554039874082,
+ "grad_norm": 0.1554814469217892,
+ "learning_rate": 1.302465622318549e-05,
+ "loss": 0.4422169327735901,
+ "step": 1377
+ },
+ {
+ "epoch": 0.7229800629590766,
+ "grad_norm": 0.15375370214966386,
+ "learning_rate": 1.2996283878356947e-05,
+ "loss": 0.4564763903617859,
+ "step": 1378
+ },
+ {
+ "epoch": 0.723504721930745,
+ "grad_norm": 0.15942841969395802,
+ "learning_rate": 1.2967950936878311e-05,
+ "loss": 0.45121875405311584,
+ "step": 1379
+ },
+ {
+ "epoch": 0.7240293809024134,
+ "grad_norm": 0.15731308716774903,
+ "learning_rate": 1.2939657475723844e-05,
+ "loss": 0.4285027086734772,
+ "step": 1380
+ },
+ {
+ "epoch": 0.7245540398740818,
+ "grad_norm": 0.15997753613932147,
+ "learning_rate": 1.291140357176054e-05,
+ "loss": 0.4566386342048645,
+ "step": 1381
+ },
+ {
+ "epoch": 0.7250786988457503,
+ "grad_norm": 0.15333715267473727,
+ "learning_rate": 1.2883189301747927e-05,
+ "loss": 0.45093846321105957,
+ "step": 1382
+ },
+ {
+ "epoch": 0.7256033578174187,
+ "grad_norm": 0.1514510304709061,
+ "learning_rate": 1.2855014742337874e-05,
+ "loss": 0.44256967306137085,
+ "step": 1383
+ },
+ {
+ "epoch": 0.7261280167890871,
+ "grad_norm": 0.15624301298995746,
+ "learning_rate": 1.2826879970074346e-05,
+ "loss": 0.4361875653266907,
+ "step": 1384
+ },
+ {
+ "epoch": 0.7266526757607555,
+ "grad_norm": 0.15767850122006694,
+ "learning_rate": 1.279878506139322e-05,
+ "loss": 0.4477671980857849,
+ "step": 1385
+ },
+ {
+ "epoch": 0.727177334732424,
+ "grad_norm": 0.15537676969694064,
+ "learning_rate": 1.2770730092622072e-05,
+ "loss": 0.4339832663536072,
+ "step": 1386
+ },
+ {
+ "epoch": 0.7277019937040924,
+ "grad_norm": 0.15944113633281293,
+ "learning_rate": 1.2742715139979965e-05,
+ "loss": 0.45407935976982117,
+ "step": 1387
+ },
+ {
+ "epoch": 0.7282266526757608,
+ "grad_norm": 0.16763519518606218,
+ "learning_rate": 1.271474027957727e-05,
+ "loss": 0.46442392468452454,
+ "step": 1388
+ },
+ {
+ "epoch": 0.7287513116474291,
+ "grad_norm": 0.1564572774601902,
+ "learning_rate": 1.2686805587415423e-05,
+ "loss": 0.4463968276977539,
+ "step": 1389
+ },
+ {
+ "epoch": 0.7292759706190975,
+ "grad_norm": 0.16384626314986128,
+ "learning_rate": 1.2658911139386709e-05,
+ "loss": 0.45519885420799255,
+ "step": 1390
+ },
+ {
+ "epoch": 0.729800629590766,
+ "grad_norm": 0.1562023802820964,
+ "learning_rate": 1.2631057011274106e-05,
+ "loss": 0.4365222454071045,
+ "step": 1391
+ },
+ {
+ "epoch": 0.7303252885624344,
+ "grad_norm": 0.15597287232940926,
+ "learning_rate": 1.2603243278751042e-05,
+ "loss": 0.4328266680240631,
+ "step": 1392
+ },
+ {
+ "epoch": 0.7308499475341028,
+ "grad_norm": 0.14762833706132167,
+ "learning_rate": 1.2575470017381214e-05,
+ "loss": 0.43902426958084106,
+ "step": 1393
+ },
+ {
+ "epoch": 0.7313746065057712,
+ "grad_norm": 0.1688371515856315,
+ "learning_rate": 1.2547737302618351e-05,
+ "loss": 0.4649963676929474,
+ "step": 1394
+ },
+ {
+ "epoch": 0.7318992654774397,
+ "grad_norm": 0.15501514509586015,
+ "learning_rate": 1.2520045209806024e-05,
+ "loss": 0.4501558542251587,
+ "step": 1395
+ },
+ {
+ "epoch": 0.7324239244491081,
+ "grad_norm": 0.15516748917693537,
+ "learning_rate": 1.2492393814177457e-05,
+ "loss": 0.4405498206615448,
+ "step": 1396
+ },
+ {
+ "epoch": 0.7329485834207765,
+ "grad_norm": 0.17115400468199243,
+ "learning_rate": 1.2464783190855295e-05,
+ "loss": 0.4445379674434662,
+ "step": 1397
+ },
+ {
+ "epoch": 0.7334732423924449,
+ "grad_norm": 0.15966386341604186,
+ "learning_rate": 1.243721341485143e-05,
+ "loss": 0.4769889712333679,
+ "step": 1398
+ },
+ {
+ "epoch": 0.7339979013641134,
+ "grad_norm": 0.15614706359781472,
+ "learning_rate": 1.2409684561066766e-05,
+ "loss": 0.43643918633461,
+ "step": 1399
+ },
+ {
+ "epoch": 0.7345225603357818,
+ "grad_norm": 0.17467968960064226,
+ "learning_rate": 1.238219670429104e-05,
+ "loss": 0.45413005352020264,
+ "step": 1400
+ },
+ {
+ "epoch": 0.7350472193074502,
+ "grad_norm": 0.1544517430810819,
+ "learning_rate": 1.2354749919202598e-05,
+ "loss": 0.45218425989151,
+ "step": 1401
+ },
+ {
+ "epoch": 0.7355718782791185,
+ "grad_norm": 0.16551083673374484,
+ "learning_rate": 1.2327344280368217e-05,
+ "loss": 0.4459628462791443,
+ "step": 1402
+ },
+ {
+ "epoch": 0.7360965372507869,
+ "grad_norm": 0.17614490459613263,
+ "learning_rate": 1.2299979862242882e-05,
+ "loss": 0.4406280815601349,
+ "step": 1403
+ },
+ {
+ "epoch": 0.7366211962224554,
+ "grad_norm": 0.15554925394435032,
+ "learning_rate": 1.2272656739169588e-05,
+ "loss": 0.4380159378051758,
+ "step": 1404
+ },
+ {
+ "epoch": 0.7371458551941238,
+ "grad_norm": 0.16475466799870916,
+ "learning_rate": 1.2245374985379147e-05,
+ "loss": 0.4442812204360962,
+ "step": 1405
+ },
+ {
+ "epoch": 0.7376705141657922,
+ "grad_norm": 0.16877301168875541,
+ "learning_rate": 1.2218134674989978e-05,
+ "loss": 0.4589429199695587,
+ "step": 1406
+ },
+ {
+ "epoch": 0.7381951731374606,
+ "grad_norm": 0.1583278589457438,
+ "learning_rate": 1.2190935882007894e-05,
+ "loss": 0.44896745681762695,
+ "step": 1407
+ },
+ {
+ "epoch": 0.7387198321091291,
+ "grad_norm": 0.1582022541656192,
+ "learning_rate": 1.2163778680325947e-05,
+ "loss": 0.4367997646331787,
+ "step": 1408
+ },
+ {
+ "epoch": 0.7392444910807975,
+ "grad_norm": 0.1669573982300068,
+ "learning_rate": 1.2136663143724163e-05,
+ "loss": 0.4503386616706848,
+ "step": 1409
+ },
+ {
+ "epoch": 0.7397691500524659,
+ "grad_norm": 0.1637877523924539,
+ "learning_rate": 1.2109589345869387e-05,
+ "loss": 0.4660584330558777,
+ "step": 1410
+ },
+ {
+ "epoch": 0.7402938090241343,
+ "grad_norm": 0.1640136159943701,
+ "learning_rate": 1.2082557360315065e-05,
+ "loss": 0.45206528902053833,
+ "step": 1411
+ },
+ {
+ "epoch": 0.7408184679958028,
+ "grad_norm": 0.1596580168682554,
+ "learning_rate": 1.2055567260501054e-05,
+ "loss": 0.4405035376548767,
+ "step": 1412
+ },
+ {
+ "epoch": 0.7413431269674712,
+ "grad_norm": 0.15224511741945293,
+ "learning_rate": 1.2028619119753402e-05,
+ "loss": 0.42810291051864624,
+ "step": 1413
+ },
+ {
+ "epoch": 0.7418677859391396,
+ "grad_norm": 0.15631381931817098,
+ "learning_rate": 1.2001713011284193e-05,
+ "loss": 0.4421626925468445,
+ "step": 1414
+ },
+ {
+ "epoch": 0.742392444910808,
+ "grad_norm": 0.15721819981265503,
+ "learning_rate": 1.1974849008191288e-05,
+ "loss": 0.4540696144104004,
+ "step": 1415
+ },
+ {
+ "epoch": 0.7429171038824763,
+ "grad_norm": 0.15904478578744463,
+ "learning_rate": 1.1948027183458182e-05,
+ "loss": 0.4601830244064331,
+ "step": 1416
+ },
+ {
+ "epoch": 0.7434417628541448,
+ "grad_norm": 0.1530755605235123,
+ "learning_rate": 1.1921247609953747e-05,
+ "loss": 0.4552589952945709,
+ "step": 1417
+ },
+ {
+ "epoch": 0.7439664218258132,
+ "grad_norm": 0.15375751072258462,
+ "learning_rate": 1.1894510360432106e-05,
+ "loss": 0.42965373396873474,
+ "step": 1418
+ },
+ {
+ "epoch": 0.7444910807974816,
+ "grad_norm": 0.15678686023937333,
+ "learning_rate": 1.1867815507532377e-05,
+ "loss": 0.44684016704559326,
+ "step": 1419
+ },
+ {
+ "epoch": 0.74501573976915,
+ "grad_norm": 0.15486952983783606,
+ "learning_rate": 1.1841163123778502e-05,
+ "loss": 0.44415023922920227,
+ "step": 1420
+ },
+ {
+ "epoch": 0.7455403987408185,
+ "grad_norm": 0.15342611490545593,
+ "learning_rate": 1.1814553281579034e-05,
+ "loss": 0.4501304626464844,
+ "step": 1421
+ },
+ {
+ "epoch": 0.7460650577124869,
+ "grad_norm": 0.15374260012788166,
+ "learning_rate": 1.1787986053226954e-05,
+ "loss": 0.4521749019622803,
+ "step": 1422
+ },
+ {
+ "epoch": 0.7465897166841553,
+ "grad_norm": 0.1543553258370361,
+ "learning_rate": 1.176146151089949e-05,
+ "loss": 0.44753748178482056,
+ "step": 1423
+ },
+ {
+ "epoch": 0.7471143756558237,
+ "grad_norm": 0.15921670193848306,
+ "learning_rate": 1.1734979726657878e-05,
+ "loss": 0.4473114311695099,
+ "step": 1424
+ },
+ {
+ "epoch": 0.7476390346274921,
+ "grad_norm": 0.15023498686186612,
+ "learning_rate": 1.17085407724472e-05,
+ "loss": 0.4568778872489929,
+ "step": 1425
+ },
+ {
+ "epoch": 0.7481636935991606,
+ "grad_norm": 0.14601156920274583,
+ "learning_rate": 1.1682144720096168e-05,
+ "loss": 0.4389464855194092,
+ "step": 1426
+ },
+ {
+ "epoch": 0.748688352570829,
+ "grad_norm": 0.1520548095149398,
+ "learning_rate": 1.1655791641316954e-05,
+ "loss": 0.440733402967453,
+ "step": 1427
+ },
+ {
+ "epoch": 0.7492130115424974,
+ "grad_norm": 0.1591915163097993,
+ "learning_rate": 1.1629481607704963e-05,
+ "loss": 0.46245887875556946,
+ "step": 1428
+ },
+ {
+ "epoch": 0.7497376705141658,
+ "grad_norm": 0.14867295658447263,
+ "learning_rate": 1.1603214690738683e-05,
+ "loss": 0.4332764744758606,
+ "step": 1429
+ },
+ {
+ "epoch": 0.7502623294858342,
+ "grad_norm": 0.1498957762903134,
+ "learning_rate": 1.1576990961779447e-05,
+ "loss": 0.45215466618537903,
+ "step": 1430
+ },
+ {
+ "epoch": 0.7507869884575026,
+ "grad_norm": 0.15430555803491824,
+ "learning_rate": 1.1550810492071243e-05,
+ "loss": 0.4454396963119507,
+ "step": 1431
+ },
+ {
+ "epoch": 0.751311647429171,
+ "grad_norm": 0.1517461021936533,
+ "learning_rate": 1.1524673352740543e-05,
+ "loss": 0.43869835138320923,
+ "step": 1432
+ },
+ {
+ "epoch": 0.7518363064008394,
+ "grad_norm": 0.14541657994692936,
+ "learning_rate": 1.1498579614796121e-05,
+ "loss": 0.4373054504394531,
+ "step": 1433
+ },
+ {
+ "epoch": 0.7523609653725079,
+ "grad_norm": 0.15539513763786977,
+ "learning_rate": 1.1472529349128812e-05,
+ "loss": 0.45451128482818604,
+ "step": 1434
+ },
+ {
+ "epoch": 0.7528856243441763,
+ "grad_norm": 0.1482043113690946,
+ "learning_rate": 1.144652262651136e-05,
+ "loss": 0.43644773960113525,
+ "step": 1435
+ },
+ {
+ "epoch": 0.7534102833158447,
+ "grad_norm": 0.16293628094190815,
+ "learning_rate": 1.1420559517598205e-05,
+ "loss": 0.4687269330024719,
+ "step": 1436
+ },
+ {
+ "epoch": 0.7539349422875131,
+ "grad_norm": 0.1553393974390287,
+ "learning_rate": 1.1394640092925302e-05,
+ "loss": 0.4501398205757141,
+ "step": 1437
+ },
+ {
+ "epoch": 0.7544596012591815,
+ "grad_norm": 0.16214837694535802,
+ "learning_rate": 1.1368764422909919e-05,
+ "loss": 0.4542577862739563,
+ "step": 1438
+ },
+ {
+ "epoch": 0.75498426023085,
+ "grad_norm": 0.15838878372428297,
+ "learning_rate": 1.1342932577850473e-05,
+ "loss": 0.46228161454200745,
+ "step": 1439
+ },
+ {
+ "epoch": 0.7555089192025184,
+ "grad_norm": 0.15090419497122687,
+ "learning_rate": 1.13171446279263e-05,
+ "loss": 0.4495932161808014,
+ "step": 1440
+ },
+ {
+ "epoch": 0.7560335781741868,
+ "grad_norm": 0.1543376786458394,
+ "learning_rate": 1.1291400643197484e-05,
+ "loss": 0.451630175113678,
+ "step": 1441
+ },
+ {
+ "epoch": 0.7565582371458552,
+ "grad_norm": 0.1539470751610696,
+ "learning_rate": 1.1265700693604668e-05,
+ "loss": 0.447547972202301,
+ "step": 1442
+ },
+ {
+ "epoch": 0.7570828961175237,
+ "grad_norm": 0.15761047703305445,
+ "learning_rate": 1.1240044848968856e-05,
+ "loss": 0.4516204595565796,
+ "step": 1443
+ },
+ {
+ "epoch": 0.757607555089192,
+ "grad_norm": 0.15675279683704735,
+ "learning_rate": 1.1214433178991258e-05,
+ "loss": 0.4597971737384796,
+ "step": 1444
+ },
+ {
+ "epoch": 0.7581322140608604,
+ "grad_norm": 0.15592110994904113,
+ "learning_rate": 1.1188865753253022e-05,
+ "loss": 0.443898469209671,
+ "step": 1445
+ },
+ {
+ "epoch": 0.7586568730325288,
+ "grad_norm": 0.15597501691030619,
+ "learning_rate": 1.116334264121513e-05,
+ "loss": 0.4581306278705597,
+ "step": 1446
+ },
+ {
+ "epoch": 0.7591815320041972,
+ "grad_norm": 0.15552304728589958,
+ "learning_rate": 1.1137863912218154e-05,
+ "loss": 0.4475254416465759,
+ "step": 1447
+ },
+ {
+ "epoch": 0.7597061909758657,
+ "grad_norm": 0.15759990123898035,
+ "learning_rate": 1.1112429635482118e-05,
+ "loss": 0.43585264682769775,
+ "step": 1448
+ },
+ {
+ "epoch": 0.7602308499475341,
+ "grad_norm": 0.15964657264676635,
+ "learning_rate": 1.1087039880106247e-05,
+ "loss": 0.46786290407180786,
+ "step": 1449
+ },
+ {
+ "epoch": 0.7607555089192025,
+ "grad_norm": 0.15155804051218852,
+ "learning_rate": 1.106169471506882e-05,
+ "loss": 0.42292726039886475,
+ "step": 1450
+ },
+ {
+ "epoch": 0.7612801678908709,
+ "grad_norm": 0.15519270966930884,
+ "learning_rate": 1.103639420922698e-05,
+ "loss": 0.4248972237110138,
+ "step": 1451
+ },
+ {
+ "epoch": 0.7618048268625394,
+ "grad_norm": 0.1526330511128821,
+ "learning_rate": 1.1011138431316537e-05,
+ "loss": 0.4317784309387207,
+ "step": 1452
+ },
+ {
+ "epoch": 0.7623294858342078,
+ "grad_norm": 0.15797682979272226,
+ "learning_rate": 1.0985927449951779e-05,
+ "loss": 0.44506698846817017,
+ "step": 1453
+ },
+ {
+ "epoch": 0.7628541448058762,
+ "grad_norm": 0.15479979838505462,
+ "learning_rate": 1.096076133362531e-05,
+ "loss": 0.4471631348133087,
+ "step": 1454
+ },
+ {
+ "epoch": 0.7633788037775446,
+ "grad_norm": 0.15712267274447342,
+ "learning_rate": 1.093564015070783e-05,
+ "loss": 0.4321915805339813,
+ "step": 1455
+ },
+ {
+ "epoch": 0.7639034627492131,
+ "grad_norm": 0.15857431494180016,
+ "learning_rate": 1.0910563969447969e-05,
+ "loss": 0.44455766677856445,
+ "step": 1456
+ },
+ {
+ "epoch": 0.7644281217208815,
+ "grad_norm": 0.15377057080997789,
+ "learning_rate": 1.0885532857972097e-05,
+ "loss": 0.4567924439907074,
+ "step": 1457
+ },
+ {
+ "epoch": 0.7649527806925498,
+ "grad_norm": 0.15544844005009104,
+ "learning_rate": 1.0860546884284143e-05,
+ "loss": 0.4491572976112366,
+ "step": 1458
+ },
+ {
+ "epoch": 0.7654774396642182,
+ "grad_norm": 0.16210735566479478,
+ "learning_rate": 1.08356061162654e-05,
+ "loss": 0.43823331594467163,
+ "step": 1459
+ },
+ {
+ "epoch": 0.7660020986358866,
+ "grad_norm": 0.15213559486379452,
+ "learning_rate": 1.0810710621674355e-05,
+ "loss": 0.45375770330429077,
+ "step": 1460
+ },
+ {
+ "epoch": 0.7665267576075551,
+ "grad_norm": 0.1727832306402992,
+ "learning_rate": 1.0785860468146494e-05,
+ "loss": 0.4639572501182556,
+ "step": 1461
+ },
+ {
+ "epoch": 0.7670514165792235,
+ "grad_norm": 0.1637555902013663,
+ "learning_rate": 1.076105572319412e-05,
+ "loss": 0.4355923533439636,
+ "step": 1462
+ },
+ {
+ "epoch": 0.7675760755508919,
+ "grad_norm": 0.16646936030470866,
+ "learning_rate": 1.0736296454206166e-05,
+ "loss": 0.4650861322879791,
+ "step": 1463
+ },
+ {
+ "epoch": 0.7681007345225603,
+ "grad_norm": 0.14987620615431385,
+ "learning_rate": 1.0711582728448033e-05,
+ "loss": 0.45445430278778076,
+ "step": 1464
+ },
+ {
+ "epoch": 0.7686253934942288,
+ "grad_norm": 0.16428618403547932,
+ "learning_rate": 1.0686914613061377e-05,
+ "loss": 0.46050935983657837,
+ "step": 1465
+ },
+ {
+ "epoch": 0.7691500524658972,
+ "grad_norm": 0.15052999053128632,
+ "learning_rate": 1.0662292175063949e-05,
+ "loss": 0.43385741114616394,
+ "step": 1466
+ },
+ {
+ "epoch": 0.7696747114375656,
+ "grad_norm": 0.1654949402740152,
+ "learning_rate": 1.0637715481349397e-05,
+ "loss": 0.4434410333633423,
+ "step": 1467
+ },
+ {
+ "epoch": 0.770199370409234,
+ "grad_norm": 0.15923836508490036,
+ "learning_rate": 1.0613184598687088e-05,
+ "loss": 0.4410185217857361,
+ "step": 1468
+ },
+ {
+ "epoch": 0.7707240293809025,
+ "grad_norm": 0.15813007111170696,
+ "learning_rate": 1.0588699593721957e-05,
+ "loss": 0.43588417768478394,
+ "step": 1469
+ },
+ {
+ "epoch": 0.7712486883525709,
+ "grad_norm": 0.16978817898431764,
+ "learning_rate": 1.0564260532974274e-05,
+ "loss": 0.4514843225479126,
+ "step": 1470
+ },
+ {
+ "epoch": 0.7717733473242392,
+ "grad_norm": 0.1613844852697828,
+ "learning_rate": 1.0539867482839499e-05,
+ "loss": 0.4554520845413208,
+ "step": 1471
+ },
+ {
+ "epoch": 0.7722980062959076,
+ "grad_norm": 0.1587788181984633,
+ "learning_rate": 1.0515520509588075e-05,
+ "loss": 0.43200671672821045,
+ "step": 1472
+ },
+ {
+ "epoch": 0.772822665267576,
+ "grad_norm": 0.1539930824827353,
+ "learning_rate": 1.0491219679365283e-05,
+ "loss": 0.4343116283416748,
+ "step": 1473
+ },
+ {
+ "epoch": 0.7733473242392445,
+ "grad_norm": 0.1561612537773709,
+ "learning_rate": 1.0466965058191049e-05,
+ "loss": 0.4477815330028534,
+ "step": 1474
+ },
+ {
+ "epoch": 0.7738719832109129,
+ "grad_norm": 0.169990777691657,
+ "learning_rate": 1.0442756711959745e-05,
+ "loss": 0.4476611614227295,
+ "step": 1475
+ },
+ {
+ "epoch": 0.7743966421825813,
+ "grad_norm": 0.1624313049193793,
+ "learning_rate": 1.041859470644002e-05,
+ "loss": 0.47136399149894714,
+ "step": 1476
+ },
+ {
+ "epoch": 0.7749213011542497,
+ "grad_norm": 0.15993315458994328,
+ "learning_rate": 1.0394479107274645e-05,
+ "loss": 0.4672670364379883,
+ "step": 1477
+ },
+ {
+ "epoch": 0.7754459601259182,
+ "grad_norm": 0.16755349272347217,
+ "learning_rate": 1.0370409979980294e-05,
+ "loss": 0.45135706663131714,
+ "step": 1478
+ },
+ {
+ "epoch": 0.7759706190975866,
+ "grad_norm": 0.15665059815830362,
+ "learning_rate": 1.0346387389947407e-05,
+ "loss": 0.45178359746932983,
+ "step": 1479
+ },
+ {
+ "epoch": 0.776495278069255,
+ "grad_norm": 0.1554612731452571,
+ "learning_rate": 1.0322411402439983e-05,
+ "loss": 0.4412590265274048,
+ "step": 1480
+ },
+ {
+ "epoch": 0.7770199370409234,
+ "grad_norm": 0.1532496869312065,
+ "learning_rate": 1.0298482082595408e-05,
+ "loss": 0.44104987382888794,
+ "step": 1481
+ },
+ {
+ "epoch": 0.7775445960125918,
+ "grad_norm": 0.15065868316960426,
+ "learning_rate": 1.0274599495424286e-05,
+ "loss": 0.4336603283882141,
+ "step": 1482
+ },
+ {
+ "epoch": 0.7780692549842603,
+ "grad_norm": 0.1548373981760097,
+ "learning_rate": 1.0250763705810267e-05,
+ "loss": 0.4658375680446625,
+ "step": 1483
+ },
+ {
+ "epoch": 0.7785939139559287,
+ "grad_norm": 0.15470181175866582,
+ "learning_rate": 1.0226974778509846e-05,
+ "loss": 0.45232075452804565,
+ "step": 1484
+ },
+ {
+ "epoch": 0.779118572927597,
+ "grad_norm": 0.15507371187727106,
+ "learning_rate": 1.0203232778152228e-05,
+ "loss": 0.43640458583831787,
+ "step": 1485
+ },
+ {
+ "epoch": 0.7796432318992654,
+ "grad_norm": 0.15269440736544085,
+ "learning_rate": 1.01795377692391e-05,
+ "loss": 0.4400939345359802,
+ "step": 1486
+ },
+ {
+ "epoch": 0.7801678908709339,
+ "grad_norm": 0.15610899550395085,
+ "learning_rate": 1.0155889816144495e-05,
+ "loss": 0.4546602666378021,
+ "step": 1487
+ },
+ {
+ "epoch": 0.7806925498426023,
+ "grad_norm": 0.1559328242601668,
+ "learning_rate": 1.0132288983114611e-05,
+ "loss": 0.45510709285736084,
+ "step": 1488
+ },
+ {
+ "epoch": 0.7812172088142707,
+ "grad_norm": 0.1537255194600787,
+ "learning_rate": 1.0108735334267639e-05,
+ "loss": 0.4475342631340027,
+ "step": 1489
+ },
+ {
+ "epoch": 0.7817418677859391,
+ "grad_norm": 0.14616575061664402,
+ "learning_rate": 1.0085228933593558e-05,
+ "loss": 0.4378190040588379,
+ "step": 1490
+ },
+ {
+ "epoch": 0.7822665267576075,
+ "grad_norm": 0.1624360967746598,
+ "learning_rate": 1.0061769844953999e-05,
+ "loss": 0.4439539909362793,
+ "step": 1491
+ },
+ {
+ "epoch": 0.782791185729276,
+ "grad_norm": 0.15407336125298413,
+ "learning_rate": 1.003835813208206e-05,
+ "loss": 0.44074907898902893,
+ "step": 1492
+ },
+ {
+ "epoch": 0.7833158447009444,
+ "grad_norm": 0.15786980710136125,
+ "learning_rate": 1.001499385858211e-05,
+ "loss": 0.4491667151451111,
+ "step": 1493
+ },
+ {
+ "epoch": 0.7838405036726128,
+ "grad_norm": 0.15338836738460088,
+ "learning_rate": 9.991677087929665e-06,
+ "loss": 0.446003794670105,
+ "step": 1494
+ },
+ {
+ "epoch": 0.7843651626442812,
+ "grad_norm": 0.156710894892,
+ "learning_rate": 9.968407883471165e-06,
+ "loss": 0.43604689836502075,
+ "step": 1495
+ },
+ {
+ "epoch": 0.7848898216159497,
+ "grad_norm": 0.16100539063191666,
+ "learning_rate": 9.945186308423832e-06,
+ "loss": 0.4431147873401642,
+ "step": 1496
+ },
+ {
+ "epoch": 0.7854144805876181,
+ "grad_norm": 0.15776018851122997,
+ "learning_rate": 9.92201242587548e-06,
+ "loss": 0.4532504975795746,
+ "step": 1497
+ },
+ {
+ "epoch": 0.7859391395592865,
+ "grad_norm": 0.1540333423540934,
+ "learning_rate": 9.898886298784368e-06,
+ "loss": 0.43605154752731323,
+ "step": 1498
+ },
+ {
+ "epoch": 0.7864637985309548,
+ "grad_norm": 0.16009602938972295,
+ "learning_rate": 9.875807989979005e-06,
+ "loss": 0.4605238437652588,
+ "step": 1499
+ },
+ {
+ "epoch": 0.7869884575026233,
+ "grad_norm": 0.15639368804171708,
+ "learning_rate": 9.852777562157983e-06,
+ "loss": 0.4509708881378174,
+ "step": 1500
+ },
+ {
+ "epoch": 0.7875131164742917,
+ "grad_norm": 0.15074011188741518,
+ "learning_rate": 9.829795077889828e-06,
+ "loss": 0.4527304172515869,
+ "step": 1501
+ },
+ {
+ "epoch": 0.7880377754459601,
+ "grad_norm": 0.15155083996490956,
+ "learning_rate": 9.806860599612799e-06,
+ "loss": 0.4306413531303406,
+ "step": 1502
+ },
+ {
+ "epoch": 0.7885624344176285,
+ "grad_norm": 0.1478956630795445,
+ "learning_rate": 9.783974189634734e-06,
+ "loss": 0.4434714913368225,
+ "step": 1503
+ },
+ {
+ "epoch": 0.789087093389297,
+ "grad_norm": 0.15327668434806432,
+ "learning_rate": 9.761135910132902e-06,
+ "loss": 0.44431227445602417,
+ "step": 1504
+ },
+ {
+ "epoch": 0.7896117523609654,
+ "grad_norm": 0.2243177585497212,
+ "learning_rate": 9.738345823153783e-06,
+ "loss": 0.4541007876396179,
+ "step": 1505
+ },
+ {
+ "epoch": 0.7901364113326338,
+ "grad_norm": 0.1600239865812282,
+ "learning_rate": 9.715603990612945e-06,
+ "loss": 0.4456910490989685,
+ "step": 1506
+ },
+ {
+ "epoch": 0.7906610703043022,
+ "grad_norm": 0.15304634075332066,
+ "learning_rate": 9.692910474294858e-06,
+ "loss": 0.43965545296669006,
+ "step": 1507
+ },
+ {
+ "epoch": 0.7911857292759706,
+ "grad_norm": 0.15657752062053032,
+ "learning_rate": 9.670265335852725e-06,
+ "loss": 0.43358081579208374,
+ "step": 1508
+ },
+ {
+ "epoch": 0.7917103882476391,
+ "grad_norm": 0.16110057783374115,
+ "learning_rate": 9.647668636808304e-06,
+ "loss": 0.45711368322372437,
+ "step": 1509
+ },
+ {
+ "epoch": 0.7922350472193075,
+ "grad_norm": 0.1551133690299623,
+ "learning_rate": 9.625120438551791e-06,
+ "loss": 0.4613519608974457,
+ "step": 1510
+ },
+ {
+ "epoch": 0.7927597061909759,
+ "grad_norm": 0.15690897380725946,
+ "learning_rate": 9.602620802341581e-06,
+ "loss": 0.4472261071205139,
+ "step": 1511
+ },
+ {
+ "epoch": 0.7932843651626443,
+ "grad_norm": 0.147796599322911,
+ "learning_rate": 9.580169789304152e-06,
+ "loss": 0.4447203576564789,
+ "step": 1512
+ },
+ {
+ "epoch": 0.7938090241343126,
+ "grad_norm": 0.152019559199633,
+ "learning_rate": 9.557767460433878e-06,
+ "loss": 0.445905864238739,
+ "step": 1513
+ },
+ {
+ "epoch": 0.7943336831059811,
+ "grad_norm": 0.1574763924872037,
+ "learning_rate": 9.535413876592872e-06,
+ "loss": 0.4502032995223999,
+ "step": 1514
+ },
+ {
+ "epoch": 0.7948583420776495,
+ "grad_norm": 0.1524218315499056,
+ "learning_rate": 9.513109098510816e-06,
+ "loss": 0.43753111362457275,
+ "step": 1515
+ },
+ {
+ "epoch": 0.7953830010493179,
+ "grad_norm": 0.15589147437937434,
+ "learning_rate": 9.490853186784801e-06,
+ "loss": 0.44524312019348145,
+ "step": 1516
+ },
+ {
+ "epoch": 0.7959076600209863,
+ "grad_norm": 0.15604009714081407,
+ "learning_rate": 9.468646201879156e-06,
+ "loss": 0.44141292572021484,
+ "step": 1517
+ },
+ {
+ "epoch": 0.7964323189926548,
+ "grad_norm": 0.1562432554004435,
+ "learning_rate": 9.44648820412529e-06,
+ "loss": 0.461681067943573,
+ "step": 1518
+ },
+ {
+ "epoch": 0.7969569779643232,
+ "grad_norm": 0.15233588531495665,
+ "learning_rate": 9.424379253721513e-06,
+ "loss": 0.4529497027397156,
+ "step": 1519
+ },
+ {
+ "epoch": 0.7974816369359916,
+ "grad_norm": 0.16856991685290923,
+ "learning_rate": 9.40231941073291e-06,
+ "loss": 0.4483596086502075,
+ "step": 1520
+ },
+ {
+ "epoch": 0.79800629590766,
+ "grad_norm": 0.1546876712981077,
+ "learning_rate": 9.380308735091125e-06,
+ "loss": 0.43846625089645386,
+ "step": 1521
+ },
+ {
+ "epoch": 0.7985309548793285,
+ "grad_norm": 0.15198637556890449,
+ "learning_rate": 9.35834728659424e-06,
+ "loss": 0.4625553488731384,
+ "step": 1522
+ },
+ {
+ "epoch": 0.7990556138509969,
+ "grad_norm": 0.1519777007198701,
+ "learning_rate": 9.336435124906592e-06,
+ "loss": 0.44183820486068726,
+ "step": 1523
+ },
+ {
+ "epoch": 0.7995802728226653,
+ "grad_norm": 0.16125823074979284,
+ "learning_rate": 9.314572309558616e-06,
+ "loss": 0.46057307720184326,
+ "step": 1524
+ },
+ {
+ "epoch": 0.8001049317943337,
+ "grad_norm": 0.16283323624848034,
+ "learning_rate": 9.2927588999467e-06,
+ "loss": 0.4530155062675476,
+ "step": 1525
+ },
+ {
+ "epoch": 0.8006295907660022,
+ "grad_norm": 0.14998654489027738,
+ "learning_rate": 9.270994955332984e-06,
+ "loss": 0.43155500292778015,
+ "step": 1526
+ },
+ {
+ "epoch": 0.8011542497376705,
+ "grad_norm": 0.15683095030848093,
+ "learning_rate": 9.249280534845241e-06,
+ "loss": 0.45097050070762634,
+ "step": 1527
+ },
+ {
+ "epoch": 0.8016789087093389,
+ "grad_norm": 0.1561578254492378,
+ "learning_rate": 9.227615697476674e-06,
+ "loss": 0.42797762155532837,
+ "step": 1528
+ },
+ {
+ "epoch": 0.8022035676810073,
+ "grad_norm": 0.15642385261469652,
+ "learning_rate": 9.206000502085813e-06,
+ "loss": 0.4560159742832184,
+ "step": 1529
+ },
+ {
+ "epoch": 0.8027282266526757,
+ "grad_norm": 0.1623126279097499,
+ "learning_rate": 9.1844350073963e-06,
+ "loss": 0.4704165458679199,
+ "step": 1530
+ },
+ {
+ "epoch": 0.8032528856243442,
+ "grad_norm": 0.15090934786253773,
+ "learning_rate": 9.162919271996755e-06,
+ "loss": 0.44642555713653564,
+ "step": 1531
+ },
+ {
+ "epoch": 0.8037775445960126,
+ "grad_norm": 0.15305216978534084,
+ "learning_rate": 9.14145335434062e-06,
+ "loss": 0.4634283781051636,
+ "step": 1532
+ },
+ {
+ "epoch": 0.804302203567681,
+ "grad_norm": 0.14653219861345698,
+ "learning_rate": 9.120037312745979e-06,
+ "loss": 0.44544070959091187,
+ "step": 1533
+ },
+ {
+ "epoch": 0.8048268625393494,
+ "grad_norm": 0.16247397938502464,
+ "learning_rate": 9.098671205395422e-06,
+ "loss": 0.4552367925643921,
+ "step": 1534
+ },
+ {
+ "epoch": 0.8053515215110179,
+ "grad_norm": 0.14963174165902007,
+ "learning_rate": 9.077355090335892e-06,
+ "loss": 0.43211787939071655,
+ "step": 1535
+ },
+ {
+ "epoch": 0.8058761804826863,
+ "grad_norm": 0.1554853846772283,
+ "learning_rate": 9.056089025478495e-06,
+ "loss": 0.4319286644458771,
+ "step": 1536
+ },
+ {
+ "epoch": 0.8064008394543547,
+ "grad_norm": 0.15840895821310472,
+ "learning_rate": 9.03487306859837e-06,
+ "loss": 0.4494994878768921,
+ "step": 1537
+ },
+ {
+ "epoch": 0.8069254984260231,
+ "grad_norm": 0.1543496716800982,
+ "learning_rate": 9.013707277334515e-06,
+ "loss": 0.45500510931015015,
+ "step": 1538
+ },
+ {
+ "epoch": 0.8074501573976915,
+ "grad_norm": 0.14797282318550045,
+ "learning_rate": 8.992591709189651e-06,
+ "loss": 0.4385896325111389,
+ "step": 1539
+ },
+ {
+ "epoch": 0.8079748163693599,
+ "grad_norm": 0.17045570682787758,
+ "learning_rate": 8.971526421530066e-06,
+ "loss": 0.4585988521575928,
+ "step": 1540
+ },
+ {
+ "epoch": 0.8084994753410283,
+ "grad_norm": 0.14758220713044884,
+ "learning_rate": 8.950511471585407e-06,
+ "loss": 0.4295082092285156,
+ "step": 1541
+ },
+ {
+ "epoch": 0.8090241343126967,
+ "grad_norm": 0.16005752068175036,
+ "learning_rate": 8.929546916448597e-06,
+ "loss": 0.4507679343223572,
+ "step": 1542
+ },
+ {
+ "epoch": 0.8095487932843651,
+ "grad_norm": 0.14884684083313074,
+ "learning_rate": 8.908632813075641e-06,
+ "loss": 0.43547600507736206,
+ "step": 1543
+ },
+ {
+ "epoch": 0.8100734522560336,
+ "grad_norm": 0.14860107595788136,
+ "learning_rate": 8.88776921828547e-06,
+ "loss": 0.42700642347335815,
+ "step": 1544
+ },
+ {
+ "epoch": 0.810598111227702,
+ "grad_norm": 0.1472630611843838,
+ "learning_rate": 8.86695618875981e-06,
+ "loss": 0.4473339319229126,
+ "step": 1545
+ },
+ {
+ "epoch": 0.8111227701993704,
+ "grad_norm": 0.15105523788317893,
+ "learning_rate": 8.846193781043e-06,
+ "loss": 0.44923150539398193,
+ "step": 1546
+ },
+ {
+ "epoch": 0.8116474291710388,
+ "grad_norm": 0.16121409015263036,
+ "learning_rate": 8.82548205154185e-06,
+ "loss": 0.4653576612472534,
+ "step": 1547
+ },
+ {
+ "epoch": 0.8121720881427072,
+ "grad_norm": 0.1595195150959421,
+ "learning_rate": 8.804821056525497e-06,
+ "loss": 0.4522424638271332,
+ "step": 1548
+ },
+ {
+ "epoch": 0.8126967471143757,
+ "grad_norm": 0.15163510106316194,
+ "learning_rate": 8.784210852125232e-06,
+ "loss": 0.43657630681991577,
+ "step": 1549
+ },
+ {
+ "epoch": 0.8132214060860441,
+ "grad_norm": 0.15155109347013793,
+ "learning_rate": 8.763651494334377e-06,
+ "loss": 0.43183624744415283,
+ "step": 1550
+ },
+ {
+ "epoch": 0.8137460650577125,
+ "grad_norm": 0.14930024958763324,
+ "learning_rate": 8.7431430390081e-06,
+ "loss": 0.43563947081565857,
+ "step": 1551
+ },
+ {
+ "epoch": 0.8142707240293809,
+ "grad_norm": 0.15361200153430712,
+ "learning_rate": 8.722685541863283e-06,
+ "loss": 0.4387405514717102,
+ "step": 1552
+ },
+ {
+ "epoch": 0.8147953830010494,
+ "grad_norm": 0.15549528796193493,
+ "learning_rate": 8.702279058478366e-06,
+ "loss": 0.44347938895225525,
+ "step": 1553
+ },
+ {
+ "epoch": 0.8153200419727177,
+ "grad_norm": 0.16243042867861712,
+ "learning_rate": 8.681923644293194e-06,
+ "loss": 0.44852834939956665,
+ "step": 1554
+ },
+ {
+ "epoch": 0.8158447009443861,
+ "grad_norm": 0.15706129652944026,
+ "learning_rate": 8.661619354608873e-06,
+ "loss": 0.45419150590896606,
+ "step": 1555
+ },
+ {
+ "epoch": 0.8163693599160545,
+ "grad_norm": 0.14854258573134127,
+ "learning_rate": 8.641366244587611e-06,
+ "loss": 0.44866859912872314,
+ "step": 1556
+ },
+ {
+ "epoch": 0.816894018887723,
+ "grad_norm": 0.14896937678017533,
+ "learning_rate": 8.62116436925258e-06,
+ "loss": 0.44596076011657715,
+ "step": 1557
+ },
+ {
+ "epoch": 0.8174186778593914,
+ "grad_norm": 0.15614508495620508,
+ "learning_rate": 8.601013783487746e-06,
+ "loss": 0.45645207166671753,
+ "step": 1558
+ },
+ {
+ "epoch": 0.8179433368310598,
+ "grad_norm": 0.1486538543050606,
+ "learning_rate": 8.580914542037734e-06,
+ "loss": 0.4592300355434418,
+ "step": 1559
+ },
+ {
+ "epoch": 0.8184679958027282,
+ "grad_norm": 0.15475320756830518,
+ "learning_rate": 8.560866699507702e-06,
+ "loss": 0.4605134129524231,
+ "step": 1560
+ },
+ {
+ "epoch": 0.8189926547743966,
+ "grad_norm": 0.14754406728057914,
+ "learning_rate": 8.540870310363137e-06,
+ "loss": 0.43651673197746277,
+ "step": 1561
+ },
+ {
+ "epoch": 0.8195173137460651,
+ "grad_norm": 0.14871990701077825,
+ "learning_rate": 8.520925428929755e-06,
+ "loss": 0.44830837845802307,
+ "step": 1562
+ },
+ {
+ "epoch": 0.8200419727177335,
+ "grad_norm": 0.15358241317712973,
+ "learning_rate": 8.501032109393334e-06,
+ "loss": 0.4546644687652588,
+ "step": 1563
+ },
+ {
+ "epoch": 0.8205666316894019,
+ "grad_norm": 0.15650856948126138,
+ "learning_rate": 8.481190405799569e-06,
+ "loss": 0.4423891305923462,
+ "step": 1564
+ },
+ {
+ "epoch": 0.8210912906610703,
+ "grad_norm": 0.15474996746242287,
+ "learning_rate": 8.461400372053923e-06,
+ "loss": 0.44242313504219055,
+ "step": 1565
+ },
+ {
+ "epoch": 0.8216159496327388,
+ "grad_norm": 0.16544926591962897,
+ "learning_rate": 8.441662061921497e-06,
+ "loss": 0.4487020969390869,
+ "step": 1566
+ },
+ {
+ "epoch": 0.8221406086044072,
+ "grad_norm": 0.1489733586650647,
+ "learning_rate": 8.421975529026855e-06,
+ "loss": 0.4371356964111328,
+ "step": 1567
+ },
+ {
+ "epoch": 0.8226652675760755,
+ "grad_norm": 0.15995455652906124,
+ "learning_rate": 8.402340826853907e-06,
+ "loss": 0.4566021263599396,
+ "step": 1568
+ },
+ {
+ "epoch": 0.8231899265477439,
+ "grad_norm": 0.1500716152622102,
+ "learning_rate": 8.382758008745726e-06,
+ "loss": 0.43888187408447266,
+ "step": 1569
+ },
+ {
+ "epoch": 0.8237145855194123,
+ "grad_norm": 0.15554904485712012,
+ "learning_rate": 8.363227127904464e-06,
+ "loss": 0.4584001898765564,
+ "step": 1570
+ },
+ {
+ "epoch": 0.8242392444910808,
+ "grad_norm": 0.14820917815884277,
+ "learning_rate": 8.343748237391144e-06,
+ "loss": 0.43289071321487427,
+ "step": 1571
+ },
+ {
+ "epoch": 0.8247639034627492,
+ "grad_norm": 0.15801200747859714,
+ "learning_rate": 8.324321390125547e-06,
+ "loss": 0.44196009635925293,
+ "step": 1572
+ },
+ {
+ "epoch": 0.8252885624344176,
+ "grad_norm": 0.15323277975380797,
+ "learning_rate": 8.304946638886078e-06,
+ "loss": 0.43244463205337524,
+ "step": 1573
+ },
+ {
+ "epoch": 0.825813221406086,
+ "grad_norm": 0.15076940176344772,
+ "learning_rate": 8.285624036309583e-06,
+ "loss": 0.44401073455810547,
+ "step": 1574
+ },
+ {
+ "epoch": 0.8263378803777545,
+ "grad_norm": 0.1544958300475393,
+ "learning_rate": 8.266353634891266e-06,
+ "loss": 0.45198434591293335,
+ "step": 1575
+ },
+ {
+ "epoch": 0.8268625393494229,
+ "grad_norm": 0.15093528292068448,
+ "learning_rate": 8.247135486984486e-06,
+ "loss": 0.4403747618198395,
+ "step": 1576
+ },
+ {
+ "epoch": 0.8273871983210913,
+ "grad_norm": 0.14680662784017148,
+ "learning_rate": 8.227969644800648e-06,
+ "loss": 0.4375043213367462,
+ "step": 1577
+ },
+ {
+ "epoch": 0.8279118572927597,
+ "grad_norm": 0.14791704009291914,
+ "learning_rate": 8.208856160409055e-06,
+ "loss": 0.43811577558517456,
+ "step": 1578
+ },
+ {
+ "epoch": 0.8284365162644282,
+ "grad_norm": 0.14924185710518617,
+ "learning_rate": 8.189795085736765e-06,
+ "loss": 0.43853428959846497,
+ "step": 1579
+ },
+ {
+ "epoch": 0.8289611752360966,
+ "grad_norm": 0.15420576936040198,
+ "learning_rate": 8.170786472568451e-06,
+ "loss": 0.445467084646225,
+ "step": 1580
+ },
+ {
+ "epoch": 0.829485834207765,
+ "grad_norm": 0.15577901772605834,
+ "learning_rate": 8.151830372546258e-06,
+ "loss": 0.4529452621936798,
+ "step": 1581
+ },
+ {
+ "epoch": 0.8300104931794333,
+ "grad_norm": 0.15387840488797552,
+ "learning_rate": 8.132926837169676e-06,
+ "loss": 0.46009817719459534,
+ "step": 1582
+ },
+ {
+ "epoch": 0.8305351521511017,
+ "grad_norm": 0.1527913309957562,
+ "learning_rate": 8.11407591779537e-06,
+ "loss": 0.4513983726501465,
+ "step": 1583
+ },
+ {
+ "epoch": 0.8310598111227702,
+ "grad_norm": 0.1537090286922398,
+ "learning_rate": 8.095277665637061e-06,
+ "loss": 0.4537739157676697,
+ "step": 1584
+ },
+ {
+ "epoch": 0.8315844700944386,
+ "grad_norm": 0.15722962107654076,
+ "learning_rate": 8.076532131765408e-06,
+ "loss": 0.45232969522476196,
+ "step": 1585
+ },
+ {
+ "epoch": 0.832109129066107,
+ "grad_norm": 0.1450142396611787,
+ "learning_rate": 8.057839367107817e-06,
+ "loss": 0.4270634651184082,
+ "step": 1586
+ },
+ {
+ "epoch": 0.8326337880377754,
+ "grad_norm": 0.150264001570592,
+ "learning_rate": 8.039199422448355e-06,
+ "loss": 0.43563878536224365,
+ "step": 1587
+ },
+ {
+ "epoch": 0.8331584470094439,
+ "grad_norm": 0.14904943307590535,
+ "learning_rate": 8.020612348427572e-06,
+ "loss": 0.4401509165763855,
+ "step": 1588
+ },
+ {
+ "epoch": 0.8336831059811123,
+ "grad_norm": 0.15453821662860034,
+ "learning_rate": 8.002078195542393e-06,
+ "loss": 0.44545161724090576,
+ "step": 1589
+ },
+ {
+ "epoch": 0.8342077649527807,
+ "grad_norm": 0.1570809678830383,
+ "learning_rate": 7.983597014145951e-06,
+ "loss": 0.45228466391563416,
+ "step": 1590
+ },
+ {
+ "epoch": 0.8347324239244491,
+ "grad_norm": 0.14741240769629818,
+ "learning_rate": 7.965168854447493e-06,
+ "loss": 0.4485381245613098,
+ "step": 1591
+ },
+ {
+ "epoch": 0.8352570828961176,
+ "grad_norm": 0.15362267095533516,
+ "learning_rate": 7.946793766512195e-06,
+ "loss": 0.4358059763908386,
+ "step": 1592
+ },
+ {
+ "epoch": 0.835781741867786,
+ "grad_norm": 0.1544754586495605,
+ "learning_rate": 7.92847180026106e-06,
+ "loss": 0.43327778577804565,
+ "step": 1593
+ },
+ {
+ "epoch": 0.8363064008394544,
+ "grad_norm": 0.15023225026164963,
+ "learning_rate": 7.91020300547077e-06,
+ "loss": 0.44359731674194336,
+ "step": 1594
+ },
+ {
+ "epoch": 0.8368310598111228,
+ "grad_norm": 0.15665530392095345,
+ "learning_rate": 7.891987431773545e-06,
+ "loss": 0.44503387808799744,
+ "step": 1595
+ },
+ {
+ "epoch": 0.8373557187827911,
+ "grad_norm": 0.14717614926392292,
+ "learning_rate": 7.873825128657033e-06,
+ "loss": 0.45314061641693115,
+ "step": 1596
+ },
+ {
+ "epoch": 0.8378803777544596,
+ "grad_norm": 0.15207464934030612,
+ "learning_rate": 7.855716145464131e-06,
+ "loss": 0.4526442885398865,
+ "step": 1597
+ },
+ {
+ "epoch": 0.838405036726128,
+ "grad_norm": 0.15129881541794873,
+ "learning_rate": 7.837660531392904e-06,
+ "loss": 0.45127853751182556,
+ "step": 1598
+ },
+ {
+ "epoch": 0.8389296956977964,
+ "grad_norm": 0.15623362715417285,
+ "learning_rate": 7.819658335496405e-06,
+ "loss": 0.46432793140411377,
+ "step": 1599
+ },
+ {
+ "epoch": 0.8394543546694648,
+ "grad_norm": 0.1529219879642125,
+ "learning_rate": 7.801709606682573e-06,
+ "loss": 0.4475358724594116,
+ "step": 1600
+ },
+ {
+ "epoch": 0.8399790136411333,
+ "grad_norm": 0.14649139854087698,
+ "learning_rate": 7.78381439371409e-06,
+ "loss": 0.4363391101360321,
+ "step": 1601
+ },
+ {
+ "epoch": 0.8405036726128017,
+ "grad_norm": 0.14571233251948087,
+ "learning_rate": 7.765972745208243e-06,
+ "loss": 0.44419488310813904,
+ "step": 1602
+ },
+ {
+ "epoch": 0.8410283315844701,
+ "grad_norm": 0.15168460066406128,
+ "learning_rate": 7.748184709636791e-06,
+ "loss": 0.4433790445327759,
+ "step": 1603
+ },
+ {
+ "epoch": 0.8415529905561385,
+ "grad_norm": 0.14852860876513757,
+ "learning_rate": 7.730450335325853e-06,
+ "loss": 0.4493539333343506,
+ "step": 1604
+ },
+ {
+ "epoch": 0.842077649527807,
+ "grad_norm": 0.1505615491087377,
+ "learning_rate": 7.712769670455741e-06,
+ "loss": 0.4646318554878235,
+ "step": 1605
+ },
+ {
+ "epoch": 0.8426023084994754,
+ "grad_norm": 0.15407610486660114,
+ "learning_rate": 7.695142763060878e-06,
+ "loss": 0.4484049677848816,
+ "step": 1606
+ },
+ {
+ "epoch": 0.8431269674711438,
+ "grad_norm": 0.1521449709584124,
+ "learning_rate": 7.67756966102962e-06,
+ "loss": 0.43366891145706177,
+ "step": 1607
+ },
+ {
+ "epoch": 0.8436516264428122,
+ "grad_norm": 0.14747762647596166,
+ "learning_rate": 7.660050412104153e-06,
+ "loss": 0.43950092792510986,
+ "step": 1608
+ },
+ {
+ "epoch": 0.8441762854144806,
+ "grad_norm": 0.15512610809240027,
+ "learning_rate": 7.642585063880351e-06,
+ "loss": 0.46813055872917175,
+ "step": 1609
+ },
+ {
+ "epoch": 0.844700944386149,
+ "grad_norm": 0.15825480008734216,
+ "learning_rate": 7.625173663807655e-06,
+ "loss": 0.4348755478858948,
+ "step": 1610
+ },
+ {
+ "epoch": 0.8452256033578174,
+ "grad_norm": 0.1573968981854278,
+ "learning_rate": 7.6078162591889434e-06,
+ "loss": 0.45048123598098755,
+ "step": 1611
+ },
+ {
+ "epoch": 0.8457502623294858,
+ "grad_norm": 0.1517038960096417,
+ "learning_rate": 7.590512897180399e-06,
+ "loss": 0.43563610315322876,
+ "step": 1612
+ },
+ {
+ "epoch": 0.8462749213011542,
+ "grad_norm": 0.15503576044724665,
+ "learning_rate": 7.573263624791382e-06,
+ "loss": 0.4467726945877075,
+ "step": 1613
+ },
+ {
+ "epoch": 0.8467995802728226,
+ "grad_norm": 0.1560549927192375,
+ "learning_rate": 7.556068488884304e-06,
+ "loss": 0.4320477545261383,
+ "step": 1614
+ },
+ {
+ "epoch": 0.8473242392444911,
+ "grad_norm": 0.1679830356844654,
+ "learning_rate": 7.538927536174498e-06,
+ "loss": 0.4476034343242645,
+ "step": 1615
+ },
+ {
+ "epoch": 0.8478488982161595,
+ "grad_norm": 0.15169257383162213,
+ "learning_rate": 7.521840813230098e-06,
+ "loss": 0.4325854778289795,
+ "step": 1616
+ },
+ {
+ "epoch": 0.8483735571878279,
+ "grad_norm": 0.15980226324329827,
+ "learning_rate": 7.504808366471909e-06,
+ "loss": 0.4472721815109253,
+ "step": 1617
+ },
+ {
+ "epoch": 0.8488982161594963,
+ "grad_norm": 0.16932727986736365,
+ "learning_rate": 7.48783024217327e-06,
+ "loss": 0.45428287982940674,
+ "step": 1618
+ },
+ {
+ "epoch": 0.8494228751311648,
+ "grad_norm": 0.15065676306311376,
+ "learning_rate": 7.470906486459946e-06,
+ "loss": 0.44418883323669434,
+ "step": 1619
+ },
+ {
+ "epoch": 0.8499475341028332,
+ "grad_norm": 0.1533992998314183,
+ "learning_rate": 7.454037145309986e-06,
+ "loss": 0.43552401661872864,
+ "step": 1620
+ },
+ {
+ "epoch": 0.8504721930745016,
+ "grad_norm": 0.15479480515120017,
+ "learning_rate": 7.437222264553628e-06,
+ "loss": 0.444465696811676,
+ "step": 1621
+ },
+ {
+ "epoch": 0.85099685204617,
+ "grad_norm": 0.15845678337942112,
+ "learning_rate": 7.420461889873129e-06,
+ "loss": 0.44896578788757324,
+ "step": 1622
+ },
+ {
+ "epoch": 0.8515215110178384,
+ "grad_norm": 0.15038940533473208,
+ "learning_rate": 7.4037560668026804e-06,
+ "loss": 0.43062180280685425,
+ "step": 1623
+ },
+ {
+ "epoch": 0.8520461699895068,
+ "grad_norm": 0.1480017115109342,
+ "learning_rate": 7.387104840728259e-06,
+ "loss": 0.4630432724952698,
+ "step": 1624
+ },
+ {
+ "epoch": 0.8525708289611752,
+ "grad_norm": 0.15321305032824092,
+ "learning_rate": 7.370508256887514e-06,
+ "loss": 0.4419487714767456,
+ "step": 1625
+ },
+ {
+ "epoch": 0.8530954879328436,
+ "grad_norm": 0.1568246735699272,
+ "learning_rate": 7.353966360369659e-06,
+ "loss": 0.4483609199523926,
+ "step": 1626
+ },
+ {
+ "epoch": 0.853620146904512,
+ "grad_norm": 0.15539536583814875,
+ "learning_rate": 7.337479196115317e-06,
+ "loss": 0.4567753076553345,
+ "step": 1627
+ },
+ {
+ "epoch": 0.8541448058761805,
+ "grad_norm": 0.15528695146633797,
+ "learning_rate": 7.321046808916426e-06,
+ "loss": 0.4560334086418152,
+ "step": 1628
+ },
+ {
+ "epoch": 0.8546694648478489,
+ "grad_norm": 0.15005728554864303,
+ "learning_rate": 7.304669243416102e-06,
+ "loss": 0.43833765387535095,
+ "step": 1629
+ },
+ {
+ "epoch": 0.8551941238195173,
+ "grad_norm": 0.15959100645189242,
+ "learning_rate": 7.288346544108518e-06,
+ "loss": 0.4713831841945648,
+ "step": 1630
+ },
+ {
+ "epoch": 0.8557187827911857,
+ "grad_norm": 0.14898121463872632,
+ "learning_rate": 7.272078755338807e-06,
+ "loss": 0.44003745913505554,
+ "step": 1631
+ },
+ {
+ "epoch": 0.8562434417628542,
+ "grad_norm": 0.15083110802343333,
+ "learning_rate": 7.255865921302903e-06,
+ "loss": 0.429004043340683,
+ "step": 1632
+ },
+ {
+ "epoch": 0.8567681007345226,
+ "grad_norm": 0.1522757086796616,
+ "learning_rate": 7.239708086047445e-06,
+ "loss": 0.4403839707374573,
+ "step": 1633
+ },
+ {
+ "epoch": 0.857292759706191,
+ "grad_norm": 0.15719824838921245,
+ "learning_rate": 7.223605293469654e-06,
+ "loss": 0.4377821087837219,
+ "step": 1634
+ },
+ {
+ "epoch": 0.8578174186778594,
+ "grad_norm": 0.1450616202708373,
+ "learning_rate": 7.207557587317216e-06,
+ "loss": 0.4357878267765045,
+ "step": 1635
+ },
+ {
+ "epoch": 0.8583420776495279,
+ "grad_norm": 0.15703162220646427,
+ "learning_rate": 7.191565011188149e-06,
+ "loss": 0.44412243366241455,
+ "step": 1636
+ },
+ {
+ "epoch": 0.8588667366211962,
+ "grad_norm": 0.150175313334118,
+ "learning_rate": 7.175627608530716e-06,
+ "loss": 0.4543853998184204,
+ "step": 1637
+ },
+ {
+ "epoch": 0.8593913955928646,
+ "grad_norm": 0.15091711492716975,
+ "learning_rate": 7.159745422643261e-06,
+ "loss": 0.4268660545349121,
+ "step": 1638
+ },
+ {
+ "epoch": 0.859916054564533,
+ "grad_norm": 0.15449835654964011,
+ "learning_rate": 7.1439184966741295e-06,
+ "loss": 0.4507150650024414,
+ "step": 1639
+ },
+ {
+ "epoch": 0.8604407135362014,
+ "grad_norm": 0.15689612593442026,
+ "learning_rate": 7.128146873621533e-06,
+ "loss": 0.45520636439323425,
+ "step": 1640
+ },
+ {
+ "epoch": 0.8609653725078699,
+ "grad_norm": 0.1565226785234465,
+ "learning_rate": 7.11243059633345e-06,
+ "loss": 0.45102012157440186,
+ "step": 1641
+ },
+ {
+ "epoch": 0.8614900314795383,
+ "grad_norm": 0.15500877922230677,
+ "learning_rate": 7.096769707507481e-06,
+ "loss": 0.447115421295166,
+ "step": 1642
+ },
+ {
+ "epoch": 0.8620146904512067,
+ "grad_norm": 0.1595248428000243,
+ "learning_rate": 7.081164249690761e-06,
+ "loss": 0.4627235233783722,
+ "step": 1643
+ },
+ {
+ "epoch": 0.8625393494228751,
+ "grad_norm": 0.15958177510163488,
+ "learning_rate": 7.065614265279818e-06,
+ "loss": 0.4386388063430786,
+ "step": 1644
+ },
+ {
+ "epoch": 0.8630640083945436,
+ "grad_norm": 0.15152834825396286,
+ "learning_rate": 7.050119796520482e-06,
+ "loss": 0.4531296491622925,
+ "step": 1645
+ },
+ {
+ "epoch": 0.863588667366212,
+ "grad_norm": 0.15635733905345023,
+ "learning_rate": 7.034680885507752e-06,
+ "loss": 0.45573747158050537,
+ "step": 1646
+ },
+ {
+ "epoch": 0.8641133263378804,
+ "grad_norm": 0.15666918852633857,
+ "learning_rate": 7.019297574185703e-06,
+ "loss": 0.4499792158603668,
+ "step": 1647
+ },
+ {
+ "epoch": 0.8646379853095488,
+ "grad_norm": 0.156444627477174,
+ "learning_rate": 7.0039699043473405e-06,
+ "loss": 0.4545009136199951,
+ "step": 1648
+ },
+ {
+ "epoch": 0.8651626442812173,
+ "grad_norm": 0.14638989347446527,
+ "learning_rate": 6.988697917634515e-06,
+ "loss": 0.4397326409816742,
+ "step": 1649
+ },
+ {
+ "epoch": 0.8656873032528857,
+ "grad_norm": 0.15492813701343372,
+ "learning_rate": 6.973481655537793e-06,
+ "loss": 0.4402945041656494,
+ "step": 1650
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 4.366570907867546e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1650/video_preprocessor_config.json b/checkpoints/checkpoint-1650/video_preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..a933f0ec6941222c33aeba7980bd5acf8dd98e91
--- /dev/null
+++ b/checkpoints/checkpoint-1650/video_preprocessor_config.json
@@ -0,0 +1,30 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "do_sample_frames": true,
+ "fps": 2,
+ "image_mean": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "image_std": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "max_frames": 768,
+ "merge_size": 2,
+ "min_frames": 4,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 786432,
+ "shortest_edge": 131072
+ },
+ "temporal_patch_size": 2,
+ "video_processor_type": "Qwen3VLVideoProcessor"
+}
diff --git a/checkpoints/checkpoint-1700/chat_template.jinja b/checkpoints/checkpoint-1700/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1700/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1700/config.json b/checkpoints/checkpoint-1700/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1700/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1700/expert_prune_metadata.json b/checkpoints/checkpoint-1700/expert_prune_metadata.json
new file mode 100644
index 0000000000000000000000000000000000000000..16e01c8a10636331e334732fa9629579c68df44e
--- /dev/null
+++ b/checkpoints/checkpoint-1700/expert_prune_metadata.json
@@ -0,0 +1,14 @@
+{
+ "source_model": "/mnt/data/wuwei/model_list/Qwen3.6-35B-A3B/models--Qwen--Qwen3.6-35B-A3B/snapshots/995ad96eacd98c81ed38be0c5b274b04031597b0",
+ "output_model": "/mnt/data/wuwei/model_list/qwen3-19b-a3b",
+ "total_parameters": 18990568816,
+ "parameter_billions_decimal": 18.990568816,
+ "directory_size_label": "19B",
+ "original_num_experts": 256,
+ "kept_experts_per_layer": 128,
+ "num_profiled_layers": 40,
+ "mtp_weights": "removed because Qwen3_5MoeForConditionalGeneration does not instantiate or use the source checkpoint's mtp.* training-only tensors",
+ "expert_order": "ascending original expert index; router rows and expert tensors use the same remapping",
+ "router_statistics": "router_stats.json and router_stats.pt",
+ "selection_score": "top-8 router hit frequency * mean(abs(actual expert output after down_proj and before routing-weight multiplication))"
+}
diff --git a/checkpoints/checkpoint-1700/generation_config.json b/checkpoints/checkpoint-1700/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1700/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1700/model.safetensors b/checkpoints/checkpoint-1700/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..6922d801f9b270ab8e141817f7a0f871fa931431
--- /dev/null
+++ b/checkpoints/checkpoint-1700/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:427592c9a52ad75a9d6224940993e2b7f6ef9c1026101a6b111203c0a91cadeb
+size 37983421000
diff --git a/checkpoints/checkpoint-1700/preprocessor_config.json b/checkpoints/checkpoint-1700/preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ec43d1f859ecadf800ebae92fda4a812d34975d3
--- /dev/null
+++ b/checkpoints/checkpoint-1700/preprocessor_config.json
@@ -0,0 +1,26 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "image_mean": [
+ 0.48145466,
+ 0.4578275,
+ 0.40821073
+ ],
+ "image_processor_type": "Qwen2VLImageProcessorFast",
+ "image_std": [
+ 0.26862954,
+ 0.26130258,
+ 0.27577711
+ ],
+ "merge_size": 2,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 1003520,
+ "shortest_edge": 3136
+ },
+ "temporal_patch_size": 2
+}
diff --git a/checkpoints/checkpoint-1700/tokenizer.json b/checkpoints/checkpoint-1700/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1700/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1700/tokenizer_config.json b/checkpoints/checkpoint-1700/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1700/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1700/trainer_state.json b/checkpoints/checkpoint-1700/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..d7afbe8dc91b667a419badea1144f491286b7a3b
--- /dev/null
+++ b/checkpoints/checkpoint-1700/trainer_state.json
@@ -0,0 +1,11934 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.8919202518363064,
+ "eval_steps": 500,
+ "global_step": 1700,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ },
+ {
+ "epoch": 0.5776495278069255,
+ "grad_norm": 0.1862053032854098,
+ "learning_rate": 2.2065553964033126e-05,
+ "loss": 0.4816875457763672,
+ "step": 1101
+ },
+ {
+ "epoch": 0.5781741867785939,
+ "grad_norm": 0.16996015894363117,
+ "learning_rate": 2.20295734045532e-05,
+ "loss": 0.4471197724342346,
+ "step": 1102
+ },
+ {
+ "epoch": 0.5786988457502623,
+ "grad_norm": 0.16889227008390692,
+ "learning_rate": 2.1993607706997725e-05,
+ "loss": 0.468391478061676,
+ "step": 1103
+ },
+ {
+ "epoch": 0.5792235047219307,
+ "grad_norm": 0.16670931681562373,
+ "learning_rate": 2.195765696907745e-05,
+ "loss": 0.44934162497520447,
+ "step": 1104
+ },
+ {
+ "epoch": 0.5797481636935992,
+ "grad_norm": 0.16328418741223716,
+ "learning_rate": 2.1921721288462482e-05,
+ "loss": 0.4558270275592804,
+ "step": 1105
+ },
+ {
+ "epoch": 0.5802728226652676,
+ "grad_norm": 0.17364660408703936,
+ "learning_rate": 2.188580076278203e-05,
+ "loss": 0.4597751796245575,
+ "step": 1106
+ },
+ {
+ "epoch": 0.580797481636936,
+ "grad_norm": 0.1628618453800601,
+ "learning_rate": 2.18498954896241e-05,
+ "loss": 0.4521859288215637,
+ "step": 1107
+ },
+ {
+ "epoch": 0.5813221406086044,
+ "grad_norm": 0.16813817951871482,
+ "learning_rate": 2.1814005566535302e-05,
+ "loss": 0.4581005573272705,
+ "step": 1108
+ },
+ {
+ "epoch": 0.5818467995802729,
+ "grad_norm": 0.16825923629726025,
+ "learning_rate": 2.1778131091020516e-05,
+ "loss": 0.4433916211128235,
+ "step": 1109
+ },
+ {
+ "epoch": 0.5823714585519413,
+ "grad_norm": 0.17194743467883944,
+ "learning_rate": 2.1742272160542654e-05,
+ "loss": 0.47296425700187683,
+ "step": 1110
+ },
+ {
+ "epoch": 0.5828961175236097,
+ "grad_norm": 0.16947730546039969,
+ "learning_rate": 2.170642887252241e-05,
+ "loss": 0.4815937876701355,
+ "step": 1111
+ },
+ {
+ "epoch": 0.583420776495278,
+ "grad_norm": 0.1657449928668473,
+ "learning_rate": 2.167060132433798e-05,
+ "loss": 0.4479164779186249,
+ "step": 1112
+ },
+ {
+ "epoch": 0.5839454354669464,
+ "grad_norm": 0.1658952516607364,
+ "learning_rate": 2.1634789613324778e-05,
+ "loss": 0.4637463688850403,
+ "step": 1113
+ },
+ {
+ "epoch": 0.5844700944386149,
+ "grad_norm": 0.16777002188998236,
+ "learning_rate": 2.1598993836775218e-05,
+ "loss": 0.457524836063385,
+ "step": 1114
+ },
+ {
+ "epoch": 0.5849947534102833,
+ "grad_norm": 0.15837566447341322,
+ "learning_rate": 2.1563214091938405e-05,
+ "loss": 0.44938087463378906,
+ "step": 1115
+ },
+ {
+ "epoch": 0.5855194123819517,
+ "grad_norm": 0.17162206617035314,
+ "learning_rate": 2.1527450476019896e-05,
+ "loss": 0.4393450617790222,
+ "step": 1116
+ },
+ {
+ "epoch": 0.5860440713536201,
+ "grad_norm": 0.16224101921433814,
+ "learning_rate": 2.1491703086181426e-05,
+ "loss": 0.451913982629776,
+ "step": 1117
+ },
+ {
+ "epoch": 0.5865687303252886,
+ "grad_norm": 0.1736253504575144,
+ "learning_rate": 2.145597201954066e-05,
+ "loss": 0.4611290395259857,
+ "step": 1118
+ },
+ {
+ "epoch": 0.587093389296957,
+ "grad_norm": 0.16308408333737234,
+ "learning_rate": 2.1420257373170903e-05,
+ "loss": 0.45898568630218506,
+ "step": 1119
+ },
+ {
+ "epoch": 0.5876180482686254,
+ "grad_norm": 0.16948634794175524,
+ "learning_rate": 2.1384559244100848e-05,
+ "loss": 0.45982882380485535,
+ "step": 1120
+ },
+ {
+ "epoch": 0.5881427072402938,
+ "grad_norm": 0.16676366538565252,
+ "learning_rate": 2.1348877729314336e-05,
+ "loss": 0.44826576113700867,
+ "step": 1121
+ },
+ {
+ "epoch": 0.5886673662119623,
+ "grad_norm": 0.16431125630712295,
+ "learning_rate": 2.1313212925750037e-05,
+ "loss": 0.4616445302963257,
+ "step": 1122
+ },
+ {
+ "epoch": 0.5891920251836307,
+ "grad_norm": 0.1631659061616955,
+ "learning_rate": 2.1277564930301252e-05,
+ "loss": 0.45651188492774963,
+ "step": 1123
+ },
+ {
+ "epoch": 0.5897166841552991,
+ "grad_norm": 0.15805950502689237,
+ "learning_rate": 2.1241933839815604e-05,
+ "loss": 0.4486045837402344,
+ "step": 1124
+ },
+ {
+ "epoch": 0.5902413431269675,
+ "grad_norm": 0.16530288038603502,
+ "learning_rate": 2.1206319751094773e-05,
+ "loss": 0.4485066533088684,
+ "step": 1125
+ },
+ {
+ "epoch": 0.5907660020986358,
+ "grad_norm": 0.16061903194239047,
+ "learning_rate": 2.1170722760894295e-05,
+ "loss": 0.4565330743789673,
+ "step": 1126
+ },
+ {
+ "epoch": 0.5912906610703043,
+ "grad_norm": 0.1643346999353627,
+ "learning_rate": 2.1135142965923195e-05,
+ "loss": 0.45560717582702637,
+ "step": 1127
+ },
+ {
+ "epoch": 0.5918153200419727,
+ "grad_norm": 0.17484644315215753,
+ "learning_rate": 2.109958046284382e-05,
+ "loss": 0.4795891046524048,
+ "step": 1128
+ },
+ {
+ "epoch": 0.5923399790136411,
+ "grad_norm": 0.1623762678622651,
+ "learning_rate": 2.1064035348271528e-05,
+ "loss": 0.47395604848861694,
+ "step": 1129
+ },
+ {
+ "epoch": 0.5928646379853095,
+ "grad_norm": 0.16888824241669015,
+ "learning_rate": 2.1028507718774456e-05,
+ "loss": 0.4601455330848694,
+ "step": 1130
+ },
+ {
+ "epoch": 0.593389296956978,
+ "grad_norm": 0.16291744568806302,
+ "learning_rate": 2.0992997670873193e-05,
+ "loss": 0.46248236298561096,
+ "step": 1131
+ },
+ {
+ "epoch": 0.5939139559286464,
+ "grad_norm": 0.16185598691236988,
+ "learning_rate": 2.0957505301040603e-05,
+ "loss": 0.45085301995277405,
+ "step": 1132
+ },
+ {
+ "epoch": 0.5944386149003148,
+ "grad_norm": 0.1678000806117814,
+ "learning_rate": 2.0922030705701518e-05,
+ "loss": 0.44939014315605164,
+ "step": 1133
+ },
+ {
+ "epoch": 0.5949632738719832,
+ "grad_norm": 0.16104474282288994,
+ "learning_rate": 2.088657398123246e-05,
+ "loss": 0.4313976764678955,
+ "step": 1134
+ },
+ {
+ "epoch": 0.5954879328436516,
+ "grad_norm": 0.1631128968199669,
+ "learning_rate": 2.085113522396141e-05,
+ "loss": 0.4508278965950012,
+ "step": 1135
+ },
+ {
+ "epoch": 0.5960125918153201,
+ "grad_norm": 0.16307688195844533,
+ "learning_rate": 2.0815714530167557e-05,
+ "loss": 0.4467576742172241,
+ "step": 1136
+ },
+ {
+ "epoch": 0.5965372507869885,
+ "grad_norm": 0.16456832304532207,
+ "learning_rate": 2.0780311996080976e-05,
+ "loss": 0.4621703624725342,
+ "step": 1137
+ },
+ {
+ "epoch": 0.5970619097586569,
+ "grad_norm": 0.16346217748325367,
+ "learning_rate": 2.0744927717882435e-05,
+ "loss": 0.4589449465274811,
+ "step": 1138
+ },
+ {
+ "epoch": 0.5975865687303253,
+ "grad_norm": 0.16180691501076994,
+ "learning_rate": 2.0709561791703097e-05,
+ "loss": 0.43698400259017944,
+ "step": 1139
+ },
+ {
+ "epoch": 0.5981112277019937,
+ "grad_norm": 0.15695459976898835,
+ "learning_rate": 2.067421431362427e-05,
+ "loss": 0.452271044254303,
+ "step": 1140
+ },
+ {
+ "epoch": 0.5986358866736621,
+ "grad_norm": 0.15930893553721054,
+ "learning_rate": 2.063888537967713e-05,
+ "loss": 0.44868963956832886,
+ "step": 1141
+ },
+ {
+ "epoch": 0.5991605456453305,
+ "grad_norm": 0.15963802284017928,
+ "learning_rate": 2.0603575085842485e-05,
+ "loss": 0.4376864731311798,
+ "step": 1142
+ },
+ {
+ "epoch": 0.5996852046169989,
+ "grad_norm": 0.16610266900507178,
+ "learning_rate": 2.05682835280505e-05,
+ "loss": 0.44676727056503296,
+ "step": 1143
+ },
+ {
+ "epoch": 0.6002098635886673,
+ "grad_norm": 0.15978253236669612,
+ "learning_rate": 2.053301080218044e-05,
+ "loss": 0.434773713350296,
+ "step": 1144
+ },
+ {
+ "epoch": 0.6007345225603358,
+ "grad_norm": 0.16305141803526324,
+ "learning_rate": 2.0497757004060403e-05,
+ "loss": 0.4529780149459839,
+ "step": 1145
+ },
+ {
+ "epoch": 0.6012591815320042,
+ "grad_norm": 0.1649737864101215,
+ "learning_rate": 2.046252222946707e-05,
+ "loss": 0.4571434259414673,
+ "step": 1146
+ },
+ {
+ "epoch": 0.6017838405036726,
+ "grad_norm": 0.15738824968326598,
+ "learning_rate": 2.0427306574125427e-05,
+ "loss": 0.4446289539337158,
+ "step": 1147
+ },
+ {
+ "epoch": 0.602308499475341,
+ "grad_norm": 0.1731139535185187,
+ "learning_rate": 2.039211013370853e-05,
+ "loss": 0.4551090598106384,
+ "step": 1148
+ },
+ {
+ "epoch": 0.6028331584470095,
+ "grad_norm": 0.15857251900043037,
+ "learning_rate": 2.0356933003837237e-05,
+ "loss": 0.4644010663032532,
+ "step": 1149
+ },
+ {
+ "epoch": 0.6033578174186779,
+ "grad_norm": 0.1589582861134211,
+ "learning_rate": 2.032177528007992e-05,
+ "loss": 0.45741650462150574,
+ "step": 1150
+ },
+ {
+ "epoch": 0.6038824763903463,
+ "grad_norm": 0.15639092706575122,
+ "learning_rate": 2.0286637057952257e-05,
+ "loss": 0.4375547766685486,
+ "step": 1151
+ },
+ {
+ "epoch": 0.6044071353620147,
+ "grad_norm": 0.16335605581337187,
+ "learning_rate": 2.0251518432916933e-05,
+ "loss": 0.4636140465736389,
+ "step": 1152
+ },
+ {
+ "epoch": 0.6049317943336832,
+ "grad_norm": 0.15437437113359356,
+ "learning_rate": 2.0216419500383372e-05,
+ "loss": 0.44632142782211304,
+ "step": 1153
+ },
+ {
+ "epoch": 0.6054564533053515,
+ "grad_norm": 0.15522039198698664,
+ "learning_rate": 2.0181340355707545e-05,
+ "loss": 0.449049174785614,
+ "step": 1154
+ },
+ {
+ "epoch": 0.6059811122770199,
+ "grad_norm": 0.1599951917909323,
+ "learning_rate": 2.0146281094191604e-05,
+ "loss": 0.44395962357521057,
+ "step": 1155
+ },
+ {
+ "epoch": 0.6065057712486883,
+ "grad_norm": 0.16866621620527716,
+ "learning_rate": 2.011124181108373e-05,
+ "loss": 0.45693716406822205,
+ "step": 1156
+ },
+ {
+ "epoch": 0.6070304302203567,
+ "grad_norm": 0.15508935650135075,
+ "learning_rate": 2.0076222601577815e-05,
+ "loss": 0.4525151252746582,
+ "step": 1157
+ },
+ {
+ "epoch": 0.6075550891920252,
+ "grad_norm": 0.16508998118005971,
+ "learning_rate": 2.004122356081321e-05,
+ "loss": 0.466006875038147,
+ "step": 1158
+ },
+ {
+ "epoch": 0.6080797481636936,
+ "grad_norm": 0.15817418864022237,
+ "learning_rate": 2.0006244783874466e-05,
+ "loss": 0.4496110677719116,
+ "step": 1159
+ },
+ {
+ "epoch": 0.608604407135362,
+ "grad_norm": 0.16127270259749824,
+ "learning_rate": 1.997128636579109e-05,
+ "loss": 0.45918628573417664,
+ "step": 1160
+ },
+ {
+ "epoch": 0.6091290661070304,
+ "grad_norm": 0.1702599924159677,
+ "learning_rate": 1.9936348401537292e-05,
+ "loss": 0.47218555212020874,
+ "step": 1161
+ },
+ {
+ "epoch": 0.6096537250786989,
+ "grad_norm": 0.15690176248846127,
+ "learning_rate": 1.9901430986031695e-05,
+ "loss": 0.4516097903251648,
+ "step": 1162
+ },
+ {
+ "epoch": 0.6101783840503673,
+ "grad_norm": 0.1713614405614144,
+ "learning_rate": 1.9866534214137085e-05,
+ "loss": 0.45030856132507324,
+ "step": 1163
+ },
+ {
+ "epoch": 0.6107030430220357,
+ "grad_norm": 0.1670865563026229,
+ "learning_rate": 1.9831658180660195e-05,
+ "loss": 0.45603489875793457,
+ "step": 1164
+ },
+ {
+ "epoch": 0.6112277019937041,
+ "grad_norm": 0.16897539113401952,
+ "learning_rate": 1.9796802980351392e-05,
+ "loss": 0.46314537525177,
+ "step": 1165
+ },
+ {
+ "epoch": 0.6117523609653726,
+ "grad_norm": 0.15832868870697644,
+ "learning_rate": 1.976196870790446e-05,
+ "loss": 0.44784611463546753,
+ "step": 1166
+ },
+ {
+ "epoch": 0.612277019937041,
+ "grad_norm": 0.16595365551891114,
+ "learning_rate": 1.9727155457956315e-05,
+ "loss": 0.4409624934196472,
+ "step": 1167
+ },
+ {
+ "epoch": 0.6128016789087093,
+ "grad_norm": 0.1642630411212017,
+ "learning_rate": 1.969236332508677e-05,
+ "loss": 0.46120351552963257,
+ "step": 1168
+ },
+ {
+ "epoch": 0.6133263378803777,
+ "grad_norm": 0.16481796135868812,
+ "learning_rate": 1.9657592403818253e-05,
+ "loss": 0.45544904470443726,
+ "step": 1169
+ },
+ {
+ "epoch": 0.6138509968520461,
+ "grad_norm": 0.1629832779852364,
+ "learning_rate": 1.962284278861557e-05,
+ "loss": 0.4578462839126587,
+ "step": 1170
+ },
+ {
+ "epoch": 0.6143756558237146,
+ "grad_norm": 0.15916219613603688,
+ "learning_rate": 1.9588114573885664e-05,
+ "loss": 0.4546983540058136,
+ "step": 1171
+ },
+ {
+ "epoch": 0.614900314795383,
+ "grad_norm": 0.1660520165166075,
+ "learning_rate": 1.95534078539773e-05,
+ "loss": 0.4601942300796509,
+ "step": 1172
+ },
+ {
+ "epoch": 0.6154249737670514,
+ "grad_norm": 0.16924794578577768,
+ "learning_rate": 1.9518722723180877e-05,
+ "loss": 0.45997846126556396,
+ "step": 1173
+ },
+ {
+ "epoch": 0.6159496327387198,
+ "grad_norm": 0.16597646133045382,
+ "learning_rate": 1.9484059275728127e-05,
+ "loss": 0.4594605565071106,
+ "step": 1174
+ },
+ {
+ "epoch": 0.6164742917103883,
+ "grad_norm": 0.16964243792519748,
+ "learning_rate": 1.9449417605791876e-05,
+ "loss": 0.4638690650463104,
+ "step": 1175
+ },
+ {
+ "epoch": 0.6169989506820567,
+ "grad_norm": 0.1758591426791696,
+ "learning_rate": 1.9414797807485796e-05,
+ "loss": 0.4666447639465332,
+ "step": 1176
+ },
+ {
+ "epoch": 0.6175236096537251,
+ "grad_norm": 0.16217661227891753,
+ "learning_rate": 1.9380199974864126e-05,
+ "loss": 0.44252535700798035,
+ "step": 1177
+ },
+ {
+ "epoch": 0.6180482686253935,
+ "grad_norm": 0.1576446184499019,
+ "learning_rate": 1.934562420192142e-05,
+ "loss": 0.4499641954898834,
+ "step": 1178
+ },
+ {
+ "epoch": 0.618572927597062,
+ "grad_norm": 0.16231008092636912,
+ "learning_rate": 1.9311070582592327e-05,
+ "loss": 0.4561542570590973,
+ "step": 1179
+ },
+ {
+ "epoch": 0.6190975865687304,
+ "grad_norm": 0.16790382249577043,
+ "learning_rate": 1.9276539210751292e-05,
+ "loss": 0.4484325647354126,
+ "step": 1180
+ },
+ {
+ "epoch": 0.6196222455403987,
+ "grad_norm": 0.18794729932605855,
+ "learning_rate": 1.9242030180212323e-05,
+ "loss": 0.44169026613235474,
+ "step": 1181
+ },
+ {
+ "epoch": 0.6201469045120671,
+ "grad_norm": 0.16948051243889814,
+ "learning_rate": 1.920754358472872e-05,
+ "loss": 0.4592876434326172,
+ "step": 1182
+ },
+ {
+ "epoch": 0.6206715634837355,
+ "grad_norm": 0.16762558710029374,
+ "learning_rate": 1.917307951799286e-05,
+ "loss": 0.4402095079421997,
+ "step": 1183
+ },
+ {
+ "epoch": 0.621196222455404,
+ "grad_norm": 0.18170280116387003,
+ "learning_rate": 1.9138638073635885e-05,
+ "loss": 0.4633026123046875,
+ "step": 1184
+ },
+ {
+ "epoch": 0.6217208814270724,
+ "grad_norm": 0.1718051709136108,
+ "learning_rate": 1.9104219345227486e-05,
+ "loss": 0.4510066509246826,
+ "step": 1185
+ },
+ {
+ "epoch": 0.6222455403987408,
+ "grad_norm": 0.16829066326345576,
+ "learning_rate": 1.906982342627565e-05,
+ "loss": 0.4632919430732727,
+ "step": 1186
+ },
+ {
+ "epoch": 0.6227701993704092,
+ "grad_norm": 0.16669141808548274,
+ "learning_rate": 1.9035450410226395e-05,
+ "loss": 0.455544650554657,
+ "step": 1187
+ },
+ {
+ "epoch": 0.6232948583420777,
+ "grad_norm": 0.16679996559722232,
+ "learning_rate": 1.90011003904635e-05,
+ "loss": 0.459994375705719,
+ "step": 1188
+ },
+ {
+ "epoch": 0.6238195173137461,
+ "grad_norm": 0.17379754713278164,
+ "learning_rate": 1.8966773460308284e-05,
+ "loss": 0.4576749801635742,
+ "step": 1189
+ },
+ {
+ "epoch": 0.6243441762854145,
+ "grad_norm": 0.17088539733389246,
+ "learning_rate": 1.8932469713019335e-05,
+ "loss": 0.44853052496910095,
+ "step": 1190
+ },
+ {
+ "epoch": 0.6248688352570829,
+ "grad_norm": 0.15556883889201015,
+ "learning_rate": 1.8898189241792248e-05,
+ "loss": 0.44060447812080383,
+ "step": 1191
+ },
+ {
+ "epoch": 0.6253934942287513,
+ "grad_norm": 0.1640093917439215,
+ "learning_rate": 1.8863932139759388e-05,
+ "loss": 0.4531911611557007,
+ "step": 1192
+ },
+ {
+ "epoch": 0.6259181532004198,
+ "grad_norm": 0.16444284826591576,
+ "learning_rate": 1.882969849998964e-05,
+ "loss": 0.44176363945007324,
+ "step": 1193
+ },
+ {
+ "epoch": 0.6264428121720882,
+ "grad_norm": 0.1696462520995642,
+ "learning_rate": 1.8795488415488143e-05,
+ "loss": 0.44516652822494507,
+ "step": 1194
+ },
+ {
+ "epoch": 0.6269674711437565,
+ "grad_norm": 0.16810724705191796,
+ "learning_rate": 1.8761301979196032e-05,
+ "loss": 0.4542563855648041,
+ "step": 1195
+ },
+ {
+ "epoch": 0.6274921301154249,
+ "grad_norm": 0.17334813313855824,
+ "learning_rate": 1.8727139283990204e-05,
+ "loss": 0.47769248485565186,
+ "step": 1196
+ },
+ {
+ "epoch": 0.6280167890870934,
+ "grad_norm": 0.16308212035303235,
+ "learning_rate": 1.8693000422683043e-05,
+ "loss": 0.4407799243927002,
+ "step": 1197
+ },
+ {
+ "epoch": 0.6285414480587618,
+ "grad_norm": 0.16853319139928977,
+ "learning_rate": 1.8658885488022204e-05,
+ "loss": 0.4565497040748596,
+ "step": 1198
+ },
+ {
+ "epoch": 0.6290661070304302,
+ "grad_norm": 0.16694968781284703,
+ "learning_rate": 1.8624794572690328e-05,
+ "loss": 0.44522374868392944,
+ "step": 1199
+ },
+ {
+ "epoch": 0.6295907660020986,
+ "grad_norm": 0.16418579468339103,
+ "learning_rate": 1.8590727769304788e-05,
+ "loss": 0.4441048502922058,
+ "step": 1200
+ },
+ {
+ "epoch": 0.630115424973767,
+ "grad_norm": 0.1661930777124156,
+ "learning_rate": 1.8556685170417486e-05,
+ "loss": 0.457327663898468,
+ "step": 1201
+ },
+ {
+ "epoch": 0.6306400839454355,
+ "grad_norm": 0.16536630027039814,
+ "learning_rate": 1.852266686851452e-05,
+ "loss": 0.462494432926178,
+ "step": 1202
+ },
+ {
+ "epoch": 0.6311647429171039,
+ "grad_norm": 0.16611474088307454,
+ "learning_rate": 1.8488672956016013e-05,
+ "loss": 0.4557701647281647,
+ "step": 1203
+ },
+ {
+ "epoch": 0.6316894018887723,
+ "grad_norm": 0.16759343079749697,
+ "learning_rate": 1.8454703525275813e-05,
+ "loss": 0.45069995522499084,
+ "step": 1204
+ },
+ {
+ "epoch": 0.6322140608604407,
+ "grad_norm": 0.16186676294485336,
+ "learning_rate": 1.8420758668581255e-05,
+ "loss": 0.44954800605773926,
+ "step": 1205
+ },
+ {
+ "epoch": 0.6327387198321092,
+ "grad_norm": 0.15687159835368553,
+ "learning_rate": 1.8386838478152916e-05,
+ "loss": 0.4490429759025574,
+ "step": 1206
+ },
+ {
+ "epoch": 0.6332633788037776,
+ "grad_norm": 0.16242951976115577,
+ "learning_rate": 1.8352943046144372e-05,
+ "loss": 0.45819586515426636,
+ "step": 1207
+ },
+ {
+ "epoch": 0.633788037775446,
+ "grad_norm": 0.16516087266517232,
+ "learning_rate": 1.831907246464191e-05,
+ "loss": 0.44769883155822754,
+ "step": 1208
+ },
+ {
+ "epoch": 0.6343126967471143,
+ "grad_norm": 0.16709475871512644,
+ "learning_rate": 1.8285226825664332e-05,
+ "loss": 0.4699128568172455,
+ "step": 1209
+ },
+ {
+ "epoch": 0.6348373557187827,
+ "grad_norm": 0.16288623150595774,
+ "learning_rate": 1.8251406221162636e-05,
+ "loss": 0.45322558283805847,
+ "step": 1210
+ },
+ {
+ "epoch": 0.6353620146904512,
+ "grad_norm": 0.1675722848054237,
+ "learning_rate": 1.821761074301986e-05,
+ "loss": 0.4752347469329834,
+ "step": 1211
+ },
+ {
+ "epoch": 0.6358866736621196,
+ "grad_norm": 0.163285963514504,
+ "learning_rate": 1.818384048305074e-05,
+ "loss": 0.441958487033844,
+ "step": 1212
+ },
+ {
+ "epoch": 0.636411332633788,
+ "grad_norm": 0.17078102287557764,
+ "learning_rate": 1.8150095533001517e-05,
+ "loss": 0.4551333487033844,
+ "step": 1213
+ },
+ {
+ "epoch": 0.6369359916054564,
+ "grad_norm": 0.16188540746097138,
+ "learning_rate": 1.811637598454966e-05,
+ "loss": 0.4470779597759247,
+ "step": 1214
+ },
+ {
+ "epoch": 0.6374606505771249,
+ "grad_norm": 0.15235628218958286,
+ "learning_rate": 1.808268192930364e-05,
+ "loss": 0.43604153394699097,
+ "step": 1215
+ },
+ {
+ "epoch": 0.6379853095487933,
+ "grad_norm": 0.1604601949959104,
+ "learning_rate": 1.8049013458802665e-05,
+ "loss": 0.4445713758468628,
+ "step": 1216
+ },
+ {
+ "epoch": 0.6385099685204617,
+ "grad_norm": 0.15511797131215901,
+ "learning_rate": 1.8015370664516427e-05,
+ "loss": 0.4455301761627197,
+ "step": 1217
+ },
+ {
+ "epoch": 0.6390346274921301,
+ "grad_norm": 0.16815669546944095,
+ "learning_rate": 1.798175363784487e-05,
+ "loss": 0.46135014295578003,
+ "step": 1218
+ },
+ {
+ "epoch": 0.6395592864637986,
+ "grad_norm": 0.161558177688922,
+ "learning_rate": 1.7948162470117917e-05,
+ "loss": 0.4519304037094116,
+ "step": 1219
+ },
+ {
+ "epoch": 0.640083945435467,
+ "grad_norm": 0.16677842130048232,
+ "learning_rate": 1.7914597252595265e-05,
+ "loss": 0.4550550878047943,
+ "step": 1220
+ },
+ {
+ "epoch": 0.6406086044071354,
+ "grad_norm": 0.1597331480423138,
+ "learning_rate": 1.7881058076466095e-05,
+ "loss": 0.441120445728302,
+ "step": 1221
+ },
+ {
+ "epoch": 0.6411332633788038,
+ "grad_norm": 0.15951106545246824,
+ "learning_rate": 1.7847545032848844e-05,
+ "loss": 0.4471382200717926,
+ "step": 1222
+ },
+ {
+ "epoch": 0.6416579223504721,
+ "grad_norm": 0.16157787674667565,
+ "learning_rate": 1.7814058212790947e-05,
+ "loss": 0.4550648331642151,
+ "step": 1223
+ },
+ {
+ "epoch": 0.6421825813221406,
+ "grad_norm": 0.16989186763352415,
+ "learning_rate": 1.7780597707268583e-05,
+ "loss": 0.4515089988708496,
+ "step": 1224
+ },
+ {
+ "epoch": 0.642707240293809,
+ "grad_norm": 0.15952153223099508,
+ "learning_rate": 1.7747163607186468e-05,
+ "loss": 0.4658638834953308,
+ "step": 1225
+ },
+ {
+ "epoch": 0.6432318992654774,
+ "grad_norm": 0.15205926699647218,
+ "learning_rate": 1.7713756003377567e-05,
+ "loss": 0.4330868422985077,
+ "step": 1226
+ },
+ {
+ "epoch": 0.6437565582371458,
+ "grad_norm": 0.15784806480936794,
+ "learning_rate": 1.7680374986602865e-05,
+ "loss": 0.4506068527698517,
+ "step": 1227
+ },
+ {
+ "epoch": 0.6442812172088143,
+ "grad_norm": 0.16006709280686687,
+ "learning_rate": 1.7647020647551105e-05,
+ "loss": 0.446941077709198,
+ "step": 1228
+ },
+ {
+ "epoch": 0.6448058761804827,
+ "grad_norm": 0.16064631158222062,
+ "learning_rate": 1.761369307683856e-05,
+ "loss": 0.4418525695800781,
+ "step": 1229
+ },
+ {
+ "epoch": 0.6453305351521511,
+ "grad_norm": 0.16454136480444392,
+ "learning_rate": 1.7580392365008767e-05,
+ "loss": 0.46268391609191895,
+ "step": 1230
+ },
+ {
+ "epoch": 0.6458551941238195,
+ "grad_norm": 0.15401246535771462,
+ "learning_rate": 1.754711860253233e-05,
+ "loss": 0.4372989535331726,
+ "step": 1231
+ },
+ {
+ "epoch": 0.646379853095488,
+ "grad_norm": 0.1544567975028383,
+ "learning_rate": 1.7513871879806603e-05,
+ "loss": 0.44806575775146484,
+ "step": 1232
+ },
+ {
+ "epoch": 0.6469045120671564,
+ "grad_norm": 0.16411225112730965,
+ "learning_rate": 1.748065228715549e-05,
+ "loss": 0.43782058358192444,
+ "step": 1233
+ },
+ {
+ "epoch": 0.6474291710388248,
+ "grad_norm": 0.1552767832972093,
+ "learning_rate": 1.7447459914829183e-05,
+ "loss": 0.44531378149986267,
+ "step": 1234
+ },
+ {
+ "epoch": 0.6479538300104932,
+ "grad_norm": 0.16214695257123513,
+ "learning_rate": 1.741429485300392e-05,
+ "loss": 0.47399842739105225,
+ "step": 1235
+ },
+ {
+ "epoch": 0.6484784889821616,
+ "grad_norm": 0.16684861111484112,
+ "learning_rate": 1.7381157191781766e-05,
+ "loss": 0.4654282331466675,
+ "step": 1236
+ },
+ {
+ "epoch": 0.64900314795383,
+ "grad_norm": 0.15901864178630273,
+ "learning_rate": 1.7348047021190334e-05,
+ "loss": 0.45128414034843445,
+ "step": 1237
+ },
+ {
+ "epoch": 0.6495278069254984,
+ "grad_norm": 0.15360991961742707,
+ "learning_rate": 1.7314964431182523e-05,
+ "loss": 0.4303375482559204,
+ "step": 1238
+ },
+ {
+ "epoch": 0.6500524658971668,
+ "grad_norm": 0.1633151574385741,
+ "learning_rate": 1.728190951163633e-05,
+ "loss": 0.45568403601646423,
+ "step": 1239
+ },
+ {
+ "epoch": 0.6505771248688352,
+ "grad_norm": 0.16377853571058645,
+ "learning_rate": 1.724888235235457e-05,
+ "loss": 0.46503543853759766,
+ "step": 1240
+ },
+ {
+ "epoch": 0.6511017838405037,
+ "grad_norm": 0.1595023801275368,
+ "learning_rate": 1.7215883043064655e-05,
+ "loss": 0.46337294578552246,
+ "step": 1241
+ },
+ {
+ "epoch": 0.6516264428121721,
+ "grad_norm": 0.16324796182212292,
+ "learning_rate": 1.7182911673418313e-05,
+ "loss": 0.46197354793548584,
+ "step": 1242
+ },
+ {
+ "epoch": 0.6521511017838405,
+ "grad_norm": 0.16353734830879765,
+ "learning_rate": 1.714996833299138e-05,
+ "loss": 0.45741671323776245,
+ "step": 1243
+ },
+ {
+ "epoch": 0.6526757607555089,
+ "grad_norm": 0.15941587678263386,
+ "learning_rate": 1.7117053111283528e-05,
+ "loss": 0.4514503479003906,
+ "step": 1244
+ },
+ {
+ "epoch": 0.6532004197271774,
+ "grad_norm": 0.1642653230145008,
+ "learning_rate": 1.708416609771804e-05,
+ "loss": 0.4437655210494995,
+ "step": 1245
+ },
+ {
+ "epoch": 0.6537250786988458,
+ "grad_norm": 0.15814155519130152,
+ "learning_rate": 1.7051307381641587e-05,
+ "loss": 0.4314163327217102,
+ "step": 1246
+ },
+ {
+ "epoch": 0.6542497376705142,
+ "grad_norm": 0.16086078370768056,
+ "learning_rate": 1.701847705232394e-05,
+ "loss": 0.4526430666446686,
+ "step": 1247
+ },
+ {
+ "epoch": 0.6547743966421826,
+ "grad_norm": 0.16389360190715183,
+ "learning_rate": 1.698567519895775e-05,
+ "loss": 0.4278966188430786,
+ "step": 1248
+ },
+ {
+ "epoch": 0.655299055613851,
+ "grad_norm": 0.16303142287804262,
+ "learning_rate": 1.6952901910658316e-05,
+ "loss": 0.45778220891952515,
+ "step": 1249
+ },
+ {
+ "epoch": 0.6558237145855194,
+ "grad_norm": 0.1675209956904048,
+ "learning_rate": 1.6920157276463318e-05,
+ "loss": 0.4489403963088989,
+ "step": 1250
+ },
+ {
+ "epoch": 0.6563483735571878,
+ "grad_norm": 0.16228311551074323,
+ "learning_rate": 1.6887441385332604e-05,
+ "loss": 0.4534342885017395,
+ "step": 1251
+ },
+ {
+ "epoch": 0.6568730325288562,
+ "grad_norm": 0.16408957851046913,
+ "learning_rate": 1.6854754326147916e-05,
+ "loss": 0.4549208879470825,
+ "step": 1252
+ },
+ {
+ "epoch": 0.6573976915005246,
+ "grad_norm": 0.15967922111432845,
+ "learning_rate": 1.682209618771268e-05,
+ "loss": 0.4451141357421875,
+ "step": 1253
+ },
+ {
+ "epoch": 0.657922350472193,
+ "grad_norm": 0.16440698722204208,
+ "learning_rate": 1.678946705875176e-05,
+ "loss": 0.454200804233551,
+ "step": 1254
+ },
+ {
+ "epoch": 0.6584470094438615,
+ "grad_norm": 0.16251120184938497,
+ "learning_rate": 1.675686702791117e-05,
+ "loss": 0.4580027461051941,
+ "step": 1255
+ },
+ {
+ "epoch": 0.6589716684155299,
+ "grad_norm": 0.1643717386150457,
+ "learning_rate": 1.6724296183757924e-05,
+ "loss": 0.4422311782836914,
+ "step": 1256
+ },
+ {
+ "epoch": 0.6594963273871983,
+ "grad_norm": 0.1644586414378722,
+ "learning_rate": 1.66917546147797e-05,
+ "loss": 0.4593179225921631,
+ "step": 1257
+ },
+ {
+ "epoch": 0.6600209863588667,
+ "grad_norm": 0.16751074186967685,
+ "learning_rate": 1.6659242409384657e-05,
+ "loss": 0.457930326461792,
+ "step": 1258
+ },
+ {
+ "epoch": 0.6605456453305352,
+ "grad_norm": 0.15398418609155334,
+ "learning_rate": 1.6626759655901182e-05,
+ "loss": 0.44002899527549744,
+ "step": 1259
+ },
+ {
+ "epoch": 0.6610703043022036,
+ "grad_norm": 0.16328113175790226,
+ "learning_rate": 1.659430644257763e-05,
+ "loss": 0.4500438868999481,
+ "step": 1260
+ },
+ {
+ "epoch": 0.661594963273872,
+ "grad_norm": 0.1694244340615106,
+ "learning_rate": 1.656188285758214e-05,
+ "loss": 0.4818427264690399,
+ "step": 1261
+ },
+ {
+ "epoch": 0.6621196222455404,
+ "grad_norm": 0.1578320701754368,
+ "learning_rate": 1.652948898900232e-05,
+ "loss": 0.4492436647415161,
+ "step": 1262
+ },
+ {
+ "epoch": 0.6626442812172089,
+ "grad_norm": 0.16538865917828469,
+ "learning_rate": 1.6497124924845064e-05,
+ "loss": 0.45070338249206543,
+ "step": 1263
+ },
+ {
+ "epoch": 0.6631689401888772,
+ "grad_norm": 0.15670817848256852,
+ "learning_rate": 1.6464790753036296e-05,
+ "loss": 0.45439285039901733,
+ "step": 1264
+ },
+ {
+ "epoch": 0.6636935991605456,
+ "grad_norm": 0.16184786585965885,
+ "learning_rate": 1.64324865614207e-05,
+ "loss": 0.446522980928421,
+ "step": 1265
+ },
+ {
+ "epoch": 0.664218258132214,
+ "grad_norm": 0.15719726462251984,
+ "learning_rate": 1.6400212437761554e-05,
+ "loss": 0.4495280981063843,
+ "step": 1266
+ },
+ {
+ "epoch": 0.6647429171038824,
+ "grad_norm": 0.1680804414134405,
+ "learning_rate": 1.6367968469740416e-05,
+ "loss": 0.4652596712112427,
+ "step": 1267
+ },
+ {
+ "epoch": 0.6652675760755509,
+ "grad_norm": 0.15919541846683566,
+ "learning_rate": 1.6335754744956937e-05,
+ "loss": 0.43842798471450806,
+ "step": 1268
+ },
+ {
+ "epoch": 0.6657922350472193,
+ "grad_norm": 0.16322634187139418,
+ "learning_rate": 1.630357135092859e-05,
+ "loss": 0.45242664217948914,
+ "step": 1269
+ },
+ {
+ "epoch": 0.6663168940188877,
+ "grad_norm": 0.15644867734734783,
+ "learning_rate": 1.6271418375090445e-05,
+ "loss": 0.44801655411720276,
+ "step": 1270
+ },
+ {
+ "epoch": 0.6668415529905561,
+ "grad_norm": 0.1514717416124423,
+ "learning_rate": 1.623929590479496e-05,
+ "loss": 0.4506194293498993,
+ "step": 1271
+ },
+ {
+ "epoch": 0.6673662119622246,
+ "grad_norm": 0.1740345085495889,
+ "learning_rate": 1.6207204027311683e-05,
+ "loss": 0.4818177819252014,
+ "step": 1272
+ },
+ {
+ "epoch": 0.667890870933893,
+ "grad_norm": 0.15797701900765457,
+ "learning_rate": 1.6175142829827067e-05,
+ "loss": 0.4639382064342499,
+ "step": 1273
+ },
+ {
+ "epoch": 0.6684155299055614,
+ "grad_norm": 0.15663812316653875,
+ "learning_rate": 1.6143112399444198e-05,
+ "loss": 0.46217501163482666,
+ "step": 1274
+ },
+ {
+ "epoch": 0.6689401888772298,
+ "grad_norm": 0.15474018977043758,
+ "learning_rate": 1.6111112823182592e-05,
+ "loss": 0.43691688776016235,
+ "step": 1275
+ },
+ {
+ "epoch": 0.6694648478488983,
+ "grad_norm": 0.1615634942529431,
+ "learning_rate": 1.6079144187977923e-05,
+ "loss": 0.4412601590156555,
+ "step": 1276
+ },
+ {
+ "epoch": 0.6699895068205667,
+ "grad_norm": 0.15440022842754444,
+ "learning_rate": 1.604720658068184e-05,
+ "loss": 0.4523273706436157,
+ "step": 1277
+ },
+ {
+ "epoch": 0.670514165792235,
+ "grad_norm": 0.1634985665776242,
+ "learning_rate": 1.6015300088061664e-05,
+ "loss": 0.46264398097991943,
+ "step": 1278
+ },
+ {
+ "epoch": 0.6710388247639034,
+ "grad_norm": 0.15037886370689527,
+ "learning_rate": 1.598342479680018e-05,
+ "loss": 0.451465904712677,
+ "step": 1279
+ },
+ {
+ "epoch": 0.6715634837355718,
+ "grad_norm": 0.16895242437746613,
+ "learning_rate": 1.595158079349542e-05,
+ "loss": 0.4466145634651184,
+ "step": 1280
+ },
+ {
+ "epoch": 0.6720881427072403,
+ "grad_norm": 0.1629548326999642,
+ "learning_rate": 1.5919768164660425e-05,
+ "loss": 0.45415782928466797,
+ "step": 1281
+ },
+ {
+ "epoch": 0.6726128016789087,
+ "grad_norm": 0.1563872102268991,
+ "learning_rate": 1.588798699672298e-05,
+ "loss": 0.4386737048625946,
+ "step": 1282
+ },
+ {
+ "epoch": 0.6731374606505771,
+ "grad_norm": 0.16038172773344717,
+ "learning_rate": 1.5856237376025403e-05,
+ "loss": 0.446260929107666,
+ "step": 1283
+ },
+ {
+ "epoch": 0.6736621196222455,
+ "grad_norm": 0.16189780714407623,
+ "learning_rate": 1.582451938882431e-05,
+ "loss": 0.45458072423934937,
+ "step": 1284
+ },
+ {
+ "epoch": 0.674186778593914,
+ "grad_norm": 0.15204521785049285,
+ "learning_rate": 1.5792833121290367e-05,
+ "loss": 0.4425617754459381,
+ "step": 1285
+ },
+ {
+ "epoch": 0.6747114375655824,
+ "grad_norm": 0.16237376704792778,
+ "learning_rate": 1.5761178659508063e-05,
+ "loss": 0.44631800055503845,
+ "step": 1286
+ },
+ {
+ "epoch": 0.6752360965372508,
+ "grad_norm": 0.1582063618466069,
+ "learning_rate": 1.57295560894755e-05,
+ "loss": 0.44280534982681274,
+ "step": 1287
+ },
+ {
+ "epoch": 0.6757607555089192,
+ "grad_norm": 0.16587032349847394,
+ "learning_rate": 1.5697965497104116e-05,
+ "loss": 0.4495997130870819,
+ "step": 1288
+ },
+ {
+ "epoch": 0.6762854144805877,
+ "grad_norm": 0.1557269262755563,
+ "learning_rate": 1.566640696821848e-05,
+ "loss": 0.4493359625339508,
+ "step": 1289
+ },
+ {
+ "epoch": 0.6768100734522561,
+ "grad_norm": 0.16038490357012256,
+ "learning_rate": 1.5634880588556043e-05,
+ "loss": 0.4447544813156128,
+ "step": 1290
+ },
+ {
+ "epoch": 0.6773347324239245,
+ "grad_norm": 0.15355201908546406,
+ "learning_rate": 1.560338644376692e-05,
+ "loss": 0.4468432664871216,
+ "step": 1291
+ },
+ {
+ "epoch": 0.6778593913955928,
+ "grad_norm": 0.15077009515589776,
+ "learning_rate": 1.5571924619413668e-05,
+ "loss": 0.44959044456481934,
+ "step": 1292
+ },
+ {
+ "epoch": 0.6783840503672612,
+ "grad_norm": 0.1583038146613965,
+ "learning_rate": 1.5540495200971012e-05,
+ "loss": 0.42450425028800964,
+ "step": 1293
+ },
+ {
+ "epoch": 0.6789087093389297,
+ "grad_norm": 0.15950687190598087,
+ "learning_rate": 1.5509098273825635e-05,
+ "loss": 0.4527238607406616,
+ "step": 1294
+ },
+ {
+ "epoch": 0.6794333683105981,
+ "grad_norm": 0.1526262208847808,
+ "learning_rate": 1.5477733923275968e-05,
+ "loss": 0.4379724860191345,
+ "step": 1295
+ },
+ {
+ "epoch": 0.6799580272822665,
+ "grad_norm": 0.15541579650021778,
+ "learning_rate": 1.5446402234531934e-05,
+ "loss": 0.4462166130542755,
+ "step": 1296
+ },
+ {
+ "epoch": 0.6804826862539349,
+ "grad_norm": 0.15483930734313786,
+ "learning_rate": 1.5415103292714718e-05,
+ "loss": 0.4390607178211212,
+ "step": 1297
+ },
+ {
+ "epoch": 0.6810073452256034,
+ "grad_norm": 0.1578005731149154,
+ "learning_rate": 1.5383837182856532e-05,
+ "loss": 0.44462254643440247,
+ "step": 1298
+ },
+ {
+ "epoch": 0.6815320041972718,
+ "grad_norm": 0.15830183884958823,
+ "learning_rate": 1.5352603989900406e-05,
+ "loss": 0.4563537538051605,
+ "step": 1299
+ },
+ {
+ "epoch": 0.6820566631689402,
+ "grad_norm": 0.15605472163445522,
+ "learning_rate": 1.5321403798699923e-05,
+ "loss": 0.43852412700653076,
+ "step": 1300
+ },
+ {
+ "epoch": 0.6825813221406086,
+ "grad_norm": 0.15770470227141278,
+ "learning_rate": 1.5290236694019017e-05,
+ "loss": 0.4438081979751587,
+ "step": 1301
+ },
+ {
+ "epoch": 0.683105981112277,
+ "grad_norm": 0.16175339845012077,
+ "learning_rate": 1.5259102760531747e-05,
+ "loss": 0.46670734882354736,
+ "step": 1302
+ },
+ {
+ "epoch": 0.6836306400839455,
+ "grad_norm": 0.16359504716154477,
+ "learning_rate": 1.5228002082822036e-05,
+ "loss": 0.4590187668800354,
+ "step": 1303
+ },
+ {
+ "epoch": 0.6841552990556139,
+ "grad_norm": 0.15422879500254713,
+ "learning_rate": 1.5196934745383457e-05,
+ "loss": 0.45287445187568665,
+ "step": 1304
+ },
+ {
+ "epoch": 0.6846799580272823,
+ "grad_norm": 0.15929902143574254,
+ "learning_rate": 1.5165900832619015e-05,
+ "loss": 0.44286859035491943,
+ "step": 1305
+ },
+ {
+ "epoch": 0.6852046169989506,
+ "grad_norm": 0.14838868714016124,
+ "learning_rate": 1.5134900428840904e-05,
+ "loss": 0.45155346393585205,
+ "step": 1306
+ },
+ {
+ "epoch": 0.6857292759706191,
+ "grad_norm": 0.15494288640397672,
+ "learning_rate": 1.5103933618270283e-05,
+ "loss": 0.44010141491889954,
+ "step": 1307
+ },
+ {
+ "epoch": 0.6862539349422875,
+ "grad_norm": 0.1686262360598338,
+ "learning_rate": 1.5073000485037042e-05,
+ "loss": 0.446555495262146,
+ "step": 1308
+ },
+ {
+ "epoch": 0.6867785939139559,
+ "grad_norm": 0.16122110823982602,
+ "learning_rate": 1.5042101113179579e-05,
+ "loss": 0.44640374183654785,
+ "step": 1309
+ },
+ {
+ "epoch": 0.6873032528856243,
+ "grad_norm": 0.16849937797919867,
+ "learning_rate": 1.5011235586644572e-05,
+ "loss": 0.4465785026550293,
+ "step": 1310
+ },
+ {
+ "epoch": 0.6878279118572928,
+ "grad_norm": 0.16062971262800754,
+ "learning_rate": 1.498040398928674e-05,
+ "loss": 0.445321649312973,
+ "step": 1311
+ },
+ {
+ "epoch": 0.6883525708289612,
+ "grad_norm": 0.1608953185993034,
+ "learning_rate": 1.4949606404868638e-05,
+ "loss": 0.46518227458000183,
+ "step": 1312
+ },
+ {
+ "epoch": 0.6888772298006296,
+ "grad_norm": 0.15880255672830915,
+ "learning_rate": 1.491884291706041e-05,
+ "loss": 0.44582217931747437,
+ "step": 1313
+ },
+ {
+ "epoch": 0.689401888772298,
+ "grad_norm": 0.16715145200789822,
+ "learning_rate": 1.4888113609439563e-05,
+ "loss": 0.4435983896255493,
+ "step": 1314
+ },
+ {
+ "epoch": 0.6899265477439664,
+ "grad_norm": 0.15263740340694454,
+ "learning_rate": 1.485741856549075e-05,
+ "loss": 0.4405066967010498,
+ "step": 1315
+ },
+ {
+ "epoch": 0.6904512067156349,
+ "grad_norm": 0.16247034518813966,
+ "learning_rate": 1.4826757868605517e-05,
+ "loss": 0.4544215500354767,
+ "step": 1316
+ },
+ {
+ "epoch": 0.6909758656873033,
+ "grad_norm": 0.15649146705978162,
+ "learning_rate": 1.4796131602082142e-05,
+ "loss": 0.4454413056373596,
+ "step": 1317
+ },
+ {
+ "epoch": 0.6915005246589717,
+ "grad_norm": 0.16286718636058178,
+ "learning_rate": 1.4765539849125318e-05,
+ "loss": 0.4497183561325073,
+ "step": 1318
+ },
+ {
+ "epoch": 0.6920251836306401,
+ "grad_norm": 0.15812342085005582,
+ "learning_rate": 1.473498269284599e-05,
+ "loss": 0.445032000541687,
+ "step": 1319
+ },
+ {
+ "epoch": 0.6925498426023085,
+ "grad_norm": 0.1558101290511678,
+ "learning_rate": 1.4704460216261123e-05,
+ "loss": 0.45769619941711426,
+ "step": 1320
+ },
+ {
+ "epoch": 0.6930745015739769,
+ "grad_norm": 0.15642889059495207,
+ "learning_rate": 1.4673972502293428e-05,
+ "loss": 0.44788360595703125,
+ "step": 1321
+ },
+ {
+ "epoch": 0.6935991605456453,
+ "grad_norm": 0.15722513518113074,
+ "learning_rate": 1.4643519633771222e-05,
+ "loss": 0.449631929397583,
+ "step": 1322
+ },
+ {
+ "epoch": 0.6941238195173137,
+ "grad_norm": 0.15587633748081542,
+ "learning_rate": 1.4613101693428125e-05,
+ "loss": 0.4450187087059021,
+ "step": 1323
+ },
+ {
+ "epoch": 0.6946484784889821,
+ "grad_norm": 0.1577107074882479,
+ "learning_rate": 1.458271876390287e-05,
+ "loss": 0.45108890533447266,
+ "step": 1324
+ },
+ {
+ "epoch": 0.6951731374606506,
+ "grad_norm": 0.15649839486990696,
+ "learning_rate": 1.4552370927739073e-05,
+ "loss": 0.46430301666259766,
+ "step": 1325
+ },
+ {
+ "epoch": 0.695697796432319,
+ "grad_norm": 0.18573600420690636,
+ "learning_rate": 1.4522058267385008e-05,
+ "loss": 0.4701972007751465,
+ "step": 1326
+ },
+ {
+ "epoch": 0.6962224554039874,
+ "grad_norm": 0.16296497084037542,
+ "learning_rate": 1.4491780865193402e-05,
+ "loss": 0.4607710838317871,
+ "step": 1327
+ },
+ {
+ "epoch": 0.6967471143756558,
+ "grad_norm": 0.161124007217343,
+ "learning_rate": 1.446153880342117e-05,
+ "loss": 0.45095115900039673,
+ "step": 1328
+ },
+ {
+ "epoch": 0.6972717733473243,
+ "grad_norm": 0.1537884039166766,
+ "learning_rate": 1.4431332164229222e-05,
+ "loss": 0.424068808555603,
+ "step": 1329
+ },
+ {
+ "epoch": 0.6977964323189927,
+ "grad_norm": 0.16588834503456087,
+ "learning_rate": 1.4401161029682237e-05,
+ "loss": 0.4622839093208313,
+ "step": 1330
+ },
+ {
+ "epoch": 0.6983210912906611,
+ "grad_norm": 0.16100510604015544,
+ "learning_rate": 1.4371025481748435e-05,
+ "loss": 0.4660705626010895,
+ "step": 1331
+ },
+ {
+ "epoch": 0.6988457502623295,
+ "grad_norm": 0.15559888940590075,
+ "learning_rate": 1.4340925602299337e-05,
+ "loss": 0.4484976530075073,
+ "step": 1332
+ },
+ {
+ "epoch": 0.6993704092339978,
+ "grad_norm": 0.15503821966213488,
+ "learning_rate": 1.4310861473109616e-05,
+ "loss": 0.45636600255966187,
+ "step": 1333
+ },
+ {
+ "epoch": 0.6998950682056663,
+ "grad_norm": 0.15539074896109334,
+ "learning_rate": 1.4280833175856743e-05,
+ "loss": 0.45863068103790283,
+ "step": 1334
+ },
+ {
+ "epoch": 0.7004197271773347,
+ "grad_norm": 0.15664821009154872,
+ "learning_rate": 1.4250840792120896e-05,
+ "loss": 0.44043463468551636,
+ "step": 1335
+ },
+ {
+ "epoch": 0.7009443861490031,
+ "grad_norm": 0.15825478808520285,
+ "learning_rate": 1.4220884403384655e-05,
+ "loss": 0.46522051095962524,
+ "step": 1336
+ },
+ {
+ "epoch": 0.7014690451206715,
+ "grad_norm": 0.16004268568141544,
+ "learning_rate": 1.4190964091032843e-05,
+ "loss": 0.46114686131477356,
+ "step": 1337
+ },
+ {
+ "epoch": 0.70199370409234,
+ "grad_norm": 0.16413139949945635,
+ "learning_rate": 1.4161079936352234e-05,
+ "loss": 0.4457515478134155,
+ "step": 1338
+ },
+ {
+ "epoch": 0.7025183630640084,
+ "grad_norm": 0.1538890577972702,
+ "learning_rate": 1.4131232020531398e-05,
+ "loss": 0.4616093337535858,
+ "step": 1339
+ },
+ {
+ "epoch": 0.7030430220356768,
+ "grad_norm": 0.16240737854323284,
+ "learning_rate": 1.4101420424660434e-05,
+ "loss": 0.44631537795066833,
+ "step": 1340
+ },
+ {
+ "epoch": 0.7035676810073452,
+ "grad_norm": 0.15984708603406073,
+ "learning_rate": 1.4071645229730762e-05,
+ "loss": 0.4366717040538788,
+ "step": 1341
+ },
+ {
+ "epoch": 0.7040923399790137,
+ "grad_norm": 0.15149695695635904,
+ "learning_rate": 1.4041906516634946e-05,
+ "loss": 0.44129428267478943,
+ "step": 1342
+ },
+ {
+ "epoch": 0.7046169989506821,
+ "grad_norm": 0.15041167501768415,
+ "learning_rate": 1.4012204366166399e-05,
+ "loss": 0.43654775619506836,
+ "step": 1343
+ },
+ {
+ "epoch": 0.7051416579223505,
+ "grad_norm": 0.1596097627963756,
+ "learning_rate": 1.3982538859019217e-05,
+ "loss": 0.4652445614337921,
+ "step": 1344
+ },
+ {
+ "epoch": 0.7056663168940189,
+ "grad_norm": 0.15621704205027884,
+ "learning_rate": 1.3952910075787941e-05,
+ "loss": 0.44577041268348694,
+ "step": 1345
+ },
+ {
+ "epoch": 0.7061909758656874,
+ "grad_norm": 0.16073041098008783,
+ "learning_rate": 1.3923318096967341e-05,
+ "loss": 0.4391976594924927,
+ "step": 1346
+ },
+ {
+ "epoch": 0.7067156348373557,
+ "grad_norm": 0.14793853946372448,
+ "learning_rate": 1.3893763002952204e-05,
+ "loss": 0.4463799297809601,
+ "step": 1347
+ },
+ {
+ "epoch": 0.7072402938090241,
+ "grad_norm": 0.16176080358867354,
+ "learning_rate": 1.3864244874037097e-05,
+ "loss": 0.45183271169662476,
+ "step": 1348
+ },
+ {
+ "epoch": 0.7077649527806925,
+ "grad_norm": 0.16372771179026924,
+ "learning_rate": 1.3834763790416171e-05,
+ "loss": 0.44205576181411743,
+ "step": 1349
+ },
+ {
+ "epoch": 0.7082896117523609,
+ "grad_norm": 0.15135222516127125,
+ "learning_rate": 1.3805319832182937e-05,
+ "loss": 0.4390261173248291,
+ "step": 1350
+ },
+ {
+ "epoch": 0.7088142707240294,
+ "grad_norm": 0.15493183225792365,
+ "learning_rate": 1.377591307933002e-05,
+ "loss": 0.45837974548339844,
+ "step": 1351
+ },
+ {
+ "epoch": 0.7093389296956978,
+ "grad_norm": 0.16405906906516504,
+ "learning_rate": 1.3746543611749008e-05,
+ "loss": 0.45032668113708496,
+ "step": 1352
+ },
+ {
+ "epoch": 0.7098635886673662,
+ "grad_norm": 0.15725677976524505,
+ "learning_rate": 1.3717211509230155e-05,
+ "loss": 0.4486224949359894,
+ "step": 1353
+ },
+ {
+ "epoch": 0.7103882476390346,
+ "grad_norm": 0.15781469066983164,
+ "learning_rate": 1.3687916851462227e-05,
+ "loss": 0.4502638578414917,
+ "step": 1354
+ },
+ {
+ "epoch": 0.710912906610703,
+ "grad_norm": 0.16115007729054298,
+ "learning_rate": 1.3658659718032241e-05,
+ "loss": 0.4508914351463318,
+ "step": 1355
+ },
+ {
+ "epoch": 0.7114375655823715,
+ "grad_norm": 0.16755060475968744,
+ "learning_rate": 1.3629440188425286e-05,
+ "loss": 0.45653945207595825,
+ "step": 1356
+ },
+ {
+ "epoch": 0.7119622245540399,
+ "grad_norm": 0.1574994294756123,
+ "learning_rate": 1.3600258342024268e-05,
+ "loss": 0.4349849820137024,
+ "step": 1357
+ },
+ {
+ "epoch": 0.7124868835257083,
+ "grad_norm": 0.16104440922657715,
+ "learning_rate": 1.3571114258109758e-05,
+ "loss": 0.4501233696937561,
+ "step": 1358
+ },
+ {
+ "epoch": 0.7130115424973767,
+ "grad_norm": 0.15163184290939022,
+ "learning_rate": 1.3542008015859691e-05,
+ "loss": 0.44171035289764404,
+ "step": 1359
+ },
+ {
+ "epoch": 0.7135362014690452,
+ "grad_norm": 0.15955896674456582,
+ "learning_rate": 1.3512939694349211e-05,
+ "loss": 0.4558671712875366,
+ "step": 1360
+ },
+ {
+ "epoch": 0.7140608604407135,
+ "grad_norm": 0.1589376751536802,
+ "learning_rate": 1.3483909372550446e-05,
+ "loss": 0.4430835247039795,
+ "step": 1361
+ },
+ {
+ "epoch": 0.7145855194123819,
+ "grad_norm": 0.15979088397822155,
+ "learning_rate": 1.3454917129332279e-05,
+ "loss": 0.4426245391368866,
+ "step": 1362
+ },
+ {
+ "epoch": 0.7151101783840503,
+ "grad_norm": 0.15493179714072167,
+ "learning_rate": 1.3425963043460143e-05,
+ "loss": 0.4550474286079407,
+ "step": 1363
+ },
+ {
+ "epoch": 0.7156348373557188,
+ "grad_norm": 0.15396498968995032,
+ "learning_rate": 1.3397047193595808e-05,
+ "loss": 0.44220536947250366,
+ "step": 1364
+ },
+ {
+ "epoch": 0.7161594963273872,
+ "grad_norm": 0.1489346822667273,
+ "learning_rate": 1.336816965829717e-05,
+ "loss": 0.4437323212623596,
+ "step": 1365
+ },
+ {
+ "epoch": 0.7166841552990556,
+ "grad_norm": 0.16311964931734477,
+ "learning_rate": 1.333933051601802e-05,
+ "loss": 0.45308777689933777,
+ "step": 1366
+ },
+ {
+ "epoch": 0.717208814270724,
+ "grad_norm": 0.1534354619021179,
+ "learning_rate": 1.331052984510785e-05,
+ "loss": 0.4334886372089386,
+ "step": 1367
+ },
+ {
+ "epoch": 0.7177334732423925,
+ "grad_norm": 0.15541089758065404,
+ "learning_rate": 1.328176772381165e-05,
+ "loss": 0.45248323678970337,
+ "step": 1368
+ },
+ {
+ "epoch": 0.7182581322140609,
+ "grad_norm": 0.1617921305700596,
+ "learning_rate": 1.3253044230269651e-05,
+ "loss": 0.44092947244644165,
+ "step": 1369
+ },
+ {
+ "epoch": 0.7187827911857293,
+ "grad_norm": 0.15047353793760151,
+ "learning_rate": 1.3224359442517165e-05,
+ "loss": 0.4424414336681366,
+ "step": 1370
+ },
+ {
+ "epoch": 0.7193074501573977,
+ "grad_norm": 0.15958089102258743,
+ "learning_rate": 1.319571343848433e-05,
+ "loss": 0.4458580017089844,
+ "step": 1371
+ },
+ {
+ "epoch": 0.7198321091290661,
+ "grad_norm": 0.15349258306968733,
+ "learning_rate": 1.3167106295995924e-05,
+ "loss": 0.4285386800765991,
+ "step": 1372
+ },
+ {
+ "epoch": 0.7203567681007346,
+ "grad_norm": 0.16042529563759886,
+ "learning_rate": 1.3138538092771164e-05,
+ "loss": 0.4497690200805664,
+ "step": 1373
+ },
+ {
+ "epoch": 0.720881427072403,
+ "grad_norm": 0.1565113736588754,
+ "learning_rate": 1.3110008906423454e-05,
+ "loss": 0.43562617897987366,
+ "step": 1374
+ },
+ {
+ "epoch": 0.7214060860440713,
+ "grad_norm": 0.15177291086871506,
+ "learning_rate": 1.3081518814460215e-05,
+ "loss": 0.44456908106803894,
+ "step": 1375
+ },
+ {
+ "epoch": 0.7219307450157397,
+ "grad_norm": 0.15928490480439225,
+ "learning_rate": 1.3053067894282622e-05,
+ "loss": 0.46170780062675476,
+ "step": 1376
+ },
+ {
+ "epoch": 0.7224554039874082,
+ "grad_norm": 0.1554814469217892,
+ "learning_rate": 1.302465622318549e-05,
+ "loss": 0.4422169327735901,
+ "step": 1377
+ },
+ {
+ "epoch": 0.7229800629590766,
+ "grad_norm": 0.15375370214966386,
+ "learning_rate": 1.2996283878356947e-05,
+ "loss": 0.4564763903617859,
+ "step": 1378
+ },
+ {
+ "epoch": 0.723504721930745,
+ "grad_norm": 0.15942841969395802,
+ "learning_rate": 1.2967950936878311e-05,
+ "loss": 0.45121875405311584,
+ "step": 1379
+ },
+ {
+ "epoch": 0.7240293809024134,
+ "grad_norm": 0.15731308716774903,
+ "learning_rate": 1.2939657475723844e-05,
+ "loss": 0.4285027086734772,
+ "step": 1380
+ },
+ {
+ "epoch": 0.7245540398740818,
+ "grad_norm": 0.15997753613932147,
+ "learning_rate": 1.291140357176054e-05,
+ "loss": 0.4566386342048645,
+ "step": 1381
+ },
+ {
+ "epoch": 0.7250786988457503,
+ "grad_norm": 0.15333715267473727,
+ "learning_rate": 1.2883189301747927e-05,
+ "loss": 0.45093846321105957,
+ "step": 1382
+ },
+ {
+ "epoch": 0.7256033578174187,
+ "grad_norm": 0.1514510304709061,
+ "learning_rate": 1.2855014742337874e-05,
+ "loss": 0.44256967306137085,
+ "step": 1383
+ },
+ {
+ "epoch": 0.7261280167890871,
+ "grad_norm": 0.15624301298995746,
+ "learning_rate": 1.2826879970074346e-05,
+ "loss": 0.4361875653266907,
+ "step": 1384
+ },
+ {
+ "epoch": 0.7266526757607555,
+ "grad_norm": 0.15767850122006694,
+ "learning_rate": 1.279878506139322e-05,
+ "loss": 0.4477671980857849,
+ "step": 1385
+ },
+ {
+ "epoch": 0.727177334732424,
+ "grad_norm": 0.15537676969694064,
+ "learning_rate": 1.2770730092622072e-05,
+ "loss": 0.4339832663536072,
+ "step": 1386
+ },
+ {
+ "epoch": 0.7277019937040924,
+ "grad_norm": 0.15944113633281293,
+ "learning_rate": 1.2742715139979965e-05,
+ "loss": 0.45407935976982117,
+ "step": 1387
+ },
+ {
+ "epoch": 0.7282266526757608,
+ "grad_norm": 0.16763519518606218,
+ "learning_rate": 1.271474027957727e-05,
+ "loss": 0.46442392468452454,
+ "step": 1388
+ },
+ {
+ "epoch": 0.7287513116474291,
+ "grad_norm": 0.1564572774601902,
+ "learning_rate": 1.2686805587415423e-05,
+ "loss": 0.4463968276977539,
+ "step": 1389
+ },
+ {
+ "epoch": 0.7292759706190975,
+ "grad_norm": 0.16384626314986128,
+ "learning_rate": 1.2658911139386709e-05,
+ "loss": 0.45519885420799255,
+ "step": 1390
+ },
+ {
+ "epoch": 0.729800629590766,
+ "grad_norm": 0.1562023802820964,
+ "learning_rate": 1.2631057011274106e-05,
+ "loss": 0.4365222454071045,
+ "step": 1391
+ },
+ {
+ "epoch": 0.7303252885624344,
+ "grad_norm": 0.15597287232940926,
+ "learning_rate": 1.2603243278751042e-05,
+ "loss": 0.4328266680240631,
+ "step": 1392
+ },
+ {
+ "epoch": 0.7308499475341028,
+ "grad_norm": 0.14762833706132167,
+ "learning_rate": 1.2575470017381214e-05,
+ "loss": 0.43902426958084106,
+ "step": 1393
+ },
+ {
+ "epoch": 0.7313746065057712,
+ "grad_norm": 0.1688371515856315,
+ "learning_rate": 1.2547737302618351e-05,
+ "loss": 0.4649963676929474,
+ "step": 1394
+ },
+ {
+ "epoch": 0.7318992654774397,
+ "grad_norm": 0.15501514509586015,
+ "learning_rate": 1.2520045209806024e-05,
+ "loss": 0.4501558542251587,
+ "step": 1395
+ },
+ {
+ "epoch": 0.7324239244491081,
+ "grad_norm": 0.15516748917693537,
+ "learning_rate": 1.2492393814177457e-05,
+ "loss": 0.4405498206615448,
+ "step": 1396
+ },
+ {
+ "epoch": 0.7329485834207765,
+ "grad_norm": 0.17115400468199243,
+ "learning_rate": 1.2464783190855295e-05,
+ "loss": 0.4445379674434662,
+ "step": 1397
+ },
+ {
+ "epoch": 0.7334732423924449,
+ "grad_norm": 0.15966386341604186,
+ "learning_rate": 1.243721341485143e-05,
+ "loss": 0.4769889712333679,
+ "step": 1398
+ },
+ {
+ "epoch": 0.7339979013641134,
+ "grad_norm": 0.15614706359781472,
+ "learning_rate": 1.2409684561066766e-05,
+ "loss": 0.43643918633461,
+ "step": 1399
+ },
+ {
+ "epoch": 0.7345225603357818,
+ "grad_norm": 0.17467968960064226,
+ "learning_rate": 1.238219670429104e-05,
+ "loss": 0.45413005352020264,
+ "step": 1400
+ },
+ {
+ "epoch": 0.7350472193074502,
+ "grad_norm": 0.1544517430810819,
+ "learning_rate": 1.2354749919202598e-05,
+ "loss": 0.45218425989151,
+ "step": 1401
+ },
+ {
+ "epoch": 0.7355718782791185,
+ "grad_norm": 0.16551083673374484,
+ "learning_rate": 1.2327344280368217e-05,
+ "loss": 0.4459628462791443,
+ "step": 1402
+ },
+ {
+ "epoch": 0.7360965372507869,
+ "grad_norm": 0.17614490459613263,
+ "learning_rate": 1.2299979862242882e-05,
+ "loss": 0.4406280815601349,
+ "step": 1403
+ },
+ {
+ "epoch": 0.7366211962224554,
+ "grad_norm": 0.15554925394435032,
+ "learning_rate": 1.2272656739169588e-05,
+ "loss": 0.4380159378051758,
+ "step": 1404
+ },
+ {
+ "epoch": 0.7371458551941238,
+ "grad_norm": 0.16475466799870916,
+ "learning_rate": 1.2245374985379147e-05,
+ "loss": 0.4442812204360962,
+ "step": 1405
+ },
+ {
+ "epoch": 0.7376705141657922,
+ "grad_norm": 0.16877301168875541,
+ "learning_rate": 1.2218134674989978e-05,
+ "loss": 0.4589429199695587,
+ "step": 1406
+ },
+ {
+ "epoch": 0.7381951731374606,
+ "grad_norm": 0.1583278589457438,
+ "learning_rate": 1.2190935882007894e-05,
+ "loss": 0.44896745681762695,
+ "step": 1407
+ },
+ {
+ "epoch": 0.7387198321091291,
+ "grad_norm": 0.1582022541656192,
+ "learning_rate": 1.2163778680325947e-05,
+ "loss": 0.4367997646331787,
+ "step": 1408
+ },
+ {
+ "epoch": 0.7392444910807975,
+ "grad_norm": 0.1669573982300068,
+ "learning_rate": 1.2136663143724163e-05,
+ "loss": 0.4503386616706848,
+ "step": 1409
+ },
+ {
+ "epoch": 0.7397691500524659,
+ "grad_norm": 0.1637877523924539,
+ "learning_rate": 1.2109589345869387e-05,
+ "loss": 0.4660584330558777,
+ "step": 1410
+ },
+ {
+ "epoch": 0.7402938090241343,
+ "grad_norm": 0.1640136159943701,
+ "learning_rate": 1.2082557360315065e-05,
+ "loss": 0.45206528902053833,
+ "step": 1411
+ },
+ {
+ "epoch": 0.7408184679958028,
+ "grad_norm": 0.1596580168682554,
+ "learning_rate": 1.2055567260501054e-05,
+ "loss": 0.4405035376548767,
+ "step": 1412
+ },
+ {
+ "epoch": 0.7413431269674712,
+ "grad_norm": 0.15224511741945293,
+ "learning_rate": 1.2028619119753402e-05,
+ "loss": 0.42810291051864624,
+ "step": 1413
+ },
+ {
+ "epoch": 0.7418677859391396,
+ "grad_norm": 0.15631381931817098,
+ "learning_rate": 1.2001713011284193e-05,
+ "loss": 0.4421626925468445,
+ "step": 1414
+ },
+ {
+ "epoch": 0.742392444910808,
+ "grad_norm": 0.15721819981265503,
+ "learning_rate": 1.1974849008191288e-05,
+ "loss": 0.4540696144104004,
+ "step": 1415
+ },
+ {
+ "epoch": 0.7429171038824763,
+ "grad_norm": 0.15904478578744463,
+ "learning_rate": 1.1948027183458182e-05,
+ "loss": 0.4601830244064331,
+ "step": 1416
+ },
+ {
+ "epoch": 0.7434417628541448,
+ "grad_norm": 0.1530755605235123,
+ "learning_rate": 1.1921247609953747e-05,
+ "loss": 0.4552589952945709,
+ "step": 1417
+ },
+ {
+ "epoch": 0.7439664218258132,
+ "grad_norm": 0.15375751072258462,
+ "learning_rate": 1.1894510360432106e-05,
+ "loss": 0.42965373396873474,
+ "step": 1418
+ },
+ {
+ "epoch": 0.7444910807974816,
+ "grad_norm": 0.15678686023937333,
+ "learning_rate": 1.1867815507532377e-05,
+ "loss": 0.44684016704559326,
+ "step": 1419
+ },
+ {
+ "epoch": 0.74501573976915,
+ "grad_norm": 0.15486952983783606,
+ "learning_rate": 1.1841163123778502e-05,
+ "loss": 0.44415023922920227,
+ "step": 1420
+ },
+ {
+ "epoch": 0.7455403987408185,
+ "grad_norm": 0.15342611490545593,
+ "learning_rate": 1.1814553281579034e-05,
+ "loss": 0.4501304626464844,
+ "step": 1421
+ },
+ {
+ "epoch": 0.7460650577124869,
+ "grad_norm": 0.15374260012788166,
+ "learning_rate": 1.1787986053226954e-05,
+ "loss": 0.4521749019622803,
+ "step": 1422
+ },
+ {
+ "epoch": 0.7465897166841553,
+ "grad_norm": 0.1543553258370361,
+ "learning_rate": 1.176146151089949e-05,
+ "loss": 0.44753748178482056,
+ "step": 1423
+ },
+ {
+ "epoch": 0.7471143756558237,
+ "grad_norm": 0.15921670193848306,
+ "learning_rate": 1.1734979726657878e-05,
+ "loss": 0.4473114311695099,
+ "step": 1424
+ },
+ {
+ "epoch": 0.7476390346274921,
+ "grad_norm": 0.15023498686186612,
+ "learning_rate": 1.17085407724472e-05,
+ "loss": 0.4568778872489929,
+ "step": 1425
+ },
+ {
+ "epoch": 0.7481636935991606,
+ "grad_norm": 0.14601156920274583,
+ "learning_rate": 1.1682144720096168e-05,
+ "loss": 0.4389464855194092,
+ "step": 1426
+ },
+ {
+ "epoch": 0.748688352570829,
+ "grad_norm": 0.1520548095149398,
+ "learning_rate": 1.1655791641316954e-05,
+ "loss": 0.440733402967453,
+ "step": 1427
+ },
+ {
+ "epoch": 0.7492130115424974,
+ "grad_norm": 0.1591915163097993,
+ "learning_rate": 1.1629481607704963e-05,
+ "loss": 0.46245887875556946,
+ "step": 1428
+ },
+ {
+ "epoch": 0.7497376705141658,
+ "grad_norm": 0.14867295658447263,
+ "learning_rate": 1.1603214690738683e-05,
+ "loss": 0.4332764744758606,
+ "step": 1429
+ },
+ {
+ "epoch": 0.7502623294858342,
+ "grad_norm": 0.1498957762903134,
+ "learning_rate": 1.1576990961779447e-05,
+ "loss": 0.45215466618537903,
+ "step": 1430
+ },
+ {
+ "epoch": 0.7507869884575026,
+ "grad_norm": 0.15430555803491824,
+ "learning_rate": 1.1550810492071243e-05,
+ "loss": 0.4454396963119507,
+ "step": 1431
+ },
+ {
+ "epoch": 0.751311647429171,
+ "grad_norm": 0.1517461021936533,
+ "learning_rate": 1.1524673352740543e-05,
+ "loss": 0.43869835138320923,
+ "step": 1432
+ },
+ {
+ "epoch": 0.7518363064008394,
+ "grad_norm": 0.14541657994692936,
+ "learning_rate": 1.1498579614796121e-05,
+ "loss": 0.4373054504394531,
+ "step": 1433
+ },
+ {
+ "epoch": 0.7523609653725079,
+ "grad_norm": 0.15539513763786977,
+ "learning_rate": 1.1472529349128812e-05,
+ "loss": 0.45451128482818604,
+ "step": 1434
+ },
+ {
+ "epoch": 0.7528856243441763,
+ "grad_norm": 0.1482043113690946,
+ "learning_rate": 1.144652262651136e-05,
+ "loss": 0.43644773960113525,
+ "step": 1435
+ },
+ {
+ "epoch": 0.7534102833158447,
+ "grad_norm": 0.16293628094190815,
+ "learning_rate": 1.1420559517598205e-05,
+ "loss": 0.4687269330024719,
+ "step": 1436
+ },
+ {
+ "epoch": 0.7539349422875131,
+ "grad_norm": 0.1553393974390287,
+ "learning_rate": 1.1394640092925302e-05,
+ "loss": 0.4501398205757141,
+ "step": 1437
+ },
+ {
+ "epoch": 0.7544596012591815,
+ "grad_norm": 0.16214837694535802,
+ "learning_rate": 1.1368764422909919e-05,
+ "loss": 0.4542577862739563,
+ "step": 1438
+ },
+ {
+ "epoch": 0.75498426023085,
+ "grad_norm": 0.15838878372428297,
+ "learning_rate": 1.1342932577850473e-05,
+ "loss": 0.46228161454200745,
+ "step": 1439
+ },
+ {
+ "epoch": 0.7555089192025184,
+ "grad_norm": 0.15090419497122687,
+ "learning_rate": 1.13171446279263e-05,
+ "loss": 0.4495932161808014,
+ "step": 1440
+ },
+ {
+ "epoch": 0.7560335781741868,
+ "grad_norm": 0.1543376786458394,
+ "learning_rate": 1.1291400643197484e-05,
+ "loss": 0.451630175113678,
+ "step": 1441
+ },
+ {
+ "epoch": 0.7565582371458552,
+ "grad_norm": 0.1539470751610696,
+ "learning_rate": 1.1265700693604668e-05,
+ "loss": 0.447547972202301,
+ "step": 1442
+ },
+ {
+ "epoch": 0.7570828961175237,
+ "grad_norm": 0.15761047703305445,
+ "learning_rate": 1.1240044848968856e-05,
+ "loss": 0.4516204595565796,
+ "step": 1443
+ },
+ {
+ "epoch": 0.757607555089192,
+ "grad_norm": 0.15675279683704735,
+ "learning_rate": 1.1214433178991258e-05,
+ "loss": 0.4597971737384796,
+ "step": 1444
+ },
+ {
+ "epoch": 0.7581322140608604,
+ "grad_norm": 0.15592110994904113,
+ "learning_rate": 1.1188865753253022e-05,
+ "loss": 0.443898469209671,
+ "step": 1445
+ },
+ {
+ "epoch": 0.7586568730325288,
+ "grad_norm": 0.15597501691030619,
+ "learning_rate": 1.116334264121513e-05,
+ "loss": 0.4581306278705597,
+ "step": 1446
+ },
+ {
+ "epoch": 0.7591815320041972,
+ "grad_norm": 0.15552304728589958,
+ "learning_rate": 1.1137863912218154e-05,
+ "loss": 0.4475254416465759,
+ "step": 1447
+ },
+ {
+ "epoch": 0.7597061909758657,
+ "grad_norm": 0.15759990123898035,
+ "learning_rate": 1.1112429635482118e-05,
+ "loss": 0.43585264682769775,
+ "step": 1448
+ },
+ {
+ "epoch": 0.7602308499475341,
+ "grad_norm": 0.15964657264676635,
+ "learning_rate": 1.1087039880106247e-05,
+ "loss": 0.46786290407180786,
+ "step": 1449
+ },
+ {
+ "epoch": 0.7607555089192025,
+ "grad_norm": 0.15155804051218852,
+ "learning_rate": 1.106169471506882e-05,
+ "loss": 0.42292726039886475,
+ "step": 1450
+ },
+ {
+ "epoch": 0.7612801678908709,
+ "grad_norm": 0.15519270966930884,
+ "learning_rate": 1.103639420922698e-05,
+ "loss": 0.4248972237110138,
+ "step": 1451
+ },
+ {
+ "epoch": 0.7618048268625394,
+ "grad_norm": 0.1526330511128821,
+ "learning_rate": 1.1011138431316537e-05,
+ "loss": 0.4317784309387207,
+ "step": 1452
+ },
+ {
+ "epoch": 0.7623294858342078,
+ "grad_norm": 0.15797682979272226,
+ "learning_rate": 1.0985927449951779e-05,
+ "loss": 0.44506698846817017,
+ "step": 1453
+ },
+ {
+ "epoch": 0.7628541448058762,
+ "grad_norm": 0.15479979838505462,
+ "learning_rate": 1.096076133362531e-05,
+ "loss": 0.4471631348133087,
+ "step": 1454
+ },
+ {
+ "epoch": 0.7633788037775446,
+ "grad_norm": 0.15712267274447342,
+ "learning_rate": 1.093564015070783e-05,
+ "loss": 0.4321915805339813,
+ "step": 1455
+ },
+ {
+ "epoch": 0.7639034627492131,
+ "grad_norm": 0.15857431494180016,
+ "learning_rate": 1.0910563969447969e-05,
+ "loss": 0.44455766677856445,
+ "step": 1456
+ },
+ {
+ "epoch": 0.7644281217208815,
+ "grad_norm": 0.15377057080997789,
+ "learning_rate": 1.0885532857972097e-05,
+ "loss": 0.4567924439907074,
+ "step": 1457
+ },
+ {
+ "epoch": 0.7649527806925498,
+ "grad_norm": 0.15544844005009104,
+ "learning_rate": 1.0860546884284143e-05,
+ "loss": 0.4491572976112366,
+ "step": 1458
+ },
+ {
+ "epoch": 0.7654774396642182,
+ "grad_norm": 0.16210735566479478,
+ "learning_rate": 1.08356061162654e-05,
+ "loss": 0.43823331594467163,
+ "step": 1459
+ },
+ {
+ "epoch": 0.7660020986358866,
+ "grad_norm": 0.15213559486379452,
+ "learning_rate": 1.0810710621674355e-05,
+ "loss": 0.45375770330429077,
+ "step": 1460
+ },
+ {
+ "epoch": 0.7665267576075551,
+ "grad_norm": 0.1727832306402992,
+ "learning_rate": 1.0785860468146494e-05,
+ "loss": 0.4639572501182556,
+ "step": 1461
+ },
+ {
+ "epoch": 0.7670514165792235,
+ "grad_norm": 0.1637555902013663,
+ "learning_rate": 1.076105572319412e-05,
+ "loss": 0.4355923533439636,
+ "step": 1462
+ },
+ {
+ "epoch": 0.7675760755508919,
+ "grad_norm": 0.16646936030470866,
+ "learning_rate": 1.0736296454206166e-05,
+ "loss": 0.4650861322879791,
+ "step": 1463
+ },
+ {
+ "epoch": 0.7681007345225603,
+ "grad_norm": 0.14987620615431385,
+ "learning_rate": 1.0711582728448033e-05,
+ "loss": 0.45445430278778076,
+ "step": 1464
+ },
+ {
+ "epoch": 0.7686253934942288,
+ "grad_norm": 0.16428618403547932,
+ "learning_rate": 1.0686914613061377e-05,
+ "loss": 0.46050935983657837,
+ "step": 1465
+ },
+ {
+ "epoch": 0.7691500524658972,
+ "grad_norm": 0.15052999053128632,
+ "learning_rate": 1.0662292175063949e-05,
+ "loss": 0.43385741114616394,
+ "step": 1466
+ },
+ {
+ "epoch": 0.7696747114375656,
+ "grad_norm": 0.1654949402740152,
+ "learning_rate": 1.0637715481349397e-05,
+ "loss": 0.4434410333633423,
+ "step": 1467
+ },
+ {
+ "epoch": 0.770199370409234,
+ "grad_norm": 0.15923836508490036,
+ "learning_rate": 1.0613184598687088e-05,
+ "loss": 0.4410185217857361,
+ "step": 1468
+ },
+ {
+ "epoch": 0.7707240293809025,
+ "grad_norm": 0.15813007111170696,
+ "learning_rate": 1.0588699593721957e-05,
+ "loss": 0.43588417768478394,
+ "step": 1469
+ },
+ {
+ "epoch": 0.7712486883525709,
+ "grad_norm": 0.16978817898431764,
+ "learning_rate": 1.0564260532974274e-05,
+ "loss": 0.4514843225479126,
+ "step": 1470
+ },
+ {
+ "epoch": 0.7717733473242392,
+ "grad_norm": 0.1613844852697828,
+ "learning_rate": 1.0539867482839499e-05,
+ "loss": 0.4554520845413208,
+ "step": 1471
+ },
+ {
+ "epoch": 0.7722980062959076,
+ "grad_norm": 0.1587788181984633,
+ "learning_rate": 1.0515520509588075e-05,
+ "loss": 0.43200671672821045,
+ "step": 1472
+ },
+ {
+ "epoch": 0.772822665267576,
+ "grad_norm": 0.1539930824827353,
+ "learning_rate": 1.0491219679365283e-05,
+ "loss": 0.4343116283416748,
+ "step": 1473
+ },
+ {
+ "epoch": 0.7733473242392445,
+ "grad_norm": 0.1561612537773709,
+ "learning_rate": 1.0466965058191049e-05,
+ "loss": 0.4477815330028534,
+ "step": 1474
+ },
+ {
+ "epoch": 0.7738719832109129,
+ "grad_norm": 0.169990777691657,
+ "learning_rate": 1.0442756711959745e-05,
+ "loss": 0.4476611614227295,
+ "step": 1475
+ },
+ {
+ "epoch": 0.7743966421825813,
+ "grad_norm": 0.1624313049193793,
+ "learning_rate": 1.041859470644002e-05,
+ "loss": 0.47136399149894714,
+ "step": 1476
+ },
+ {
+ "epoch": 0.7749213011542497,
+ "grad_norm": 0.15993315458994328,
+ "learning_rate": 1.0394479107274645e-05,
+ "loss": 0.4672670364379883,
+ "step": 1477
+ },
+ {
+ "epoch": 0.7754459601259182,
+ "grad_norm": 0.16755349272347217,
+ "learning_rate": 1.0370409979980294e-05,
+ "loss": 0.45135706663131714,
+ "step": 1478
+ },
+ {
+ "epoch": 0.7759706190975866,
+ "grad_norm": 0.15665059815830362,
+ "learning_rate": 1.0346387389947407e-05,
+ "loss": 0.45178359746932983,
+ "step": 1479
+ },
+ {
+ "epoch": 0.776495278069255,
+ "grad_norm": 0.1554612731452571,
+ "learning_rate": 1.0322411402439983e-05,
+ "loss": 0.4412590265274048,
+ "step": 1480
+ },
+ {
+ "epoch": 0.7770199370409234,
+ "grad_norm": 0.1532496869312065,
+ "learning_rate": 1.0298482082595408e-05,
+ "loss": 0.44104987382888794,
+ "step": 1481
+ },
+ {
+ "epoch": 0.7775445960125918,
+ "grad_norm": 0.15065868316960426,
+ "learning_rate": 1.0274599495424286e-05,
+ "loss": 0.4336603283882141,
+ "step": 1482
+ },
+ {
+ "epoch": 0.7780692549842603,
+ "grad_norm": 0.1548373981760097,
+ "learning_rate": 1.0250763705810267e-05,
+ "loss": 0.4658375680446625,
+ "step": 1483
+ },
+ {
+ "epoch": 0.7785939139559287,
+ "grad_norm": 0.15470181175866582,
+ "learning_rate": 1.0226974778509846e-05,
+ "loss": 0.45232075452804565,
+ "step": 1484
+ },
+ {
+ "epoch": 0.779118572927597,
+ "grad_norm": 0.15507371187727106,
+ "learning_rate": 1.0203232778152228e-05,
+ "loss": 0.43640458583831787,
+ "step": 1485
+ },
+ {
+ "epoch": 0.7796432318992654,
+ "grad_norm": 0.15269440736544085,
+ "learning_rate": 1.01795377692391e-05,
+ "loss": 0.4400939345359802,
+ "step": 1486
+ },
+ {
+ "epoch": 0.7801678908709339,
+ "grad_norm": 0.15610899550395085,
+ "learning_rate": 1.0155889816144495e-05,
+ "loss": 0.4546602666378021,
+ "step": 1487
+ },
+ {
+ "epoch": 0.7806925498426023,
+ "grad_norm": 0.1559328242601668,
+ "learning_rate": 1.0132288983114611e-05,
+ "loss": 0.45510709285736084,
+ "step": 1488
+ },
+ {
+ "epoch": 0.7812172088142707,
+ "grad_norm": 0.1537255194600787,
+ "learning_rate": 1.0108735334267639e-05,
+ "loss": 0.4475342631340027,
+ "step": 1489
+ },
+ {
+ "epoch": 0.7817418677859391,
+ "grad_norm": 0.14616575061664402,
+ "learning_rate": 1.0085228933593558e-05,
+ "loss": 0.4378190040588379,
+ "step": 1490
+ },
+ {
+ "epoch": 0.7822665267576075,
+ "grad_norm": 0.1624360967746598,
+ "learning_rate": 1.0061769844953999e-05,
+ "loss": 0.4439539909362793,
+ "step": 1491
+ },
+ {
+ "epoch": 0.782791185729276,
+ "grad_norm": 0.15407336125298413,
+ "learning_rate": 1.003835813208206e-05,
+ "loss": 0.44074907898902893,
+ "step": 1492
+ },
+ {
+ "epoch": 0.7833158447009444,
+ "grad_norm": 0.15786980710136125,
+ "learning_rate": 1.001499385858211e-05,
+ "loss": 0.4491667151451111,
+ "step": 1493
+ },
+ {
+ "epoch": 0.7838405036726128,
+ "grad_norm": 0.15338836738460088,
+ "learning_rate": 9.991677087929665e-06,
+ "loss": 0.446003794670105,
+ "step": 1494
+ },
+ {
+ "epoch": 0.7843651626442812,
+ "grad_norm": 0.156710894892,
+ "learning_rate": 9.968407883471165e-06,
+ "loss": 0.43604689836502075,
+ "step": 1495
+ },
+ {
+ "epoch": 0.7848898216159497,
+ "grad_norm": 0.16100539063191666,
+ "learning_rate": 9.945186308423832e-06,
+ "loss": 0.4431147873401642,
+ "step": 1496
+ },
+ {
+ "epoch": 0.7854144805876181,
+ "grad_norm": 0.15776018851122997,
+ "learning_rate": 9.92201242587548e-06,
+ "loss": 0.4532504975795746,
+ "step": 1497
+ },
+ {
+ "epoch": 0.7859391395592865,
+ "grad_norm": 0.1540333423540934,
+ "learning_rate": 9.898886298784368e-06,
+ "loss": 0.43605154752731323,
+ "step": 1498
+ },
+ {
+ "epoch": 0.7864637985309548,
+ "grad_norm": 0.16009602938972295,
+ "learning_rate": 9.875807989979005e-06,
+ "loss": 0.4605238437652588,
+ "step": 1499
+ },
+ {
+ "epoch": 0.7869884575026233,
+ "grad_norm": 0.15639368804171708,
+ "learning_rate": 9.852777562157983e-06,
+ "loss": 0.4509708881378174,
+ "step": 1500
+ },
+ {
+ "epoch": 0.7875131164742917,
+ "grad_norm": 0.15074011188741518,
+ "learning_rate": 9.829795077889828e-06,
+ "loss": 0.4527304172515869,
+ "step": 1501
+ },
+ {
+ "epoch": 0.7880377754459601,
+ "grad_norm": 0.15155083996490956,
+ "learning_rate": 9.806860599612799e-06,
+ "loss": 0.4306413531303406,
+ "step": 1502
+ },
+ {
+ "epoch": 0.7885624344176285,
+ "grad_norm": 0.1478956630795445,
+ "learning_rate": 9.783974189634734e-06,
+ "loss": 0.4434714913368225,
+ "step": 1503
+ },
+ {
+ "epoch": 0.789087093389297,
+ "grad_norm": 0.15327668434806432,
+ "learning_rate": 9.761135910132902e-06,
+ "loss": 0.44431227445602417,
+ "step": 1504
+ },
+ {
+ "epoch": 0.7896117523609654,
+ "grad_norm": 0.2243177585497212,
+ "learning_rate": 9.738345823153783e-06,
+ "loss": 0.4541007876396179,
+ "step": 1505
+ },
+ {
+ "epoch": 0.7901364113326338,
+ "grad_norm": 0.1600239865812282,
+ "learning_rate": 9.715603990612945e-06,
+ "loss": 0.4456910490989685,
+ "step": 1506
+ },
+ {
+ "epoch": 0.7906610703043022,
+ "grad_norm": 0.15304634075332066,
+ "learning_rate": 9.692910474294858e-06,
+ "loss": 0.43965545296669006,
+ "step": 1507
+ },
+ {
+ "epoch": 0.7911857292759706,
+ "grad_norm": 0.15657752062053032,
+ "learning_rate": 9.670265335852725e-06,
+ "loss": 0.43358081579208374,
+ "step": 1508
+ },
+ {
+ "epoch": 0.7917103882476391,
+ "grad_norm": 0.16110057783374115,
+ "learning_rate": 9.647668636808304e-06,
+ "loss": 0.45711368322372437,
+ "step": 1509
+ },
+ {
+ "epoch": 0.7922350472193075,
+ "grad_norm": 0.1551133690299623,
+ "learning_rate": 9.625120438551791e-06,
+ "loss": 0.4613519608974457,
+ "step": 1510
+ },
+ {
+ "epoch": 0.7927597061909759,
+ "grad_norm": 0.15690897380725946,
+ "learning_rate": 9.602620802341581e-06,
+ "loss": 0.4472261071205139,
+ "step": 1511
+ },
+ {
+ "epoch": 0.7932843651626443,
+ "grad_norm": 0.147796599322911,
+ "learning_rate": 9.580169789304152e-06,
+ "loss": 0.4447203576564789,
+ "step": 1512
+ },
+ {
+ "epoch": 0.7938090241343126,
+ "grad_norm": 0.152019559199633,
+ "learning_rate": 9.557767460433878e-06,
+ "loss": 0.445905864238739,
+ "step": 1513
+ },
+ {
+ "epoch": 0.7943336831059811,
+ "grad_norm": 0.1574763924872037,
+ "learning_rate": 9.535413876592872e-06,
+ "loss": 0.4502032995223999,
+ "step": 1514
+ },
+ {
+ "epoch": 0.7948583420776495,
+ "grad_norm": 0.1524218315499056,
+ "learning_rate": 9.513109098510816e-06,
+ "loss": 0.43753111362457275,
+ "step": 1515
+ },
+ {
+ "epoch": 0.7953830010493179,
+ "grad_norm": 0.15589147437937434,
+ "learning_rate": 9.490853186784801e-06,
+ "loss": 0.44524312019348145,
+ "step": 1516
+ },
+ {
+ "epoch": 0.7959076600209863,
+ "grad_norm": 0.15604009714081407,
+ "learning_rate": 9.468646201879156e-06,
+ "loss": 0.44141292572021484,
+ "step": 1517
+ },
+ {
+ "epoch": 0.7964323189926548,
+ "grad_norm": 0.1562432554004435,
+ "learning_rate": 9.44648820412529e-06,
+ "loss": 0.461681067943573,
+ "step": 1518
+ },
+ {
+ "epoch": 0.7969569779643232,
+ "grad_norm": 0.15233588531495665,
+ "learning_rate": 9.424379253721513e-06,
+ "loss": 0.4529497027397156,
+ "step": 1519
+ },
+ {
+ "epoch": 0.7974816369359916,
+ "grad_norm": 0.16856991685290923,
+ "learning_rate": 9.40231941073291e-06,
+ "loss": 0.4483596086502075,
+ "step": 1520
+ },
+ {
+ "epoch": 0.79800629590766,
+ "grad_norm": 0.1546876712981077,
+ "learning_rate": 9.380308735091125e-06,
+ "loss": 0.43846625089645386,
+ "step": 1521
+ },
+ {
+ "epoch": 0.7985309548793285,
+ "grad_norm": 0.15198637556890449,
+ "learning_rate": 9.35834728659424e-06,
+ "loss": 0.4625553488731384,
+ "step": 1522
+ },
+ {
+ "epoch": 0.7990556138509969,
+ "grad_norm": 0.1519777007198701,
+ "learning_rate": 9.336435124906592e-06,
+ "loss": 0.44183820486068726,
+ "step": 1523
+ },
+ {
+ "epoch": 0.7995802728226653,
+ "grad_norm": 0.16125823074979284,
+ "learning_rate": 9.314572309558616e-06,
+ "loss": 0.46057307720184326,
+ "step": 1524
+ },
+ {
+ "epoch": 0.8001049317943337,
+ "grad_norm": 0.16283323624848034,
+ "learning_rate": 9.2927588999467e-06,
+ "loss": 0.4530155062675476,
+ "step": 1525
+ },
+ {
+ "epoch": 0.8006295907660022,
+ "grad_norm": 0.14998654489027738,
+ "learning_rate": 9.270994955332984e-06,
+ "loss": 0.43155500292778015,
+ "step": 1526
+ },
+ {
+ "epoch": 0.8011542497376705,
+ "grad_norm": 0.15683095030848093,
+ "learning_rate": 9.249280534845241e-06,
+ "loss": 0.45097050070762634,
+ "step": 1527
+ },
+ {
+ "epoch": 0.8016789087093389,
+ "grad_norm": 0.1561578254492378,
+ "learning_rate": 9.227615697476674e-06,
+ "loss": 0.42797762155532837,
+ "step": 1528
+ },
+ {
+ "epoch": 0.8022035676810073,
+ "grad_norm": 0.15642385261469652,
+ "learning_rate": 9.206000502085813e-06,
+ "loss": 0.4560159742832184,
+ "step": 1529
+ },
+ {
+ "epoch": 0.8027282266526757,
+ "grad_norm": 0.1623126279097499,
+ "learning_rate": 9.1844350073963e-06,
+ "loss": 0.4704165458679199,
+ "step": 1530
+ },
+ {
+ "epoch": 0.8032528856243442,
+ "grad_norm": 0.15090934786253773,
+ "learning_rate": 9.162919271996755e-06,
+ "loss": 0.44642555713653564,
+ "step": 1531
+ },
+ {
+ "epoch": 0.8037775445960126,
+ "grad_norm": 0.15305216978534084,
+ "learning_rate": 9.14145335434062e-06,
+ "loss": 0.4634283781051636,
+ "step": 1532
+ },
+ {
+ "epoch": 0.804302203567681,
+ "grad_norm": 0.14653219861345698,
+ "learning_rate": 9.120037312745979e-06,
+ "loss": 0.44544070959091187,
+ "step": 1533
+ },
+ {
+ "epoch": 0.8048268625393494,
+ "grad_norm": 0.16247397938502464,
+ "learning_rate": 9.098671205395422e-06,
+ "loss": 0.4552367925643921,
+ "step": 1534
+ },
+ {
+ "epoch": 0.8053515215110179,
+ "grad_norm": 0.14963174165902007,
+ "learning_rate": 9.077355090335892e-06,
+ "loss": 0.43211787939071655,
+ "step": 1535
+ },
+ {
+ "epoch": 0.8058761804826863,
+ "grad_norm": 0.1554853846772283,
+ "learning_rate": 9.056089025478495e-06,
+ "loss": 0.4319286644458771,
+ "step": 1536
+ },
+ {
+ "epoch": 0.8064008394543547,
+ "grad_norm": 0.15840895821310472,
+ "learning_rate": 9.03487306859837e-06,
+ "loss": 0.4494994878768921,
+ "step": 1537
+ },
+ {
+ "epoch": 0.8069254984260231,
+ "grad_norm": 0.1543496716800982,
+ "learning_rate": 9.013707277334515e-06,
+ "loss": 0.45500510931015015,
+ "step": 1538
+ },
+ {
+ "epoch": 0.8074501573976915,
+ "grad_norm": 0.14797282318550045,
+ "learning_rate": 8.992591709189651e-06,
+ "loss": 0.4385896325111389,
+ "step": 1539
+ },
+ {
+ "epoch": 0.8079748163693599,
+ "grad_norm": 0.17045570682787758,
+ "learning_rate": 8.971526421530066e-06,
+ "loss": 0.4585988521575928,
+ "step": 1540
+ },
+ {
+ "epoch": 0.8084994753410283,
+ "grad_norm": 0.14758220713044884,
+ "learning_rate": 8.950511471585407e-06,
+ "loss": 0.4295082092285156,
+ "step": 1541
+ },
+ {
+ "epoch": 0.8090241343126967,
+ "grad_norm": 0.16005752068175036,
+ "learning_rate": 8.929546916448597e-06,
+ "loss": 0.4507679343223572,
+ "step": 1542
+ },
+ {
+ "epoch": 0.8095487932843651,
+ "grad_norm": 0.14884684083313074,
+ "learning_rate": 8.908632813075641e-06,
+ "loss": 0.43547600507736206,
+ "step": 1543
+ },
+ {
+ "epoch": 0.8100734522560336,
+ "grad_norm": 0.14860107595788136,
+ "learning_rate": 8.88776921828547e-06,
+ "loss": 0.42700642347335815,
+ "step": 1544
+ },
+ {
+ "epoch": 0.810598111227702,
+ "grad_norm": 0.1472630611843838,
+ "learning_rate": 8.86695618875981e-06,
+ "loss": 0.4473339319229126,
+ "step": 1545
+ },
+ {
+ "epoch": 0.8111227701993704,
+ "grad_norm": 0.15105523788317893,
+ "learning_rate": 8.846193781043e-06,
+ "loss": 0.44923150539398193,
+ "step": 1546
+ },
+ {
+ "epoch": 0.8116474291710388,
+ "grad_norm": 0.16121409015263036,
+ "learning_rate": 8.82548205154185e-06,
+ "loss": 0.4653576612472534,
+ "step": 1547
+ },
+ {
+ "epoch": 0.8121720881427072,
+ "grad_norm": 0.1595195150959421,
+ "learning_rate": 8.804821056525497e-06,
+ "loss": 0.4522424638271332,
+ "step": 1548
+ },
+ {
+ "epoch": 0.8126967471143757,
+ "grad_norm": 0.15163510106316194,
+ "learning_rate": 8.784210852125232e-06,
+ "loss": 0.43657630681991577,
+ "step": 1549
+ },
+ {
+ "epoch": 0.8132214060860441,
+ "grad_norm": 0.15155109347013793,
+ "learning_rate": 8.763651494334377e-06,
+ "loss": 0.43183624744415283,
+ "step": 1550
+ },
+ {
+ "epoch": 0.8137460650577125,
+ "grad_norm": 0.14930024958763324,
+ "learning_rate": 8.7431430390081e-06,
+ "loss": 0.43563947081565857,
+ "step": 1551
+ },
+ {
+ "epoch": 0.8142707240293809,
+ "grad_norm": 0.15361200153430712,
+ "learning_rate": 8.722685541863283e-06,
+ "loss": 0.4387405514717102,
+ "step": 1552
+ },
+ {
+ "epoch": 0.8147953830010494,
+ "grad_norm": 0.15549528796193493,
+ "learning_rate": 8.702279058478366e-06,
+ "loss": 0.44347938895225525,
+ "step": 1553
+ },
+ {
+ "epoch": 0.8153200419727177,
+ "grad_norm": 0.16243042867861712,
+ "learning_rate": 8.681923644293194e-06,
+ "loss": 0.44852834939956665,
+ "step": 1554
+ },
+ {
+ "epoch": 0.8158447009443861,
+ "grad_norm": 0.15706129652944026,
+ "learning_rate": 8.661619354608873e-06,
+ "loss": 0.45419150590896606,
+ "step": 1555
+ },
+ {
+ "epoch": 0.8163693599160545,
+ "grad_norm": 0.14854258573134127,
+ "learning_rate": 8.641366244587611e-06,
+ "loss": 0.44866859912872314,
+ "step": 1556
+ },
+ {
+ "epoch": 0.816894018887723,
+ "grad_norm": 0.14896937678017533,
+ "learning_rate": 8.62116436925258e-06,
+ "loss": 0.44596076011657715,
+ "step": 1557
+ },
+ {
+ "epoch": 0.8174186778593914,
+ "grad_norm": 0.15614508495620508,
+ "learning_rate": 8.601013783487746e-06,
+ "loss": 0.45645207166671753,
+ "step": 1558
+ },
+ {
+ "epoch": 0.8179433368310598,
+ "grad_norm": 0.1486538543050606,
+ "learning_rate": 8.580914542037734e-06,
+ "loss": 0.4592300355434418,
+ "step": 1559
+ },
+ {
+ "epoch": 0.8184679958027282,
+ "grad_norm": 0.15475320756830518,
+ "learning_rate": 8.560866699507702e-06,
+ "loss": 0.4605134129524231,
+ "step": 1560
+ },
+ {
+ "epoch": 0.8189926547743966,
+ "grad_norm": 0.14754406728057914,
+ "learning_rate": 8.540870310363137e-06,
+ "loss": 0.43651673197746277,
+ "step": 1561
+ },
+ {
+ "epoch": 0.8195173137460651,
+ "grad_norm": 0.14871990701077825,
+ "learning_rate": 8.520925428929755e-06,
+ "loss": 0.44830837845802307,
+ "step": 1562
+ },
+ {
+ "epoch": 0.8200419727177335,
+ "grad_norm": 0.15358241317712973,
+ "learning_rate": 8.501032109393334e-06,
+ "loss": 0.4546644687652588,
+ "step": 1563
+ },
+ {
+ "epoch": 0.8205666316894019,
+ "grad_norm": 0.15650856948126138,
+ "learning_rate": 8.481190405799569e-06,
+ "loss": 0.4423891305923462,
+ "step": 1564
+ },
+ {
+ "epoch": 0.8210912906610703,
+ "grad_norm": 0.15474996746242287,
+ "learning_rate": 8.461400372053923e-06,
+ "loss": 0.44242313504219055,
+ "step": 1565
+ },
+ {
+ "epoch": 0.8216159496327388,
+ "grad_norm": 0.16544926591962897,
+ "learning_rate": 8.441662061921497e-06,
+ "loss": 0.4487020969390869,
+ "step": 1566
+ },
+ {
+ "epoch": 0.8221406086044072,
+ "grad_norm": 0.1489733586650647,
+ "learning_rate": 8.421975529026855e-06,
+ "loss": 0.4371356964111328,
+ "step": 1567
+ },
+ {
+ "epoch": 0.8226652675760755,
+ "grad_norm": 0.15995455652906124,
+ "learning_rate": 8.402340826853907e-06,
+ "loss": 0.4566021263599396,
+ "step": 1568
+ },
+ {
+ "epoch": 0.8231899265477439,
+ "grad_norm": 0.1500716152622102,
+ "learning_rate": 8.382758008745726e-06,
+ "loss": 0.43888187408447266,
+ "step": 1569
+ },
+ {
+ "epoch": 0.8237145855194123,
+ "grad_norm": 0.15554904485712012,
+ "learning_rate": 8.363227127904464e-06,
+ "loss": 0.4584001898765564,
+ "step": 1570
+ },
+ {
+ "epoch": 0.8242392444910808,
+ "grad_norm": 0.14820917815884277,
+ "learning_rate": 8.343748237391144e-06,
+ "loss": 0.43289071321487427,
+ "step": 1571
+ },
+ {
+ "epoch": 0.8247639034627492,
+ "grad_norm": 0.15801200747859714,
+ "learning_rate": 8.324321390125547e-06,
+ "loss": 0.44196009635925293,
+ "step": 1572
+ },
+ {
+ "epoch": 0.8252885624344176,
+ "grad_norm": 0.15323277975380797,
+ "learning_rate": 8.304946638886078e-06,
+ "loss": 0.43244463205337524,
+ "step": 1573
+ },
+ {
+ "epoch": 0.825813221406086,
+ "grad_norm": 0.15076940176344772,
+ "learning_rate": 8.285624036309583e-06,
+ "loss": 0.44401073455810547,
+ "step": 1574
+ },
+ {
+ "epoch": 0.8263378803777545,
+ "grad_norm": 0.1544958300475393,
+ "learning_rate": 8.266353634891266e-06,
+ "loss": 0.45198434591293335,
+ "step": 1575
+ },
+ {
+ "epoch": 0.8268625393494229,
+ "grad_norm": 0.15093528292068448,
+ "learning_rate": 8.247135486984486e-06,
+ "loss": 0.4403747618198395,
+ "step": 1576
+ },
+ {
+ "epoch": 0.8273871983210913,
+ "grad_norm": 0.14680662784017148,
+ "learning_rate": 8.227969644800648e-06,
+ "loss": 0.4375043213367462,
+ "step": 1577
+ },
+ {
+ "epoch": 0.8279118572927597,
+ "grad_norm": 0.14791704009291914,
+ "learning_rate": 8.208856160409055e-06,
+ "loss": 0.43811577558517456,
+ "step": 1578
+ },
+ {
+ "epoch": 0.8284365162644282,
+ "grad_norm": 0.14924185710518617,
+ "learning_rate": 8.189795085736765e-06,
+ "loss": 0.43853428959846497,
+ "step": 1579
+ },
+ {
+ "epoch": 0.8289611752360966,
+ "grad_norm": 0.15420576936040198,
+ "learning_rate": 8.170786472568451e-06,
+ "loss": 0.445467084646225,
+ "step": 1580
+ },
+ {
+ "epoch": 0.829485834207765,
+ "grad_norm": 0.15577901772605834,
+ "learning_rate": 8.151830372546258e-06,
+ "loss": 0.4529452621936798,
+ "step": 1581
+ },
+ {
+ "epoch": 0.8300104931794333,
+ "grad_norm": 0.15387840488797552,
+ "learning_rate": 8.132926837169676e-06,
+ "loss": 0.46009817719459534,
+ "step": 1582
+ },
+ {
+ "epoch": 0.8305351521511017,
+ "grad_norm": 0.1527913309957562,
+ "learning_rate": 8.11407591779537e-06,
+ "loss": 0.4513983726501465,
+ "step": 1583
+ },
+ {
+ "epoch": 0.8310598111227702,
+ "grad_norm": 0.1537090286922398,
+ "learning_rate": 8.095277665637061e-06,
+ "loss": 0.4537739157676697,
+ "step": 1584
+ },
+ {
+ "epoch": 0.8315844700944386,
+ "grad_norm": 0.15722962107654076,
+ "learning_rate": 8.076532131765408e-06,
+ "loss": 0.45232969522476196,
+ "step": 1585
+ },
+ {
+ "epoch": 0.832109129066107,
+ "grad_norm": 0.1450142396611787,
+ "learning_rate": 8.057839367107817e-06,
+ "loss": 0.4270634651184082,
+ "step": 1586
+ },
+ {
+ "epoch": 0.8326337880377754,
+ "grad_norm": 0.150264001570592,
+ "learning_rate": 8.039199422448355e-06,
+ "loss": 0.43563878536224365,
+ "step": 1587
+ },
+ {
+ "epoch": 0.8331584470094439,
+ "grad_norm": 0.14904943307590535,
+ "learning_rate": 8.020612348427572e-06,
+ "loss": 0.4401509165763855,
+ "step": 1588
+ },
+ {
+ "epoch": 0.8336831059811123,
+ "grad_norm": 0.15453821662860034,
+ "learning_rate": 8.002078195542393e-06,
+ "loss": 0.44545161724090576,
+ "step": 1589
+ },
+ {
+ "epoch": 0.8342077649527807,
+ "grad_norm": 0.1570809678830383,
+ "learning_rate": 7.983597014145951e-06,
+ "loss": 0.45228466391563416,
+ "step": 1590
+ },
+ {
+ "epoch": 0.8347324239244491,
+ "grad_norm": 0.14741240769629818,
+ "learning_rate": 7.965168854447493e-06,
+ "loss": 0.4485381245613098,
+ "step": 1591
+ },
+ {
+ "epoch": 0.8352570828961176,
+ "grad_norm": 0.15362267095533516,
+ "learning_rate": 7.946793766512195e-06,
+ "loss": 0.4358059763908386,
+ "step": 1592
+ },
+ {
+ "epoch": 0.835781741867786,
+ "grad_norm": 0.1544754586495605,
+ "learning_rate": 7.92847180026106e-06,
+ "loss": 0.43327778577804565,
+ "step": 1593
+ },
+ {
+ "epoch": 0.8363064008394544,
+ "grad_norm": 0.15023225026164963,
+ "learning_rate": 7.91020300547077e-06,
+ "loss": 0.44359731674194336,
+ "step": 1594
+ },
+ {
+ "epoch": 0.8368310598111228,
+ "grad_norm": 0.15665530392095345,
+ "learning_rate": 7.891987431773545e-06,
+ "loss": 0.44503387808799744,
+ "step": 1595
+ },
+ {
+ "epoch": 0.8373557187827911,
+ "grad_norm": 0.14717614926392292,
+ "learning_rate": 7.873825128657033e-06,
+ "loss": 0.45314061641693115,
+ "step": 1596
+ },
+ {
+ "epoch": 0.8378803777544596,
+ "grad_norm": 0.15207464934030612,
+ "learning_rate": 7.855716145464131e-06,
+ "loss": 0.4526442885398865,
+ "step": 1597
+ },
+ {
+ "epoch": 0.838405036726128,
+ "grad_norm": 0.15129881541794873,
+ "learning_rate": 7.837660531392904e-06,
+ "loss": 0.45127853751182556,
+ "step": 1598
+ },
+ {
+ "epoch": 0.8389296956977964,
+ "grad_norm": 0.15623362715417285,
+ "learning_rate": 7.819658335496405e-06,
+ "loss": 0.46432793140411377,
+ "step": 1599
+ },
+ {
+ "epoch": 0.8394543546694648,
+ "grad_norm": 0.1529219879642125,
+ "learning_rate": 7.801709606682573e-06,
+ "loss": 0.4475358724594116,
+ "step": 1600
+ },
+ {
+ "epoch": 0.8399790136411333,
+ "grad_norm": 0.14649139854087698,
+ "learning_rate": 7.78381439371409e-06,
+ "loss": 0.4363391101360321,
+ "step": 1601
+ },
+ {
+ "epoch": 0.8405036726128017,
+ "grad_norm": 0.14571233251948087,
+ "learning_rate": 7.765972745208243e-06,
+ "loss": 0.44419488310813904,
+ "step": 1602
+ },
+ {
+ "epoch": 0.8410283315844701,
+ "grad_norm": 0.15168460066406128,
+ "learning_rate": 7.748184709636791e-06,
+ "loss": 0.4433790445327759,
+ "step": 1603
+ },
+ {
+ "epoch": 0.8415529905561385,
+ "grad_norm": 0.14852860876513757,
+ "learning_rate": 7.730450335325853e-06,
+ "loss": 0.4493539333343506,
+ "step": 1604
+ },
+ {
+ "epoch": 0.842077649527807,
+ "grad_norm": 0.1505615491087377,
+ "learning_rate": 7.712769670455741e-06,
+ "loss": 0.4646318554878235,
+ "step": 1605
+ },
+ {
+ "epoch": 0.8426023084994754,
+ "grad_norm": 0.15407610486660114,
+ "learning_rate": 7.695142763060878e-06,
+ "loss": 0.4484049677848816,
+ "step": 1606
+ },
+ {
+ "epoch": 0.8431269674711438,
+ "grad_norm": 0.1521449709584124,
+ "learning_rate": 7.67756966102962e-06,
+ "loss": 0.43366891145706177,
+ "step": 1607
+ },
+ {
+ "epoch": 0.8436516264428122,
+ "grad_norm": 0.14747762647596166,
+ "learning_rate": 7.660050412104153e-06,
+ "loss": 0.43950092792510986,
+ "step": 1608
+ },
+ {
+ "epoch": 0.8441762854144806,
+ "grad_norm": 0.15512610809240027,
+ "learning_rate": 7.642585063880351e-06,
+ "loss": 0.46813055872917175,
+ "step": 1609
+ },
+ {
+ "epoch": 0.844700944386149,
+ "grad_norm": 0.15825480008734216,
+ "learning_rate": 7.625173663807655e-06,
+ "loss": 0.4348755478858948,
+ "step": 1610
+ },
+ {
+ "epoch": 0.8452256033578174,
+ "grad_norm": 0.1573968981854278,
+ "learning_rate": 7.6078162591889434e-06,
+ "loss": 0.45048123598098755,
+ "step": 1611
+ },
+ {
+ "epoch": 0.8457502623294858,
+ "grad_norm": 0.1517038960096417,
+ "learning_rate": 7.590512897180399e-06,
+ "loss": 0.43563610315322876,
+ "step": 1612
+ },
+ {
+ "epoch": 0.8462749213011542,
+ "grad_norm": 0.15503576044724665,
+ "learning_rate": 7.573263624791382e-06,
+ "loss": 0.4467726945877075,
+ "step": 1613
+ },
+ {
+ "epoch": 0.8467995802728226,
+ "grad_norm": 0.1560549927192375,
+ "learning_rate": 7.556068488884304e-06,
+ "loss": 0.4320477545261383,
+ "step": 1614
+ },
+ {
+ "epoch": 0.8473242392444911,
+ "grad_norm": 0.1679830356844654,
+ "learning_rate": 7.538927536174498e-06,
+ "loss": 0.4476034343242645,
+ "step": 1615
+ },
+ {
+ "epoch": 0.8478488982161595,
+ "grad_norm": 0.15169257383162213,
+ "learning_rate": 7.521840813230098e-06,
+ "loss": 0.4325854778289795,
+ "step": 1616
+ },
+ {
+ "epoch": 0.8483735571878279,
+ "grad_norm": 0.15980226324329827,
+ "learning_rate": 7.504808366471909e-06,
+ "loss": 0.4472721815109253,
+ "step": 1617
+ },
+ {
+ "epoch": 0.8488982161594963,
+ "grad_norm": 0.16932727986736365,
+ "learning_rate": 7.48783024217327e-06,
+ "loss": 0.45428287982940674,
+ "step": 1618
+ },
+ {
+ "epoch": 0.8494228751311648,
+ "grad_norm": 0.15065676306311376,
+ "learning_rate": 7.470906486459946e-06,
+ "loss": 0.44418883323669434,
+ "step": 1619
+ },
+ {
+ "epoch": 0.8499475341028332,
+ "grad_norm": 0.1533992998314183,
+ "learning_rate": 7.454037145309986e-06,
+ "loss": 0.43552401661872864,
+ "step": 1620
+ },
+ {
+ "epoch": 0.8504721930745016,
+ "grad_norm": 0.15479480515120017,
+ "learning_rate": 7.437222264553628e-06,
+ "loss": 0.444465696811676,
+ "step": 1621
+ },
+ {
+ "epoch": 0.85099685204617,
+ "grad_norm": 0.15845678337942112,
+ "learning_rate": 7.420461889873129e-06,
+ "loss": 0.44896578788757324,
+ "step": 1622
+ },
+ {
+ "epoch": 0.8515215110178384,
+ "grad_norm": 0.15038940533473208,
+ "learning_rate": 7.4037560668026804e-06,
+ "loss": 0.43062180280685425,
+ "step": 1623
+ },
+ {
+ "epoch": 0.8520461699895068,
+ "grad_norm": 0.1480017115109342,
+ "learning_rate": 7.387104840728259e-06,
+ "loss": 0.4630432724952698,
+ "step": 1624
+ },
+ {
+ "epoch": 0.8525708289611752,
+ "grad_norm": 0.15321305032824092,
+ "learning_rate": 7.370508256887514e-06,
+ "loss": 0.4419487714767456,
+ "step": 1625
+ },
+ {
+ "epoch": 0.8530954879328436,
+ "grad_norm": 0.1568246735699272,
+ "learning_rate": 7.353966360369659e-06,
+ "loss": 0.4483609199523926,
+ "step": 1626
+ },
+ {
+ "epoch": 0.853620146904512,
+ "grad_norm": 0.15539536583814875,
+ "learning_rate": 7.337479196115317e-06,
+ "loss": 0.4567753076553345,
+ "step": 1627
+ },
+ {
+ "epoch": 0.8541448058761805,
+ "grad_norm": 0.15528695146633797,
+ "learning_rate": 7.321046808916426e-06,
+ "loss": 0.4560334086418152,
+ "step": 1628
+ },
+ {
+ "epoch": 0.8546694648478489,
+ "grad_norm": 0.15005728554864303,
+ "learning_rate": 7.304669243416102e-06,
+ "loss": 0.43833765387535095,
+ "step": 1629
+ },
+ {
+ "epoch": 0.8551941238195173,
+ "grad_norm": 0.15959100645189242,
+ "learning_rate": 7.288346544108518e-06,
+ "loss": 0.4713831841945648,
+ "step": 1630
+ },
+ {
+ "epoch": 0.8557187827911857,
+ "grad_norm": 0.14898121463872632,
+ "learning_rate": 7.272078755338807e-06,
+ "loss": 0.44003745913505554,
+ "step": 1631
+ },
+ {
+ "epoch": 0.8562434417628542,
+ "grad_norm": 0.15083110802343333,
+ "learning_rate": 7.255865921302903e-06,
+ "loss": 0.429004043340683,
+ "step": 1632
+ },
+ {
+ "epoch": 0.8567681007345226,
+ "grad_norm": 0.1522757086796616,
+ "learning_rate": 7.239708086047445e-06,
+ "loss": 0.4403839707374573,
+ "step": 1633
+ },
+ {
+ "epoch": 0.857292759706191,
+ "grad_norm": 0.15719824838921245,
+ "learning_rate": 7.223605293469654e-06,
+ "loss": 0.4377821087837219,
+ "step": 1634
+ },
+ {
+ "epoch": 0.8578174186778594,
+ "grad_norm": 0.1450616202708373,
+ "learning_rate": 7.207557587317216e-06,
+ "loss": 0.4357878267765045,
+ "step": 1635
+ },
+ {
+ "epoch": 0.8583420776495279,
+ "grad_norm": 0.15703162220646427,
+ "learning_rate": 7.191565011188149e-06,
+ "loss": 0.44412243366241455,
+ "step": 1636
+ },
+ {
+ "epoch": 0.8588667366211962,
+ "grad_norm": 0.150175313334118,
+ "learning_rate": 7.175627608530716e-06,
+ "loss": 0.4543853998184204,
+ "step": 1637
+ },
+ {
+ "epoch": 0.8593913955928646,
+ "grad_norm": 0.15091711492716975,
+ "learning_rate": 7.159745422643261e-06,
+ "loss": 0.4268660545349121,
+ "step": 1638
+ },
+ {
+ "epoch": 0.859916054564533,
+ "grad_norm": 0.15449835654964011,
+ "learning_rate": 7.1439184966741295e-06,
+ "loss": 0.4507150650024414,
+ "step": 1639
+ },
+ {
+ "epoch": 0.8604407135362014,
+ "grad_norm": 0.15689612593442026,
+ "learning_rate": 7.128146873621533e-06,
+ "loss": 0.45520636439323425,
+ "step": 1640
+ },
+ {
+ "epoch": 0.8609653725078699,
+ "grad_norm": 0.1565226785234465,
+ "learning_rate": 7.11243059633345e-06,
+ "loss": 0.45102012157440186,
+ "step": 1641
+ },
+ {
+ "epoch": 0.8614900314795383,
+ "grad_norm": 0.15500877922230677,
+ "learning_rate": 7.096769707507481e-06,
+ "loss": 0.447115421295166,
+ "step": 1642
+ },
+ {
+ "epoch": 0.8620146904512067,
+ "grad_norm": 0.1595248428000243,
+ "learning_rate": 7.081164249690761e-06,
+ "loss": 0.4627235233783722,
+ "step": 1643
+ },
+ {
+ "epoch": 0.8625393494228751,
+ "grad_norm": 0.15958177510163488,
+ "learning_rate": 7.065614265279818e-06,
+ "loss": 0.4386388063430786,
+ "step": 1644
+ },
+ {
+ "epoch": 0.8630640083945436,
+ "grad_norm": 0.15152834825396286,
+ "learning_rate": 7.050119796520482e-06,
+ "loss": 0.4531296491622925,
+ "step": 1645
+ },
+ {
+ "epoch": 0.863588667366212,
+ "grad_norm": 0.15635733905345023,
+ "learning_rate": 7.034680885507752e-06,
+ "loss": 0.45573747158050537,
+ "step": 1646
+ },
+ {
+ "epoch": 0.8641133263378804,
+ "grad_norm": 0.15666918852633857,
+ "learning_rate": 7.019297574185703e-06,
+ "loss": 0.4499792158603668,
+ "step": 1647
+ },
+ {
+ "epoch": 0.8646379853095488,
+ "grad_norm": 0.156444627477174,
+ "learning_rate": 7.0039699043473405e-06,
+ "loss": 0.4545009136199951,
+ "step": 1648
+ },
+ {
+ "epoch": 0.8651626442812173,
+ "grad_norm": 0.14638989347446527,
+ "learning_rate": 6.988697917634515e-06,
+ "loss": 0.4397326409816742,
+ "step": 1649
+ },
+ {
+ "epoch": 0.8656873032528857,
+ "grad_norm": 0.15492813701343372,
+ "learning_rate": 6.973481655537793e-06,
+ "loss": 0.4402945041656494,
+ "step": 1650
+ },
+ {
+ "epoch": 0.866211962224554,
+ "grad_norm": 0.14466615385057716,
+ "learning_rate": 6.958321159396352e-06,
+ "loss": 0.4479280114173889,
+ "step": 1651
+ },
+ {
+ "epoch": 0.8667366211962224,
+ "grad_norm": 0.15865895330952856,
+ "learning_rate": 6.943216470397871e-06,
+ "loss": 0.4553218483924866,
+ "step": 1652
+ },
+ {
+ "epoch": 0.8672612801678908,
+ "grad_norm": 0.15293327582467067,
+ "learning_rate": 6.928167629578407e-06,
+ "loss": 0.44509196281433105,
+ "step": 1653
+ },
+ {
+ "epoch": 0.8677859391395593,
+ "grad_norm": 0.1515507946195896,
+ "learning_rate": 6.91317467782229e-06,
+ "loss": 0.4498351216316223,
+ "step": 1654
+ },
+ {
+ "epoch": 0.8683105981112277,
+ "grad_norm": 0.15313372799407085,
+ "learning_rate": 6.898237655862007e-06,
+ "loss": 0.4445887804031372,
+ "step": 1655
+ },
+ {
+ "epoch": 0.8688352570828961,
+ "grad_norm": 0.1491970050887588,
+ "learning_rate": 6.883356604278115e-06,
+ "loss": 0.4323808252811432,
+ "step": 1656
+ },
+ {
+ "epoch": 0.8693599160545645,
+ "grad_norm": 0.14449269685472518,
+ "learning_rate": 6.868531563499096e-06,
+ "loss": 0.42508652806282043,
+ "step": 1657
+ },
+ {
+ "epoch": 0.869884575026233,
+ "grad_norm": 0.14997328847830618,
+ "learning_rate": 6.853762573801264e-06,
+ "loss": 0.43655943870544434,
+ "step": 1658
+ },
+ {
+ "epoch": 0.8704092339979014,
+ "grad_norm": 0.15288208370561934,
+ "learning_rate": 6.839049675308662e-06,
+ "loss": 0.4505232274532318,
+ "step": 1659
+ },
+ {
+ "epoch": 0.8709338929695698,
+ "grad_norm": 0.14416201723141706,
+ "learning_rate": 6.824392907992938e-06,
+ "loss": 0.42444318532943726,
+ "step": 1660
+ },
+ {
+ "epoch": 0.8714585519412382,
+ "grad_norm": 0.14468507187223545,
+ "learning_rate": 6.809792311673244e-06,
+ "loss": 0.4251434803009033,
+ "step": 1661
+ },
+ {
+ "epoch": 0.8719832109129066,
+ "grad_norm": 0.15406295659798216,
+ "learning_rate": 6.795247926016142e-06,
+ "loss": 0.45068371295928955,
+ "step": 1662
+ },
+ {
+ "epoch": 0.8725078698845751,
+ "grad_norm": 0.1516721215663676,
+ "learning_rate": 6.780759790535465e-06,
+ "loss": 0.441638320684433,
+ "step": 1663
+ },
+ {
+ "epoch": 0.8730325288562435,
+ "grad_norm": 0.1537661139082594,
+ "learning_rate": 6.766327944592239e-06,
+ "loss": 0.44044923782348633,
+ "step": 1664
+ },
+ {
+ "epoch": 0.8735571878279118,
+ "grad_norm": 0.15175839438218855,
+ "learning_rate": 6.751952427394553e-06,
+ "loss": 0.44454342126846313,
+ "step": 1665
+ },
+ {
+ "epoch": 0.8740818467995802,
+ "grad_norm": 0.1531195282211392,
+ "learning_rate": 6.737633277997471e-06,
+ "loss": 0.4540765881538391,
+ "step": 1666
+ },
+ {
+ "epoch": 0.8746065057712487,
+ "grad_norm": 0.14961744003585584,
+ "learning_rate": 6.723370535302915e-06,
+ "loss": 0.44223541021347046,
+ "step": 1667
+ },
+ {
+ "epoch": 0.8751311647429171,
+ "grad_norm": 0.14816820299503658,
+ "learning_rate": 6.709164238059568e-06,
+ "loss": 0.43701833486557007,
+ "step": 1668
+ },
+ {
+ "epoch": 0.8756558237145855,
+ "grad_norm": 0.14622402294164918,
+ "learning_rate": 6.695014424862756e-06,
+ "loss": 0.42922335863113403,
+ "step": 1669
+ },
+ {
+ "epoch": 0.8761804826862539,
+ "grad_norm": 0.1547291062309004,
+ "learning_rate": 6.680921134154355e-06,
+ "loss": 0.4233800768852234,
+ "step": 1670
+ },
+ {
+ "epoch": 0.8767051416579223,
+ "grad_norm": 0.15378903523440587,
+ "learning_rate": 6.666884404222678e-06,
+ "loss": 0.4432950019836426,
+ "step": 1671
+ },
+ {
+ "epoch": 0.8772298006295908,
+ "grad_norm": 0.15973037860999773,
+ "learning_rate": 6.652904273202384e-06,
+ "loss": 0.4560699462890625,
+ "step": 1672
+ },
+ {
+ "epoch": 0.8777544596012592,
+ "grad_norm": 0.15430233088028492,
+ "learning_rate": 6.638980779074358e-06,
+ "loss": 0.45776498317718506,
+ "step": 1673
+ },
+ {
+ "epoch": 0.8782791185729276,
+ "grad_norm": 0.1587664730992995,
+ "learning_rate": 6.625113959665614e-06,
+ "loss": 0.43579810857772827,
+ "step": 1674
+ },
+ {
+ "epoch": 0.878803777544596,
+ "grad_norm": 0.14370919039693034,
+ "learning_rate": 6.6113038526491946e-06,
+ "loss": 0.4426189064979553,
+ "step": 1675
+ },
+ {
+ "epoch": 0.8793284365162645,
+ "grad_norm": 0.14933009355373217,
+ "learning_rate": 6.59755049554407e-06,
+ "loss": 0.44747287034988403,
+ "step": 1676
+ },
+ {
+ "epoch": 0.8798530954879329,
+ "grad_norm": 0.15713697891738065,
+ "learning_rate": 6.583853925715039e-06,
+ "loss": 0.46320077776908875,
+ "step": 1677
+ },
+ {
+ "epoch": 0.8803777544596013,
+ "grad_norm": 0.1542096833927892,
+ "learning_rate": 6.570214180372609e-06,
+ "loss": 0.4473888874053955,
+ "step": 1678
+ },
+ {
+ "epoch": 0.8809024134312696,
+ "grad_norm": 0.16187259968193946,
+ "learning_rate": 6.556631296572922e-06,
+ "loss": 0.4421869218349457,
+ "step": 1679
+ },
+ {
+ "epoch": 0.881427072402938,
+ "grad_norm": 0.15978032996583555,
+ "learning_rate": 6.54310531121762e-06,
+ "loss": 0.4305853843688965,
+ "step": 1680
+ },
+ {
+ "epoch": 0.8819517313746065,
+ "grad_norm": 0.15631580579775123,
+ "learning_rate": 6.529636261053784e-06,
+ "loss": 0.4340512454509735,
+ "step": 1681
+ },
+ {
+ "epoch": 0.8824763903462749,
+ "grad_norm": 0.150084161390103,
+ "learning_rate": 6.516224182673812e-06,
+ "loss": 0.4383493661880493,
+ "step": 1682
+ },
+ {
+ "epoch": 0.8830010493179433,
+ "grad_norm": 0.15527917215838966,
+ "learning_rate": 6.502869112515314e-06,
+ "loss": 0.45441877841949463,
+ "step": 1683
+ },
+ {
+ "epoch": 0.8835257082896117,
+ "grad_norm": 0.15164994594035522,
+ "learning_rate": 6.489571086861027e-06,
+ "loss": 0.43363988399505615,
+ "step": 1684
+ },
+ {
+ "epoch": 0.8840503672612802,
+ "grad_norm": 0.1467230535452683,
+ "learning_rate": 6.47633014183871e-06,
+ "loss": 0.4263266623020172,
+ "step": 1685
+ },
+ {
+ "epoch": 0.8845750262329486,
+ "grad_norm": 0.14597001769185472,
+ "learning_rate": 6.463146313421044e-06,
+ "loss": 0.4223391115665436,
+ "step": 1686
+ },
+ {
+ "epoch": 0.885099685204617,
+ "grad_norm": 0.15050446002714588,
+ "learning_rate": 6.4500196374255435e-06,
+ "loss": 0.43097931146621704,
+ "step": 1687
+ },
+ {
+ "epoch": 0.8856243441762854,
+ "grad_norm": 0.14918033614543294,
+ "learning_rate": 6.436950149514448e-06,
+ "loss": 0.4587080478668213,
+ "step": 1688
+ },
+ {
+ "epoch": 0.8861490031479539,
+ "grad_norm": 0.15469492243631364,
+ "learning_rate": 6.423937885194631e-06,
+ "loss": 0.45661866664886475,
+ "step": 1689
+ },
+ {
+ "epoch": 0.8866736621196223,
+ "grad_norm": 0.15552798640177423,
+ "learning_rate": 6.410982879817498e-06,
+ "loss": 0.44370236992836,
+ "step": 1690
+ },
+ {
+ "epoch": 0.8871983210912907,
+ "grad_norm": 0.15595593365088142,
+ "learning_rate": 6.398085168578901e-06,
+ "loss": 0.4291878342628479,
+ "step": 1691
+ },
+ {
+ "epoch": 0.887722980062959,
+ "grad_norm": 0.14708046948755318,
+ "learning_rate": 6.385244786519036e-06,
+ "loss": 0.4422011375427246,
+ "step": 1692
+ },
+ {
+ "epoch": 0.8882476390346274,
+ "grad_norm": 0.1499253382866992,
+ "learning_rate": 6.3724617685223404e-06,
+ "loss": 0.4595015048980713,
+ "step": 1693
+ },
+ {
+ "epoch": 0.8887722980062959,
+ "grad_norm": 0.14702868569252941,
+ "learning_rate": 6.3597361493174205e-06,
+ "loss": 0.4347558319568634,
+ "step": 1694
+ },
+ {
+ "epoch": 0.8892969569779643,
+ "grad_norm": 0.14977289247327386,
+ "learning_rate": 6.347067963476928e-06,
+ "loss": 0.43877455592155457,
+ "step": 1695
+ },
+ {
+ "epoch": 0.8898216159496327,
+ "grad_norm": 0.1453902077891248,
+ "learning_rate": 6.334457245417491e-06,
+ "loss": 0.4282415807247162,
+ "step": 1696
+ },
+ {
+ "epoch": 0.8903462749213011,
+ "grad_norm": 0.1545933676895574,
+ "learning_rate": 6.321904029399607e-06,
+ "loss": 0.4429236054420471,
+ "step": 1697
+ },
+ {
+ "epoch": 0.8908709338929696,
+ "grad_norm": 0.15625844051167467,
+ "learning_rate": 6.3094083495275565e-06,
+ "loss": 0.4284421503543854,
+ "step": 1698
+ },
+ {
+ "epoch": 0.891395592864638,
+ "grad_norm": 0.1484963883381567,
+ "learning_rate": 6.2969702397493e-06,
+ "loss": 0.4545463025569916,
+ "step": 1699
+ },
+ {
+ "epoch": 0.8919202518363064,
+ "grad_norm": 0.15122402806869387,
+ "learning_rate": 6.284589733856403e-06,
+ "loss": 0.43840622901916504,
+ "step": 1700
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 4.499834106648986e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1700/video_preprocessor_config.json b/checkpoints/checkpoint-1700/video_preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..a933f0ec6941222c33aeba7980bd5acf8dd98e91
--- /dev/null
+++ b/checkpoints/checkpoint-1700/video_preprocessor_config.json
@@ -0,0 +1,30 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "do_sample_frames": true,
+ "fps": 2,
+ "image_mean": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "image_std": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "max_frames": 768,
+ "merge_size": 2,
+ "min_frames": 4,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 786432,
+ "shortest_edge": 131072
+ },
+ "temporal_patch_size": 2,
+ "video_processor_type": "Qwen3VLVideoProcessor"
+}
diff --git a/checkpoints/checkpoint-1750/chat_template.jinja b/checkpoints/checkpoint-1750/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1750/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1750/config.json b/checkpoints/checkpoint-1750/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1750/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1750/expert_prune_metadata.json b/checkpoints/checkpoint-1750/expert_prune_metadata.json
new file mode 100644
index 0000000000000000000000000000000000000000..16e01c8a10636331e334732fa9629579c68df44e
--- /dev/null
+++ b/checkpoints/checkpoint-1750/expert_prune_metadata.json
@@ -0,0 +1,14 @@
+{
+ "source_model": "/mnt/data/wuwei/model_list/Qwen3.6-35B-A3B/models--Qwen--Qwen3.6-35B-A3B/snapshots/995ad96eacd98c81ed38be0c5b274b04031597b0",
+ "output_model": "/mnt/data/wuwei/model_list/qwen3-19b-a3b",
+ "total_parameters": 18990568816,
+ "parameter_billions_decimal": 18.990568816,
+ "directory_size_label": "19B",
+ "original_num_experts": 256,
+ "kept_experts_per_layer": 128,
+ "num_profiled_layers": 40,
+ "mtp_weights": "removed because Qwen3_5MoeForConditionalGeneration does not instantiate or use the source checkpoint's mtp.* training-only tensors",
+ "expert_order": "ascending original expert index; router rows and expert tensors use the same remapping",
+ "router_statistics": "router_stats.json and router_stats.pt",
+ "selection_score": "top-8 router hit frequency * mean(abs(actual expert output after down_proj and before routing-weight multiplication))"
+}
diff --git a/checkpoints/checkpoint-1750/generation_config.json b/checkpoints/checkpoint-1750/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1750/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1750/model.safetensors b/checkpoints/checkpoint-1750/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..8cb150b041e2c329478b9aaa7f6abf1a8d66d301
--- /dev/null
+++ b/checkpoints/checkpoint-1750/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f00a8c1358ee98c2d0cc099ddd540104da454bb7792cf33a44a6b0499f280aa
+size 37983421000
diff --git a/checkpoints/checkpoint-1750/preprocessor_config.json b/checkpoints/checkpoint-1750/preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ec43d1f859ecadf800ebae92fda4a812d34975d3
--- /dev/null
+++ b/checkpoints/checkpoint-1750/preprocessor_config.json
@@ -0,0 +1,26 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "image_mean": [
+ 0.48145466,
+ 0.4578275,
+ 0.40821073
+ ],
+ "image_processor_type": "Qwen2VLImageProcessorFast",
+ "image_std": [
+ 0.26862954,
+ 0.26130258,
+ 0.27577711
+ ],
+ "merge_size": 2,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 1003520,
+ "shortest_edge": 3136
+ },
+ "temporal_patch_size": 2
+}
diff --git a/checkpoints/checkpoint-1750/tokenizer.json b/checkpoints/checkpoint-1750/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1750/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1750/tokenizer_config.json b/checkpoints/checkpoint-1750/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1750/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1750/trainer_state.json b/checkpoints/checkpoint-1750/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..dd408d91dc50b127772c4577b090e688158a6040
--- /dev/null
+++ b/checkpoints/checkpoint-1750/trainer_state.json
@@ -0,0 +1,12284 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.9181532004197271,
+ "eval_steps": 500,
+ "global_step": 1750,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ },
+ {
+ "epoch": 0.5776495278069255,
+ "grad_norm": 0.1862053032854098,
+ "learning_rate": 2.2065553964033126e-05,
+ "loss": 0.4816875457763672,
+ "step": 1101
+ },
+ {
+ "epoch": 0.5781741867785939,
+ "grad_norm": 0.16996015894363117,
+ "learning_rate": 2.20295734045532e-05,
+ "loss": 0.4471197724342346,
+ "step": 1102
+ },
+ {
+ "epoch": 0.5786988457502623,
+ "grad_norm": 0.16889227008390692,
+ "learning_rate": 2.1993607706997725e-05,
+ "loss": 0.468391478061676,
+ "step": 1103
+ },
+ {
+ "epoch": 0.5792235047219307,
+ "grad_norm": 0.16670931681562373,
+ "learning_rate": 2.195765696907745e-05,
+ "loss": 0.44934162497520447,
+ "step": 1104
+ },
+ {
+ "epoch": 0.5797481636935992,
+ "grad_norm": 0.16328418741223716,
+ "learning_rate": 2.1921721288462482e-05,
+ "loss": 0.4558270275592804,
+ "step": 1105
+ },
+ {
+ "epoch": 0.5802728226652676,
+ "grad_norm": 0.17364660408703936,
+ "learning_rate": 2.188580076278203e-05,
+ "loss": 0.4597751796245575,
+ "step": 1106
+ },
+ {
+ "epoch": 0.580797481636936,
+ "grad_norm": 0.1628618453800601,
+ "learning_rate": 2.18498954896241e-05,
+ "loss": 0.4521859288215637,
+ "step": 1107
+ },
+ {
+ "epoch": 0.5813221406086044,
+ "grad_norm": 0.16813817951871482,
+ "learning_rate": 2.1814005566535302e-05,
+ "loss": 0.4581005573272705,
+ "step": 1108
+ },
+ {
+ "epoch": 0.5818467995802729,
+ "grad_norm": 0.16825923629726025,
+ "learning_rate": 2.1778131091020516e-05,
+ "loss": 0.4433916211128235,
+ "step": 1109
+ },
+ {
+ "epoch": 0.5823714585519413,
+ "grad_norm": 0.17194743467883944,
+ "learning_rate": 2.1742272160542654e-05,
+ "loss": 0.47296425700187683,
+ "step": 1110
+ },
+ {
+ "epoch": 0.5828961175236097,
+ "grad_norm": 0.16947730546039969,
+ "learning_rate": 2.170642887252241e-05,
+ "loss": 0.4815937876701355,
+ "step": 1111
+ },
+ {
+ "epoch": 0.583420776495278,
+ "grad_norm": 0.1657449928668473,
+ "learning_rate": 2.167060132433798e-05,
+ "loss": 0.4479164779186249,
+ "step": 1112
+ },
+ {
+ "epoch": 0.5839454354669464,
+ "grad_norm": 0.1658952516607364,
+ "learning_rate": 2.1634789613324778e-05,
+ "loss": 0.4637463688850403,
+ "step": 1113
+ },
+ {
+ "epoch": 0.5844700944386149,
+ "grad_norm": 0.16777002188998236,
+ "learning_rate": 2.1598993836775218e-05,
+ "loss": 0.457524836063385,
+ "step": 1114
+ },
+ {
+ "epoch": 0.5849947534102833,
+ "grad_norm": 0.15837566447341322,
+ "learning_rate": 2.1563214091938405e-05,
+ "loss": 0.44938087463378906,
+ "step": 1115
+ },
+ {
+ "epoch": 0.5855194123819517,
+ "grad_norm": 0.17162206617035314,
+ "learning_rate": 2.1527450476019896e-05,
+ "loss": 0.4393450617790222,
+ "step": 1116
+ },
+ {
+ "epoch": 0.5860440713536201,
+ "grad_norm": 0.16224101921433814,
+ "learning_rate": 2.1491703086181426e-05,
+ "loss": 0.451913982629776,
+ "step": 1117
+ },
+ {
+ "epoch": 0.5865687303252886,
+ "grad_norm": 0.1736253504575144,
+ "learning_rate": 2.145597201954066e-05,
+ "loss": 0.4611290395259857,
+ "step": 1118
+ },
+ {
+ "epoch": 0.587093389296957,
+ "grad_norm": 0.16308408333737234,
+ "learning_rate": 2.1420257373170903e-05,
+ "loss": 0.45898568630218506,
+ "step": 1119
+ },
+ {
+ "epoch": 0.5876180482686254,
+ "grad_norm": 0.16948634794175524,
+ "learning_rate": 2.1384559244100848e-05,
+ "loss": 0.45982882380485535,
+ "step": 1120
+ },
+ {
+ "epoch": 0.5881427072402938,
+ "grad_norm": 0.16676366538565252,
+ "learning_rate": 2.1348877729314336e-05,
+ "loss": 0.44826576113700867,
+ "step": 1121
+ },
+ {
+ "epoch": 0.5886673662119623,
+ "grad_norm": 0.16431125630712295,
+ "learning_rate": 2.1313212925750037e-05,
+ "loss": 0.4616445302963257,
+ "step": 1122
+ },
+ {
+ "epoch": 0.5891920251836307,
+ "grad_norm": 0.1631659061616955,
+ "learning_rate": 2.1277564930301252e-05,
+ "loss": 0.45651188492774963,
+ "step": 1123
+ },
+ {
+ "epoch": 0.5897166841552991,
+ "grad_norm": 0.15805950502689237,
+ "learning_rate": 2.1241933839815604e-05,
+ "loss": 0.4486045837402344,
+ "step": 1124
+ },
+ {
+ "epoch": 0.5902413431269675,
+ "grad_norm": 0.16530288038603502,
+ "learning_rate": 2.1206319751094773e-05,
+ "loss": 0.4485066533088684,
+ "step": 1125
+ },
+ {
+ "epoch": 0.5907660020986358,
+ "grad_norm": 0.16061903194239047,
+ "learning_rate": 2.1170722760894295e-05,
+ "loss": 0.4565330743789673,
+ "step": 1126
+ },
+ {
+ "epoch": 0.5912906610703043,
+ "grad_norm": 0.1643346999353627,
+ "learning_rate": 2.1135142965923195e-05,
+ "loss": 0.45560717582702637,
+ "step": 1127
+ },
+ {
+ "epoch": 0.5918153200419727,
+ "grad_norm": 0.17484644315215753,
+ "learning_rate": 2.109958046284382e-05,
+ "loss": 0.4795891046524048,
+ "step": 1128
+ },
+ {
+ "epoch": 0.5923399790136411,
+ "grad_norm": 0.1623762678622651,
+ "learning_rate": 2.1064035348271528e-05,
+ "loss": 0.47395604848861694,
+ "step": 1129
+ },
+ {
+ "epoch": 0.5928646379853095,
+ "grad_norm": 0.16888824241669015,
+ "learning_rate": 2.1028507718774456e-05,
+ "loss": 0.4601455330848694,
+ "step": 1130
+ },
+ {
+ "epoch": 0.593389296956978,
+ "grad_norm": 0.16291744568806302,
+ "learning_rate": 2.0992997670873193e-05,
+ "loss": 0.46248236298561096,
+ "step": 1131
+ },
+ {
+ "epoch": 0.5939139559286464,
+ "grad_norm": 0.16185598691236988,
+ "learning_rate": 2.0957505301040603e-05,
+ "loss": 0.45085301995277405,
+ "step": 1132
+ },
+ {
+ "epoch": 0.5944386149003148,
+ "grad_norm": 0.1678000806117814,
+ "learning_rate": 2.0922030705701518e-05,
+ "loss": 0.44939014315605164,
+ "step": 1133
+ },
+ {
+ "epoch": 0.5949632738719832,
+ "grad_norm": 0.16104474282288994,
+ "learning_rate": 2.088657398123246e-05,
+ "loss": 0.4313976764678955,
+ "step": 1134
+ },
+ {
+ "epoch": 0.5954879328436516,
+ "grad_norm": 0.1631128968199669,
+ "learning_rate": 2.085113522396141e-05,
+ "loss": 0.4508278965950012,
+ "step": 1135
+ },
+ {
+ "epoch": 0.5960125918153201,
+ "grad_norm": 0.16307688195844533,
+ "learning_rate": 2.0815714530167557e-05,
+ "loss": 0.4467576742172241,
+ "step": 1136
+ },
+ {
+ "epoch": 0.5965372507869885,
+ "grad_norm": 0.16456832304532207,
+ "learning_rate": 2.0780311996080976e-05,
+ "loss": 0.4621703624725342,
+ "step": 1137
+ },
+ {
+ "epoch": 0.5970619097586569,
+ "grad_norm": 0.16346217748325367,
+ "learning_rate": 2.0744927717882435e-05,
+ "loss": 0.4589449465274811,
+ "step": 1138
+ },
+ {
+ "epoch": 0.5975865687303253,
+ "grad_norm": 0.16180691501076994,
+ "learning_rate": 2.0709561791703097e-05,
+ "loss": 0.43698400259017944,
+ "step": 1139
+ },
+ {
+ "epoch": 0.5981112277019937,
+ "grad_norm": 0.15695459976898835,
+ "learning_rate": 2.067421431362427e-05,
+ "loss": 0.452271044254303,
+ "step": 1140
+ },
+ {
+ "epoch": 0.5986358866736621,
+ "grad_norm": 0.15930893553721054,
+ "learning_rate": 2.063888537967713e-05,
+ "loss": 0.44868963956832886,
+ "step": 1141
+ },
+ {
+ "epoch": 0.5991605456453305,
+ "grad_norm": 0.15963802284017928,
+ "learning_rate": 2.0603575085842485e-05,
+ "loss": 0.4376864731311798,
+ "step": 1142
+ },
+ {
+ "epoch": 0.5996852046169989,
+ "grad_norm": 0.16610266900507178,
+ "learning_rate": 2.05682835280505e-05,
+ "loss": 0.44676727056503296,
+ "step": 1143
+ },
+ {
+ "epoch": 0.6002098635886673,
+ "grad_norm": 0.15978253236669612,
+ "learning_rate": 2.053301080218044e-05,
+ "loss": 0.434773713350296,
+ "step": 1144
+ },
+ {
+ "epoch": 0.6007345225603358,
+ "grad_norm": 0.16305141803526324,
+ "learning_rate": 2.0497757004060403e-05,
+ "loss": 0.4529780149459839,
+ "step": 1145
+ },
+ {
+ "epoch": 0.6012591815320042,
+ "grad_norm": 0.1649737864101215,
+ "learning_rate": 2.046252222946707e-05,
+ "loss": 0.4571434259414673,
+ "step": 1146
+ },
+ {
+ "epoch": 0.6017838405036726,
+ "grad_norm": 0.15738824968326598,
+ "learning_rate": 2.0427306574125427e-05,
+ "loss": 0.4446289539337158,
+ "step": 1147
+ },
+ {
+ "epoch": 0.602308499475341,
+ "grad_norm": 0.1731139535185187,
+ "learning_rate": 2.039211013370853e-05,
+ "loss": 0.4551090598106384,
+ "step": 1148
+ },
+ {
+ "epoch": 0.6028331584470095,
+ "grad_norm": 0.15857251900043037,
+ "learning_rate": 2.0356933003837237e-05,
+ "loss": 0.4644010663032532,
+ "step": 1149
+ },
+ {
+ "epoch": 0.6033578174186779,
+ "grad_norm": 0.1589582861134211,
+ "learning_rate": 2.032177528007992e-05,
+ "loss": 0.45741650462150574,
+ "step": 1150
+ },
+ {
+ "epoch": 0.6038824763903463,
+ "grad_norm": 0.15639092706575122,
+ "learning_rate": 2.0286637057952257e-05,
+ "loss": 0.4375547766685486,
+ "step": 1151
+ },
+ {
+ "epoch": 0.6044071353620147,
+ "grad_norm": 0.16335605581337187,
+ "learning_rate": 2.0251518432916933e-05,
+ "loss": 0.4636140465736389,
+ "step": 1152
+ },
+ {
+ "epoch": 0.6049317943336832,
+ "grad_norm": 0.15437437113359356,
+ "learning_rate": 2.0216419500383372e-05,
+ "loss": 0.44632142782211304,
+ "step": 1153
+ },
+ {
+ "epoch": 0.6054564533053515,
+ "grad_norm": 0.15522039198698664,
+ "learning_rate": 2.0181340355707545e-05,
+ "loss": 0.449049174785614,
+ "step": 1154
+ },
+ {
+ "epoch": 0.6059811122770199,
+ "grad_norm": 0.1599951917909323,
+ "learning_rate": 2.0146281094191604e-05,
+ "loss": 0.44395962357521057,
+ "step": 1155
+ },
+ {
+ "epoch": 0.6065057712486883,
+ "grad_norm": 0.16866621620527716,
+ "learning_rate": 2.011124181108373e-05,
+ "loss": 0.45693716406822205,
+ "step": 1156
+ },
+ {
+ "epoch": 0.6070304302203567,
+ "grad_norm": 0.15508935650135075,
+ "learning_rate": 2.0076222601577815e-05,
+ "loss": 0.4525151252746582,
+ "step": 1157
+ },
+ {
+ "epoch": 0.6075550891920252,
+ "grad_norm": 0.16508998118005971,
+ "learning_rate": 2.004122356081321e-05,
+ "loss": 0.466006875038147,
+ "step": 1158
+ },
+ {
+ "epoch": 0.6080797481636936,
+ "grad_norm": 0.15817418864022237,
+ "learning_rate": 2.0006244783874466e-05,
+ "loss": 0.4496110677719116,
+ "step": 1159
+ },
+ {
+ "epoch": 0.608604407135362,
+ "grad_norm": 0.16127270259749824,
+ "learning_rate": 1.997128636579109e-05,
+ "loss": 0.45918628573417664,
+ "step": 1160
+ },
+ {
+ "epoch": 0.6091290661070304,
+ "grad_norm": 0.1702599924159677,
+ "learning_rate": 1.9936348401537292e-05,
+ "loss": 0.47218555212020874,
+ "step": 1161
+ },
+ {
+ "epoch": 0.6096537250786989,
+ "grad_norm": 0.15690176248846127,
+ "learning_rate": 1.9901430986031695e-05,
+ "loss": 0.4516097903251648,
+ "step": 1162
+ },
+ {
+ "epoch": 0.6101783840503673,
+ "grad_norm": 0.1713614405614144,
+ "learning_rate": 1.9866534214137085e-05,
+ "loss": 0.45030856132507324,
+ "step": 1163
+ },
+ {
+ "epoch": 0.6107030430220357,
+ "grad_norm": 0.1670865563026229,
+ "learning_rate": 1.9831658180660195e-05,
+ "loss": 0.45603489875793457,
+ "step": 1164
+ },
+ {
+ "epoch": 0.6112277019937041,
+ "grad_norm": 0.16897539113401952,
+ "learning_rate": 1.9796802980351392e-05,
+ "loss": 0.46314537525177,
+ "step": 1165
+ },
+ {
+ "epoch": 0.6117523609653726,
+ "grad_norm": 0.15832868870697644,
+ "learning_rate": 1.976196870790446e-05,
+ "loss": 0.44784611463546753,
+ "step": 1166
+ },
+ {
+ "epoch": 0.612277019937041,
+ "grad_norm": 0.16595365551891114,
+ "learning_rate": 1.9727155457956315e-05,
+ "loss": 0.4409624934196472,
+ "step": 1167
+ },
+ {
+ "epoch": 0.6128016789087093,
+ "grad_norm": 0.1642630411212017,
+ "learning_rate": 1.969236332508677e-05,
+ "loss": 0.46120351552963257,
+ "step": 1168
+ },
+ {
+ "epoch": 0.6133263378803777,
+ "grad_norm": 0.16481796135868812,
+ "learning_rate": 1.9657592403818253e-05,
+ "loss": 0.45544904470443726,
+ "step": 1169
+ },
+ {
+ "epoch": 0.6138509968520461,
+ "grad_norm": 0.1629832779852364,
+ "learning_rate": 1.962284278861557e-05,
+ "loss": 0.4578462839126587,
+ "step": 1170
+ },
+ {
+ "epoch": 0.6143756558237146,
+ "grad_norm": 0.15916219613603688,
+ "learning_rate": 1.9588114573885664e-05,
+ "loss": 0.4546983540058136,
+ "step": 1171
+ },
+ {
+ "epoch": 0.614900314795383,
+ "grad_norm": 0.1660520165166075,
+ "learning_rate": 1.95534078539773e-05,
+ "loss": 0.4601942300796509,
+ "step": 1172
+ },
+ {
+ "epoch": 0.6154249737670514,
+ "grad_norm": 0.16924794578577768,
+ "learning_rate": 1.9518722723180877e-05,
+ "loss": 0.45997846126556396,
+ "step": 1173
+ },
+ {
+ "epoch": 0.6159496327387198,
+ "grad_norm": 0.16597646133045382,
+ "learning_rate": 1.9484059275728127e-05,
+ "loss": 0.4594605565071106,
+ "step": 1174
+ },
+ {
+ "epoch": 0.6164742917103883,
+ "grad_norm": 0.16964243792519748,
+ "learning_rate": 1.9449417605791876e-05,
+ "loss": 0.4638690650463104,
+ "step": 1175
+ },
+ {
+ "epoch": 0.6169989506820567,
+ "grad_norm": 0.1758591426791696,
+ "learning_rate": 1.9414797807485796e-05,
+ "loss": 0.4666447639465332,
+ "step": 1176
+ },
+ {
+ "epoch": 0.6175236096537251,
+ "grad_norm": 0.16217661227891753,
+ "learning_rate": 1.9380199974864126e-05,
+ "loss": 0.44252535700798035,
+ "step": 1177
+ },
+ {
+ "epoch": 0.6180482686253935,
+ "grad_norm": 0.1576446184499019,
+ "learning_rate": 1.934562420192142e-05,
+ "loss": 0.4499641954898834,
+ "step": 1178
+ },
+ {
+ "epoch": 0.618572927597062,
+ "grad_norm": 0.16231008092636912,
+ "learning_rate": 1.9311070582592327e-05,
+ "loss": 0.4561542570590973,
+ "step": 1179
+ },
+ {
+ "epoch": 0.6190975865687304,
+ "grad_norm": 0.16790382249577043,
+ "learning_rate": 1.9276539210751292e-05,
+ "loss": 0.4484325647354126,
+ "step": 1180
+ },
+ {
+ "epoch": 0.6196222455403987,
+ "grad_norm": 0.18794729932605855,
+ "learning_rate": 1.9242030180212323e-05,
+ "loss": 0.44169026613235474,
+ "step": 1181
+ },
+ {
+ "epoch": 0.6201469045120671,
+ "grad_norm": 0.16948051243889814,
+ "learning_rate": 1.920754358472872e-05,
+ "loss": 0.4592876434326172,
+ "step": 1182
+ },
+ {
+ "epoch": 0.6206715634837355,
+ "grad_norm": 0.16762558710029374,
+ "learning_rate": 1.917307951799286e-05,
+ "loss": 0.4402095079421997,
+ "step": 1183
+ },
+ {
+ "epoch": 0.621196222455404,
+ "grad_norm": 0.18170280116387003,
+ "learning_rate": 1.9138638073635885e-05,
+ "loss": 0.4633026123046875,
+ "step": 1184
+ },
+ {
+ "epoch": 0.6217208814270724,
+ "grad_norm": 0.1718051709136108,
+ "learning_rate": 1.9104219345227486e-05,
+ "loss": 0.4510066509246826,
+ "step": 1185
+ },
+ {
+ "epoch": 0.6222455403987408,
+ "grad_norm": 0.16829066326345576,
+ "learning_rate": 1.906982342627565e-05,
+ "loss": 0.4632919430732727,
+ "step": 1186
+ },
+ {
+ "epoch": 0.6227701993704092,
+ "grad_norm": 0.16669141808548274,
+ "learning_rate": 1.9035450410226395e-05,
+ "loss": 0.455544650554657,
+ "step": 1187
+ },
+ {
+ "epoch": 0.6232948583420777,
+ "grad_norm": 0.16679996559722232,
+ "learning_rate": 1.90011003904635e-05,
+ "loss": 0.459994375705719,
+ "step": 1188
+ },
+ {
+ "epoch": 0.6238195173137461,
+ "grad_norm": 0.17379754713278164,
+ "learning_rate": 1.8966773460308284e-05,
+ "loss": 0.4576749801635742,
+ "step": 1189
+ },
+ {
+ "epoch": 0.6243441762854145,
+ "grad_norm": 0.17088539733389246,
+ "learning_rate": 1.8932469713019335e-05,
+ "loss": 0.44853052496910095,
+ "step": 1190
+ },
+ {
+ "epoch": 0.6248688352570829,
+ "grad_norm": 0.15556883889201015,
+ "learning_rate": 1.8898189241792248e-05,
+ "loss": 0.44060447812080383,
+ "step": 1191
+ },
+ {
+ "epoch": 0.6253934942287513,
+ "grad_norm": 0.1640093917439215,
+ "learning_rate": 1.8863932139759388e-05,
+ "loss": 0.4531911611557007,
+ "step": 1192
+ },
+ {
+ "epoch": 0.6259181532004198,
+ "grad_norm": 0.16444284826591576,
+ "learning_rate": 1.882969849998964e-05,
+ "loss": 0.44176363945007324,
+ "step": 1193
+ },
+ {
+ "epoch": 0.6264428121720882,
+ "grad_norm": 0.1696462520995642,
+ "learning_rate": 1.8795488415488143e-05,
+ "loss": 0.44516652822494507,
+ "step": 1194
+ },
+ {
+ "epoch": 0.6269674711437565,
+ "grad_norm": 0.16810724705191796,
+ "learning_rate": 1.8761301979196032e-05,
+ "loss": 0.4542563855648041,
+ "step": 1195
+ },
+ {
+ "epoch": 0.6274921301154249,
+ "grad_norm": 0.17334813313855824,
+ "learning_rate": 1.8727139283990204e-05,
+ "loss": 0.47769248485565186,
+ "step": 1196
+ },
+ {
+ "epoch": 0.6280167890870934,
+ "grad_norm": 0.16308212035303235,
+ "learning_rate": 1.8693000422683043e-05,
+ "loss": 0.4407799243927002,
+ "step": 1197
+ },
+ {
+ "epoch": 0.6285414480587618,
+ "grad_norm": 0.16853319139928977,
+ "learning_rate": 1.8658885488022204e-05,
+ "loss": 0.4565497040748596,
+ "step": 1198
+ },
+ {
+ "epoch": 0.6290661070304302,
+ "grad_norm": 0.16694968781284703,
+ "learning_rate": 1.8624794572690328e-05,
+ "loss": 0.44522374868392944,
+ "step": 1199
+ },
+ {
+ "epoch": 0.6295907660020986,
+ "grad_norm": 0.16418579468339103,
+ "learning_rate": 1.8590727769304788e-05,
+ "loss": 0.4441048502922058,
+ "step": 1200
+ },
+ {
+ "epoch": 0.630115424973767,
+ "grad_norm": 0.1661930777124156,
+ "learning_rate": 1.8556685170417486e-05,
+ "loss": 0.457327663898468,
+ "step": 1201
+ },
+ {
+ "epoch": 0.6306400839454355,
+ "grad_norm": 0.16536630027039814,
+ "learning_rate": 1.852266686851452e-05,
+ "loss": 0.462494432926178,
+ "step": 1202
+ },
+ {
+ "epoch": 0.6311647429171039,
+ "grad_norm": 0.16611474088307454,
+ "learning_rate": 1.8488672956016013e-05,
+ "loss": 0.4557701647281647,
+ "step": 1203
+ },
+ {
+ "epoch": 0.6316894018887723,
+ "grad_norm": 0.16759343079749697,
+ "learning_rate": 1.8454703525275813e-05,
+ "loss": 0.45069995522499084,
+ "step": 1204
+ },
+ {
+ "epoch": 0.6322140608604407,
+ "grad_norm": 0.16186676294485336,
+ "learning_rate": 1.8420758668581255e-05,
+ "loss": 0.44954800605773926,
+ "step": 1205
+ },
+ {
+ "epoch": 0.6327387198321092,
+ "grad_norm": 0.15687159835368553,
+ "learning_rate": 1.8386838478152916e-05,
+ "loss": 0.4490429759025574,
+ "step": 1206
+ },
+ {
+ "epoch": 0.6332633788037776,
+ "grad_norm": 0.16242951976115577,
+ "learning_rate": 1.8352943046144372e-05,
+ "loss": 0.45819586515426636,
+ "step": 1207
+ },
+ {
+ "epoch": 0.633788037775446,
+ "grad_norm": 0.16516087266517232,
+ "learning_rate": 1.831907246464191e-05,
+ "loss": 0.44769883155822754,
+ "step": 1208
+ },
+ {
+ "epoch": 0.6343126967471143,
+ "grad_norm": 0.16709475871512644,
+ "learning_rate": 1.8285226825664332e-05,
+ "loss": 0.4699128568172455,
+ "step": 1209
+ },
+ {
+ "epoch": 0.6348373557187827,
+ "grad_norm": 0.16288623150595774,
+ "learning_rate": 1.8251406221162636e-05,
+ "loss": 0.45322558283805847,
+ "step": 1210
+ },
+ {
+ "epoch": 0.6353620146904512,
+ "grad_norm": 0.1675722848054237,
+ "learning_rate": 1.821761074301986e-05,
+ "loss": 0.4752347469329834,
+ "step": 1211
+ },
+ {
+ "epoch": 0.6358866736621196,
+ "grad_norm": 0.163285963514504,
+ "learning_rate": 1.818384048305074e-05,
+ "loss": 0.441958487033844,
+ "step": 1212
+ },
+ {
+ "epoch": 0.636411332633788,
+ "grad_norm": 0.17078102287557764,
+ "learning_rate": 1.8150095533001517e-05,
+ "loss": 0.4551333487033844,
+ "step": 1213
+ },
+ {
+ "epoch": 0.6369359916054564,
+ "grad_norm": 0.16188540746097138,
+ "learning_rate": 1.811637598454966e-05,
+ "loss": 0.4470779597759247,
+ "step": 1214
+ },
+ {
+ "epoch": 0.6374606505771249,
+ "grad_norm": 0.15235628218958286,
+ "learning_rate": 1.808268192930364e-05,
+ "loss": 0.43604153394699097,
+ "step": 1215
+ },
+ {
+ "epoch": 0.6379853095487933,
+ "grad_norm": 0.1604601949959104,
+ "learning_rate": 1.8049013458802665e-05,
+ "loss": 0.4445713758468628,
+ "step": 1216
+ },
+ {
+ "epoch": 0.6385099685204617,
+ "grad_norm": 0.15511797131215901,
+ "learning_rate": 1.8015370664516427e-05,
+ "loss": 0.4455301761627197,
+ "step": 1217
+ },
+ {
+ "epoch": 0.6390346274921301,
+ "grad_norm": 0.16815669546944095,
+ "learning_rate": 1.798175363784487e-05,
+ "loss": 0.46135014295578003,
+ "step": 1218
+ },
+ {
+ "epoch": 0.6395592864637986,
+ "grad_norm": 0.161558177688922,
+ "learning_rate": 1.7948162470117917e-05,
+ "loss": 0.4519304037094116,
+ "step": 1219
+ },
+ {
+ "epoch": 0.640083945435467,
+ "grad_norm": 0.16677842130048232,
+ "learning_rate": 1.7914597252595265e-05,
+ "loss": 0.4550550878047943,
+ "step": 1220
+ },
+ {
+ "epoch": 0.6406086044071354,
+ "grad_norm": 0.1597331480423138,
+ "learning_rate": 1.7881058076466095e-05,
+ "loss": 0.441120445728302,
+ "step": 1221
+ },
+ {
+ "epoch": 0.6411332633788038,
+ "grad_norm": 0.15951106545246824,
+ "learning_rate": 1.7847545032848844e-05,
+ "loss": 0.4471382200717926,
+ "step": 1222
+ },
+ {
+ "epoch": 0.6416579223504721,
+ "grad_norm": 0.16157787674667565,
+ "learning_rate": 1.7814058212790947e-05,
+ "loss": 0.4550648331642151,
+ "step": 1223
+ },
+ {
+ "epoch": 0.6421825813221406,
+ "grad_norm": 0.16989186763352415,
+ "learning_rate": 1.7780597707268583e-05,
+ "loss": 0.4515089988708496,
+ "step": 1224
+ },
+ {
+ "epoch": 0.642707240293809,
+ "grad_norm": 0.15952153223099508,
+ "learning_rate": 1.7747163607186468e-05,
+ "loss": 0.4658638834953308,
+ "step": 1225
+ },
+ {
+ "epoch": 0.6432318992654774,
+ "grad_norm": 0.15205926699647218,
+ "learning_rate": 1.7713756003377567e-05,
+ "loss": 0.4330868422985077,
+ "step": 1226
+ },
+ {
+ "epoch": 0.6437565582371458,
+ "grad_norm": 0.15784806480936794,
+ "learning_rate": 1.7680374986602865e-05,
+ "loss": 0.4506068527698517,
+ "step": 1227
+ },
+ {
+ "epoch": 0.6442812172088143,
+ "grad_norm": 0.16006709280686687,
+ "learning_rate": 1.7647020647551105e-05,
+ "loss": 0.446941077709198,
+ "step": 1228
+ },
+ {
+ "epoch": 0.6448058761804827,
+ "grad_norm": 0.16064631158222062,
+ "learning_rate": 1.761369307683856e-05,
+ "loss": 0.4418525695800781,
+ "step": 1229
+ },
+ {
+ "epoch": 0.6453305351521511,
+ "grad_norm": 0.16454136480444392,
+ "learning_rate": 1.7580392365008767e-05,
+ "loss": 0.46268391609191895,
+ "step": 1230
+ },
+ {
+ "epoch": 0.6458551941238195,
+ "grad_norm": 0.15401246535771462,
+ "learning_rate": 1.754711860253233e-05,
+ "loss": 0.4372989535331726,
+ "step": 1231
+ },
+ {
+ "epoch": 0.646379853095488,
+ "grad_norm": 0.1544567975028383,
+ "learning_rate": 1.7513871879806603e-05,
+ "loss": 0.44806575775146484,
+ "step": 1232
+ },
+ {
+ "epoch": 0.6469045120671564,
+ "grad_norm": 0.16411225112730965,
+ "learning_rate": 1.748065228715549e-05,
+ "loss": 0.43782058358192444,
+ "step": 1233
+ },
+ {
+ "epoch": 0.6474291710388248,
+ "grad_norm": 0.1552767832972093,
+ "learning_rate": 1.7447459914829183e-05,
+ "loss": 0.44531378149986267,
+ "step": 1234
+ },
+ {
+ "epoch": 0.6479538300104932,
+ "grad_norm": 0.16214695257123513,
+ "learning_rate": 1.741429485300392e-05,
+ "loss": 0.47399842739105225,
+ "step": 1235
+ },
+ {
+ "epoch": 0.6484784889821616,
+ "grad_norm": 0.16684861111484112,
+ "learning_rate": 1.7381157191781766e-05,
+ "loss": 0.4654282331466675,
+ "step": 1236
+ },
+ {
+ "epoch": 0.64900314795383,
+ "grad_norm": 0.15901864178630273,
+ "learning_rate": 1.7348047021190334e-05,
+ "loss": 0.45128414034843445,
+ "step": 1237
+ },
+ {
+ "epoch": 0.6495278069254984,
+ "grad_norm": 0.15360991961742707,
+ "learning_rate": 1.7314964431182523e-05,
+ "loss": 0.4303375482559204,
+ "step": 1238
+ },
+ {
+ "epoch": 0.6500524658971668,
+ "grad_norm": 0.1633151574385741,
+ "learning_rate": 1.728190951163633e-05,
+ "loss": 0.45568403601646423,
+ "step": 1239
+ },
+ {
+ "epoch": 0.6505771248688352,
+ "grad_norm": 0.16377853571058645,
+ "learning_rate": 1.724888235235457e-05,
+ "loss": 0.46503543853759766,
+ "step": 1240
+ },
+ {
+ "epoch": 0.6511017838405037,
+ "grad_norm": 0.1595023801275368,
+ "learning_rate": 1.7215883043064655e-05,
+ "loss": 0.46337294578552246,
+ "step": 1241
+ },
+ {
+ "epoch": 0.6516264428121721,
+ "grad_norm": 0.16324796182212292,
+ "learning_rate": 1.7182911673418313e-05,
+ "loss": 0.46197354793548584,
+ "step": 1242
+ },
+ {
+ "epoch": 0.6521511017838405,
+ "grad_norm": 0.16353734830879765,
+ "learning_rate": 1.714996833299138e-05,
+ "loss": 0.45741671323776245,
+ "step": 1243
+ },
+ {
+ "epoch": 0.6526757607555089,
+ "grad_norm": 0.15941587678263386,
+ "learning_rate": 1.7117053111283528e-05,
+ "loss": 0.4514503479003906,
+ "step": 1244
+ },
+ {
+ "epoch": 0.6532004197271774,
+ "grad_norm": 0.1642653230145008,
+ "learning_rate": 1.708416609771804e-05,
+ "loss": 0.4437655210494995,
+ "step": 1245
+ },
+ {
+ "epoch": 0.6537250786988458,
+ "grad_norm": 0.15814155519130152,
+ "learning_rate": 1.7051307381641587e-05,
+ "loss": 0.4314163327217102,
+ "step": 1246
+ },
+ {
+ "epoch": 0.6542497376705142,
+ "grad_norm": 0.16086078370768056,
+ "learning_rate": 1.701847705232394e-05,
+ "loss": 0.4526430666446686,
+ "step": 1247
+ },
+ {
+ "epoch": 0.6547743966421826,
+ "grad_norm": 0.16389360190715183,
+ "learning_rate": 1.698567519895775e-05,
+ "loss": 0.4278966188430786,
+ "step": 1248
+ },
+ {
+ "epoch": 0.655299055613851,
+ "grad_norm": 0.16303142287804262,
+ "learning_rate": 1.6952901910658316e-05,
+ "loss": 0.45778220891952515,
+ "step": 1249
+ },
+ {
+ "epoch": 0.6558237145855194,
+ "grad_norm": 0.1675209956904048,
+ "learning_rate": 1.6920157276463318e-05,
+ "loss": 0.4489403963088989,
+ "step": 1250
+ },
+ {
+ "epoch": 0.6563483735571878,
+ "grad_norm": 0.16228311551074323,
+ "learning_rate": 1.6887441385332604e-05,
+ "loss": 0.4534342885017395,
+ "step": 1251
+ },
+ {
+ "epoch": 0.6568730325288562,
+ "grad_norm": 0.16408957851046913,
+ "learning_rate": 1.6854754326147916e-05,
+ "loss": 0.4549208879470825,
+ "step": 1252
+ },
+ {
+ "epoch": 0.6573976915005246,
+ "grad_norm": 0.15967922111432845,
+ "learning_rate": 1.682209618771268e-05,
+ "loss": 0.4451141357421875,
+ "step": 1253
+ },
+ {
+ "epoch": 0.657922350472193,
+ "grad_norm": 0.16440698722204208,
+ "learning_rate": 1.678946705875176e-05,
+ "loss": 0.454200804233551,
+ "step": 1254
+ },
+ {
+ "epoch": 0.6584470094438615,
+ "grad_norm": 0.16251120184938497,
+ "learning_rate": 1.675686702791117e-05,
+ "loss": 0.4580027461051941,
+ "step": 1255
+ },
+ {
+ "epoch": 0.6589716684155299,
+ "grad_norm": 0.1643717386150457,
+ "learning_rate": 1.6724296183757924e-05,
+ "loss": 0.4422311782836914,
+ "step": 1256
+ },
+ {
+ "epoch": 0.6594963273871983,
+ "grad_norm": 0.1644586414378722,
+ "learning_rate": 1.66917546147797e-05,
+ "loss": 0.4593179225921631,
+ "step": 1257
+ },
+ {
+ "epoch": 0.6600209863588667,
+ "grad_norm": 0.16751074186967685,
+ "learning_rate": 1.6659242409384657e-05,
+ "loss": 0.457930326461792,
+ "step": 1258
+ },
+ {
+ "epoch": 0.6605456453305352,
+ "grad_norm": 0.15398418609155334,
+ "learning_rate": 1.6626759655901182e-05,
+ "loss": 0.44002899527549744,
+ "step": 1259
+ },
+ {
+ "epoch": 0.6610703043022036,
+ "grad_norm": 0.16328113175790226,
+ "learning_rate": 1.659430644257763e-05,
+ "loss": 0.4500438868999481,
+ "step": 1260
+ },
+ {
+ "epoch": 0.661594963273872,
+ "grad_norm": 0.1694244340615106,
+ "learning_rate": 1.656188285758214e-05,
+ "loss": 0.4818427264690399,
+ "step": 1261
+ },
+ {
+ "epoch": 0.6621196222455404,
+ "grad_norm": 0.1578320701754368,
+ "learning_rate": 1.652948898900232e-05,
+ "loss": 0.4492436647415161,
+ "step": 1262
+ },
+ {
+ "epoch": 0.6626442812172089,
+ "grad_norm": 0.16538865917828469,
+ "learning_rate": 1.6497124924845064e-05,
+ "loss": 0.45070338249206543,
+ "step": 1263
+ },
+ {
+ "epoch": 0.6631689401888772,
+ "grad_norm": 0.15670817848256852,
+ "learning_rate": 1.6464790753036296e-05,
+ "loss": 0.45439285039901733,
+ "step": 1264
+ },
+ {
+ "epoch": 0.6636935991605456,
+ "grad_norm": 0.16184786585965885,
+ "learning_rate": 1.64324865614207e-05,
+ "loss": 0.446522980928421,
+ "step": 1265
+ },
+ {
+ "epoch": 0.664218258132214,
+ "grad_norm": 0.15719726462251984,
+ "learning_rate": 1.6400212437761554e-05,
+ "loss": 0.4495280981063843,
+ "step": 1266
+ },
+ {
+ "epoch": 0.6647429171038824,
+ "grad_norm": 0.1680804414134405,
+ "learning_rate": 1.6367968469740416e-05,
+ "loss": 0.4652596712112427,
+ "step": 1267
+ },
+ {
+ "epoch": 0.6652675760755509,
+ "grad_norm": 0.15919541846683566,
+ "learning_rate": 1.6335754744956937e-05,
+ "loss": 0.43842798471450806,
+ "step": 1268
+ },
+ {
+ "epoch": 0.6657922350472193,
+ "grad_norm": 0.16322634187139418,
+ "learning_rate": 1.630357135092859e-05,
+ "loss": 0.45242664217948914,
+ "step": 1269
+ },
+ {
+ "epoch": 0.6663168940188877,
+ "grad_norm": 0.15644867734734783,
+ "learning_rate": 1.6271418375090445e-05,
+ "loss": 0.44801655411720276,
+ "step": 1270
+ },
+ {
+ "epoch": 0.6668415529905561,
+ "grad_norm": 0.1514717416124423,
+ "learning_rate": 1.623929590479496e-05,
+ "loss": 0.4506194293498993,
+ "step": 1271
+ },
+ {
+ "epoch": 0.6673662119622246,
+ "grad_norm": 0.1740345085495889,
+ "learning_rate": 1.6207204027311683e-05,
+ "loss": 0.4818177819252014,
+ "step": 1272
+ },
+ {
+ "epoch": 0.667890870933893,
+ "grad_norm": 0.15797701900765457,
+ "learning_rate": 1.6175142829827067e-05,
+ "loss": 0.4639382064342499,
+ "step": 1273
+ },
+ {
+ "epoch": 0.6684155299055614,
+ "grad_norm": 0.15663812316653875,
+ "learning_rate": 1.6143112399444198e-05,
+ "loss": 0.46217501163482666,
+ "step": 1274
+ },
+ {
+ "epoch": 0.6689401888772298,
+ "grad_norm": 0.15474018977043758,
+ "learning_rate": 1.6111112823182592e-05,
+ "loss": 0.43691688776016235,
+ "step": 1275
+ },
+ {
+ "epoch": 0.6694648478488983,
+ "grad_norm": 0.1615634942529431,
+ "learning_rate": 1.6079144187977923e-05,
+ "loss": 0.4412601590156555,
+ "step": 1276
+ },
+ {
+ "epoch": 0.6699895068205667,
+ "grad_norm": 0.15440022842754444,
+ "learning_rate": 1.604720658068184e-05,
+ "loss": 0.4523273706436157,
+ "step": 1277
+ },
+ {
+ "epoch": 0.670514165792235,
+ "grad_norm": 0.1634985665776242,
+ "learning_rate": 1.6015300088061664e-05,
+ "loss": 0.46264398097991943,
+ "step": 1278
+ },
+ {
+ "epoch": 0.6710388247639034,
+ "grad_norm": 0.15037886370689527,
+ "learning_rate": 1.598342479680018e-05,
+ "loss": 0.451465904712677,
+ "step": 1279
+ },
+ {
+ "epoch": 0.6715634837355718,
+ "grad_norm": 0.16895242437746613,
+ "learning_rate": 1.595158079349542e-05,
+ "loss": 0.4466145634651184,
+ "step": 1280
+ },
+ {
+ "epoch": 0.6720881427072403,
+ "grad_norm": 0.1629548326999642,
+ "learning_rate": 1.5919768164660425e-05,
+ "loss": 0.45415782928466797,
+ "step": 1281
+ },
+ {
+ "epoch": 0.6726128016789087,
+ "grad_norm": 0.1563872102268991,
+ "learning_rate": 1.588798699672298e-05,
+ "loss": 0.4386737048625946,
+ "step": 1282
+ },
+ {
+ "epoch": 0.6731374606505771,
+ "grad_norm": 0.16038172773344717,
+ "learning_rate": 1.5856237376025403e-05,
+ "loss": 0.446260929107666,
+ "step": 1283
+ },
+ {
+ "epoch": 0.6736621196222455,
+ "grad_norm": 0.16189780714407623,
+ "learning_rate": 1.582451938882431e-05,
+ "loss": 0.45458072423934937,
+ "step": 1284
+ },
+ {
+ "epoch": 0.674186778593914,
+ "grad_norm": 0.15204521785049285,
+ "learning_rate": 1.5792833121290367e-05,
+ "loss": 0.4425617754459381,
+ "step": 1285
+ },
+ {
+ "epoch": 0.6747114375655824,
+ "grad_norm": 0.16237376704792778,
+ "learning_rate": 1.5761178659508063e-05,
+ "loss": 0.44631800055503845,
+ "step": 1286
+ },
+ {
+ "epoch": 0.6752360965372508,
+ "grad_norm": 0.1582063618466069,
+ "learning_rate": 1.57295560894755e-05,
+ "loss": 0.44280534982681274,
+ "step": 1287
+ },
+ {
+ "epoch": 0.6757607555089192,
+ "grad_norm": 0.16587032349847394,
+ "learning_rate": 1.5697965497104116e-05,
+ "loss": 0.4495997130870819,
+ "step": 1288
+ },
+ {
+ "epoch": 0.6762854144805877,
+ "grad_norm": 0.1557269262755563,
+ "learning_rate": 1.566640696821848e-05,
+ "loss": 0.4493359625339508,
+ "step": 1289
+ },
+ {
+ "epoch": 0.6768100734522561,
+ "grad_norm": 0.16038490357012256,
+ "learning_rate": 1.5634880588556043e-05,
+ "loss": 0.4447544813156128,
+ "step": 1290
+ },
+ {
+ "epoch": 0.6773347324239245,
+ "grad_norm": 0.15355201908546406,
+ "learning_rate": 1.560338644376692e-05,
+ "loss": 0.4468432664871216,
+ "step": 1291
+ },
+ {
+ "epoch": 0.6778593913955928,
+ "grad_norm": 0.15077009515589776,
+ "learning_rate": 1.5571924619413668e-05,
+ "loss": 0.44959044456481934,
+ "step": 1292
+ },
+ {
+ "epoch": 0.6783840503672612,
+ "grad_norm": 0.1583038146613965,
+ "learning_rate": 1.5540495200971012e-05,
+ "loss": 0.42450425028800964,
+ "step": 1293
+ },
+ {
+ "epoch": 0.6789087093389297,
+ "grad_norm": 0.15950687190598087,
+ "learning_rate": 1.5509098273825635e-05,
+ "loss": 0.4527238607406616,
+ "step": 1294
+ },
+ {
+ "epoch": 0.6794333683105981,
+ "grad_norm": 0.1526262208847808,
+ "learning_rate": 1.5477733923275968e-05,
+ "loss": 0.4379724860191345,
+ "step": 1295
+ },
+ {
+ "epoch": 0.6799580272822665,
+ "grad_norm": 0.15541579650021778,
+ "learning_rate": 1.5446402234531934e-05,
+ "loss": 0.4462166130542755,
+ "step": 1296
+ },
+ {
+ "epoch": 0.6804826862539349,
+ "grad_norm": 0.15483930734313786,
+ "learning_rate": 1.5415103292714718e-05,
+ "loss": 0.4390607178211212,
+ "step": 1297
+ },
+ {
+ "epoch": 0.6810073452256034,
+ "grad_norm": 0.1578005731149154,
+ "learning_rate": 1.5383837182856532e-05,
+ "loss": 0.44462254643440247,
+ "step": 1298
+ },
+ {
+ "epoch": 0.6815320041972718,
+ "grad_norm": 0.15830183884958823,
+ "learning_rate": 1.5352603989900406e-05,
+ "loss": 0.4563537538051605,
+ "step": 1299
+ },
+ {
+ "epoch": 0.6820566631689402,
+ "grad_norm": 0.15605472163445522,
+ "learning_rate": 1.5321403798699923e-05,
+ "loss": 0.43852412700653076,
+ "step": 1300
+ },
+ {
+ "epoch": 0.6825813221406086,
+ "grad_norm": 0.15770470227141278,
+ "learning_rate": 1.5290236694019017e-05,
+ "loss": 0.4438081979751587,
+ "step": 1301
+ },
+ {
+ "epoch": 0.683105981112277,
+ "grad_norm": 0.16175339845012077,
+ "learning_rate": 1.5259102760531747e-05,
+ "loss": 0.46670734882354736,
+ "step": 1302
+ },
+ {
+ "epoch": 0.6836306400839455,
+ "grad_norm": 0.16359504716154477,
+ "learning_rate": 1.5228002082822036e-05,
+ "loss": 0.4590187668800354,
+ "step": 1303
+ },
+ {
+ "epoch": 0.6841552990556139,
+ "grad_norm": 0.15422879500254713,
+ "learning_rate": 1.5196934745383457e-05,
+ "loss": 0.45287445187568665,
+ "step": 1304
+ },
+ {
+ "epoch": 0.6846799580272823,
+ "grad_norm": 0.15929902143574254,
+ "learning_rate": 1.5165900832619015e-05,
+ "loss": 0.44286859035491943,
+ "step": 1305
+ },
+ {
+ "epoch": 0.6852046169989506,
+ "grad_norm": 0.14838868714016124,
+ "learning_rate": 1.5134900428840904e-05,
+ "loss": 0.45155346393585205,
+ "step": 1306
+ },
+ {
+ "epoch": 0.6857292759706191,
+ "grad_norm": 0.15494288640397672,
+ "learning_rate": 1.5103933618270283e-05,
+ "loss": 0.44010141491889954,
+ "step": 1307
+ },
+ {
+ "epoch": 0.6862539349422875,
+ "grad_norm": 0.1686262360598338,
+ "learning_rate": 1.5073000485037042e-05,
+ "loss": 0.446555495262146,
+ "step": 1308
+ },
+ {
+ "epoch": 0.6867785939139559,
+ "grad_norm": 0.16122110823982602,
+ "learning_rate": 1.5042101113179579e-05,
+ "loss": 0.44640374183654785,
+ "step": 1309
+ },
+ {
+ "epoch": 0.6873032528856243,
+ "grad_norm": 0.16849937797919867,
+ "learning_rate": 1.5011235586644572e-05,
+ "loss": 0.4465785026550293,
+ "step": 1310
+ },
+ {
+ "epoch": 0.6878279118572928,
+ "grad_norm": 0.16062971262800754,
+ "learning_rate": 1.498040398928674e-05,
+ "loss": 0.445321649312973,
+ "step": 1311
+ },
+ {
+ "epoch": 0.6883525708289612,
+ "grad_norm": 0.1608953185993034,
+ "learning_rate": 1.4949606404868638e-05,
+ "loss": 0.46518227458000183,
+ "step": 1312
+ },
+ {
+ "epoch": 0.6888772298006296,
+ "grad_norm": 0.15880255672830915,
+ "learning_rate": 1.491884291706041e-05,
+ "loss": 0.44582217931747437,
+ "step": 1313
+ },
+ {
+ "epoch": 0.689401888772298,
+ "grad_norm": 0.16715145200789822,
+ "learning_rate": 1.4888113609439563e-05,
+ "loss": 0.4435983896255493,
+ "step": 1314
+ },
+ {
+ "epoch": 0.6899265477439664,
+ "grad_norm": 0.15263740340694454,
+ "learning_rate": 1.485741856549075e-05,
+ "loss": 0.4405066967010498,
+ "step": 1315
+ },
+ {
+ "epoch": 0.6904512067156349,
+ "grad_norm": 0.16247034518813966,
+ "learning_rate": 1.4826757868605517e-05,
+ "loss": 0.4544215500354767,
+ "step": 1316
+ },
+ {
+ "epoch": 0.6909758656873033,
+ "grad_norm": 0.15649146705978162,
+ "learning_rate": 1.4796131602082142e-05,
+ "loss": 0.4454413056373596,
+ "step": 1317
+ },
+ {
+ "epoch": 0.6915005246589717,
+ "grad_norm": 0.16286718636058178,
+ "learning_rate": 1.4765539849125318e-05,
+ "loss": 0.4497183561325073,
+ "step": 1318
+ },
+ {
+ "epoch": 0.6920251836306401,
+ "grad_norm": 0.15812342085005582,
+ "learning_rate": 1.473498269284599e-05,
+ "loss": 0.445032000541687,
+ "step": 1319
+ },
+ {
+ "epoch": 0.6925498426023085,
+ "grad_norm": 0.1558101290511678,
+ "learning_rate": 1.4704460216261123e-05,
+ "loss": 0.45769619941711426,
+ "step": 1320
+ },
+ {
+ "epoch": 0.6930745015739769,
+ "grad_norm": 0.15642889059495207,
+ "learning_rate": 1.4673972502293428e-05,
+ "loss": 0.44788360595703125,
+ "step": 1321
+ },
+ {
+ "epoch": 0.6935991605456453,
+ "grad_norm": 0.15722513518113074,
+ "learning_rate": 1.4643519633771222e-05,
+ "loss": 0.449631929397583,
+ "step": 1322
+ },
+ {
+ "epoch": 0.6941238195173137,
+ "grad_norm": 0.15587633748081542,
+ "learning_rate": 1.4613101693428125e-05,
+ "loss": 0.4450187087059021,
+ "step": 1323
+ },
+ {
+ "epoch": 0.6946484784889821,
+ "grad_norm": 0.1577107074882479,
+ "learning_rate": 1.458271876390287e-05,
+ "loss": 0.45108890533447266,
+ "step": 1324
+ },
+ {
+ "epoch": 0.6951731374606506,
+ "grad_norm": 0.15649839486990696,
+ "learning_rate": 1.4552370927739073e-05,
+ "loss": 0.46430301666259766,
+ "step": 1325
+ },
+ {
+ "epoch": 0.695697796432319,
+ "grad_norm": 0.18573600420690636,
+ "learning_rate": 1.4522058267385008e-05,
+ "loss": 0.4701972007751465,
+ "step": 1326
+ },
+ {
+ "epoch": 0.6962224554039874,
+ "grad_norm": 0.16296497084037542,
+ "learning_rate": 1.4491780865193402e-05,
+ "loss": 0.4607710838317871,
+ "step": 1327
+ },
+ {
+ "epoch": 0.6967471143756558,
+ "grad_norm": 0.161124007217343,
+ "learning_rate": 1.446153880342117e-05,
+ "loss": 0.45095115900039673,
+ "step": 1328
+ },
+ {
+ "epoch": 0.6972717733473243,
+ "grad_norm": 0.1537884039166766,
+ "learning_rate": 1.4431332164229222e-05,
+ "loss": 0.424068808555603,
+ "step": 1329
+ },
+ {
+ "epoch": 0.6977964323189927,
+ "grad_norm": 0.16588834503456087,
+ "learning_rate": 1.4401161029682237e-05,
+ "loss": 0.4622839093208313,
+ "step": 1330
+ },
+ {
+ "epoch": 0.6983210912906611,
+ "grad_norm": 0.16100510604015544,
+ "learning_rate": 1.4371025481748435e-05,
+ "loss": 0.4660705626010895,
+ "step": 1331
+ },
+ {
+ "epoch": 0.6988457502623295,
+ "grad_norm": 0.15559888940590075,
+ "learning_rate": 1.4340925602299337e-05,
+ "loss": 0.4484976530075073,
+ "step": 1332
+ },
+ {
+ "epoch": 0.6993704092339978,
+ "grad_norm": 0.15503821966213488,
+ "learning_rate": 1.4310861473109616e-05,
+ "loss": 0.45636600255966187,
+ "step": 1333
+ },
+ {
+ "epoch": 0.6998950682056663,
+ "grad_norm": 0.15539074896109334,
+ "learning_rate": 1.4280833175856743e-05,
+ "loss": 0.45863068103790283,
+ "step": 1334
+ },
+ {
+ "epoch": 0.7004197271773347,
+ "grad_norm": 0.15664821009154872,
+ "learning_rate": 1.4250840792120896e-05,
+ "loss": 0.44043463468551636,
+ "step": 1335
+ },
+ {
+ "epoch": 0.7009443861490031,
+ "grad_norm": 0.15825478808520285,
+ "learning_rate": 1.4220884403384655e-05,
+ "loss": 0.46522051095962524,
+ "step": 1336
+ },
+ {
+ "epoch": 0.7014690451206715,
+ "grad_norm": 0.16004268568141544,
+ "learning_rate": 1.4190964091032843e-05,
+ "loss": 0.46114686131477356,
+ "step": 1337
+ },
+ {
+ "epoch": 0.70199370409234,
+ "grad_norm": 0.16413139949945635,
+ "learning_rate": 1.4161079936352234e-05,
+ "loss": 0.4457515478134155,
+ "step": 1338
+ },
+ {
+ "epoch": 0.7025183630640084,
+ "grad_norm": 0.1538890577972702,
+ "learning_rate": 1.4131232020531398e-05,
+ "loss": 0.4616093337535858,
+ "step": 1339
+ },
+ {
+ "epoch": 0.7030430220356768,
+ "grad_norm": 0.16240737854323284,
+ "learning_rate": 1.4101420424660434e-05,
+ "loss": 0.44631537795066833,
+ "step": 1340
+ },
+ {
+ "epoch": 0.7035676810073452,
+ "grad_norm": 0.15984708603406073,
+ "learning_rate": 1.4071645229730762e-05,
+ "loss": 0.4366717040538788,
+ "step": 1341
+ },
+ {
+ "epoch": 0.7040923399790137,
+ "grad_norm": 0.15149695695635904,
+ "learning_rate": 1.4041906516634946e-05,
+ "loss": 0.44129428267478943,
+ "step": 1342
+ },
+ {
+ "epoch": 0.7046169989506821,
+ "grad_norm": 0.15041167501768415,
+ "learning_rate": 1.4012204366166399e-05,
+ "loss": 0.43654775619506836,
+ "step": 1343
+ },
+ {
+ "epoch": 0.7051416579223505,
+ "grad_norm": 0.1596097627963756,
+ "learning_rate": 1.3982538859019217e-05,
+ "loss": 0.4652445614337921,
+ "step": 1344
+ },
+ {
+ "epoch": 0.7056663168940189,
+ "grad_norm": 0.15621704205027884,
+ "learning_rate": 1.3952910075787941e-05,
+ "loss": 0.44577041268348694,
+ "step": 1345
+ },
+ {
+ "epoch": 0.7061909758656874,
+ "grad_norm": 0.16073041098008783,
+ "learning_rate": 1.3923318096967341e-05,
+ "loss": 0.4391976594924927,
+ "step": 1346
+ },
+ {
+ "epoch": 0.7067156348373557,
+ "grad_norm": 0.14793853946372448,
+ "learning_rate": 1.3893763002952204e-05,
+ "loss": 0.4463799297809601,
+ "step": 1347
+ },
+ {
+ "epoch": 0.7072402938090241,
+ "grad_norm": 0.16176080358867354,
+ "learning_rate": 1.3864244874037097e-05,
+ "loss": 0.45183271169662476,
+ "step": 1348
+ },
+ {
+ "epoch": 0.7077649527806925,
+ "grad_norm": 0.16372771179026924,
+ "learning_rate": 1.3834763790416171e-05,
+ "loss": 0.44205576181411743,
+ "step": 1349
+ },
+ {
+ "epoch": 0.7082896117523609,
+ "grad_norm": 0.15135222516127125,
+ "learning_rate": 1.3805319832182937e-05,
+ "loss": 0.4390261173248291,
+ "step": 1350
+ },
+ {
+ "epoch": 0.7088142707240294,
+ "grad_norm": 0.15493183225792365,
+ "learning_rate": 1.377591307933002e-05,
+ "loss": 0.45837974548339844,
+ "step": 1351
+ },
+ {
+ "epoch": 0.7093389296956978,
+ "grad_norm": 0.16405906906516504,
+ "learning_rate": 1.3746543611749008e-05,
+ "loss": 0.45032668113708496,
+ "step": 1352
+ },
+ {
+ "epoch": 0.7098635886673662,
+ "grad_norm": 0.15725677976524505,
+ "learning_rate": 1.3717211509230155e-05,
+ "loss": 0.4486224949359894,
+ "step": 1353
+ },
+ {
+ "epoch": 0.7103882476390346,
+ "grad_norm": 0.15781469066983164,
+ "learning_rate": 1.3687916851462227e-05,
+ "loss": 0.4502638578414917,
+ "step": 1354
+ },
+ {
+ "epoch": 0.710912906610703,
+ "grad_norm": 0.16115007729054298,
+ "learning_rate": 1.3658659718032241e-05,
+ "loss": 0.4508914351463318,
+ "step": 1355
+ },
+ {
+ "epoch": 0.7114375655823715,
+ "grad_norm": 0.16755060475968744,
+ "learning_rate": 1.3629440188425286e-05,
+ "loss": 0.45653945207595825,
+ "step": 1356
+ },
+ {
+ "epoch": 0.7119622245540399,
+ "grad_norm": 0.1574994294756123,
+ "learning_rate": 1.3600258342024268e-05,
+ "loss": 0.4349849820137024,
+ "step": 1357
+ },
+ {
+ "epoch": 0.7124868835257083,
+ "grad_norm": 0.16104440922657715,
+ "learning_rate": 1.3571114258109758e-05,
+ "loss": 0.4501233696937561,
+ "step": 1358
+ },
+ {
+ "epoch": 0.7130115424973767,
+ "grad_norm": 0.15163184290939022,
+ "learning_rate": 1.3542008015859691e-05,
+ "loss": 0.44171035289764404,
+ "step": 1359
+ },
+ {
+ "epoch": 0.7135362014690452,
+ "grad_norm": 0.15955896674456582,
+ "learning_rate": 1.3512939694349211e-05,
+ "loss": 0.4558671712875366,
+ "step": 1360
+ },
+ {
+ "epoch": 0.7140608604407135,
+ "grad_norm": 0.1589376751536802,
+ "learning_rate": 1.3483909372550446e-05,
+ "loss": 0.4430835247039795,
+ "step": 1361
+ },
+ {
+ "epoch": 0.7145855194123819,
+ "grad_norm": 0.15979088397822155,
+ "learning_rate": 1.3454917129332279e-05,
+ "loss": 0.4426245391368866,
+ "step": 1362
+ },
+ {
+ "epoch": 0.7151101783840503,
+ "grad_norm": 0.15493179714072167,
+ "learning_rate": 1.3425963043460143e-05,
+ "loss": 0.4550474286079407,
+ "step": 1363
+ },
+ {
+ "epoch": 0.7156348373557188,
+ "grad_norm": 0.15396498968995032,
+ "learning_rate": 1.3397047193595808e-05,
+ "loss": 0.44220536947250366,
+ "step": 1364
+ },
+ {
+ "epoch": 0.7161594963273872,
+ "grad_norm": 0.1489346822667273,
+ "learning_rate": 1.336816965829717e-05,
+ "loss": 0.4437323212623596,
+ "step": 1365
+ },
+ {
+ "epoch": 0.7166841552990556,
+ "grad_norm": 0.16311964931734477,
+ "learning_rate": 1.333933051601802e-05,
+ "loss": 0.45308777689933777,
+ "step": 1366
+ },
+ {
+ "epoch": 0.717208814270724,
+ "grad_norm": 0.1534354619021179,
+ "learning_rate": 1.331052984510785e-05,
+ "loss": 0.4334886372089386,
+ "step": 1367
+ },
+ {
+ "epoch": 0.7177334732423925,
+ "grad_norm": 0.15541089758065404,
+ "learning_rate": 1.328176772381165e-05,
+ "loss": 0.45248323678970337,
+ "step": 1368
+ },
+ {
+ "epoch": 0.7182581322140609,
+ "grad_norm": 0.1617921305700596,
+ "learning_rate": 1.3253044230269651e-05,
+ "loss": 0.44092947244644165,
+ "step": 1369
+ },
+ {
+ "epoch": 0.7187827911857293,
+ "grad_norm": 0.15047353793760151,
+ "learning_rate": 1.3224359442517165e-05,
+ "loss": 0.4424414336681366,
+ "step": 1370
+ },
+ {
+ "epoch": 0.7193074501573977,
+ "grad_norm": 0.15958089102258743,
+ "learning_rate": 1.319571343848433e-05,
+ "loss": 0.4458580017089844,
+ "step": 1371
+ },
+ {
+ "epoch": 0.7198321091290661,
+ "grad_norm": 0.15349258306968733,
+ "learning_rate": 1.3167106295995924e-05,
+ "loss": 0.4285386800765991,
+ "step": 1372
+ },
+ {
+ "epoch": 0.7203567681007346,
+ "grad_norm": 0.16042529563759886,
+ "learning_rate": 1.3138538092771164e-05,
+ "loss": 0.4497690200805664,
+ "step": 1373
+ },
+ {
+ "epoch": 0.720881427072403,
+ "grad_norm": 0.1565113736588754,
+ "learning_rate": 1.3110008906423454e-05,
+ "loss": 0.43562617897987366,
+ "step": 1374
+ },
+ {
+ "epoch": 0.7214060860440713,
+ "grad_norm": 0.15177291086871506,
+ "learning_rate": 1.3081518814460215e-05,
+ "loss": 0.44456908106803894,
+ "step": 1375
+ },
+ {
+ "epoch": 0.7219307450157397,
+ "grad_norm": 0.15928490480439225,
+ "learning_rate": 1.3053067894282622e-05,
+ "loss": 0.46170780062675476,
+ "step": 1376
+ },
+ {
+ "epoch": 0.7224554039874082,
+ "grad_norm": 0.1554814469217892,
+ "learning_rate": 1.302465622318549e-05,
+ "loss": 0.4422169327735901,
+ "step": 1377
+ },
+ {
+ "epoch": 0.7229800629590766,
+ "grad_norm": 0.15375370214966386,
+ "learning_rate": 1.2996283878356947e-05,
+ "loss": 0.4564763903617859,
+ "step": 1378
+ },
+ {
+ "epoch": 0.723504721930745,
+ "grad_norm": 0.15942841969395802,
+ "learning_rate": 1.2967950936878311e-05,
+ "loss": 0.45121875405311584,
+ "step": 1379
+ },
+ {
+ "epoch": 0.7240293809024134,
+ "grad_norm": 0.15731308716774903,
+ "learning_rate": 1.2939657475723844e-05,
+ "loss": 0.4285027086734772,
+ "step": 1380
+ },
+ {
+ "epoch": 0.7245540398740818,
+ "grad_norm": 0.15997753613932147,
+ "learning_rate": 1.291140357176054e-05,
+ "loss": 0.4566386342048645,
+ "step": 1381
+ },
+ {
+ "epoch": 0.7250786988457503,
+ "grad_norm": 0.15333715267473727,
+ "learning_rate": 1.2883189301747927e-05,
+ "loss": 0.45093846321105957,
+ "step": 1382
+ },
+ {
+ "epoch": 0.7256033578174187,
+ "grad_norm": 0.1514510304709061,
+ "learning_rate": 1.2855014742337874e-05,
+ "loss": 0.44256967306137085,
+ "step": 1383
+ },
+ {
+ "epoch": 0.7261280167890871,
+ "grad_norm": 0.15624301298995746,
+ "learning_rate": 1.2826879970074346e-05,
+ "loss": 0.4361875653266907,
+ "step": 1384
+ },
+ {
+ "epoch": 0.7266526757607555,
+ "grad_norm": 0.15767850122006694,
+ "learning_rate": 1.279878506139322e-05,
+ "loss": 0.4477671980857849,
+ "step": 1385
+ },
+ {
+ "epoch": 0.727177334732424,
+ "grad_norm": 0.15537676969694064,
+ "learning_rate": 1.2770730092622072e-05,
+ "loss": 0.4339832663536072,
+ "step": 1386
+ },
+ {
+ "epoch": 0.7277019937040924,
+ "grad_norm": 0.15944113633281293,
+ "learning_rate": 1.2742715139979965e-05,
+ "loss": 0.45407935976982117,
+ "step": 1387
+ },
+ {
+ "epoch": 0.7282266526757608,
+ "grad_norm": 0.16763519518606218,
+ "learning_rate": 1.271474027957727e-05,
+ "loss": 0.46442392468452454,
+ "step": 1388
+ },
+ {
+ "epoch": 0.7287513116474291,
+ "grad_norm": 0.1564572774601902,
+ "learning_rate": 1.2686805587415423e-05,
+ "loss": 0.4463968276977539,
+ "step": 1389
+ },
+ {
+ "epoch": 0.7292759706190975,
+ "grad_norm": 0.16384626314986128,
+ "learning_rate": 1.2658911139386709e-05,
+ "loss": 0.45519885420799255,
+ "step": 1390
+ },
+ {
+ "epoch": 0.729800629590766,
+ "grad_norm": 0.1562023802820964,
+ "learning_rate": 1.2631057011274106e-05,
+ "loss": 0.4365222454071045,
+ "step": 1391
+ },
+ {
+ "epoch": 0.7303252885624344,
+ "grad_norm": 0.15597287232940926,
+ "learning_rate": 1.2603243278751042e-05,
+ "loss": 0.4328266680240631,
+ "step": 1392
+ },
+ {
+ "epoch": 0.7308499475341028,
+ "grad_norm": 0.14762833706132167,
+ "learning_rate": 1.2575470017381214e-05,
+ "loss": 0.43902426958084106,
+ "step": 1393
+ },
+ {
+ "epoch": 0.7313746065057712,
+ "grad_norm": 0.1688371515856315,
+ "learning_rate": 1.2547737302618351e-05,
+ "loss": 0.4649963676929474,
+ "step": 1394
+ },
+ {
+ "epoch": 0.7318992654774397,
+ "grad_norm": 0.15501514509586015,
+ "learning_rate": 1.2520045209806024e-05,
+ "loss": 0.4501558542251587,
+ "step": 1395
+ },
+ {
+ "epoch": 0.7324239244491081,
+ "grad_norm": 0.15516748917693537,
+ "learning_rate": 1.2492393814177457e-05,
+ "loss": 0.4405498206615448,
+ "step": 1396
+ },
+ {
+ "epoch": 0.7329485834207765,
+ "grad_norm": 0.17115400468199243,
+ "learning_rate": 1.2464783190855295e-05,
+ "loss": 0.4445379674434662,
+ "step": 1397
+ },
+ {
+ "epoch": 0.7334732423924449,
+ "grad_norm": 0.15966386341604186,
+ "learning_rate": 1.243721341485143e-05,
+ "loss": 0.4769889712333679,
+ "step": 1398
+ },
+ {
+ "epoch": 0.7339979013641134,
+ "grad_norm": 0.15614706359781472,
+ "learning_rate": 1.2409684561066766e-05,
+ "loss": 0.43643918633461,
+ "step": 1399
+ },
+ {
+ "epoch": 0.7345225603357818,
+ "grad_norm": 0.17467968960064226,
+ "learning_rate": 1.238219670429104e-05,
+ "loss": 0.45413005352020264,
+ "step": 1400
+ },
+ {
+ "epoch": 0.7350472193074502,
+ "grad_norm": 0.1544517430810819,
+ "learning_rate": 1.2354749919202598e-05,
+ "loss": 0.45218425989151,
+ "step": 1401
+ },
+ {
+ "epoch": 0.7355718782791185,
+ "grad_norm": 0.16551083673374484,
+ "learning_rate": 1.2327344280368217e-05,
+ "loss": 0.4459628462791443,
+ "step": 1402
+ },
+ {
+ "epoch": 0.7360965372507869,
+ "grad_norm": 0.17614490459613263,
+ "learning_rate": 1.2299979862242882e-05,
+ "loss": 0.4406280815601349,
+ "step": 1403
+ },
+ {
+ "epoch": 0.7366211962224554,
+ "grad_norm": 0.15554925394435032,
+ "learning_rate": 1.2272656739169588e-05,
+ "loss": 0.4380159378051758,
+ "step": 1404
+ },
+ {
+ "epoch": 0.7371458551941238,
+ "grad_norm": 0.16475466799870916,
+ "learning_rate": 1.2245374985379147e-05,
+ "loss": 0.4442812204360962,
+ "step": 1405
+ },
+ {
+ "epoch": 0.7376705141657922,
+ "grad_norm": 0.16877301168875541,
+ "learning_rate": 1.2218134674989978e-05,
+ "loss": 0.4589429199695587,
+ "step": 1406
+ },
+ {
+ "epoch": 0.7381951731374606,
+ "grad_norm": 0.1583278589457438,
+ "learning_rate": 1.2190935882007894e-05,
+ "loss": 0.44896745681762695,
+ "step": 1407
+ },
+ {
+ "epoch": 0.7387198321091291,
+ "grad_norm": 0.1582022541656192,
+ "learning_rate": 1.2163778680325947e-05,
+ "loss": 0.4367997646331787,
+ "step": 1408
+ },
+ {
+ "epoch": 0.7392444910807975,
+ "grad_norm": 0.1669573982300068,
+ "learning_rate": 1.2136663143724163e-05,
+ "loss": 0.4503386616706848,
+ "step": 1409
+ },
+ {
+ "epoch": 0.7397691500524659,
+ "grad_norm": 0.1637877523924539,
+ "learning_rate": 1.2109589345869387e-05,
+ "loss": 0.4660584330558777,
+ "step": 1410
+ },
+ {
+ "epoch": 0.7402938090241343,
+ "grad_norm": 0.1640136159943701,
+ "learning_rate": 1.2082557360315065e-05,
+ "loss": 0.45206528902053833,
+ "step": 1411
+ },
+ {
+ "epoch": 0.7408184679958028,
+ "grad_norm": 0.1596580168682554,
+ "learning_rate": 1.2055567260501054e-05,
+ "loss": 0.4405035376548767,
+ "step": 1412
+ },
+ {
+ "epoch": 0.7413431269674712,
+ "grad_norm": 0.15224511741945293,
+ "learning_rate": 1.2028619119753402e-05,
+ "loss": 0.42810291051864624,
+ "step": 1413
+ },
+ {
+ "epoch": 0.7418677859391396,
+ "grad_norm": 0.15631381931817098,
+ "learning_rate": 1.2001713011284193e-05,
+ "loss": 0.4421626925468445,
+ "step": 1414
+ },
+ {
+ "epoch": 0.742392444910808,
+ "grad_norm": 0.15721819981265503,
+ "learning_rate": 1.1974849008191288e-05,
+ "loss": 0.4540696144104004,
+ "step": 1415
+ },
+ {
+ "epoch": 0.7429171038824763,
+ "grad_norm": 0.15904478578744463,
+ "learning_rate": 1.1948027183458182e-05,
+ "loss": 0.4601830244064331,
+ "step": 1416
+ },
+ {
+ "epoch": 0.7434417628541448,
+ "grad_norm": 0.1530755605235123,
+ "learning_rate": 1.1921247609953747e-05,
+ "loss": 0.4552589952945709,
+ "step": 1417
+ },
+ {
+ "epoch": 0.7439664218258132,
+ "grad_norm": 0.15375751072258462,
+ "learning_rate": 1.1894510360432106e-05,
+ "loss": 0.42965373396873474,
+ "step": 1418
+ },
+ {
+ "epoch": 0.7444910807974816,
+ "grad_norm": 0.15678686023937333,
+ "learning_rate": 1.1867815507532377e-05,
+ "loss": 0.44684016704559326,
+ "step": 1419
+ },
+ {
+ "epoch": 0.74501573976915,
+ "grad_norm": 0.15486952983783606,
+ "learning_rate": 1.1841163123778502e-05,
+ "loss": 0.44415023922920227,
+ "step": 1420
+ },
+ {
+ "epoch": 0.7455403987408185,
+ "grad_norm": 0.15342611490545593,
+ "learning_rate": 1.1814553281579034e-05,
+ "loss": 0.4501304626464844,
+ "step": 1421
+ },
+ {
+ "epoch": 0.7460650577124869,
+ "grad_norm": 0.15374260012788166,
+ "learning_rate": 1.1787986053226954e-05,
+ "loss": 0.4521749019622803,
+ "step": 1422
+ },
+ {
+ "epoch": 0.7465897166841553,
+ "grad_norm": 0.1543553258370361,
+ "learning_rate": 1.176146151089949e-05,
+ "loss": 0.44753748178482056,
+ "step": 1423
+ },
+ {
+ "epoch": 0.7471143756558237,
+ "grad_norm": 0.15921670193848306,
+ "learning_rate": 1.1734979726657878e-05,
+ "loss": 0.4473114311695099,
+ "step": 1424
+ },
+ {
+ "epoch": 0.7476390346274921,
+ "grad_norm": 0.15023498686186612,
+ "learning_rate": 1.17085407724472e-05,
+ "loss": 0.4568778872489929,
+ "step": 1425
+ },
+ {
+ "epoch": 0.7481636935991606,
+ "grad_norm": 0.14601156920274583,
+ "learning_rate": 1.1682144720096168e-05,
+ "loss": 0.4389464855194092,
+ "step": 1426
+ },
+ {
+ "epoch": 0.748688352570829,
+ "grad_norm": 0.1520548095149398,
+ "learning_rate": 1.1655791641316954e-05,
+ "loss": 0.440733402967453,
+ "step": 1427
+ },
+ {
+ "epoch": 0.7492130115424974,
+ "grad_norm": 0.1591915163097993,
+ "learning_rate": 1.1629481607704963e-05,
+ "loss": 0.46245887875556946,
+ "step": 1428
+ },
+ {
+ "epoch": 0.7497376705141658,
+ "grad_norm": 0.14867295658447263,
+ "learning_rate": 1.1603214690738683e-05,
+ "loss": 0.4332764744758606,
+ "step": 1429
+ },
+ {
+ "epoch": 0.7502623294858342,
+ "grad_norm": 0.1498957762903134,
+ "learning_rate": 1.1576990961779447e-05,
+ "loss": 0.45215466618537903,
+ "step": 1430
+ },
+ {
+ "epoch": 0.7507869884575026,
+ "grad_norm": 0.15430555803491824,
+ "learning_rate": 1.1550810492071243e-05,
+ "loss": 0.4454396963119507,
+ "step": 1431
+ },
+ {
+ "epoch": 0.751311647429171,
+ "grad_norm": 0.1517461021936533,
+ "learning_rate": 1.1524673352740543e-05,
+ "loss": 0.43869835138320923,
+ "step": 1432
+ },
+ {
+ "epoch": 0.7518363064008394,
+ "grad_norm": 0.14541657994692936,
+ "learning_rate": 1.1498579614796121e-05,
+ "loss": 0.4373054504394531,
+ "step": 1433
+ },
+ {
+ "epoch": 0.7523609653725079,
+ "grad_norm": 0.15539513763786977,
+ "learning_rate": 1.1472529349128812e-05,
+ "loss": 0.45451128482818604,
+ "step": 1434
+ },
+ {
+ "epoch": 0.7528856243441763,
+ "grad_norm": 0.1482043113690946,
+ "learning_rate": 1.144652262651136e-05,
+ "loss": 0.43644773960113525,
+ "step": 1435
+ },
+ {
+ "epoch": 0.7534102833158447,
+ "grad_norm": 0.16293628094190815,
+ "learning_rate": 1.1420559517598205e-05,
+ "loss": 0.4687269330024719,
+ "step": 1436
+ },
+ {
+ "epoch": 0.7539349422875131,
+ "grad_norm": 0.1553393974390287,
+ "learning_rate": 1.1394640092925302e-05,
+ "loss": 0.4501398205757141,
+ "step": 1437
+ },
+ {
+ "epoch": 0.7544596012591815,
+ "grad_norm": 0.16214837694535802,
+ "learning_rate": 1.1368764422909919e-05,
+ "loss": 0.4542577862739563,
+ "step": 1438
+ },
+ {
+ "epoch": 0.75498426023085,
+ "grad_norm": 0.15838878372428297,
+ "learning_rate": 1.1342932577850473e-05,
+ "loss": 0.46228161454200745,
+ "step": 1439
+ },
+ {
+ "epoch": 0.7555089192025184,
+ "grad_norm": 0.15090419497122687,
+ "learning_rate": 1.13171446279263e-05,
+ "loss": 0.4495932161808014,
+ "step": 1440
+ },
+ {
+ "epoch": 0.7560335781741868,
+ "grad_norm": 0.1543376786458394,
+ "learning_rate": 1.1291400643197484e-05,
+ "loss": 0.451630175113678,
+ "step": 1441
+ },
+ {
+ "epoch": 0.7565582371458552,
+ "grad_norm": 0.1539470751610696,
+ "learning_rate": 1.1265700693604668e-05,
+ "loss": 0.447547972202301,
+ "step": 1442
+ },
+ {
+ "epoch": 0.7570828961175237,
+ "grad_norm": 0.15761047703305445,
+ "learning_rate": 1.1240044848968856e-05,
+ "loss": 0.4516204595565796,
+ "step": 1443
+ },
+ {
+ "epoch": 0.757607555089192,
+ "grad_norm": 0.15675279683704735,
+ "learning_rate": 1.1214433178991258e-05,
+ "loss": 0.4597971737384796,
+ "step": 1444
+ },
+ {
+ "epoch": 0.7581322140608604,
+ "grad_norm": 0.15592110994904113,
+ "learning_rate": 1.1188865753253022e-05,
+ "loss": 0.443898469209671,
+ "step": 1445
+ },
+ {
+ "epoch": 0.7586568730325288,
+ "grad_norm": 0.15597501691030619,
+ "learning_rate": 1.116334264121513e-05,
+ "loss": 0.4581306278705597,
+ "step": 1446
+ },
+ {
+ "epoch": 0.7591815320041972,
+ "grad_norm": 0.15552304728589958,
+ "learning_rate": 1.1137863912218154e-05,
+ "loss": 0.4475254416465759,
+ "step": 1447
+ },
+ {
+ "epoch": 0.7597061909758657,
+ "grad_norm": 0.15759990123898035,
+ "learning_rate": 1.1112429635482118e-05,
+ "loss": 0.43585264682769775,
+ "step": 1448
+ },
+ {
+ "epoch": 0.7602308499475341,
+ "grad_norm": 0.15964657264676635,
+ "learning_rate": 1.1087039880106247e-05,
+ "loss": 0.46786290407180786,
+ "step": 1449
+ },
+ {
+ "epoch": 0.7607555089192025,
+ "grad_norm": 0.15155804051218852,
+ "learning_rate": 1.106169471506882e-05,
+ "loss": 0.42292726039886475,
+ "step": 1450
+ },
+ {
+ "epoch": 0.7612801678908709,
+ "grad_norm": 0.15519270966930884,
+ "learning_rate": 1.103639420922698e-05,
+ "loss": 0.4248972237110138,
+ "step": 1451
+ },
+ {
+ "epoch": 0.7618048268625394,
+ "grad_norm": 0.1526330511128821,
+ "learning_rate": 1.1011138431316537e-05,
+ "loss": 0.4317784309387207,
+ "step": 1452
+ },
+ {
+ "epoch": 0.7623294858342078,
+ "grad_norm": 0.15797682979272226,
+ "learning_rate": 1.0985927449951779e-05,
+ "loss": 0.44506698846817017,
+ "step": 1453
+ },
+ {
+ "epoch": 0.7628541448058762,
+ "grad_norm": 0.15479979838505462,
+ "learning_rate": 1.096076133362531e-05,
+ "loss": 0.4471631348133087,
+ "step": 1454
+ },
+ {
+ "epoch": 0.7633788037775446,
+ "grad_norm": 0.15712267274447342,
+ "learning_rate": 1.093564015070783e-05,
+ "loss": 0.4321915805339813,
+ "step": 1455
+ },
+ {
+ "epoch": 0.7639034627492131,
+ "grad_norm": 0.15857431494180016,
+ "learning_rate": 1.0910563969447969e-05,
+ "loss": 0.44455766677856445,
+ "step": 1456
+ },
+ {
+ "epoch": 0.7644281217208815,
+ "grad_norm": 0.15377057080997789,
+ "learning_rate": 1.0885532857972097e-05,
+ "loss": 0.4567924439907074,
+ "step": 1457
+ },
+ {
+ "epoch": 0.7649527806925498,
+ "grad_norm": 0.15544844005009104,
+ "learning_rate": 1.0860546884284143e-05,
+ "loss": 0.4491572976112366,
+ "step": 1458
+ },
+ {
+ "epoch": 0.7654774396642182,
+ "grad_norm": 0.16210735566479478,
+ "learning_rate": 1.08356061162654e-05,
+ "loss": 0.43823331594467163,
+ "step": 1459
+ },
+ {
+ "epoch": 0.7660020986358866,
+ "grad_norm": 0.15213559486379452,
+ "learning_rate": 1.0810710621674355e-05,
+ "loss": 0.45375770330429077,
+ "step": 1460
+ },
+ {
+ "epoch": 0.7665267576075551,
+ "grad_norm": 0.1727832306402992,
+ "learning_rate": 1.0785860468146494e-05,
+ "loss": 0.4639572501182556,
+ "step": 1461
+ },
+ {
+ "epoch": 0.7670514165792235,
+ "grad_norm": 0.1637555902013663,
+ "learning_rate": 1.076105572319412e-05,
+ "loss": 0.4355923533439636,
+ "step": 1462
+ },
+ {
+ "epoch": 0.7675760755508919,
+ "grad_norm": 0.16646936030470866,
+ "learning_rate": 1.0736296454206166e-05,
+ "loss": 0.4650861322879791,
+ "step": 1463
+ },
+ {
+ "epoch": 0.7681007345225603,
+ "grad_norm": 0.14987620615431385,
+ "learning_rate": 1.0711582728448033e-05,
+ "loss": 0.45445430278778076,
+ "step": 1464
+ },
+ {
+ "epoch": 0.7686253934942288,
+ "grad_norm": 0.16428618403547932,
+ "learning_rate": 1.0686914613061377e-05,
+ "loss": 0.46050935983657837,
+ "step": 1465
+ },
+ {
+ "epoch": 0.7691500524658972,
+ "grad_norm": 0.15052999053128632,
+ "learning_rate": 1.0662292175063949e-05,
+ "loss": 0.43385741114616394,
+ "step": 1466
+ },
+ {
+ "epoch": 0.7696747114375656,
+ "grad_norm": 0.1654949402740152,
+ "learning_rate": 1.0637715481349397e-05,
+ "loss": 0.4434410333633423,
+ "step": 1467
+ },
+ {
+ "epoch": 0.770199370409234,
+ "grad_norm": 0.15923836508490036,
+ "learning_rate": 1.0613184598687088e-05,
+ "loss": 0.4410185217857361,
+ "step": 1468
+ },
+ {
+ "epoch": 0.7707240293809025,
+ "grad_norm": 0.15813007111170696,
+ "learning_rate": 1.0588699593721957e-05,
+ "loss": 0.43588417768478394,
+ "step": 1469
+ },
+ {
+ "epoch": 0.7712486883525709,
+ "grad_norm": 0.16978817898431764,
+ "learning_rate": 1.0564260532974274e-05,
+ "loss": 0.4514843225479126,
+ "step": 1470
+ },
+ {
+ "epoch": 0.7717733473242392,
+ "grad_norm": 0.1613844852697828,
+ "learning_rate": 1.0539867482839499e-05,
+ "loss": 0.4554520845413208,
+ "step": 1471
+ },
+ {
+ "epoch": 0.7722980062959076,
+ "grad_norm": 0.1587788181984633,
+ "learning_rate": 1.0515520509588075e-05,
+ "loss": 0.43200671672821045,
+ "step": 1472
+ },
+ {
+ "epoch": 0.772822665267576,
+ "grad_norm": 0.1539930824827353,
+ "learning_rate": 1.0491219679365283e-05,
+ "loss": 0.4343116283416748,
+ "step": 1473
+ },
+ {
+ "epoch": 0.7733473242392445,
+ "grad_norm": 0.1561612537773709,
+ "learning_rate": 1.0466965058191049e-05,
+ "loss": 0.4477815330028534,
+ "step": 1474
+ },
+ {
+ "epoch": 0.7738719832109129,
+ "grad_norm": 0.169990777691657,
+ "learning_rate": 1.0442756711959745e-05,
+ "loss": 0.4476611614227295,
+ "step": 1475
+ },
+ {
+ "epoch": 0.7743966421825813,
+ "grad_norm": 0.1624313049193793,
+ "learning_rate": 1.041859470644002e-05,
+ "loss": 0.47136399149894714,
+ "step": 1476
+ },
+ {
+ "epoch": 0.7749213011542497,
+ "grad_norm": 0.15993315458994328,
+ "learning_rate": 1.0394479107274645e-05,
+ "loss": 0.4672670364379883,
+ "step": 1477
+ },
+ {
+ "epoch": 0.7754459601259182,
+ "grad_norm": 0.16755349272347217,
+ "learning_rate": 1.0370409979980294e-05,
+ "loss": 0.45135706663131714,
+ "step": 1478
+ },
+ {
+ "epoch": 0.7759706190975866,
+ "grad_norm": 0.15665059815830362,
+ "learning_rate": 1.0346387389947407e-05,
+ "loss": 0.45178359746932983,
+ "step": 1479
+ },
+ {
+ "epoch": 0.776495278069255,
+ "grad_norm": 0.1554612731452571,
+ "learning_rate": 1.0322411402439983e-05,
+ "loss": 0.4412590265274048,
+ "step": 1480
+ },
+ {
+ "epoch": 0.7770199370409234,
+ "grad_norm": 0.1532496869312065,
+ "learning_rate": 1.0298482082595408e-05,
+ "loss": 0.44104987382888794,
+ "step": 1481
+ },
+ {
+ "epoch": 0.7775445960125918,
+ "grad_norm": 0.15065868316960426,
+ "learning_rate": 1.0274599495424286e-05,
+ "loss": 0.4336603283882141,
+ "step": 1482
+ },
+ {
+ "epoch": 0.7780692549842603,
+ "grad_norm": 0.1548373981760097,
+ "learning_rate": 1.0250763705810267e-05,
+ "loss": 0.4658375680446625,
+ "step": 1483
+ },
+ {
+ "epoch": 0.7785939139559287,
+ "grad_norm": 0.15470181175866582,
+ "learning_rate": 1.0226974778509846e-05,
+ "loss": 0.45232075452804565,
+ "step": 1484
+ },
+ {
+ "epoch": 0.779118572927597,
+ "grad_norm": 0.15507371187727106,
+ "learning_rate": 1.0203232778152228e-05,
+ "loss": 0.43640458583831787,
+ "step": 1485
+ },
+ {
+ "epoch": 0.7796432318992654,
+ "grad_norm": 0.15269440736544085,
+ "learning_rate": 1.01795377692391e-05,
+ "loss": 0.4400939345359802,
+ "step": 1486
+ },
+ {
+ "epoch": 0.7801678908709339,
+ "grad_norm": 0.15610899550395085,
+ "learning_rate": 1.0155889816144495e-05,
+ "loss": 0.4546602666378021,
+ "step": 1487
+ },
+ {
+ "epoch": 0.7806925498426023,
+ "grad_norm": 0.1559328242601668,
+ "learning_rate": 1.0132288983114611e-05,
+ "loss": 0.45510709285736084,
+ "step": 1488
+ },
+ {
+ "epoch": 0.7812172088142707,
+ "grad_norm": 0.1537255194600787,
+ "learning_rate": 1.0108735334267639e-05,
+ "loss": 0.4475342631340027,
+ "step": 1489
+ },
+ {
+ "epoch": 0.7817418677859391,
+ "grad_norm": 0.14616575061664402,
+ "learning_rate": 1.0085228933593558e-05,
+ "loss": 0.4378190040588379,
+ "step": 1490
+ },
+ {
+ "epoch": 0.7822665267576075,
+ "grad_norm": 0.1624360967746598,
+ "learning_rate": 1.0061769844953999e-05,
+ "loss": 0.4439539909362793,
+ "step": 1491
+ },
+ {
+ "epoch": 0.782791185729276,
+ "grad_norm": 0.15407336125298413,
+ "learning_rate": 1.003835813208206e-05,
+ "loss": 0.44074907898902893,
+ "step": 1492
+ },
+ {
+ "epoch": 0.7833158447009444,
+ "grad_norm": 0.15786980710136125,
+ "learning_rate": 1.001499385858211e-05,
+ "loss": 0.4491667151451111,
+ "step": 1493
+ },
+ {
+ "epoch": 0.7838405036726128,
+ "grad_norm": 0.15338836738460088,
+ "learning_rate": 9.991677087929665e-06,
+ "loss": 0.446003794670105,
+ "step": 1494
+ },
+ {
+ "epoch": 0.7843651626442812,
+ "grad_norm": 0.156710894892,
+ "learning_rate": 9.968407883471165e-06,
+ "loss": 0.43604689836502075,
+ "step": 1495
+ },
+ {
+ "epoch": 0.7848898216159497,
+ "grad_norm": 0.16100539063191666,
+ "learning_rate": 9.945186308423832e-06,
+ "loss": 0.4431147873401642,
+ "step": 1496
+ },
+ {
+ "epoch": 0.7854144805876181,
+ "grad_norm": 0.15776018851122997,
+ "learning_rate": 9.92201242587548e-06,
+ "loss": 0.4532504975795746,
+ "step": 1497
+ },
+ {
+ "epoch": 0.7859391395592865,
+ "grad_norm": 0.1540333423540934,
+ "learning_rate": 9.898886298784368e-06,
+ "loss": 0.43605154752731323,
+ "step": 1498
+ },
+ {
+ "epoch": 0.7864637985309548,
+ "grad_norm": 0.16009602938972295,
+ "learning_rate": 9.875807989979005e-06,
+ "loss": 0.4605238437652588,
+ "step": 1499
+ },
+ {
+ "epoch": 0.7869884575026233,
+ "grad_norm": 0.15639368804171708,
+ "learning_rate": 9.852777562157983e-06,
+ "loss": 0.4509708881378174,
+ "step": 1500
+ },
+ {
+ "epoch": 0.7875131164742917,
+ "grad_norm": 0.15074011188741518,
+ "learning_rate": 9.829795077889828e-06,
+ "loss": 0.4527304172515869,
+ "step": 1501
+ },
+ {
+ "epoch": 0.7880377754459601,
+ "grad_norm": 0.15155083996490956,
+ "learning_rate": 9.806860599612799e-06,
+ "loss": 0.4306413531303406,
+ "step": 1502
+ },
+ {
+ "epoch": 0.7885624344176285,
+ "grad_norm": 0.1478956630795445,
+ "learning_rate": 9.783974189634734e-06,
+ "loss": 0.4434714913368225,
+ "step": 1503
+ },
+ {
+ "epoch": 0.789087093389297,
+ "grad_norm": 0.15327668434806432,
+ "learning_rate": 9.761135910132902e-06,
+ "loss": 0.44431227445602417,
+ "step": 1504
+ },
+ {
+ "epoch": 0.7896117523609654,
+ "grad_norm": 0.2243177585497212,
+ "learning_rate": 9.738345823153783e-06,
+ "loss": 0.4541007876396179,
+ "step": 1505
+ },
+ {
+ "epoch": 0.7901364113326338,
+ "grad_norm": 0.1600239865812282,
+ "learning_rate": 9.715603990612945e-06,
+ "loss": 0.4456910490989685,
+ "step": 1506
+ },
+ {
+ "epoch": 0.7906610703043022,
+ "grad_norm": 0.15304634075332066,
+ "learning_rate": 9.692910474294858e-06,
+ "loss": 0.43965545296669006,
+ "step": 1507
+ },
+ {
+ "epoch": 0.7911857292759706,
+ "grad_norm": 0.15657752062053032,
+ "learning_rate": 9.670265335852725e-06,
+ "loss": 0.43358081579208374,
+ "step": 1508
+ },
+ {
+ "epoch": 0.7917103882476391,
+ "grad_norm": 0.16110057783374115,
+ "learning_rate": 9.647668636808304e-06,
+ "loss": 0.45711368322372437,
+ "step": 1509
+ },
+ {
+ "epoch": 0.7922350472193075,
+ "grad_norm": 0.1551133690299623,
+ "learning_rate": 9.625120438551791e-06,
+ "loss": 0.4613519608974457,
+ "step": 1510
+ },
+ {
+ "epoch": 0.7927597061909759,
+ "grad_norm": 0.15690897380725946,
+ "learning_rate": 9.602620802341581e-06,
+ "loss": 0.4472261071205139,
+ "step": 1511
+ },
+ {
+ "epoch": 0.7932843651626443,
+ "grad_norm": 0.147796599322911,
+ "learning_rate": 9.580169789304152e-06,
+ "loss": 0.4447203576564789,
+ "step": 1512
+ },
+ {
+ "epoch": 0.7938090241343126,
+ "grad_norm": 0.152019559199633,
+ "learning_rate": 9.557767460433878e-06,
+ "loss": 0.445905864238739,
+ "step": 1513
+ },
+ {
+ "epoch": 0.7943336831059811,
+ "grad_norm": 0.1574763924872037,
+ "learning_rate": 9.535413876592872e-06,
+ "loss": 0.4502032995223999,
+ "step": 1514
+ },
+ {
+ "epoch": 0.7948583420776495,
+ "grad_norm": 0.1524218315499056,
+ "learning_rate": 9.513109098510816e-06,
+ "loss": 0.43753111362457275,
+ "step": 1515
+ },
+ {
+ "epoch": 0.7953830010493179,
+ "grad_norm": 0.15589147437937434,
+ "learning_rate": 9.490853186784801e-06,
+ "loss": 0.44524312019348145,
+ "step": 1516
+ },
+ {
+ "epoch": 0.7959076600209863,
+ "grad_norm": 0.15604009714081407,
+ "learning_rate": 9.468646201879156e-06,
+ "loss": 0.44141292572021484,
+ "step": 1517
+ },
+ {
+ "epoch": 0.7964323189926548,
+ "grad_norm": 0.1562432554004435,
+ "learning_rate": 9.44648820412529e-06,
+ "loss": 0.461681067943573,
+ "step": 1518
+ },
+ {
+ "epoch": 0.7969569779643232,
+ "grad_norm": 0.15233588531495665,
+ "learning_rate": 9.424379253721513e-06,
+ "loss": 0.4529497027397156,
+ "step": 1519
+ },
+ {
+ "epoch": 0.7974816369359916,
+ "grad_norm": 0.16856991685290923,
+ "learning_rate": 9.40231941073291e-06,
+ "loss": 0.4483596086502075,
+ "step": 1520
+ },
+ {
+ "epoch": 0.79800629590766,
+ "grad_norm": 0.1546876712981077,
+ "learning_rate": 9.380308735091125e-06,
+ "loss": 0.43846625089645386,
+ "step": 1521
+ },
+ {
+ "epoch": 0.7985309548793285,
+ "grad_norm": 0.15198637556890449,
+ "learning_rate": 9.35834728659424e-06,
+ "loss": 0.4625553488731384,
+ "step": 1522
+ },
+ {
+ "epoch": 0.7990556138509969,
+ "grad_norm": 0.1519777007198701,
+ "learning_rate": 9.336435124906592e-06,
+ "loss": 0.44183820486068726,
+ "step": 1523
+ },
+ {
+ "epoch": 0.7995802728226653,
+ "grad_norm": 0.16125823074979284,
+ "learning_rate": 9.314572309558616e-06,
+ "loss": 0.46057307720184326,
+ "step": 1524
+ },
+ {
+ "epoch": 0.8001049317943337,
+ "grad_norm": 0.16283323624848034,
+ "learning_rate": 9.2927588999467e-06,
+ "loss": 0.4530155062675476,
+ "step": 1525
+ },
+ {
+ "epoch": 0.8006295907660022,
+ "grad_norm": 0.14998654489027738,
+ "learning_rate": 9.270994955332984e-06,
+ "loss": 0.43155500292778015,
+ "step": 1526
+ },
+ {
+ "epoch": 0.8011542497376705,
+ "grad_norm": 0.15683095030848093,
+ "learning_rate": 9.249280534845241e-06,
+ "loss": 0.45097050070762634,
+ "step": 1527
+ },
+ {
+ "epoch": 0.8016789087093389,
+ "grad_norm": 0.1561578254492378,
+ "learning_rate": 9.227615697476674e-06,
+ "loss": 0.42797762155532837,
+ "step": 1528
+ },
+ {
+ "epoch": 0.8022035676810073,
+ "grad_norm": 0.15642385261469652,
+ "learning_rate": 9.206000502085813e-06,
+ "loss": 0.4560159742832184,
+ "step": 1529
+ },
+ {
+ "epoch": 0.8027282266526757,
+ "grad_norm": 0.1623126279097499,
+ "learning_rate": 9.1844350073963e-06,
+ "loss": 0.4704165458679199,
+ "step": 1530
+ },
+ {
+ "epoch": 0.8032528856243442,
+ "grad_norm": 0.15090934786253773,
+ "learning_rate": 9.162919271996755e-06,
+ "loss": 0.44642555713653564,
+ "step": 1531
+ },
+ {
+ "epoch": 0.8037775445960126,
+ "grad_norm": 0.15305216978534084,
+ "learning_rate": 9.14145335434062e-06,
+ "loss": 0.4634283781051636,
+ "step": 1532
+ },
+ {
+ "epoch": 0.804302203567681,
+ "grad_norm": 0.14653219861345698,
+ "learning_rate": 9.120037312745979e-06,
+ "loss": 0.44544070959091187,
+ "step": 1533
+ },
+ {
+ "epoch": 0.8048268625393494,
+ "grad_norm": 0.16247397938502464,
+ "learning_rate": 9.098671205395422e-06,
+ "loss": 0.4552367925643921,
+ "step": 1534
+ },
+ {
+ "epoch": 0.8053515215110179,
+ "grad_norm": 0.14963174165902007,
+ "learning_rate": 9.077355090335892e-06,
+ "loss": 0.43211787939071655,
+ "step": 1535
+ },
+ {
+ "epoch": 0.8058761804826863,
+ "grad_norm": 0.1554853846772283,
+ "learning_rate": 9.056089025478495e-06,
+ "loss": 0.4319286644458771,
+ "step": 1536
+ },
+ {
+ "epoch": 0.8064008394543547,
+ "grad_norm": 0.15840895821310472,
+ "learning_rate": 9.03487306859837e-06,
+ "loss": 0.4494994878768921,
+ "step": 1537
+ },
+ {
+ "epoch": 0.8069254984260231,
+ "grad_norm": 0.1543496716800982,
+ "learning_rate": 9.013707277334515e-06,
+ "loss": 0.45500510931015015,
+ "step": 1538
+ },
+ {
+ "epoch": 0.8074501573976915,
+ "grad_norm": 0.14797282318550045,
+ "learning_rate": 8.992591709189651e-06,
+ "loss": 0.4385896325111389,
+ "step": 1539
+ },
+ {
+ "epoch": 0.8079748163693599,
+ "grad_norm": 0.17045570682787758,
+ "learning_rate": 8.971526421530066e-06,
+ "loss": 0.4585988521575928,
+ "step": 1540
+ },
+ {
+ "epoch": 0.8084994753410283,
+ "grad_norm": 0.14758220713044884,
+ "learning_rate": 8.950511471585407e-06,
+ "loss": 0.4295082092285156,
+ "step": 1541
+ },
+ {
+ "epoch": 0.8090241343126967,
+ "grad_norm": 0.16005752068175036,
+ "learning_rate": 8.929546916448597e-06,
+ "loss": 0.4507679343223572,
+ "step": 1542
+ },
+ {
+ "epoch": 0.8095487932843651,
+ "grad_norm": 0.14884684083313074,
+ "learning_rate": 8.908632813075641e-06,
+ "loss": 0.43547600507736206,
+ "step": 1543
+ },
+ {
+ "epoch": 0.8100734522560336,
+ "grad_norm": 0.14860107595788136,
+ "learning_rate": 8.88776921828547e-06,
+ "loss": 0.42700642347335815,
+ "step": 1544
+ },
+ {
+ "epoch": 0.810598111227702,
+ "grad_norm": 0.1472630611843838,
+ "learning_rate": 8.86695618875981e-06,
+ "loss": 0.4473339319229126,
+ "step": 1545
+ },
+ {
+ "epoch": 0.8111227701993704,
+ "grad_norm": 0.15105523788317893,
+ "learning_rate": 8.846193781043e-06,
+ "loss": 0.44923150539398193,
+ "step": 1546
+ },
+ {
+ "epoch": 0.8116474291710388,
+ "grad_norm": 0.16121409015263036,
+ "learning_rate": 8.82548205154185e-06,
+ "loss": 0.4653576612472534,
+ "step": 1547
+ },
+ {
+ "epoch": 0.8121720881427072,
+ "grad_norm": 0.1595195150959421,
+ "learning_rate": 8.804821056525497e-06,
+ "loss": 0.4522424638271332,
+ "step": 1548
+ },
+ {
+ "epoch": 0.8126967471143757,
+ "grad_norm": 0.15163510106316194,
+ "learning_rate": 8.784210852125232e-06,
+ "loss": 0.43657630681991577,
+ "step": 1549
+ },
+ {
+ "epoch": 0.8132214060860441,
+ "grad_norm": 0.15155109347013793,
+ "learning_rate": 8.763651494334377e-06,
+ "loss": 0.43183624744415283,
+ "step": 1550
+ },
+ {
+ "epoch": 0.8137460650577125,
+ "grad_norm": 0.14930024958763324,
+ "learning_rate": 8.7431430390081e-06,
+ "loss": 0.43563947081565857,
+ "step": 1551
+ },
+ {
+ "epoch": 0.8142707240293809,
+ "grad_norm": 0.15361200153430712,
+ "learning_rate": 8.722685541863283e-06,
+ "loss": 0.4387405514717102,
+ "step": 1552
+ },
+ {
+ "epoch": 0.8147953830010494,
+ "grad_norm": 0.15549528796193493,
+ "learning_rate": 8.702279058478366e-06,
+ "loss": 0.44347938895225525,
+ "step": 1553
+ },
+ {
+ "epoch": 0.8153200419727177,
+ "grad_norm": 0.16243042867861712,
+ "learning_rate": 8.681923644293194e-06,
+ "loss": 0.44852834939956665,
+ "step": 1554
+ },
+ {
+ "epoch": 0.8158447009443861,
+ "grad_norm": 0.15706129652944026,
+ "learning_rate": 8.661619354608873e-06,
+ "loss": 0.45419150590896606,
+ "step": 1555
+ },
+ {
+ "epoch": 0.8163693599160545,
+ "grad_norm": 0.14854258573134127,
+ "learning_rate": 8.641366244587611e-06,
+ "loss": 0.44866859912872314,
+ "step": 1556
+ },
+ {
+ "epoch": 0.816894018887723,
+ "grad_norm": 0.14896937678017533,
+ "learning_rate": 8.62116436925258e-06,
+ "loss": 0.44596076011657715,
+ "step": 1557
+ },
+ {
+ "epoch": 0.8174186778593914,
+ "grad_norm": 0.15614508495620508,
+ "learning_rate": 8.601013783487746e-06,
+ "loss": 0.45645207166671753,
+ "step": 1558
+ },
+ {
+ "epoch": 0.8179433368310598,
+ "grad_norm": 0.1486538543050606,
+ "learning_rate": 8.580914542037734e-06,
+ "loss": 0.4592300355434418,
+ "step": 1559
+ },
+ {
+ "epoch": 0.8184679958027282,
+ "grad_norm": 0.15475320756830518,
+ "learning_rate": 8.560866699507702e-06,
+ "loss": 0.4605134129524231,
+ "step": 1560
+ },
+ {
+ "epoch": 0.8189926547743966,
+ "grad_norm": 0.14754406728057914,
+ "learning_rate": 8.540870310363137e-06,
+ "loss": 0.43651673197746277,
+ "step": 1561
+ },
+ {
+ "epoch": 0.8195173137460651,
+ "grad_norm": 0.14871990701077825,
+ "learning_rate": 8.520925428929755e-06,
+ "loss": 0.44830837845802307,
+ "step": 1562
+ },
+ {
+ "epoch": 0.8200419727177335,
+ "grad_norm": 0.15358241317712973,
+ "learning_rate": 8.501032109393334e-06,
+ "loss": 0.4546644687652588,
+ "step": 1563
+ },
+ {
+ "epoch": 0.8205666316894019,
+ "grad_norm": 0.15650856948126138,
+ "learning_rate": 8.481190405799569e-06,
+ "loss": 0.4423891305923462,
+ "step": 1564
+ },
+ {
+ "epoch": 0.8210912906610703,
+ "grad_norm": 0.15474996746242287,
+ "learning_rate": 8.461400372053923e-06,
+ "loss": 0.44242313504219055,
+ "step": 1565
+ },
+ {
+ "epoch": 0.8216159496327388,
+ "grad_norm": 0.16544926591962897,
+ "learning_rate": 8.441662061921497e-06,
+ "loss": 0.4487020969390869,
+ "step": 1566
+ },
+ {
+ "epoch": 0.8221406086044072,
+ "grad_norm": 0.1489733586650647,
+ "learning_rate": 8.421975529026855e-06,
+ "loss": 0.4371356964111328,
+ "step": 1567
+ },
+ {
+ "epoch": 0.8226652675760755,
+ "grad_norm": 0.15995455652906124,
+ "learning_rate": 8.402340826853907e-06,
+ "loss": 0.4566021263599396,
+ "step": 1568
+ },
+ {
+ "epoch": 0.8231899265477439,
+ "grad_norm": 0.1500716152622102,
+ "learning_rate": 8.382758008745726e-06,
+ "loss": 0.43888187408447266,
+ "step": 1569
+ },
+ {
+ "epoch": 0.8237145855194123,
+ "grad_norm": 0.15554904485712012,
+ "learning_rate": 8.363227127904464e-06,
+ "loss": 0.4584001898765564,
+ "step": 1570
+ },
+ {
+ "epoch": 0.8242392444910808,
+ "grad_norm": 0.14820917815884277,
+ "learning_rate": 8.343748237391144e-06,
+ "loss": 0.43289071321487427,
+ "step": 1571
+ },
+ {
+ "epoch": 0.8247639034627492,
+ "grad_norm": 0.15801200747859714,
+ "learning_rate": 8.324321390125547e-06,
+ "loss": 0.44196009635925293,
+ "step": 1572
+ },
+ {
+ "epoch": 0.8252885624344176,
+ "grad_norm": 0.15323277975380797,
+ "learning_rate": 8.304946638886078e-06,
+ "loss": 0.43244463205337524,
+ "step": 1573
+ },
+ {
+ "epoch": 0.825813221406086,
+ "grad_norm": 0.15076940176344772,
+ "learning_rate": 8.285624036309583e-06,
+ "loss": 0.44401073455810547,
+ "step": 1574
+ },
+ {
+ "epoch": 0.8263378803777545,
+ "grad_norm": 0.1544958300475393,
+ "learning_rate": 8.266353634891266e-06,
+ "loss": 0.45198434591293335,
+ "step": 1575
+ },
+ {
+ "epoch": 0.8268625393494229,
+ "grad_norm": 0.15093528292068448,
+ "learning_rate": 8.247135486984486e-06,
+ "loss": 0.4403747618198395,
+ "step": 1576
+ },
+ {
+ "epoch": 0.8273871983210913,
+ "grad_norm": 0.14680662784017148,
+ "learning_rate": 8.227969644800648e-06,
+ "loss": 0.4375043213367462,
+ "step": 1577
+ },
+ {
+ "epoch": 0.8279118572927597,
+ "grad_norm": 0.14791704009291914,
+ "learning_rate": 8.208856160409055e-06,
+ "loss": 0.43811577558517456,
+ "step": 1578
+ },
+ {
+ "epoch": 0.8284365162644282,
+ "grad_norm": 0.14924185710518617,
+ "learning_rate": 8.189795085736765e-06,
+ "loss": 0.43853428959846497,
+ "step": 1579
+ },
+ {
+ "epoch": 0.8289611752360966,
+ "grad_norm": 0.15420576936040198,
+ "learning_rate": 8.170786472568451e-06,
+ "loss": 0.445467084646225,
+ "step": 1580
+ },
+ {
+ "epoch": 0.829485834207765,
+ "grad_norm": 0.15577901772605834,
+ "learning_rate": 8.151830372546258e-06,
+ "loss": 0.4529452621936798,
+ "step": 1581
+ },
+ {
+ "epoch": 0.8300104931794333,
+ "grad_norm": 0.15387840488797552,
+ "learning_rate": 8.132926837169676e-06,
+ "loss": 0.46009817719459534,
+ "step": 1582
+ },
+ {
+ "epoch": 0.8305351521511017,
+ "grad_norm": 0.1527913309957562,
+ "learning_rate": 8.11407591779537e-06,
+ "loss": 0.4513983726501465,
+ "step": 1583
+ },
+ {
+ "epoch": 0.8310598111227702,
+ "grad_norm": 0.1537090286922398,
+ "learning_rate": 8.095277665637061e-06,
+ "loss": 0.4537739157676697,
+ "step": 1584
+ },
+ {
+ "epoch": 0.8315844700944386,
+ "grad_norm": 0.15722962107654076,
+ "learning_rate": 8.076532131765408e-06,
+ "loss": 0.45232969522476196,
+ "step": 1585
+ },
+ {
+ "epoch": 0.832109129066107,
+ "grad_norm": 0.1450142396611787,
+ "learning_rate": 8.057839367107817e-06,
+ "loss": 0.4270634651184082,
+ "step": 1586
+ },
+ {
+ "epoch": 0.8326337880377754,
+ "grad_norm": 0.150264001570592,
+ "learning_rate": 8.039199422448355e-06,
+ "loss": 0.43563878536224365,
+ "step": 1587
+ },
+ {
+ "epoch": 0.8331584470094439,
+ "grad_norm": 0.14904943307590535,
+ "learning_rate": 8.020612348427572e-06,
+ "loss": 0.4401509165763855,
+ "step": 1588
+ },
+ {
+ "epoch": 0.8336831059811123,
+ "grad_norm": 0.15453821662860034,
+ "learning_rate": 8.002078195542393e-06,
+ "loss": 0.44545161724090576,
+ "step": 1589
+ },
+ {
+ "epoch": 0.8342077649527807,
+ "grad_norm": 0.1570809678830383,
+ "learning_rate": 7.983597014145951e-06,
+ "loss": 0.45228466391563416,
+ "step": 1590
+ },
+ {
+ "epoch": 0.8347324239244491,
+ "grad_norm": 0.14741240769629818,
+ "learning_rate": 7.965168854447493e-06,
+ "loss": 0.4485381245613098,
+ "step": 1591
+ },
+ {
+ "epoch": 0.8352570828961176,
+ "grad_norm": 0.15362267095533516,
+ "learning_rate": 7.946793766512195e-06,
+ "loss": 0.4358059763908386,
+ "step": 1592
+ },
+ {
+ "epoch": 0.835781741867786,
+ "grad_norm": 0.1544754586495605,
+ "learning_rate": 7.92847180026106e-06,
+ "loss": 0.43327778577804565,
+ "step": 1593
+ },
+ {
+ "epoch": 0.8363064008394544,
+ "grad_norm": 0.15023225026164963,
+ "learning_rate": 7.91020300547077e-06,
+ "loss": 0.44359731674194336,
+ "step": 1594
+ },
+ {
+ "epoch": 0.8368310598111228,
+ "grad_norm": 0.15665530392095345,
+ "learning_rate": 7.891987431773545e-06,
+ "loss": 0.44503387808799744,
+ "step": 1595
+ },
+ {
+ "epoch": 0.8373557187827911,
+ "grad_norm": 0.14717614926392292,
+ "learning_rate": 7.873825128657033e-06,
+ "loss": 0.45314061641693115,
+ "step": 1596
+ },
+ {
+ "epoch": 0.8378803777544596,
+ "grad_norm": 0.15207464934030612,
+ "learning_rate": 7.855716145464131e-06,
+ "loss": 0.4526442885398865,
+ "step": 1597
+ },
+ {
+ "epoch": 0.838405036726128,
+ "grad_norm": 0.15129881541794873,
+ "learning_rate": 7.837660531392904e-06,
+ "loss": 0.45127853751182556,
+ "step": 1598
+ },
+ {
+ "epoch": 0.8389296956977964,
+ "grad_norm": 0.15623362715417285,
+ "learning_rate": 7.819658335496405e-06,
+ "loss": 0.46432793140411377,
+ "step": 1599
+ },
+ {
+ "epoch": 0.8394543546694648,
+ "grad_norm": 0.1529219879642125,
+ "learning_rate": 7.801709606682573e-06,
+ "loss": 0.4475358724594116,
+ "step": 1600
+ },
+ {
+ "epoch": 0.8399790136411333,
+ "grad_norm": 0.14649139854087698,
+ "learning_rate": 7.78381439371409e-06,
+ "loss": 0.4363391101360321,
+ "step": 1601
+ },
+ {
+ "epoch": 0.8405036726128017,
+ "grad_norm": 0.14571233251948087,
+ "learning_rate": 7.765972745208243e-06,
+ "loss": 0.44419488310813904,
+ "step": 1602
+ },
+ {
+ "epoch": 0.8410283315844701,
+ "grad_norm": 0.15168460066406128,
+ "learning_rate": 7.748184709636791e-06,
+ "loss": 0.4433790445327759,
+ "step": 1603
+ },
+ {
+ "epoch": 0.8415529905561385,
+ "grad_norm": 0.14852860876513757,
+ "learning_rate": 7.730450335325853e-06,
+ "loss": 0.4493539333343506,
+ "step": 1604
+ },
+ {
+ "epoch": 0.842077649527807,
+ "grad_norm": 0.1505615491087377,
+ "learning_rate": 7.712769670455741e-06,
+ "loss": 0.4646318554878235,
+ "step": 1605
+ },
+ {
+ "epoch": 0.8426023084994754,
+ "grad_norm": 0.15407610486660114,
+ "learning_rate": 7.695142763060878e-06,
+ "loss": 0.4484049677848816,
+ "step": 1606
+ },
+ {
+ "epoch": 0.8431269674711438,
+ "grad_norm": 0.1521449709584124,
+ "learning_rate": 7.67756966102962e-06,
+ "loss": 0.43366891145706177,
+ "step": 1607
+ },
+ {
+ "epoch": 0.8436516264428122,
+ "grad_norm": 0.14747762647596166,
+ "learning_rate": 7.660050412104153e-06,
+ "loss": 0.43950092792510986,
+ "step": 1608
+ },
+ {
+ "epoch": 0.8441762854144806,
+ "grad_norm": 0.15512610809240027,
+ "learning_rate": 7.642585063880351e-06,
+ "loss": 0.46813055872917175,
+ "step": 1609
+ },
+ {
+ "epoch": 0.844700944386149,
+ "grad_norm": 0.15825480008734216,
+ "learning_rate": 7.625173663807655e-06,
+ "loss": 0.4348755478858948,
+ "step": 1610
+ },
+ {
+ "epoch": 0.8452256033578174,
+ "grad_norm": 0.1573968981854278,
+ "learning_rate": 7.6078162591889434e-06,
+ "loss": 0.45048123598098755,
+ "step": 1611
+ },
+ {
+ "epoch": 0.8457502623294858,
+ "grad_norm": 0.1517038960096417,
+ "learning_rate": 7.590512897180399e-06,
+ "loss": 0.43563610315322876,
+ "step": 1612
+ },
+ {
+ "epoch": 0.8462749213011542,
+ "grad_norm": 0.15503576044724665,
+ "learning_rate": 7.573263624791382e-06,
+ "loss": 0.4467726945877075,
+ "step": 1613
+ },
+ {
+ "epoch": 0.8467995802728226,
+ "grad_norm": 0.1560549927192375,
+ "learning_rate": 7.556068488884304e-06,
+ "loss": 0.4320477545261383,
+ "step": 1614
+ },
+ {
+ "epoch": 0.8473242392444911,
+ "grad_norm": 0.1679830356844654,
+ "learning_rate": 7.538927536174498e-06,
+ "loss": 0.4476034343242645,
+ "step": 1615
+ },
+ {
+ "epoch": 0.8478488982161595,
+ "grad_norm": 0.15169257383162213,
+ "learning_rate": 7.521840813230098e-06,
+ "loss": 0.4325854778289795,
+ "step": 1616
+ },
+ {
+ "epoch": 0.8483735571878279,
+ "grad_norm": 0.15980226324329827,
+ "learning_rate": 7.504808366471909e-06,
+ "loss": 0.4472721815109253,
+ "step": 1617
+ },
+ {
+ "epoch": 0.8488982161594963,
+ "grad_norm": 0.16932727986736365,
+ "learning_rate": 7.48783024217327e-06,
+ "loss": 0.45428287982940674,
+ "step": 1618
+ },
+ {
+ "epoch": 0.8494228751311648,
+ "grad_norm": 0.15065676306311376,
+ "learning_rate": 7.470906486459946e-06,
+ "loss": 0.44418883323669434,
+ "step": 1619
+ },
+ {
+ "epoch": 0.8499475341028332,
+ "grad_norm": 0.1533992998314183,
+ "learning_rate": 7.454037145309986e-06,
+ "loss": 0.43552401661872864,
+ "step": 1620
+ },
+ {
+ "epoch": 0.8504721930745016,
+ "grad_norm": 0.15479480515120017,
+ "learning_rate": 7.437222264553628e-06,
+ "loss": 0.444465696811676,
+ "step": 1621
+ },
+ {
+ "epoch": 0.85099685204617,
+ "grad_norm": 0.15845678337942112,
+ "learning_rate": 7.420461889873129e-06,
+ "loss": 0.44896578788757324,
+ "step": 1622
+ },
+ {
+ "epoch": 0.8515215110178384,
+ "grad_norm": 0.15038940533473208,
+ "learning_rate": 7.4037560668026804e-06,
+ "loss": 0.43062180280685425,
+ "step": 1623
+ },
+ {
+ "epoch": 0.8520461699895068,
+ "grad_norm": 0.1480017115109342,
+ "learning_rate": 7.387104840728259e-06,
+ "loss": 0.4630432724952698,
+ "step": 1624
+ },
+ {
+ "epoch": 0.8525708289611752,
+ "grad_norm": 0.15321305032824092,
+ "learning_rate": 7.370508256887514e-06,
+ "loss": 0.4419487714767456,
+ "step": 1625
+ },
+ {
+ "epoch": 0.8530954879328436,
+ "grad_norm": 0.1568246735699272,
+ "learning_rate": 7.353966360369659e-06,
+ "loss": 0.4483609199523926,
+ "step": 1626
+ },
+ {
+ "epoch": 0.853620146904512,
+ "grad_norm": 0.15539536583814875,
+ "learning_rate": 7.337479196115317e-06,
+ "loss": 0.4567753076553345,
+ "step": 1627
+ },
+ {
+ "epoch": 0.8541448058761805,
+ "grad_norm": 0.15528695146633797,
+ "learning_rate": 7.321046808916426e-06,
+ "loss": 0.4560334086418152,
+ "step": 1628
+ },
+ {
+ "epoch": 0.8546694648478489,
+ "grad_norm": 0.15005728554864303,
+ "learning_rate": 7.304669243416102e-06,
+ "loss": 0.43833765387535095,
+ "step": 1629
+ },
+ {
+ "epoch": 0.8551941238195173,
+ "grad_norm": 0.15959100645189242,
+ "learning_rate": 7.288346544108518e-06,
+ "loss": 0.4713831841945648,
+ "step": 1630
+ },
+ {
+ "epoch": 0.8557187827911857,
+ "grad_norm": 0.14898121463872632,
+ "learning_rate": 7.272078755338807e-06,
+ "loss": 0.44003745913505554,
+ "step": 1631
+ },
+ {
+ "epoch": 0.8562434417628542,
+ "grad_norm": 0.15083110802343333,
+ "learning_rate": 7.255865921302903e-06,
+ "loss": 0.429004043340683,
+ "step": 1632
+ },
+ {
+ "epoch": 0.8567681007345226,
+ "grad_norm": 0.1522757086796616,
+ "learning_rate": 7.239708086047445e-06,
+ "loss": 0.4403839707374573,
+ "step": 1633
+ },
+ {
+ "epoch": 0.857292759706191,
+ "grad_norm": 0.15719824838921245,
+ "learning_rate": 7.223605293469654e-06,
+ "loss": 0.4377821087837219,
+ "step": 1634
+ },
+ {
+ "epoch": 0.8578174186778594,
+ "grad_norm": 0.1450616202708373,
+ "learning_rate": 7.207557587317216e-06,
+ "loss": 0.4357878267765045,
+ "step": 1635
+ },
+ {
+ "epoch": 0.8583420776495279,
+ "grad_norm": 0.15703162220646427,
+ "learning_rate": 7.191565011188149e-06,
+ "loss": 0.44412243366241455,
+ "step": 1636
+ },
+ {
+ "epoch": 0.8588667366211962,
+ "grad_norm": 0.150175313334118,
+ "learning_rate": 7.175627608530716e-06,
+ "loss": 0.4543853998184204,
+ "step": 1637
+ },
+ {
+ "epoch": 0.8593913955928646,
+ "grad_norm": 0.15091711492716975,
+ "learning_rate": 7.159745422643261e-06,
+ "loss": 0.4268660545349121,
+ "step": 1638
+ },
+ {
+ "epoch": 0.859916054564533,
+ "grad_norm": 0.15449835654964011,
+ "learning_rate": 7.1439184966741295e-06,
+ "loss": 0.4507150650024414,
+ "step": 1639
+ },
+ {
+ "epoch": 0.8604407135362014,
+ "grad_norm": 0.15689612593442026,
+ "learning_rate": 7.128146873621533e-06,
+ "loss": 0.45520636439323425,
+ "step": 1640
+ },
+ {
+ "epoch": 0.8609653725078699,
+ "grad_norm": 0.1565226785234465,
+ "learning_rate": 7.11243059633345e-06,
+ "loss": 0.45102012157440186,
+ "step": 1641
+ },
+ {
+ "epoch": 0.8614900314795383,
+ "grad_norm": 0.15500877922230677,
+ "learning_rate": 7.096769707507481e-06,
+ "loss": 0.447115421295166,
+ "step": 1642
+ },
+ {
+ "epoch": 0.8620146904512067,
+ "grad_norm": 0.1595248428000243,
+ "learning_rate": 7.081164249690761e-06,
+ "loss": 0.4627235233783722,
+ "step": 1643
+ },
+ {
+ "epoch": 0.8625393494228751,
+ "grad_norm": 0.15958177510163488,
+ "learning_rate": 7.065614265279818e-06,
+ "loss": 0.4386388063430786,
+ "step": 1644
+ },
+ {
+ "epoch": 0.8630640083945436,
+ "grad_norm": 0.15152834825396286,
+ "learning_rate": 7.050119796520482e-06,
+ "loss": 0.4531296491622925,
+ "step": 1645
+ },
+ {
+ "epoch": 0.863588667366212,
+ "grad_norm": 0.15635733905345023,
+ "learning_rate": 7.034680885507752e-06,
+ "loss": 0.45573747158050537,
+ "step": 1646
+ },
+ {
+ "epoch": 0.8641133263378804,
+ "grad_norm": 0.15666918852633857,
+ "learning_rate": 7.019297574185703e-06,
+ "loss": 0.4499792158603668,
+ "step": 1647
+ },
+ {
+ "epoch": 0.8646379853095488,
+ "grad_norm": 0.156444627477174,
+ "learning_rate": 7.0039699043473405e-06,
+ "loss": 0.4545009136199951,
+ "step": 1648
+ },
+ {
+ "epoch": 0.8651626442812173,
+ "grad_norm": 0.14638989347446527,
+ "learning_rate": 6.988697917634515e-06,
+ "loss": 0.4397326409816742,
+ "step": 1649
+ },
+ {
+ "epoch": 0.8656873032528857,
+ "grad_norm": 0.15492813701343372,
+ "learning_rate": 6.973481655537793e-06,
+ "loss": 0.4402945041656494,
+ "step": 1650
+ },
+ {
+ "epoch": 0.866211962224554,
+ "grad_norm": 0.14466615385057716,
+ "learning_rate": 6.958321159396352e-06,
+ "loss": 0.4479280114173889,
+ "step": 1651
+ },
+ {
+ "epoch": 0.8667366211962224,
+ "grad_norm": 0.15865895330952856,
+ "learning_rate": 6.943216470397871e-06,
+ "loss": 0.4553218483924866,
+ "step": 1652
+ },
+ {
+ "epoch": 0.8672612801678908,
+ "grad_norm": 0.15293327582467067,
+ "learning_rate": 6.928167629578407e-06,
+ "loss": 0.44509196281433105,
+ "step": 1653
+ },
+ {
+ "epoch": 0.8677859391395593,
+ "grad_norm": 0.1515507946195896,
+ "learning_rate": 6.91317467782229e-06,
+ "loss": 0.4498351216316223,
+ "step": 1654
+ },
+ {
+ "epoch": 0.8683105981112277,
+ "grad_norm": 0.15313372799407085,
+ "learning_rate": 6.898237655862007e-06,
+ "loss": 0.4445887804031372,
+ "step": 1655
+ },
+ {
+ "epoch": 0.8688352570828961,
+ "grad_norm": 0.1491970050887588,
+ "learning_rate": 6.883356604278115e-06,
+ "loss": 0.4323808252811432,
+ "step": 1656
+ },
+ {
+ "epoch": 0.8693599160545645,
+ "grad_norm": 0.14449269685472518,
+ "learning_rate": 6.868531563499096e-06,
+ "loss": 0.42508652806282043,
+ "step": 1657
+ },
+ {
+ "epoch": 0.869884575026233,
+ "grad_norm": 0.14997328847830618,
+ "learning_rate": 6.853762573801264e-06,
+ "loss": 0.43655943870544434,
+ "step": 1658
+ },
+ {
+ "epoch": 0.8704092339979014,
+ "grad_norm": 0.15288208370561934,
+ "learning_rate": 6.839049675308662e-06,
+ "loss": 0.4505232274532318,
+ "step": 1659
+ },
+ {
+ "epoch": 0.8709338929695698,
+ "grad_norm": 0.14416201723141706,
+ "learning_rate": 6.824392907992938e-06,
+ "loss": 0.42444318532943726,
+ "step": 1660
+ },
+ {
+ "epoch": 0.8714585519412382,
+ "grad_norm": 0.14468507187223545,
+ "learning_rate": 6.809792311673244e-06,
+ "loss": 0.4251434803009033,
+ "step": 1661
+ },
+ {
+ "epoch": 0.8719832109129066,
+ "grad_norm": 0.15406295659798216,
+ "learning_rate": 6.795247926016142e-06,
+ "loss": 0.45068371295928955,
+ "step": 1662
+ },
+ {
+ "epoch": 0.8725078698845751,
+ "grad_norm": 0.1516721215663676,
+ "learning_rate": 6.780759790535465e-06,
+ "loss": 0.441638320684433,
+ "step": 1663
+ },
+ {
+ "epoch": 0.8730325288562435,
+ "grad_norm": 0.1537661139082594,
+ "learning_rate": 6.766327944592239e-06,
+ "loss": 0.44044923782348633,
+ "step": 1664
+ },
+ {
+ "epoch": 0.8735571878279118,
+ "grad_norm": 0.15175839438218855,
+ "learning_rate": 6.751952427394553e-06,
+ "loss": 0.44454342126846313,
+ "step": 1665
+ },
+ {
+ "epoch": 0.8740818467995802,
+ "grad_norm": 0.1531195282211392,
+ "learning_rate": 6.737633277997471e-06,
+ "loss": 0.4540765881538391,
+ "step": 1666
+ },
+ {
+ "epoch": 0.8746065057712487,
+ "grad_norm": 0.14961744003585584,
+ "learning_rate": 6.723370535302915e-06,
+ "loss": 0.44223541021347046,
+ "step": 1667
+ },
+ {
+ "epoch": 0.8751311647429171,
+ "grad_norm": 0.14816820299503658,
+ "learning_rate": 6.709164238059568e-06,
+ "loss": 0.43701833486557007,
+ "step": 1668
+ },
+ {
+ "epoch": 0.8756558237145855,
+ "grad_norm": 0.14622402294164918,
+ "learning_rate": 6.695014424862756e-06,
+ "loss": 0.42922335863113403,
+ "step": 1669
+ },
+ {
+ "epoch": 0.8761804826862539,
+ "grad_norm": 0.1547291062309004,
+ "learning_rate": 6.680921134154355e-06,
+ "loss": 0.4233800768852234,
+ "step": 1670
+ },
+ {
+ "epoch": 0.8767051416579223,
+ "grad_norm": 0.15378903523440587,
+ "learning_rate": 6.666884404222678e-06,
+ "loss": 0.4432950019836426,
+ "step": 1671
+ },
+ {
+ "epoch": 0.8772298006295908,
+ "grad_norm": 0.15973037860999773,
+ "learning_rate": 6.652904273202384e-06,
+ "loss": 0.4560699462890625,
+ "step": 1672
+ },
+ {
+ "epoch": 0.8777544596012592,
+ "grad_norm": 0.15430233088028492,
+ "learning_rate": 6.638980779074358e-06,
+ "loss": 0.45776498317718506,
+ "step": 1673
+ },
+ {
+ "epoch": 0.8782791185729276,
+ "grad_norm": 0.1587664730992995,
+ "learning_rate": 6.625113959665614e-06,
+ "loss": 0.43579810857772827,
+ "step": 1674
+ },
+ {
+ "epoch": 0.878803777544596,
+ "grad_norm": 0.14370919039693034,
+ "learning_rate": 6.6113038526491946e-06,
+ "loss": 0.4426189064979553,
+ "step": 1675
+ },
+ {
+ "epoch": 0.8793284365162645,
+ "grad_norm": 0.14933009355373217,
+ "learning_rate": 6.59755049554407e-06,
+ "loss": 0.44747287034988403,
+ "step": 1676
+ },
+ {
+ "epoch": 0.8798530954879329,
+ "grad_norm": 0.15713697891738065,
+ "learning_rate": 6.583853925715039e-06,
+ "loss": 0.46320077776908875,
+ "step": 1677
+ },
+ {
+ "epoch": 0.8803777544596013,
+ "grad_norm": 0.1542096833927892,
+ "learning_rate": 6.570214180372609e-06,
+ "loss": 0.4473888874053955,
+ "step": 1678
+ },
+ {
+ "epoch": 0.8809024134312696,
+ "grad_norm": 0.16187259968193946,
+ "learning_rate": 6.556631296572922e-06,
+ "loss": 0.4421869218349457,
+ "step": 1679
+ },
+ {
+ "epoch": 0.881427072402938,
+ "grad_norm": 0.15978032996583555,
+ "learning_rate": 6.54310531121762e-06,
+ "loss": 0.4305853843688965,
+ "step": 1680
+ },
+ {
+ "epoch": 0.8819517313746065,
+ "grad_norm": 0.15631580579775123,
+ "learning_rate": 6.529636261053784e-06,
+ "loss": 0.4340512454509735,
+ "step": 1681
+ },
+ {
+ "epoch": 0.8824763903462749,
+ "grad_norm": 0.150084161390103,
+ "learning_rate": 6.516224182673812e-06,
+ "loss": 0.4383493661880493,
+ "step": 1682
+ },
+ {
+ "epoch": 0.8830010493179433,
+ "grad_norm": 0.15527917215838966,
+ "learning_rate": 6.502869112515314e-06,
+ "loss": 0.45441877841949463,
+ "step": 1683
+ },
+ {
+ "epoch": 0.8835257082896117,
+ "grad_norm": 0.15164994594035522,
+ "learning_rate": 6.489571086861027e-06,
+ "loss": 0.43363988399505615,
+ "step": 1684
+ },
+ {
+ "epoch": 0.8840503672612802,
+ "grad_norm": 0.1467230535452683,
+ "learning_rate": 6.47633014183871e-06,
+ "loss": 0.4263266623020172,
+ "step": 1685
+ },
+ {
+ "epoch": 0.8845750262329486,
+ "grad_norm": 0.14597001769185472,
+ "learning_rate": 6.463146313421044e-06,
+ "loss": 0.4223391115665436,
+ "step": 1686
+ },
+ {
+ "epoch": 0.885099685204617,
+ "grad_norm": 0.15050446002714588,
+ "learning_rate": 6.4500196374255435e-06,
+ "loss": 0.43097931146621704,
+ "step": 1687
+ },
+ {
+ "epoch": 0.8856243441762854,
+ "grad_norm": 0.14918033614543294,
+ "learning_rate": 6.436950149514448e-06,
+ "loss": 0.4587080478668213,
+ "step": 1688
+ },
+ {
+ "epoch": 0.8861490031479539,
+ "grad_norm": 0.15469492243631364,
+ "learning_rate": 6.423937885194631e-06,
+ "loss": 0.45661866664886475,
+ "step": 1689
+ },
+ {
+ "epoch": 0.8866736621196223,
+ "grad_norm": 0.15552798640177423,
+ "learning_rate": 6.410982879817498e-06,
+ "loss": 0.44370236992836,
+ "step": 1690
+ },
+ {
+ "epoch": 0.8871983210912907,
+ "grad_norm": 0.15595593365088142,
+ "learning_rate": 6.398085168578901e-06,
+ "loss": 0.4291878342628479,
+ "step": 1691
+ },
+ {
+ "epoch": 0.887722980062959,
+ "grad_norm": 0.14708046948755318,
+ "learning_rate": 6.385244786519036e-06,
+ "loss": 0.4422011375427246,
+ "step": 1692
+ },
+ {
+ "epoch": 0.8882476390346274,
+ "grad_norm": 0.1499253382866992,
+ "learning_rate": 6.3724617685223404e-06,
+ "loss": 0.4595015048980713,
+ "step": 1693
+ },
+ {
+ "epoch": 0.8887722980062959,
+ "grad_norm": 0.14702868569252941,
+ "learning_rate": 6.3597361493174205e-06,
+ "loss": 0.4347558319568634,
+ "step": 1694
+ },
+ {
+ "epoch": 0.8892969569779643,
+ "grad_norm": 0.14977289247327386,
+ "learning_rate": 6.347067963476928e-06,
+ "loss": 0.43877455592155457,
+ "step": 1695
+ },
+ {
+ "epoch": 0.8898216159496327,
+ "grad_norm": 0.1453902077891248,
+ "learning_rate": 6.334457245417491e-06,
+ "loss": 0.4282415807247162,
+ "step": 1696
+ },
+ {
+ "epoch": 0.8903462749213011,
+ "grad_norm": 0.1545933676895574,
+ "learning_rate": 6.321904029399607e-06,
+ "loss": 0.4429236054420471,
+ "step": 1697
+ },
+ {
+ "epoch": 0.8908709338929696,
+ "grad_norm": 0.15625844051167467,
+ "learning_rate": 6.3094083495275565e-06,
+ "loss": 0.4284421503543854,
+ "step": 1698
+ },
+ {
+ "epoch": 0.891395592864638,
+ "grad_norm": 0.1484963883381567,
+ "learning_rate": 6.2969702397493e-06,
+ "loss": 0.4545463025569916,
+ "step": 1699
+ },
+ {
+ "epoch": 0.8919202518363064,
+ "grad_norm": 0.15122402806869387,
+ "learning_rate": 6.284589733856403e-06,
+ "loss": 0.43840622901916504,
+ "step": 1700
+ },
+ {
+ "epoch": 0.8924449108079748,
+ "grad_norm": 0.14889313167657334,
+ "learning_rate": 6.27226686548392e-06,
+ "loss": 0.4451153874397278,
+ "step": 1701
+ },
+ {
+ "epoch": 0.8929695697796433,
+ "grad_norm": 0.14724940941036385,
+ "learning_rate": 6.260001668110331e-06,
+ "loss": 0.434165358543396,
+ "step": 1702
+ },
+ {
+ "epoch": 0.8934942287513117,
+ "grad_norm": 0.15520838093295836,
+ "learning_rate": 6.247794175057433e-06,
+ "loss": 0.44641873240470886,
+ "step": 1703
+ },
+ {
+ "epoch": 0.8940188877229801,
+ "grad_norm": 0.159167904797364,
+ "learning_rate": 6.235644419490253e-06,
+ "loss": 0.44591957330703735,
+ "step": 1704
+ },
+ {
+ "epoch": 0.8945435466946485,
+ "grad_norm": 0.14843235881370814,
+ "learning_rate": 6.223552434416951e-06,
+ "loss": 0.43010181188583374,
+ "step": 1705
+ },
+ {
+ "epoch": 0.8950682056663168,
+ "grad_norm": 0.1590998153483183,
+ "learning_rate": 6.211518252688749e-06,
+ "loss": 0.4602118134498596,
+ "step": 1706
+ },
+ {
+ "epoch": 0.8955928646379853,
+ "grad_norm": 0.15257025539137767,
+ "learning_rate": 6.19954190699982e-06,
+ "loss": 0.45429927110671997,
+ "step": 1707
+ },
+ {
+ "epoch": 0.8961175236096537,
+ "grad_norm": 0.15304035006408193,
+ "learning_rate": 6.18762342988722e-06,
+ "loss": 0.4405020773410797,
+ "step": 1708
+ },
+ {
+ "epoch": 0.8966421825813221,
+ "grad_norm": 0.15956303033928781,
+ "learning_rate": 6.175762853730777e-06,
+ "loss": 0.4455980062484741,
+ "step": 1709
+ },
+ {
+ "epoch": 0.8971668415529905,
+ "grad_norm": 0.14973831573751925,
+ "learning_rate": 6.163960210753025e-06,
+ "loss": 0.4331505298614502,
+ "step": 1710
+ },
+ {
+ "epoch": 0.897691500524659,
+ "grad_norm": 0.14573687980573624,
+ "learning_rate": 6.1522155330190985e-06,
+ "loss": 0.43307584524154663,
+ "step": 1711
+ },
+ {
+ "epoch": 0.8982161594963274,
+ "grad_norm": 0.1556523345686865,
+ "learning_rate": 6.140528852436665e-06,
+ "loss": 0.4531600773334503,
+ "step": 1712
+ },
+ {
+ "epoch": 0.8987408184679958,
+ "grad_norm": 0.14262442470666953,
+ "learning_rate": 6.128900200755817e-06,
+ "loss": 0.4350634813308716,
+ "step": 1713
+ },
+ {
+ "epoch": 0.8992654774396642,
+ "grad_norm": 0.14709392307082428,
+ "learning_rate": 6.117329609568998e-06,
+ "loss": 0.43465620279312134,
+ "step": 1714
+ },
+ {
+ "epoch": 0.8997901364113327,
+ "grad_norm": 0.15648040137428026,
+ "learning_rate": 6.105817110310909e-06,
+ "loss": 0.4344560503959656,
+ "step": 1715
+ },
+ {
+ "epoch": 0.9003147953830011,
+ "grad_norm": 0.15758204487742147,
+ "learning_rate": 6.094362734258438e-06,
+ "loss": 0.4409951865673065,
+ "step": 1716
+ },
+ {
+ "epoch": 0.9008394543546695,
+ "grad_norm": 0.15755547382189733,
+ "learning_rate": 6.082966512530558e-06,
+ "loss": 0.44050097465515137,
+ "step": 1717
+ },
+ {
+ "epoch": 0.9013641133263379,
+ "grad_norm": 0.1536165045521179,
+ "learning_rate": 6.071628476088257e-06,
+ "loss": 0.43460220098495483,
+ "step": 1718
+ },
+ {
+ "epoch": 0.9018887722980063,
+ "grad_norm": 0.16253540769141986,
+ "learning_rate": 6.060348655734439e-06,
+ "loss": 0.4490512013435364,
+ "step": 1719
+ },
+ {
+ "epoch": 0.9024134312696747,
+ "grad_norm": 0.15003242262029914,
+ "learning_rate": 6.049127082113853e-06,
+ "loss": 0.44427937269210815,
+ "step": 1720
+ },
+ {
+ "epoch": 0.9029380902413431,
+ "grad_norm": 0.1579908758627471,
+ "learning_rate": 6.037963785712995e-06,
+ "loss": 0.4421137571334839,
+ "step": 1721
+ },
+ {
+ "epoch": 0.9034627492130115,
+ "grad_norm": 0.15111957361250286,
+ "learning_rate": 6.026858796860048e-06,
+ "loss": 0.44011324644088745,
+ "step": 1722
+ },
+ {
+ "epoch": 0.9039874081846799,
+ "grad_norm": 0.1446470746259905,
+ "learning_rate": 6.0158121457247845e-06,
+ "loss": 0.43798986077308655,
+ "step": 1723
+ },
+ {
+ "epoch": 0.9045120671563484,
+ "grad_norm": 0.14882152623490427,
+ "learning_rate": 6.004823862318474e-06,
+ "loss": 0.4265901446342468,
+ "step": 1724
+ },
+ {
+ "epoch": 0.9050367261280168,
+ "grad_norm": 0.1506889859372045,
+ "learning_rate": 5.993893976493831e-06,
+ "loss": 0.4503232538700104,
+ "step": 1725
+ },
+ {
+ "epoch": 0.9055613850996852,
+ "grad_norm": 0.16757194686045146,
+ "learning_rate": 5.983022517944901e-06,
+ "loss": 0.4509217441082001,
+ "step": 1726
+ },
+ {
+ "epoch": 0.9060860440713536,
+ "grad_norm": 0.14447292841509515,
+ "learning_rate": 5.972209516207007e-06,
+ "loss": 0.4476190507411957,
+ "step": 1727
+ },
+ {
+ "epoch": 0.906610703043022,
+ "grad_norm": 0.1536564780391058,
+ "learning_rate": 5.961455000656656e-06,
+ "loss": 0.45448166131973267,
+ "step": 1728
+ },
+ {
+ "epoch": 0.9071353620146905,
+ "grad_norm": 0.15436323862710438,
+ "learning_rate": 5.950759000511455e-06,
+ "loss": 0.45128491520881653,
+ "step": 1729
+ },
+ {
+ "epoch": 0.9076600209863589,
+ "grad_norm": 0.15379817460186834,
+ "learning_rate": 5.940121544830047e-06,
+ "loss": 0.4490806460380554,
+ "step": 1730
+ },
+ {
+ "epoch": 0.9081846799580273,
+ "grad_norm": 0.1470796082413195,
+ "learning_rate": 5.929542662512013e-06,
+ "loss": 0.4371582865715027,
+ "step": 1731
+ },
+ {
+ "epoch": 0.9087093389296957,
+ "grad_norm": 0.15107600076433048,
+ "learning_rate": 5.919022382297809e-06,
+ "loss": 0.45175206661224365,
+ "step": 1732
+ },
+ {
+ "epoch": 0.9092339979013642,
+ "grad_norm": 0.14972329317276084,
+ "learning_rate": 5.908560732768684e-06,
+ "loss": 0.43177497386932373,
+ "step": 1733
+ },
+ {
+ "epoch": 0.9097586568730325,
+ "grad_norm": 0.14437715003340176,
+ "learning_rate": 5.898157742346601e-06,
+ "loss": 0.4243622422218323,
+ "step": 1734
+ },
+ {
+ "epoch": 0.9102833158447009,
+ "grad_norm": 0.15200146823831148,
+ "learning_rate": 5.8878134392941465e-06,
+ "loss": 0.47819408774375916,
+ "step": 1735
+ },
+ {
+ "epoch": 0.9108079748163693,
+ "grad_norm": 0.156864616244505,
+ "learning_rate": 5.877527851714479e-06,
+ "loss": 0.44744759798049927,
+ "step": 1736
+ },
+ {
+ "epoch": 0.9113326337880377,
+ "grad_norm": 0.14489667207174767,
+ "learning_rate": 5.867301007551239e-06,
+ "loss": 0.4390789866447449,
+ "step": 1737
+ },
+ {
+ "epoch": 0.9118572927597062,
+ "grad_norm": 0.1528011587102748,
+ "learning_rate": 5.857132934588479e-06,
+ "loss": 0.4338226616382599,
+ "step": 1738
+ },
+ {
+ "epoch": 0.9123819517313746,
+ "grad_norm": 0.14716603285971375,
+ "learning_rate": 5.847023660450568e-06,
+ "loss": 0.43975532054901123,
+ "step": 1739
+ },
+ {
+ "epoch": 0.912906610703043,
+ "grad_norm": 0.15266609903803277,
+ "learning_rate": 5.836973212602143e-06,
+ "loss": 0.449812650680542,
+ "step": 1740
+ },
+ {
+ "epoch": 0.9134312696747114,
+ "grad_norm": 0.1466770441701569,
+ "learning_rate": 5.826981618348025e-06,
+ "loss": 0.43771812319755554,
+ "step": 1741
+ },
+ {
+ "epoch": 0.9139559286463799,
+ "grad_norm": 0.1428112225146131,
+ "learning_rate": 5.8170489048331305e-06,
+ "loss": 0.42403268814086914,
+ "step": 1742
+ },
+ {
+ "epoch": 0.9144805876180483,
+ "grad_norm": 0.1550829416480759,
+ "learning_rate": 5.8071750990424275e-06,
+ "loss": 0.45707154273986816,
+ "step": 1743
+ },
+ {
+ "epoch": 0.9150052465897167,
+ "grad_norm": 0.14687341315819044,
+ "learning_rate": 5.797360227800832e-06,
+ "loss": 0.4558846354484558,
+ "step": 1744
+ },
+ {
+ "epoch": 0.9155299055613851,
+ "grad_norm": 0.15660848922413237,
+ "learning_rate": 5.787604317773154e-06,
+ "loss": 0.44760435819625854,
+ "step": 1745
+ },
+ {
+ "epoch": 0.9160545645330536,
+ "grad_norm": 0.14891742746821762,
+ "learning_rate": 5.777907395464015e-06,
+ "loss": 0.4266620874404907,
+ "step": 1746
+ },
+ {
+ "epoch": 0.916579223504722,
+ "grad_norm": 0.14501110957005472,
+ "learning_rate": 5.768269487217777e-06,
+ "loss": 0.4388393759727478,
+ "step": 1747
+ },
+ {
+ "epoch": 0.9171038824763903,
+ "grad_norm": 0.14651438673799927,
+ "learning_rate": 5.758690619218493e-06,
+ "loss": 0.4416884183883667,
+ "step": 1748
+ },
+ {
+ "epoch": 0.9176285414480587,
+ "grad_norm": 0.14785104880890787,
+ "learning_rate": 5.749170817489792e-06,
+ "loss": 0.4368356466293335,
+ "step": 1749
+ },
+ {
+ "epoch": 0.9181532004197271,
+ "grad_norm": 0.16586567883267808,
+ "learning_rate": 5.739710107894847e-06,
+ "loss": 0.44294649362564087,
+ "step": 1750
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 4.6320582656e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1750/video_preprocessor_config.json b/checkpoints/checkpoint-1750/video_preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..a933f0ec6941222c33aeba7980bd5acf8dd98e91
--- /dev/null
+++ b/checkpoints/checkpoint-1750/video_preprocessor_config.json
@@ -0,0 +1,30 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "do_sample_frames": true,
+ "fps": 2,
+ "image_mean": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "image_std": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "max_frames": 768,
+ "merge_size": 2,
+ "min_frames": 4,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 786432,
+ "shortest_edge": 131072
+ },
+ "temporal_patch_size": 2,
+ "video_processor_type": "Qwen3VLVideoProcessor"
+}
diff --git a/checkpoints/checkpoint-1800/chat_template.jinja b/checkpoints/checkpoint-1800/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1800/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1800/config.json b/checkpoints/checkpoint-1800/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1800/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1800/expert_prune_metadata.json b/checkpoints/checkpoint-1800/expert_prune_metadata.json
new file mode 100644
index 0000000000000000000000000000000000000000..16e01c8a10636331e334732fa9629579c68df44e
--- /dev/null
+++ b/checkpoints/checkpoint-1800/expert_prune_metadata.json
@@ -0,0 +1,14 @@
+{
+ "source_model": "/mnt/data/wuwei/model_list/Qwen3.6-35B-A3B/models--Qwen--Qwen3.6-35B-A3B/snapshots/995ad96eacd98c81ed38be0c5b274b04031597b0",
+ "output_model": "/mnt/data/wuwei/model_list/qwen3-19b-a3b",
+ "total_parameters": 18990568816,
+ "parameter_billions_decimal": 18.990568816,
+ "directory_size_label": "19B",
+ "original_num_experts": 256,
+ "kept_experts_per_layer": 128,
+ "num_profiled_layers": 40,
+ "mtp_weights": "removed because Qwen3_5MoeForConditionalGeneration does not instantiate or use the source checkpoint's mtp.* training-only tensors",
+ "expert_order": "ascending original expert index; router rows and expert tensors use the same remapping",
+ "router_statistics": "router_stats.json and router_stats.pt",
+ "selection_score": "top-8 router hit frequency * mean(abs(actual expert output after down_proj and before routing-weight multiplication))"
+}
diff --git a/checkpoints/checkpoint-1800/generation_config.json b/checkpoints/checkpoint-1800/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1800/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1800/model.safetensors b/checkpoints/checkpoint-1800/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..aa468c7f50cf3a58998c869f44bcdee6deca6e91
--- /dev/null
+++ b/checkpoints/checkpoint-1800/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75a1054d79f4b9570eef80e070a0f2b1188d79a5ef5d8b87d97abe03092d06b6
+size 37983421000
diff --git a/checkpoints/checkpoint-1800/preprocessor_config.json b/checkpoints/checkpoint-1800/preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ec43d1f859ecadf800ebae92fda4a812d34975d3
--- /dev/null
+++ b/checkpoints/checkpoint-1800/preprocessor_config.json
@@ -0,0 +1,26 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "image_mean": [
+ 0.48145466,
+ 0.4578275,
+ 0.40821073
+ ],
+ "image_processor_type": "Qwen2VLImageProcessorFast",
+ "image_std": [
+ 0.26862954,
+ 0.26130258,
+ 0.27577711
+ ],
+ "merge_size": 2,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 1003520,
+ "shortest_edge": 3136
+ },
+ "temporal_patch_size": 2
+}
diff --git a/checkpoints/checkpoint-1800/tokenizer.json b/checkpoints/checkpoint-1800/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1800/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1800/tokenizer_config.json b/checkpoints/checkpoint-1800/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1800/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1800/trainer_state.json b/checkpoints/checkpoint-1800/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..1fb437082cbeaa83e427c54fca4acdfafa0ef291
--- /dev/null
+++ b/checkpoints/checkpoint-1800/trainer_state.json
@@ -0,0 +1,12634 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.944386149003148,
+ "eval_steps": 500,
+ "global_step": 1800,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ },
+ {
+ "epoch": 0.5776495278069255,
+ "grad_norm": 0.1862053032854098,
+ "learning_rate": 2.2065553964033126e-05,
+ "loss": 0.4816875457763672,
+ "step": 1101
+ },
+ {
+ "epoch": 0.5781741867785939,
+ "grad_norm": 0.16996015894363117,
+ "learning_rate": 2.20295734045532e-05,
+ "loss": 0.4471197724342346,
+ "step": 1102
+ },
+ {
+ "epoch": 0.5786988457502623,
+ "grad_norm": 0.16889227008390692,
+ "learning_rate": 2.1993607706997725e-05,
+ "loss": 0.468391478061676,
+ "step": 1103
+ },
+ {
+ "epoch": 0.5792235047219307,
+ "grad_norm": 0.16670931681562373,
+ "learning_rate": 2.195765696907745e-05,
+ "loss": 0.44934162497520447,
+ "step": 1104
+ },
+ {
+ "epoch": 0.5797481636935992,
+ "grad_norm": 0.16328418741223716,
+ "learning_rate": 2.1921721288462482e-05,
+ "loss": 0.4558270275592804,
+ "step": 1105
+ },
+ {
+ "epoch": 0.5802728226652676,
+ "grad_norm": 0.17364660408703936,
+ "learning_rate": 2.188580076278203e-05,
+ "loss": 0.4597751796245575,
+ "step": 1106
+ },
+ {
+ "epoch": 0.580797481636936,
+ "grad_norm": 0.1628618453800601,
+ "learning_rate": 2.18498954896241e-05,
+ "loss": 0.4521859288215637,
+ "step": 1107
+ },
+ {
+ "epoch": 0.5813221406086044,
+ "grad_norm": 0.16813817951871482,
+ "learning_rate": 2.1814005566535302e-05,
+ "loss": 0.4581005573272705,
+ "step": 1108
+ },
+ {
+ "epoch": 0.5818467995802729,
+ "grad_norm": 0.16825923629726025,
+ "learning_rate": 2.1778131091020516e-05,
+ "loss": 0.4433916211128235,
+ "step": 1109
+ },
+ {
+ "epoch": 0.5823714585519413,
+ "grad_norm": 0.17194743467883944,
+ "learning_rate": 2.1742272160542654e-05,
+ "loss": 0.47296425700187683,
+ "step": 1110
+ },
+ {
+ "epoch": 0.5828961175236097,
+ "grad_norm": 0.16947730546039969,
+ "learning_rate": 2.170642887252241e-05,
+ "loss": 0.4815937876701355,
+ "step": 1111
+ },
+ {
+ "epoch": 0.583420776495278,
+ "grad_norm": 0.1657449928668473,
+ "learning_rate": 2.167060132433798e-05,
+ "loss": 0.4479164779186249,
+ "step": 1112
+ },
+ {
+ "epoch": 0.5839454354669464,
+ "grad_norm": 0.1658952516607364,
+ "learning_rate": 2.1634789613324778e-05,
+ "loss": 0.4637463688850403,
+ "step": 1113
+ },
+ {
+ "epoch": 0.5844700944386149,
+ "grad_norm": 0.16777002188998236,
+ "learning_rate": 2.1598993836775218e-05,
+ "loss": 0.457524836063385,
+ "step": 1114
+ },
+ {
+ "epoch": 0.5849947534102833,
+ "grad_norm": 0.15837566447341322,
+ "learning_rate": 2.1563214091938405e-05,
+ "loss": 0.44938087463378906,
+ "step": 1115
+ },
+ {
+ "epoch": 0.5855194123819517,
+ "grad_norm": 0.17162206617035314,
+ "learning_rate": 2.1527450476019896e-05,
+ "loss": 0.4393450617790222,
+ "step": 1116
+ },
+ {
+ "epoch": 0.5860440713536201,
+ "grad_norm": 0.16224101921433814,
+ "learning_rate": 2.1491703086181426e-05,
+ "loss": 0.451913982629776,
+ "step": 1117
+ },
+ {
+ "epoch": 0.5865687303252886,
+ "grad_norm": 0.1736253504575144,
+ "learning_rate": 2.145597201954066e-05,
+ "loss": 0.4611290395259857,
+ "step": 1118
+ },
+ {
+ "epoch": 0.587093389296957,
+ "grad_norm": 0.16308408333737234,
+ "learning_rate": 2.1420257373170903e-05,
+ "loss": 0.45898568630218506,
+ "step": 1119
+ },
+ {
+ "epoch": 0.5876180482686254,
+ "grad_norm": 0.16948634794175524,
+ "learning_rate": 2.1384559244100848e-05,
+ "loss": 0.45982882380485535,
+ "step": 1120
+ },
+ {
+ "epoch": 0.5881427072402938,
+ "grad_norm": 0.16676366538565252,
+ "learning_rate": 2.1348877729314336e-05,
+ "loss": 0.44826576113700867,
+ "step": 1121
+ },
+ {
+ "epoch": 0.5886673662119623,
+ "grad_norm": 0.16431125630712295,
+ "learning_rate": 2.1313212925750037e-05,
+ "loss": 0.4616445302963257,
+ "step": 1122
+ },
+ {
+ "epoch": 0.5891920251836307,
+ "grad_norm": 0.1631659061616955,
+ "learning_rate": 2.1277564930301252e-05,
+ "loss": 0.45651188492774963,
+ "step": 1123
+ },
+ {
+ "epoch": 0.5897166841552991,
+ "grad_norm": 0.15805950502689237,
+ "learning_rate": 2.1241933839815604e-05,
+ "loss": 0.4486045837402344,
+ "step": 1124
+ },
+ {
+ "epoch": 0.5902413431269675,
+ "grad_norm": 0.16530288038603502,
+ "learning_rate": 2.1206319751094773e-05,
+ "loss": 0.4485066533088684,
+ "step": 1125
+ },
+ {
+ "epoch": 0.5907660020986358,
+ "grad_norm": 0.16061903194239047,
+ "learning_rate": 2.1170722760894295e-05,
+ "loss": 0.4565330743789673,
+ "step": 1126
+ },
+ {
+ "epoch": 0.5912906610703043,
+ "grad_norm": 0.1643346999353627,
+ "learning_rate": 2.1135142965923195e-05,
+ "loss": 0.45560717582702637,
+ "step": 1127
+ },
+ {
+ "epoch": 0.5918153200419727,
+ "grad_norm": 0.17484644315215753,
+ "learning_rate": 2.109958046284382e-05,
+ "loss": 0.4795891046524048,
+ "step": 1128
+ },
+ {
+ "epoch": 0.5923399790136411,
+ "grad_norm": 0.1623762678622651,
+ "learning_rate": 2.1064035348271528e-05,
+ "loss": 0.47395604848861694,
+ "step": 1129
+ },
+ {
+ "epoch": 0.5928646379853095,
+ "grad_norm": 0.16888824241669015,
+ "learning_rate": 2.1028507718774456e-05,
+ "loss": 0.4601455330848694,
+ "step": 1130
+ },
+ {
+ "epoch": 0.593389296956978,
+ "grad_norm": 0.16291744568806302,
+ "learning_rate": 2.0992997670873193e-05,
+ "loss": 0.46248236298561096,
+ "step": 1131
+ },
+ {
+ "epoch": 0.5939139559286464,
+ "grad_norm": 0.16185598691236988,
+ "learning_rate": 2.0957505301040603e-05,
+ "loss": 0.45085301995277405,
+ "step": 1132
+ },
+ {
+ "epoch": 0.5944386149003148,
+ "grad_norm": 0.1678000806117814,
+ "learning_rate": 2.0922030705701518e-05,
+ "loss": 0.44939014315605164,
+ "step": 1133
+ },
+ {
+ "epoch": 0.5949632738719832,
+ "grad_norm": 0.16104474282288994,
+ "learning_rate": 2.088657398123246e-05,
+ "loss": 0.4313976764678955,
+ "step": 1134
+ },
+ {
+ "epoch": 0.5954879328436516,
+ "grad_norm": 0.1631128968199669,
+ "learning_rate": 2.085113522396141e-05,
+ "loss": 0.4508278965950012,
+ "step": 1135
+ },
+ {
+ "epoch": 0.5960125918153201,
+ "grad_norm": 0.16307688195844533,
+ "learning_rate": 2.0815714530167557e-05,
+ "loss": 0.4467576742172241,
+ "step": 1136
+ },
+ {
+ "epoch": 0.5965372507869885,
+ "grad_norm": 0.16456832304532207,
+ "learning_rate": 2.0780311996080976e-05,
+ "loss": 0.4621703624725342,
+ "step": 1137
+ },
+ {
+ "epoch": 0.5970619097586569,
+ "grad_norm": 0.16346217748325367,
+ "learning_rate": 2.0744927717882435e-05,
+ "loss": 0.4589449465274811,
+ "step": 1138
+ },
+ {
+ "epoch": 0.5975865687303253,
+ "grad_norm": 0.16180691501076994,
+ "learning_rate": 2.0709561791703097e-05,
+ "loss": 0.43698400259017944,
+ "step": 1139
+ },
+ {
+ "epoch": 0.5981112277019937,
+ "grad_norm": 0.15695459976898835,
+ "learning_rate": 2.067421431362427e-05,
+ "loss": 0.452271044254303,
+ "step": 1140
+ },
+ {
+ "epoch": 0.5986358866736621,
+ "grad_norm": 0.15930893553721054,
+ "learning_rate": 2.063888537967713e-05,
+ "loss": 0.44868963956832886,
+ "step": 1141
+ },
+ {
+ "epoch": 0.5991605456453305,
+ "grad_norm": 0.15963802284017928,
+ "learning_rate": 2.0603575085842485e-05,
+ "loss": 0.4376864731311798,
+ "step": 1142
+ },
+ {
+ "epoch": 0.5996852046169989,
+ "grad_norm": 0.16610266900507178,
+ "learning_rate": 2.05682835280505e-05,
+ "loss": 0.44676727056503296,
+ "step": 1143
+ },
+ {
+ "epoch": 0.6002098635886673,
+ "grad_norm": 0.15978253236669612,
+ "learning_rate": 2.053301080218044e-05,
+ "loss": 0.434773713350296,
+ "step": 1144
+ },
+ {
+ "epoch": 0.6007345225603358,
+ "grad_norm": 0.16305141803526324,
+ "learning_rate": 2.0497757004060403e-05,
+ "loss": 0.4529780149459839,
+ "step": 1145
+ },
+ {
+ "epoch": 0.6012591815320042,
+ "grad_norm": 0.1649737864101215,
+ "learning_rate": 2.046252222946707e-05,
+ "loss": 0.4571434259414673,
+ "step": 1146
+ },
+ {
+ "epoch": 0.6017838405036726,
+ "grad_norm": 0.15738824968326598,
+ "learning_rate": 2.0427306574125427e-05,
+ "loss": 0.4446289539337158,
+ "step": 1147
+ },
+ {
+ "epoch": 0.602308499475341,
+ "grad_norm": 0.1731139535185187,
+ "learning_rate": 2.039211013370853e-05,
+ "loss": 0.4551090598106384,
+ "step": 1148
+ },
+ {
+ "epoch": 0.6028331584470095,
+ "grad_norm": 0.15857251900043037,
+ "learning_rate": 2.0356933003837237e-05,
+ "loss": 0.4644010663032532,
+ "step": 1149
+ },
+ {
+ "epoch": 0.6033578174186779,
+ "grad_norm": 0.1589582861134211,
+ "learning_rate": 2.032177528007992e-05,
+ "loss": 0.45741650462150574,
+ "step": 1150
+ },
+ {
+ "epoch": 0.6038824763903463,
+ "grad_norm": 0.15639092706575122,
+ "learning_rate": 2.0286637057952257e-05,
+ "loss": 0.4375547766685486,
+ "step": 1151
+ },
+ {
+ "epoch": 0.6044071353620147,
+ "grad_norm": 0.16335605581337187,
+ "learning_rate": 2.0251518432916933e-05,
+ "loss": 0.4636140465736389,
+ "step": 1152
+ },
+ {
+ "epoch": 0.6049317943336832,
+ "grad_norm": 0.15437437113359356,
+ "learning_rate": 2.0216419500383372e-05,
+ "loss": 0.44632142782211304,
+ "step": 1153
+ },
+ {
+ "epoch": 0.6054564533053515,
+ "grad_norm": 0.15522039198698664,
+ "learning_rate": 2.0181340355707545e-05,
+ "loss": 0.449049174785614,
+ "step": 1154
+ },
+ {
+ "epoch": 0.6059811122770199,
+ "grad_norm": 0.1599951917909323,
+ "learning_rate": 2.0146281094191604e-05,
+ "loss": 0.44395962357521057,
+ "step": 1155
+ },
+ {
+ "epoch": 0.6065057712486883,
+ "grad_norm": 0.16866621620527716,
+ "learning_rate": 2.011124181108373e-05,
+ "loss": 0.45693716406822205,
+ "step": 1156
+ },
+ {
+ "epoch": 0.6070304302203567,
+ "grad_norm": 0.15508935650135075,
+ "learning_rate": 2.0076222601577815e-05,
+ "loss": 0.4525151252746582,
+ "step": 1157
+ },
+ {
+ "epoch": 0.6075550891920252,
+ "grad_norm": 0.16508998118005971,
+ "learning_rate": 2.004122356081321e-05,
+ "loss": 0.466006875038147,
+ "step": 1158
+ },
+ {
+ "epoch": 0.6080797481636936,
+ "grad_norm": 0.15817418864022237,
+ "learning_rate": 2.0006244783874466e-05,
+ "loss": 0.4496110677719116,
+ "step": 1159
+ },
+ {
+ "epoch": 0.608604407135362,
+ "grad_norm": 0.16127270259749824,
+ "learning_rate": 1.997128636579109e-05,
+ "loss": 0.45918628573417664,
+ "step": 1160
+ },
+ {
+ "epoch": 0.6091290661070304,
+ "grad_norm": 0.1702599924159677,
+ "learning_rate": 1.9936348401537292e-05,
+ "loss": 0.47218555212020874,
+ "step": 1161
+ },
+ {
+ "epoch": 0.6096537250786989,
+ "grad_norm": 0.15690176248846127,
+ "learning_rate": 1.9901430986031695e-05,
+ "loss": 0.4516097903251648,
+ "step": 1162
+ },
+ {
+ "epoch": 0.6101783840503673,
+ "grad_norm": 0.1713614405614144,
+ "learning_rate": 1.9866534214137085e-05,
+ "loss": 0.45030856132507324,
+ "step": 1163
+ },
+ {
+ "epoch": 0.6107030430220357,
+ "grad_norm": 0.1670865563026229,
+ "learning_rate": 1.9831658180660195e-05,
+ "loss": 0.45603489875793457,
+ "step": 1164
+ },
+ {
+ "epoch": 0.6112277019937041,
+ "grad_norm": 0.16897539113401952,
+ "learning_rate": 1.9796802980351392e-05,
+ "loss": 0.46314537525177,
+ "step": 1165
+ },
+ {
+ "epoch": 0.6117523609653726,
+ "grad_norm": 0.15832868870697644,
+ "learning_rate": 1.976196870790446e-05,
+ "loss": 0.44784611463546753,
+ "step": 1166
+ },
+ {
+ "epoch": 0.612277019937041,
+ "grad_norm": 0.16595365551891114,
+ "learning_rate": 1.9727155457956315e-05,
+ "loss": 0.4409624934196472,
+ "step": 1167
+ },
+ {
+ "epoch": 0.6128016789087093,
+ "grad_norm": 0.1642630411212017,
+ "learning_rate": 1.969236332508677e-05,
+ "loss": 0.46120351552963257,
+ "step": 1168
+ },
+ {
+ "epoch": 0.6133263378803777,
+ "grad_norm": 0.16481796135868812,
+ "learning_rate": 1.9657592403818253e-05,
+ "loss": 0.45544904470443726,
+ "step": 1169
+ },
+ {
+ "epoch": 0.6138509968520461,
+ "grad_norm": 0.1629832779852364,
+ "learning_rate": 1.962284278861557e-05,
+ "loss": 0.4578462839126587,
+ "step": 1170
+ },
+ {
+ "epoch": 0.6143756558237146,
+ "grad_norm": 0.15916219613603688,
+ "learning_rate": 1.9588114573885664e-05,
+ "loss": 0.4546983540058136,
+ "step": 1171
+ },
+ {
+ "epoch": 0.614900314795383,
+ "grad_norm": 0.1660520165166075,
+ "learning_rate": 1.95534078539773e-05,
+ "loss": 0.4601942300796509,
+ "step": 1172
+ },
+ {
+ "epoch": 0.6154249737670514,
+ "grad_norm": 0.16924794578577768,
+ "learning_rate": 1.9518722723180877e-05,
+ "loss": 0.45997846126556396,
+ "step": 1173
+ },
+ {
+ "epoch": 0.6159496327387198,
+ "grad_norm": 0.16597646133045382,
+ "learning_rate": 1.9484059275728127e-05,
+ "loss": 0.4594605565071106,
+ "step": 1174
+ },
+ {
+ "epoch": 0.6164742917103883,
+ "grad_norm": 0.16964243792519748,
+ "learning_rate": 1.9449417605791876e-05,
+ "loss": 0.4638690650463104,
+ "step": 1175
+ },
+ {
+ "epoch": 0.6169989506820567,
+ "grad_norm": 0.1758591426791696,
+ "learning_rate": 1.9414797807485796e-05,
+ "loss": 0.4666447639465332,
+ "step": 1176
+ },
+ {
+ "epoch": 0.6175236096537251,
+ "grad_norm": 0.16217661227891753,
+ "learning_rate": 1.9380199974864126e-05,
+ "loss": 0.44252535700798035,
+ "step": 1177
+ },
+ {
+ "epoch": 0.6180482686253935,
+ "grad_norm": 0.1576446184499019,
+ "learning_rate": 1.934562420192142e-05,
+ "loss": 0.4499641954898834,
+ "step": 1178
+ },
+ {
+ "epoch": 0.618572927597062,
+ "grad_norm": 0.16231008092636912,
+ "learning_rate": 1.9311070582592327e-05,
+ "loss": 0.4561542570590973,
+ "step": 1179
+ },
+ {
+ "epoch": 0.6190975865687304,
+ "grad_norm": 0.16790382249577043,
+ "learning_rate": 1.9276539210751292e-05,
+ "loss": 0.4484325647354126,
+ "step": 1180
+ },
+ {
+ "epoch": 0.6196222455403987,
+ "grad_norm": 0.18794729932605855,
+ "learning_rate": 1.9242030180212323e-05,
+ "loss": 0.44169026613235474,
+ "step": 1181
+ },
+ {
+ "epoch": 0.6201469045120671,
+ "grad_norm": 0.16948051243889814,
+ "learning_rate": 1.920754358472872e-05,
+ "loss": 0.4592876434326172,
+ "step": 1182
+ },
+ {
+ "epoch": 0.6206715634837355,
+ "grad_norm": 0.16762558710029374,
+ "learning_rate": 1.917307951799286e-05,
+ "loss": 0.4402095079421997,
+ "step": 1183
+ },
+ {
+ "epoch": 0.621196222455404,
+ "grad_norm": 0.18170280116387003,
+ "learning_rate": 1.9138638073635885e-05,
+ "loss": 0.4633026123046875,
+ "step": 1184
+ },
+ {
+ "epoch": 0.6217208814270724,
+ "grad_norm": 0.1718051709136108,
+ "learning_rate": 1.9104219345227486e-05,
+ "loss": 0.4510066509246826,
+ "step": 1185
+ },
+ {
+ "epoch": 0.6222455403987408,
+ "grad_norm": 0.16829066326345576,
+ "learning_rate": 1.906982342627565e-05,
+ "loss": 0.4632919430732727,
+ "step": 1186
+ },
+ {
+ "epoch": 0.6227701993704092,
+ "grad_norm": 0.16669141808548274,
+ "learning_rate": 1.9035450410226395e-05,
+ "loss": 0.455544650554657,
+ "step": 1187
+ },
+ {
+ "epoch": 0.6232948583420777,
+ "grad_norm": 0.16679996559722232,
+ "learning_rate": 1.90011003904635e-05,
+ "loss": 0.459994375705719,
+ "step": 1188
+ },
+ {
+ "epoch": 0.6238195173137461,
+ "grad_norm": 0.17379754713278164,
+ "learning_rate": 1.8966773460308284e-05,
+ "loss": 0.4576749801635742,
+ "step": 1189
+ },
+ {
+ "epoch": 0.6243441762854145,
+ "grad_norm": 0.17088539733389246,
+ "learning_rate": 1.8932469713019335e-05,
+ "loss": 0.44853052496910095,
+ "step": 1190
+ },
+ {
+ "epoch": 0.6248688352570829,
+ "grad_norm": 0.15556883889201015,
+ "learning_rate": 1.8898189241792248e-05,
+ "loss": 0.44060447812080383,
+ "step": 1191
+ },
+ {
+ "epoch": 0.6253934942287513,
+ "grad_norm": 0.1640093917439215,
+ "learning_rate": 1.8863932139759388e-05,
+ "loss": 0.4531911611557007,
+ "step": 1192
+ },
+ {
+ "epoch": 0.6259181532004198,
+ "grad_norm": 0.16444284826591576,
+ "learning_rate": 1.882969849998964e-05,
+ "loss": 0.44176363945007324,
+ "step": 1193
+ },
+ {
+ "epoch": 0.6264428121720882,
+ "grad_norm": 0.1696462520995642,
+ "learning_rate": 1.8795488415488143e-05,
+ "loss": 0.44516652822494507,
+ "step": 1194
+ },
+ {
+ "epoch": 0.6269674711437565,
+ "grad_norm": 0.16810724705191796,
+ "learning_rate": 1.8761301979196032e-05,
+ "loss": 0.4542563855648041,
+ "step": 1195
+ },
+ {
+ "epoch": 0.6274921301154249,
+ "grad_norm": 0.17334813313855824,
+ "learning_rate": 1.8727139283990204e-05,
+ "loss": 0.47769248485565186,
+ "step": 1196
+ },
+ {
+ "epoch": 0.6280167890870934,
+ "grad_norm": 0.16308212035303235,
+ "learning_rate": 1.8693000422683043e-05,
+ "loss": 0.4407799243927002,
+ "step": 1197
+ },
+ {
+ "epoch": 0.6285414480587618,
+ "grad_norm": 0.16853319139928977,
+ "learning_rate": 1.8658885488022204e-05,
+ "loss": 0.4565497040748596,
+ "step": 1198
+ },
+ {
+ "epoch": 0.6290661070304302,
+ "grad_norm": 0.16694968781284703,
+ "learning_rate": 1.8624794572690328e-05,
+ "loss": 0.44522374868392944,
+ "step": 1199
+ },
+ {
+ "epoch": 0.6295907660020986,
+ "grad_norm": 0.16418579468339103,
+ "learning_rate": 1.8590727769304788e-05,
+ "loss": 0.4441048502922058,
+ "step": 1200
+ },
+ {
+ "epoch": 0.630115424973767,
+ "grad_norm": 0.1661930777124156,
+ "learning_rate": 1.8556685170417486e-05,
+ "loss": 0.457327663898468,
+ "step": 1201
+ },
+ {
+ "epoch": 0.6306400839454355,
+ "grad_norm": 0.16536630027039814,
+ "learning_rate": 1.852266686851452e-05,
+ "loss": 0.462494432926178,
+ "step": 1202
+ },
+ {
+ "epoch": 0.6311647429171039,
+ "grad_norm": 0.16611474088307454,
+ "learning_rate": 1.8488672956016013e-05,
+ "loss": 0.4557701647281647,
+ "step": 1203
+ },
+ {
+ "epoch": 0.6316894018887723,
+ "grad_norm": 0.16759343079749697,
+ "learning_rate": 1.8454703525275813e-05,
+ "loss": 0.45069995522499084,
+ "step": 1204
+ },
+ {
+ "epoch": 0.6322140608604407,
+ "grad_norm": 0.16186676294485336,
+ "learning_rate": 1.8420758668581255e-05,
+ "loss": 0.44954800605773926,
+ "step": 1205
+ },
+ {
+ "epoch": 0.6327387198321092,
+ "grad_norm": 0.15687159835368553,
+ "learning_rate": 1.8386838478152916e-05,
+ "loss": 0.4490429759025574,
+ "step": 1206
+ },
+ {
+ "epoch": 0.6332633788037776,
+ "grad_norm": 0.16242951976115577,
+ "learning_rate": 1.8352943046144372e-05,
+ "loss": 0.45819586515426636,
+ "step": 1207
+ },
+ {
+ "epoch": 0.633788037775446,
+ "grad_norm": 0.16516087266517232,
+ "learning_rate": 1.831907246464191e-05,
+ "loss": 0.44769883155822754,
+ "step": 1208
+ },
+ {
+ "epoch": 0.6343126967471143,
+ "grad_norm": 0.16709475871512644,
+ "learning_rate": 1.8285226825664332e-05,
+ "loss": 0.4699128568172455,
+ "step": 1209
+ },
+ {
+ "epoch": 0.6348373557187827,
+ "grad_norm": 0.16288623150595774,
+ "learning_rate": 1.8251406221162636e-05,
+ "loss": 0.45322558283805847,
+ "step": 1210
+ },
+ {
+ "epoch": 0.6353620146904512,
+ "grad_norm": 0.1675722848054237,
+ "learning_rate": 1.821761074301986e-05,
+ "loss": 0.4752347469329834,
+ "step": 1211
+ },
+ {
+ "epoch": 0.6358866736621196,
+ "grad_norm": 0.163285963514504,
+ "learning_rate": 1.818384048305074e-05,
+ "loss": 0.441958487033844,
+ "step": 1212
+ },
+ {
+ "epoch": 0.636411332633788,
+ "grad_norm": 0.17078102287557764,
+ "learning_rate": 1.8150095533001517e-05,
+ "loss": 0.4551333487033844,
+ "step": 1213
+ },
+ {
+ "epoch": 0.6369359916054564,
+ "grad_norm": 0.16188540746097138,
+ "learning_rate": 1.811637598454966e-05,
+ "loss": 0.4470779597759247,
+ "step": 1214
+ },
+ {
+ "epoch": 0.6374606505771249,
+ "grad_norm": 0.15235628218958286,
+ "learning_rate": 1.808268192930364e-05,
+ "loss": 0.43604153394699097,
+ "step": 1215
+ },
+ {
+ "epoch": 0.6379853095487933,
+ "grad_norm": 0.1604601949959104,
+ "learning_rate": 1.8049013458802665e-05,
+ "loss": 0.4445713758468628,
+ "step": 1216
+ },
+ {
+ "epoch": 0.6385099685204617,
+ "grad_norm": 0.15511797131215901,
+ "learning_rate": 1.8015370664516427e-05,
+ "loss": 0.4455301761627197,
+ "step": 1217
+ },
+ {
+ "epoch": 0.6390346274921301,
+ "grad_norm": 0.16815669546944095,
+ "learning_rate": 1.798175363784487e-05,
+ "loss": 0.46135014295578003,
+ "step": 1218
+ },
+ {
+ "epoch": 0.6395592864637986,
+ "grad_norm": 0.161558177688922,
+ "learning_rate": 1.7948162470117917e-05,
+ "loss": 0.4519304037094116,
+ "step": 1219
+ },
+ {
+ "epoch": 0.640083945435467,
+ "grad_norm": 0.16677842130048232,
+ "learning_rate": 1.7914597252595265e-05,
+ "loss": 0.4550550878047943,
+ "step": 1220
+ },
+ {
+ "epoch": 0.6406086044071354,
+ "grad_norm": 0.1597331480423138,
+ "learning_rate": 1.7881058076466095e-05,
+ "loss": 0.441120445728302,
+ "step": 1221
+ },
+ {
+ "epoch": 0.6411332633788038,
+ "grad_norm": 0.15951106545246824,
+ "learning_rate": 1.7847545032848844e-05,
+ "loss": 0.4471382200717926,
+ "step": 1222
+ },
+ {
+ "epoch": 0.6416579223504721,
+ "grad_norm": 0.16157787674667565,
+ "learning_rate": 1.7814058212790947e-05,
+ "loss": 0.4550648331642151,
+ "step": 1223
+ },
+ {
+ "epoch": 0.6421825813221406,
+ "grad_norm": 0.16989186763352415,
+ "learning_rate": 1.7780597707268583e-05,
+ "loss": 0.4515089988708496,
+ "step": 1224
+ },
+ {
+ "epoch": 0.642707240293809,
+ "grad_norm": 0.15952153223099508,
+ "learning_rate": 1.7747163607186468e-05,
+ "loss": 0.4658638834953308,
+ "step": 1225
+ },
+ {
+ "epoch": 0.6432318992654774,
+ "grad_norm": 0.15205926699647218,
+ "learning_rate": 1.7713756003377567e-05,
+ "loss": 0.4330868422985077,
+ "step": 1226
+ },
+ {
+ "epoch": 0.6437565582371458,
+ "grad_norm": 0.15784806480936794,
+ "learning_rate": 1.7680374986602865e-05,
+ "loss": 0.4506068527698517,
+ "step": 1227
+ },
+ {
+ "epoch": 0.6442812172088143,
+ "grad_norm": 0.16006709280686687,
+ "learning_rate": 1.7647020647551105e-05,
+ "loss": 0.446941077709198,
+ "step": 1228
+ },
+ {
+ "epoch": 0.6448058761804827,
+ "grad_norm": 0.16064631158222062,
+ "learning_rate": 1.761369307683856e-05,
+ "loss": 0.4418525695800781,
+ "step": 1229
+ },
+ {
+ "epoch": 0.6453305351521511,
+ "grad_norm": 0.16454136480444392,
+ "learning_rate": 1.7580392365008767e-05,
+ "loss": 0.46268391609191895,
+ "step": 1230
+ },
+ {
+ "epoch": 0.6458551941238195,
+ "grad_norm": 0.15401246535771462,
+ "learning_rate": 1.754711860253233e-05,
+ "loss": 0.4372989535331726,
+ "step": 1231
+ },
+ {
+ "epoch": 0.646379853095488,
+ "grad_norm": 0.1544567975028383,
+ "learning_rate": 1.7513871879806603e-05,
+ "loss": 0.44806575775146484,
+ "step": 1232
+ },
+ {
+ "epoch": 0.6469045120671564,
+ "grad_norm": 0.16411225112730965,
+ "learning_rate": 1.748065228715549e-05,
+ "loss": 0.43782058358192444,
+ "step": 1233
+ },
+ {
+ "epoch": 0.6474291710388248,
+ "grad_norm": 0.1552767832972093,
+ "learning_rate": 1.7447459914829183e-05,
+ "loss": 0.44531378149986267,
+ "step": 1234
+ },
+ {
+ "epoch": 0.6479538300104932,
+ "grad_norm": 0.16214695257123513,
+ "learning_rate": 1.741429485300392e-05,
+ "loss": 0.47399842739105225,
+ "step": 1235
+ },
+ {
+ "epoch": 0.6484784889821616,
+ "grad_norm": 0.16684861111484112,
+ "learning_rate": 1.7381157191781766e-05,
+ "loss": 0.4654282331466675,
+ "step": 1236
+ },
+ {
+ "epoch": 0.64900314795383,
+ "grad_norm": 0.15901864178630273,
+ "learning_rate": 1.7348047021190334e-05,
+ "loss": 0.45128414034843445,
+ "step": 1237
+ },
+ {
+ "epoch": 0.6495278069254984,
+ "grad_norm": 0.15360991961742707,
+ "learning_rate": 1.7314964431182523e-05,
+ "loss": 0.4303375482559204,
+ "step": 1238
+ },
+ {
+ "epoch": 0.6500524658971668,
+ "grad_norm": 0.1633151574385741,
+ "learning_rate": 1.728190951163633e-05,
+ "loss": 0.45568403601646423,
+ "step": 1239
+ },
+ {
+ "epoch": 0.6505771248688352,
+ "grad_norm": 0.16377853571058645,
+ "learning_rate": 1.724888235235457e-05,
+ "loss": 0.46503543853759766,
+ "step": 1240
+ },
+ {
+ "epoch": 0.6511017838405037,
+ "grad_norm": 0.1595023801275368,
+ "learning_rate": 1.7215883043064655e-05,
+ "loss": 0.46337294578552246,
+ "step": 1241
+ },
+ {
+ "epoch": 0.6516264428121721,
+ "grad_norm": 0.16324796182212292,
+ "learning_rate": 1.7182911673418313e-05,
+ "loss": 0.46197354793548584,
+ "step": 1242
+ },
+ {
+ "epoch": 0.6521511017838405,
+ "grad_norm": 0.16353734830879765,
+ "learning_rate": 1.714996833299138e-05,
+ "loss": 0.45741671323776245,
+ "step": 1243
+ },
+ {
+ "epoch": 0.6526757607555089,
+ "grad_norm": 0.15941587678263386,
+ "learning_rate": 1.7117053111283528e-05,
+ "loss": 0.4514503479003906,
+ "step": 1244
+ },
+ {
+ "epoch": 0.6532004197271774,
+ "grad_norm": 0.1642653230145008,
+ "learning_rate": 1.708416609771804e-05,
+ "loss": 0.4437655210494995,
+ "step": 1245
+ },
+ {
+ "epoch": 0.6537250786988458,
+ "grad_norm": 0.15814155519130152,
+ "learning_rate": 1.7051307381641587e-05,
+ "loss": 0.4314163327217102,
+ "step": 1246
+ },
+ {
+ "epoch": 0.6542497376705142,
+ "grad_norm": 0.16086078370768056,
+ "learning_rate": 1.701847705232394e-05,
+ "loss": 0.4526430666446686,
+ "step": 1247
+ },
+ {
+ "epoch": 0.6547743966421826,
+ "grad_norm": 0.16389360190715183,
+ "learning_rate": 1.698567519895775e-05,
+ "loss": 0.4278966188430786,
+ "step": 1248
+ },
+ {
+ "epoch": 0.655299055613851,
+ "grad_norm": 0.16303142287804262,
+ "learning_rate": 1.6952901910658316e-05,
+ "loss": 0.45778220891952515,
+ "step": 1249
+ },
+ {
+ "epoch": 0.6558237145855194,
+ "grad_norm": 0.1675209956904048,
+ "learning_rate": 1.6920157276463318e-05,
+ "loss": 0.4489403963088989,
+ "step": 1250
+ },
+ {
+ "epoch": 0.6563483735571878,
+ "grad_norm": 0.16228311551074323,
+ "learning_rate": 1.6887441385332604e-05,
+ "loss": 0.4534342885017395,
+ "step": 1251
+ },
+ {
+ "epoch": 0.6568730325288562,
+ "grad_norm": 0.16408957851046913,
+ "learning_rate": 1.6854754326147916e-05,
+ "loss": 0.4549208879470825,
+ "step": 1252
+ },
+ {
+ "epoch": 0.6573976915005246,
+ "grad_norm": 0.15967922111432845,
+ "learning_rate": 1.682209618771268e-05,
+ "loss": 0.4451141357421875,
+ "step": 1253
+ },
+ {
+ "epoch": 0.657922350472193,
+ "grad_norm": 0.16440698722204208,
+ "learning_rate": 1.678946705875176e-05,
+ "loss": 0.454200804233551,
+ "step": 1254
+ },
+ {
+ "epoch": 0.6584470094438615,
+ "grad_norm": 0.16251120184938497,
+ "learning_rate": 1.675686702791117e-05,
+ "loss": 0.4580027461051941,
+ "step": 1255
+ },
+ {
+ "epoch": 0.6589716684155299,
+ "grad_norm": 0.1643717386150457,
+ "learning_rate": 1.6724296183757924e-05,
+ "loss": 0.4422311782836914,
+ "step": 1256
+ },
+ {
+ "epoch": 0.6594963273871983,
+ "grad_norm": 0.1644586414378722,
+ "learning_rate": 1.66917546147797e-05,
+ "loss": 0.4593179225921631,
+ "step": 1257
+ },
+ {
+ "epoch": 0.6600209863588667,
+ "grad_norm": 0.16751074186967685,
+ "learning_rate": 1.6659242409384657e-05,
+ "loss": 0.457930326461792,
+ "step": 1258
+ },
+ {
+ "epoch": 0.6605456453305352,
+ "grad_norm": 0.15398418609155334,
+ "learning_rate": 1.6626759655901182e-05,
+ "loss": 0.44002899527549744,
+ "step": 1259
+ },
+ {
+ "epoch": 0.6610703043022036,
+ "grad_norm": 0.16328113175790226,
+ "learning_rate": 1.659430644257763e-05,
+ "loss": 0.4500438868999481,
+ "step": 1260
+ },
+ {
+ "epoch": 0.661594963273872,
+ "grad_norm": 0.1694244340615106,
+ "learning_rate": 1.656188285758214e-05,
+ "loss": 0.4818427264690399,
+ "step": 1261
+ },
+ {
+ "epoch": 0.6621196222455404,
+ "grad_norm": 0.1578320701754368,
+ "learning_rate": 1.652948898900232e-05,
+ "loss": 0.4492436647415161,
+ "step": 1262
+ },
+ {
+ "epoch": 0.6626442812172089,
+ "grad_norm": 0.16538865917828469,
+ "learning_rate": 1.6497124924845064e-05,
+ "loss": 0.45070338249206543,
+ "step": 1263
+ },
+ {
+ "epoch": 0.6631689401888772,
+ "grad_norm": 0.15670817848256852,
+ "learning_rate": 1.6464790753036296e-05,
+ "loss": 0.45439285039901733,
+ "step": 1264
+ },
+ {
+ "epoch": 0.6636935991605456,
+ "grad_norm": 0.16184786585965885,
+ "learning_rate": 1.64324865614207e-05,
+ "loss": 0.446522980928421,
+ "step": 1265
+ },
+ {
+ "epoch": 0.664218258132214,
+ "grad_norm": 0.15719726462251984,
+ "learning_rate": 1.6400212437761554e-05,
+ "loss": 0.4495280981063843,
+ "step": 1266
+ },
+ {
+ "epoch": 0.6647429171038824,
+ "grad_norm": 0.1680804414134405,
+ "learning_rate": 1.6367968469740416e-05,
+ "loss": 0.4652596712112427,
+ "step": 1267
+ },
+ {
+ "epoch": 0.6652675760755509,
+ "grad_norm": 0.15919541846683566,
+ "learning_rate": 1.6335754744956937e-05,
+ "loss": 0.43842798471450806,
+ "step": 1268
+ },
+ {
+ "epoch": 0.6657922350472193,
+ "grad_norm": 0.16322634187139418,
+ "learning_rate": 1.630357135092859e-05,
+ "loss": 0.45242664217948914,
+ "step": 1269
+ },
+ {
+ "epoch": 0.6663168940188877,
+ "grad_norm": 0.15644867734734783,
+ "learning_rate": 1.6271418375090445e-05,
+ "loss": 0.44801655411720276,
+ "step": 1270
+ },
+ {
+ "epoch": 0.6668415529905561,
+ "grad_norm": 0.1514717416124423,
+ "learning_rate": 1.623929590479496e-05,
+ "loss": 0.4506194293498993,
+ "step": 1271
+ },
+ {
+ "epoch": 0.6673662119622246,
+ "grad_norm": 0.1740345085495889,
+ "learning_rate": 1.6207204027311683e-05,
+ "loss": 0.4818177819252014,
+ "step": 1272
+ },
+ {
+ "epoch": 0.667890870933893,
+ "grad_norm": 0.15797701900765457,
+ "learning_rate": 1.6175142829827067e-05,
+ "loss": 0.4639382064342499,
+ "step": 1273
+ },
+ {
+ "epoch": 0.6684155299055614,
+ "grad_norm": 0.15663812316653875,
+ "learning_rate": 1.6143112399444198e-05,
+ "loss": 0.46217501163482666,
+ "step": 1274
+ },
+ {
+ "epoch": 0.6689401888772298,
+ "grad_norm": 0.15474018977043758,
+ "learning_rate": 1.6111112823182592e-05,
+ "loss": 0.43691688776016235,
+ "step": 1275
+ },
+ {
+ "epoch": 0.6694648478488983,
+ "grad_norm": 0.1615634942529431,
+ "learning_rate": 1.6079144187977923e-05,
+ "loss": 0.4412601590156555,
+ "step": 1276
+ },
+ {
+ "epoch": 0.6699895068205667,
+ "grad_norm": 0.15440022842754444,
+ "learning_rate": 1.604720658068184e-05,
+ "loss": 0.4523273706436157,
+ "step": 1277
+ },
+ {
+ "epoch": 0.670514165792235,
+ "grad_norm": 0.1634985665776242,
+ "learning_rate": 1.6015300088061664e-05,
+ "loss": 0.46264398097991943,
+ "step": 1278
+ },
+ {
+ "epoch": 0.6710388247639034,
+ "grad_norm": 0.15037886370689527,
+ "learning_rate": 1.598342479680018e-05,
+ "loss": 0.451465904712677,
+ "step": 1279
+ },
+ {
+ "epoch": 0.6715634837355718,
+ "grad_norm": 0.16895242437746613,
+ "learning_rate": 1.595158079349542e-05,
+ "loss": 0.4466145634651184,
+ "step": 1280
+ },
+ {
+ "epoch": 0.6720881427072403,
+ "grad_norm": 0.1629548326999642,
+ "learning_rate": 1.5919768164660425e-05,
+ "loss": 0.45415782928466797,
+ "step": 1281
+ },
+ {
+ "epoch": 0.6726128016789087,
+ "grad_norm": 0.1563872102268991,
+ "learning_rate": 1.588798699672298e-05,
+ "loss": 0.4386737048625946,
+ "step": 1282
+ },
+ {
+ "epoch": 0.6731374606505771,
+ "grad_norm": 0.16038172773344717,
+ "learning_rate": 1.5856237376025403e-05,
+ "loss": 0.446260929107666,
+ "step": 1283
+ },
+ {
+ "epoch": 0.6736621196222455,
+ "grad_norm": 0.16189780714407623,
+ "learning_rate": 1.582451938882431e-05,
+ "loss": 0.45458072423934937,
+ "step": 1284
+ },
+ {
+ "epoch": 0.674186778593914,
+ "grad_norm": 0.15204521785049285,
+ "learning_rate": 1.5792833121290367e-05,
+ "loss": 0.4425617754459381,
+ "step": 1285
+ },
+ {
+ "epoch": 0.6747114375655824,
+ "grad_norm": 0.16237376704792778,
+ "learning_rate": 1.5761178659508063e-05,
+ "loss": 0.44631800055503845,
+ "step": 1286
+ },
+ {
+ "epoch": 0.6752360965372508,
+ "grad_norm": 0.1582063618466069,
+ "learning_rate": 1.57295560894755e-05,
+ "loss": 0.44280534982681274,
+ "step": 1287
+ },
+ {
+ "epoch": 0.6757607555089192,
+ "grad_norm": 0.16587032349847394,
+ "learning_rate": 1.5697965497104116e-05,
+ "loss": 0.4495997130870819,
+ "step": 1288
+ },
+ {
+ "epoch": 0.6762854144805877,
+ "grad_norm": 0.1557269262755563,
+ "learning_rate": 1.566640696821848e-05,
+ "loss": 0.4493359625339508,
+ "step": 1289
+ },
+ {
+ "epoch": 0.6768100734522561,
+ "grad_norm": 0.16038490357012256,
+ "learning_rate": 1.5634880588556043e-05,
+ "loss": 0.4447544813156128,
+ "step": 1290
+ },
+ {
+ "epoch": 0.6773347324239245,
+ "grad_norm": 0.15355201908546406,
+ "learning_rate": 1.560338644376692e-05,
+ "loss": 0.4468432664871216,
+ "step": 1291
+ },
+ {
+ "epoch": 0.6778593913955928,
+ "grad_norm": 0.15077009515589776,
+ "learning_rate": 1.5571924619413668e-05,
+ "loss": 0.44959044456481934,
+ "step": 1292
+ },
+ {
+ "epoch": 0.6783840503672612,
+ "grad_norm": 0.1583038146613965,
+ "learning_rate": 1.5540495200971012e-05,
+ "loss": 0.42450425028800964,
+ "step": 1293
+ },
+ {
+ "epoch": 0.6789087093389297,
+ "grad_norm": 0.15950687190598087,
+ "learning_rate": 1.5509098273825635e-05,
+ "loss": 0.4527238607406616,
+ "step": 1294
+ },
+ {
+ "epoch": 0.6794333683105981,
+ "grad_norm": 0.1526262208847808,
+ "learning_rate": 1.5477733923275968e-05,
+ "loss": 0.4379724860191345,
+ "step": 1295
+ },
+ {
+ "epoch": 0.6799580272822665,
+ "grad_norm": 0.15541579650021778,
+ "learning_rate": 1.5446402234531934e-05,
+ "loss": 0.4462166130542755,
+ "step": 1296
+ },
+ {
+ "epoch": 0.6804826862539349,
+ "grad_norm": 0.15483930734313786,
+ "learning_rate": 1.5415103292714718e-05,
+ "loss": 0.4390607178211212,
+ "step": 1297
+ },
+ {
+ "epoch": 0.6810073452256034,
+ "grad_norm": 0.1578005731149154,
+ "learning_rate": 1.5383837182856532e-05,
+ "loss": 0.44462254643440247,
+ "step": 1298
+ },
+ {
+ "epoch": 0.6815320041972718,
+ "grad_norm": 0.15830183884958823,
+ "learning_rate": 1.5352603989900406e-05,
+ "loss": 0.4563537538051605,
+ "step": 1299
+ },
+ {
+ "epoch": 0.6820566631689402,
+ "grad_norm": 0.15605472163445522,
+ "learning_rate": 1.5321403798699923e-05,
+ "loss": 0.43852412700653076,
+ "step": 1300
+ },
+ {
+ "epoch": 0.6825813221406086,
+ "grad_norm": 0.15770470227141278,
+ "learning_rate": 1.5290236694019017e-05,
+ "loss": 0.4438081979751587,
+ "step": 1301
+ },
+ {
+ "epoch": 0.683105981112277,
+ "grad_norm": 0.16175339845012077,
+ "learning_rate": 1.5259102760531747e-05,
+ "loss": 0.46670734882354736,
+ "step": 1302
+ },
+ {
+ "epoch": 0.6836306400839455,
+ "grad_norm": 0.16359504716154477,
+ "learning_rate": 1.5228002082822036e-05,
+ "loss": 0.4590187668800354,
+ "step": 1303
+ },
+ {
+ "epoch": 0.6841552990556139,
+ "grad_norm": 0.15422879500254713,
+ "learning_rate": 1.5196934745383457e-05,
+ "loss": 0.45287445187568665,
+ "step": 1304
+ },
+ {
+ "epoch": 0.6846799580272823,
+ "grad_norm": 0.15929902143574254,
+ "learning_rate": 1.5165900832619015e-05,
+ "loss": 0.44286859035491943,
+ "step": 1305
+ },
+ {
+ "epoch": 0.6852046169989506,
+ "grad_norm": 0.14838868714016124,
+ "learning_rate": 1.5134900428840904e-05,
+ "loss": 0.45155346393585205,
+ "step": 1306
+ },
+ {
+ "epoch": 0.6857292759706191,
+ "grad_norm": 0.15494288640397672,
+ "learning_rate": 1.5103933618270283e-05,
+ "loss": 0.44010141491889954,
+ "step": 1307
+ },
+ {
+ "epoch": 0.6862539349422875,
+ "grad_norm": 0.1686262360598338,
+ "learning_rate": 1.5073000485037042e-05,
+ "loss": 0.446555495262146,
+ "step": 1308
+ },
+ {
+ "epoch": 0.6867785939139559,
+ "grad_norm": 0.16122110823982602,
+ "learning_rate": 1.5042101113179579e-05,
+ "loss": 0.44640374183654785,
+ "step": 1309
+ },
+ {
+ "epoch": 0.6873032528856243,
+ "grad_norm": 0.16849937797919867,
+ "learning_rate": 1.5011235586644572e-05,
+ "loss": 0.4465785026550293,
+ "step": 1310
+ },
+ {
+ "epoch": 0.6878279118572928,
+ "grad_norm": 0.16062971262800754,
+ "learning_rate": 1.498040398928674e-05,
+ "loss": 0.445321649312973,
+ "step": 1311
+ },
+ {
+ "epoch": 0.6883525708289612,
+ "grad_norm": 0.1608953185993034,
+ "learning_rate": 1.4949606404868638e-05,
+ "loss": 0.46518227458000183,
+ "step": 1312
+ },
+ {
+ "epoch": 0.6888772298006296,
+ "grad_norm": 0.15880255672830915,
+ "learning_rate": 1.491884291706041e-05,
+ "loss": 0.44582217931747437,
+ "step": 1313
+ },
+ {
+ "epoch": 0.689401888772298,
+ "grad_norm": 0.16715145200789822,
+ "learning_rate": 1.4888113609439563e-05,
+ "loss": 0.4435983896255493,
+ "step": 1314
+ },
+ {
+ "epoch": 0.6899265477439664,
+ "grad_norm": 0.15263740340694454,
+ "learning_rate": 1.485741856549075e-05,
+ "loss": 0.4405066967010498,
+ "step": 1315
+ },
+ {
+ "epoch": 0.6904512067156349,
+ "grad_norm": 0.16247034518813966,
+ "learning_rate": 1.4826757868605517e-05,
+ "loss": 0.4544215500354767,
+ "step": 1316
+ },
+ {
+ "epoch": 0.6909758656873033,
+ "grad_norm": 0.15649146705978162,
+ "learning_rate": 1.4796131602082142e-05,
+ "loss": 0.4454413056373596,
+ "step": 1317
+ },
+ {
+ "epoch": 0.6915005246589717,
+ "grad_norm": 0.16286718636058178,
+ "learning_rate": 1.4765539849125318e-05,
+ "loss": 0.4497183561325073,
+ "step": 1318
+ },
+ {
+ "epoch": 0.6920251836306401,
+ "grad_norm": 0.15812342085005582,
+ "learning_rate": 1.473498269284599e-05,
+ "loss": 0.445032000541687,
+ "step": 1319
+ },
+ {
+ "epoch": 0.6925498426023085,
+ "grad_norm": 0.1558101290511678,
+ "learning_rate": 1.4704460216261123e-05,
+ "loss": 0.45769619941711426,
+ "step": 1320
+ },
+ {
+ "epoch": 0.6930745015739769,
+ "grad_norm": 0.15642889059495207,
+ "learning_rate": 1.4673972502293428e-05,
+ "loss": 0.44788360595703125,
+ "step": 1321
+ },
+ {
+ "epoch": 0.6935991605456453,
+ "grad_norm": 0.15722513518113074,
+ "learning_rate": 1.4643519633771222e-05,
+ "loss": 0.449631929397583,
+ "step": 1322
+ },
+ {
+ "epoch": 0.6941238195173137,
+ "grad_norm": 0.15587633748081542,
+ "learning_rate": 1.4613101693428125e-05,
+ "loss": 0.4450187087059021,
+ "step": 1323
+ },
+ {
+ "epoch": 0.6946484784889821,
+ "grad_norm": 0.1577107074882479,
+ "learning_rate": 1.458271876390287e-05,
+ "loss": 0.45108890533447266,
+ "step": 1324
+ },
+ {
+ "epoch": 0.6951731374606506,
+ "grad_norm": 0.15649839486990696,
+ "learning_rate": 1.4552370927739073e-05,
+ "loss": 0.46430301666259766,
+ "step": 1325
+ },
+ {
+ "epoch": 0.695697796432319,
+ "grad_norm": 0.18573600420690636,
+ "learning_rate": 1.4522058267385008e-05,
+ "loss": 0.4701972007751465,
+ "step": 1326
+ },
+ {
+ "epoch": 0.6962224554039874,
+ "grad_norm": 0.16296497084037542,
+ "learning_rate": 1.4491780865193402e-05,
+ "loss": 0.4607710838317871,
+ "step": 1327
+ },
+ {
+ "epoch": 0.6967471143756558,
+ "grad_norm": 0.161124007217343,
+ "learning_rate": 1.446153880342117e-05,
+ "loss": 0.45095115900039673,
+ "step": 1328
+ },
+ {
+ "epoch": 0.6972717733473243,
+ "grad_norm": 0.1537884039166766,
+ "learning_rate": 1.4431332164229222e-05,
+ "loss": 0.424068808555603,
+ "step": 1329
+ },
+ {
+ "epoch": 0.6977964323189927,
+ "grad_norm": 0.16588834503456087,
+ "learning_rate": 1.4401161029682237e-05,
+ "loss": 0.4622839093208313,
+ "step": 1330
+ },
+ {
+ "epoch": 0.6983210912906611,
+ "grad_norm": 0.16100510604015544,
+ "learning_rate": 1.4371025481748435e-05,
+ "loss": 0.4660705626010895,
+ "step": 1331
+ },
+ {
+ "epoch": 0.6988457502623295,
+ "grad_norm": 0.15559888940590075,
+ "learning_rate": 1.4340925602299337e-05,
+ "loss": 0.4484976530075073,
+ "step": 1332
+ },
+ {
+ "epoch": 0.6993704092339978,
+ "grad_norm": 0.15503821966213488,
+ "learning_rate": 1.4310861473109616e-05,
+ "loss": 0.45636600255966187,
+ "step": 1333
+ },
+ {
+ "epoch": 0.6998950682056663,
+ "grad_norm": 0.15539074896109334,
+ "learning_rate": 1.4280833175856743e-05,
+ "loss": 0.45863068103790283,
+ "step": 1334
+ },
+ {
+ "epoch": 0.7004197271773347,
+ "grad_norm": 0.15664821009154872,
+ "learning_rate": 1.4250840792120896e-05,
+ "loss": 0.44043463468551636,
+ "step": 1335
+ },
+ {
+ "epoch": 0.7009443861490031,
+ "grad_norm": 0.15825478808520285,
+ "learning_rate": 1.4220884403384655e-05,
+ "loss": 0.46522051095962524,
+ "step": 1336
+ },
+ {
+ "epoch": 0.7014690451206715,
+ "grad_norm": 0.16004268568141544,
+ "learning_rate": 1.4190964091032843e-05,
+ "loss": 0.46114686131477356,
+ "step": 1337
+ },
+ {
+ "epoch": 0.70199370409234,
+ "grad_norm": 0.16413139949945635,
+ "learning_rate": 1.4161079936352234e-05,
+ "loss": 0.4457515478134155,
+ "step": 1338
+ },
+ {
+ "epoch": 0.7025183630640084,
+ "grad_norm": 0.1538890577972702,
+ "learning_rate": 1.4131232020531398e-05,
+ "loss": 0.4616093337535858,
+ "step": 1339
+ },
+ {
+ "epoch": 0.7030430220356768,
+ "grad_norm": 0.16240737854323284,
+ "learning_rate": 1.4101420424660434e-05,
+ "loss": 0.44631537795066833,
+ "step": 1340
+ },
+ {
+ "epoch": 0.7035676810073452,
+ "grad_norm": 0.15984708603406073,
+ "learning_rate": 1.4071645229730762e-05,
+ "loss": 0.4366717040538788,
+ "step": 1341
+ },
+ {
+ "epoch": 0.7040923399790137,
+ "grad_norm": 0.15149695695635904,
+ "learning_rate": 1.4041906516634946e-05,
+ "loss": 0.44129428267478943,
+ "step": 1342
+ },
+ {
+ "epoch": 0.7046169989506821,
+ "grad_norm": 0.15041167501768415,
+ "learning_rate": 1.4012204366166399e-05,
+ "loss": 0.43654775619506836,
+ "step": 1343
+ },
+ {
+ "epoch": 0.7051416579223505,
+ "grad_norm": 0.1596097627963756,
+ "learning_rate": 1.3982538859019217e-05,
+ "loss": 0.4652445614337921,
+ "step": 1344
+ },
+ {
+ "epoch": 0.7056663168940189,
+ "grad_norm": 0.15621704205027884,
+ "learning_rate": 1.3952910075787941e-05,
+ "loss": 0.44577041268348694,
+ "step": 1345
+ },
+ {
+ "epoch": 0.7061909758656874,
+ "grad_norm": 0.16073041098008783,
+ "learning_rate": 1.3923318096967341e-05,
+ "loss": 0.4391976594924927,
+ "step": 1346
+ },
+ {
+ "epoch": 0.7067156348373557,
+ "grad_norm": 0.14793853946372448,
+ "learning_rate": 1.3893763002952204e-05,
+ "loss": 0.4463799297809601,
+ "step": 1347
+ },
+ {
+ "epoch": 0.7072402938090241,
+ "grad_norm": 0.16176080358867354,
+ "learning_rate": 1.3864244874037097e-05,
+ "loss": 0.45183271169662476,
+ "step": 1348
+ },
+ {
+ "epoch": 0.7077649527806925,
+ "grad_norm": 0.16372771179026924,
+ "learning_rate": 1.3834763790416171e-05,
+ "loss": 0.44205576181411743,
+ "step": 1349
+ },
+ {
+ "epoch": 0.7082896117523609,
+ "grad_norm": 0.15135222516127125,
+ "learning_rate": 1.3805319832182937e-05,
+ "loss": 0.4390261173248291,
+ "step": 1350
+ },
+ {
+ "epoch": 0.7088142707240294,
+ "grad_norm": 0.15493183225792365,
+ "learning_rate": 1.377591307933002e-05,
+ "loss": 0.45837974548339844,
+ "step": 1351
+ },
+ {
+ "epoch": 0.7093389296956978,
+ "grad_norm": 0.16405906906516504,
+ "learning_rate": 1.3746543611749008e-05,
+ "loss": 0.45032668113708496,
+ "step": 1352
+ },
+ {
+ "epoch": 0.7098635886673662,
+ "grad_norm": 0.15725677976524505,
+ "learning_rate": 1.3717211509230155e-05,
+ "loss": 0.4486224949359894,
+ "step": 1353
+ },
+ {
+ "epoch": 0.7103882476390346,
+ "grad_norm": 0.15781469066983164,
+ "learning_rate": 1.3687916851462227e-05,
+ "loss": 0.4502638578414917,
+ "step": 1354
+ },
+ {
+ "epoch": 0.710912906610703,
+ "grad_norm": 0.16115007729054298,
+ "learning_rate": 1.3658659718032241e-05,
+ "loss": 0.4508914351463318,
+ "step": 1355
+ },
+ {
+ "epoch": 0.7114375655823715,
+ "grad_norm": 0.16755060475968744,
+ "learning_rate": 1.3629440188425286e-05,
+ "loss": 0.45653945207595825,
+ "step": 1356
+ },
+ {
+ "epoch": 0.7119622245540399,
+ "grad_norm": 0.1574994294756123,
+ "learning_rate": 1.3600258342024268e-05,
+ "loss": 0.4349849820137024,
+ "step": 1357
+ },
+ {
+ "epoch": 0.7124868835257083,
+ "grad_norm": 0.16104440922657715,
+ "learning_rate": 1.3571114258109758e-05,
+ "loss": 0.4501233696937561,
+ "step": 1358
+ },
+ {
+ "epoch": 0.7130115424973767,
+ "grad_norm": 0.15163184290939022,
+ "learning_rate": 1.3542008015859691e-05,
+ "loss": 0.44171035289764404,
+ "step": 1359
+ },
+ {
+ "epoch": 0.7135362014690452,
+ "grad_norm": 0.15955896674456582,
+ "learning_rate": 1.3512939694349211e-05,
+ "loss": 0.4558671712875366,
+ "step": 1360
+ },
+ {
+ "epoch": 0.7140608604407135,
+ "grad_norm": 0.1589376751536802,
+ "learning_rate": 1.3483909372550446e-05,
+ "loss": 0.4430835247039795,
+ "step": 1361
+ },
+ {
+ "epoch": 0.7145855194123819,
+ "grad_norm": 0.15979088397822155,
+ "learning_rate": 1.3454917129332279e-05,
+ "loss": 0.4426245391368866,
+ "step": 1362
+ },
+ {
+ "epoch": 0.7151101783840503,
+ "grad_norm": 0.15493179714072167,
+ "learning_rate": 1.3425963043460143e-05,
+ "loss": 0.4550474286079407,
+ "step": 1363
+ },
+ {
+ "epoch": 0.7156348373557188,
+ "grad_norm": 0.15396498968995032,
+ "learning_rate": 1.3397047193595808e-05,
+ "loss": 0.44220536947250366,
+ "step": 1364
+ },
+ {
+ "epoch": 0.7161594963273872,
+ "grad_norm": 0.1489346822667273,
+ "learning_rate": 1.336816965829717e-05,
+ "loss": 0.4437323212623596,
+ "step": 1365
+ },
+ {
+ "epoch": 0.7166841552990556,
+ "grad_norm": 0.16311964931734477,
+ "learning_rate": 1.333933051601802e-05,
+ "loss": 0.45308777689933777,
+ "step": 1366
+ },
+ {
+ "epoch": 0.717208814270724,
+ "grad_norm": 0.1534354619021179,
+ "learning_rate": 1.331052984510785e-05,
+ "loss": 0.4334886372089386,
+ "step": 1367
+ },
+ {
+ "epoch": 0.7177334732423925,
+ "grad_norm": 0.15541089758065404,
+ "learning_rate": 1.328176772381165e-05,
+ "loss": 0.45248323678970337,
+ "step": 1368
+ },
+ {
+ "epoch": 0.7182581322140609,
+ "grad_norm": 0.1617921305700596,
+ "learning_rate": 1.3253044230269651e-05,
+ "loss": 0.44092947244644165,
+ "step": 1369
+ },
+ {
+ "epoch": 0.7187827911857293,
+ "grad_norm": 0.15047353793760151,
+ "learning_rate": 1.3224359442517165e-05,
+ "loss": 0.4424414336681366,
+ "step": 1370
+ },
+ {
+ "epoch": 0.7193074501573977,
+ "grad_norm": 0.15958089102258743,
+ "learning_rate": 1.319571343848433e-05,
+ "loss": 0.4458580017089844,
+ "step": 1371
+ },
+ {
+ "epoch": 0.7198321091290661,
+ "grad_norm": 0.15349258306968733,
+ "learning_rate": 1.3167106295995924e-05,
+ "loss": 0.4285386800765991,
+ "step": 1372
+ },
+ {
+ "epoch": 0.7203567681007346,
+ "grad_norm": 0.16042529563759886,
+ "learning_rate": 1.3138538092771164e-05,
+ "loss": 0.4497690200805664,
+ "step": 1373
+ },
+ {
+ "epoch": 0.720881427072403,
+ "grad_norm": 0.1565113736588754,
+ "learning_rate": 1.3110008906423454e-05,
+ "loss": 0.43562617897987366,
+ "step": 1374
+ },
+ {
+ "epoch": 0.7214060860440713,
+ "grad_norm": 0.15177291086871506,
+ "learning_rate": 1.3081518814460215e-05,
+ "loss": 0.44456908106803894,
+ "step": 1375
+ },
+ {
+ "epoch": 0.7219307450157397,
+ "grad_norm": 0.15928490480439225,
+ "learning_rate": 1.3053067894282622e-05,
+ "loss": 0.46170780062675476,
+ "step": 1376
+ },
+ {
+ "epoch": 0.7224554039874082,
+ "grad_norm": 0.1554814469217892,
+ "learning_rate": 1.302465622318549e-05,
+ "loss": 0.4422169327735901,
+ "step": 1377
+ },
+ {
+ "epoch": 0.7229800629590766,
+ "grad_norm": 0.15375370214966386,
+ "learning_rate": 1.2996283878356947e-05,
+ "loss": 0.4564763903617859,
+ "step": 1378
+ },
+ {
+ "epoch": 0.723504721930745,
+ "grad_norm": 0.15942841969395802,
+ "learning_rate": 1.2967950936878311e-05,
+ "loss": 0.45121875405311584,
+ "step": 1379
+ },
+ {
+ "epoch": 0.7240293809024134,
+ "grad_norm": 0.15731308716774903,
+ "learning_rate": 1.2939657475723844e-05,
+ "loss": 0.4285027086734772,
+ "step": 1380
+ },
+ {
+ "epoch": 0.7245540398740818,
+ "grad_norm": 0.15997753613932147,
+ "learning_rate": 1.291140357176054e-05,
+ "loss": 0.4566386342048645,
+ "step": 1381
+ },
+ {
+ "epoch": 0.7250786988457503,
+ "grad_norm": 0.15333715267473727,
+ "learning_rate": 1.2883189301747927e-05,
+ "loss": 0.45093846321105957,
+ "step": 1382
+ },
+ {
+ "epoch": 0.7256033578174187,
+ "grad_norm": 0.1514510304709061,
+ "learning_rate": 1.2855014742337874e-05,
+ "loss": 0.44256967306137085,
+ "step": 1383
+ },
+ {
+ "epoch": 0.7261280167890871,
+ "grad_norm": 0.15624301298995746,
+ "learning_rate": 1.2826879970074346e-05,
+ "loss": 0.4361875653266907,
+ "step": 1384
+ },
+ {
+ "epoch": 0.7266526757607555,
+ "grad_norm": 0.15767850122006694,
+ "learning_rate": 1.279878506139322e-05,
+ "loss": 0.4477671980857849,
+ "step": 1385
+ },
+ {
+ "epoch": 0.727177334732424,
+ "grad_norm": 0.15537676969694064,
+ "learning_rate": 1.2770730092622072e-05,
+ "loss": 0.4339832663536072,
+ "step": 1386
+ },
+ {
+ "epoch": 0.7277019937040924,
+ "grad_norm": 0.15944113633281293,
+ "learning_rate": 1.2742715139979965e-05,
+ "loss": 0.45407935976982117,
+ "step": 1387
+ },
+ {
+ "epoch": 0.7282266526757608,
+ "grad_norm": 0.16763519518606218,
+ "learning_rate": 1.271474027957727e-05,
+ "loss": 0.46442392468452454,
+ "step": 1388
+ },
+ {
+ "epoch": 0.7287513116474291,
+ "grad_norm": 0.1564572774601902,
+ "learning_rate": 1.2686805587415423e-05,
+ "loss": 0.4463968276977539,
+ "step": 1389
+ },
+ {
+ "epoch": 0.7292759706190975,
+ "grad_norm": 0.16384626314986128,
+ "learning_rate": 1.2658911139386709e-05,
+ "loss": 0.45519885420799255,
+ "step": 1390
+ },
+ {
+ "epoch": 0.729800629590766,
+ "grad_norm": 0.1562023802820964,
+ "learning_rate": 1.2631057011274106e-05,
+ "loss": 0.4365222454071045,
+ "step": 1391
+ },
+ {
+ "epoch": 0.7303252885624344,
+ "grad_norm": 0.15597287232940926,
+ "learning_rate": 1.2603243278751042e-05,
+ "loss": 0.4328266680240631,
+ "step": 1392
+ },
+ {
+ "epoch": 0.7308499475341028,
+ "grad_norm": 0.14762833706132167,
+ "learning_rate": 1.2575470017381214e-05,
+ "loss": 0.43902426958084106,
+ "step": 1393
+ },
+ {
+ "epoch": 0.7313746065057712,
+ "grad_norm": 0.1688371515856315,
+ "learning_rate": 1.2547737302618351e-05,
+ "loss": 0.4649963676929474,
+ "step": 1394
+ },
+ {
+ "epoch": 0.7318992654774397,
+ "grad_norm": 0.15501514509586015,
+ "learning_rate": 1.2520045209806024e-05,
+ "loss": 0.4501558542251587,
+ "step": 1395
+ },
+ {
+ "epoch": 0.7324239244491081,
+ "grad_norm": 0.15516748917693537,
+ "learning_rate": 1.2492393814177457e-05,
+ "loss": 0.4405498206615448,
+ "step": 1396
+ },
+ {
+ "epoch": 0.7329485834207765,
+ "grad_norm": 0.17115400468199243,
+ "learning_rate": 1.2464783190855295e-05,
+ "loss": 0.4445379674434662,
+ "step": 1397
+ },
+ {
+ "epoch": 0.7334732423924449,
+ "grad_norm": 0.15966386341604186,
+ "learning_rate": 1.243721341485143e-05,
+ "loss": 0.4769889712333679,
+ "step": 1398
+ },
+ {
+ "epoch": 0.7339979013641134,
+ "grad_norm": 0.15614706359781472,
+ "learning_rate": 1.2409684561066766e-05,
+ "loss": 0.43643918633461,
+ "step": 1399
+ },
+ {
+ "epoch": 0.7345225603357818,
+ "grad_norm": 0.17467968960064226,
+ "learning_rate": 1.238219670429104e-05,
+ "loss": 0.45413005352020264,
+ "step": 1400
+ },
+ {
+ "epoch": 0.7350472193074502,
+ "grad_norm": 0.1544517430810819,
+ "learning_rate": 1.2354749919202598e-05,
+ "loss": 0.45218425989151,
+ "step": 1401
+ },
+ {
+ "epoch": 0.7355718782791185,
+ "grad_norm": 0.16551083673374484,
+ "learning_rate": 1.2327344280368217e-05,
+ "loss": 0.4459628462791443,
+ "step": 1402
+ },
+ {
+ "epoch": 0.7360965372507869,
+ "grad_norm": 0.17614490459613263,
+ "learning_rate": 1.2299979862242882e-05,
+ "loss": 0.4406280815601349,
+ "step": 1403
+ },
+ {
+ "epoch": 0.7366211962224554,
+ "grad_norm": 0.15554925394435032,
+ "learning_rate": 1.2272656739169588e-05,
+ "loss": 0.4380159378051758,
+ "step": 1404
+ },
+ {
+ "epoch": 0.7371458551941238,
+ "grad_norm": 0.16475466799870916,
+ "learning_rate": 1.2245374985379147e-05,
+ "loss": 0.4442812204360962,
+ "step": 1405
+ },
+ {
+ "epoch": 0.7376705141657922,
+ "grad_norm": 0.16877301168875541,
+ "learning_rate": 1.2218134674989978e-05,
+ "loss": 0.4589429199695587,
+ "step": 1406
+ },
+ {
+ "epoch": 0.7381951731374606,
+ "grad_norm": 0.1583278589457438,
+ "learning_rate": 1.2190935882007894e-05,
+ "loss": 0.44896745681762695,
+ "step": 1407
+ },
+ {
+ "epoch": 0.7387198321091291,
+ "grad_norm": 0.1582022541656192,
+ "learning_rate": 1.2163778680325947e-05,
+ "loss": 0.4367997646331787,
+ "step": 1408
+ },
+ {
+ "epoch": 0.7392444910807975,
+ "grad_norm": 0.1669573982300068,
+ "learning_rate": 1.2136663143724163e-05,
+ "loss": 0.4503386616706848,
+ "step": 1409
+ },
+ {
+ "epoch": 0.7397691500524659,
+ "grad_norm": 0.1637877523924539,
+ "learning_rate": 1.2109589345869387e-05,
+ "loss": 0.4660584330558777,
+ "step": 1410
+ },
+ {
+ "epoch": 0.7402938090241343,
+ "grad_norm": 0.1640136159943701,
+ "learning_rate": 1.2082557360315065e-05,
+ "loss": 0.45206528902053833,
+ "step": 1411
+ },
+ {
+ "epoch": 0.7408184679958028,
+ "grad_norm": 0.1596580168682554,
+ "learning_rate": 1.2055567260501054e-05,
+ "loss": 0.4405035376548767,
+ "step": 1412
+ },
+ {
+ "epoch": 0.7413431269674712,
+ "grad_norm": 0.15224511741945293,
+ "learning_rate": 1.2028619119753402e-05,
+ "loss": 0.42810291051864624,
+ "step": 1413
+ },
+ {
+ "epoch": 0.7418677859391396,
+ "grad_norm": 0.15631381931817098,
+ "learning_rate": 1.2001713011284193e-05,
+ "loss": 0.4421626925468445,
+ "step": 1414
+ },
+ {
+ "epoch": 0.742392444910808,
+ "grad_norm": 0.15721819981265503,
+ "learning_rate": 1.1974849008191288e-05,
+ "loss": 0.4540696144104004,
+ "step": 1415
+ },
+ {
+ "epoch": 0.7429171038824763,
+ "grad_norm": 0.15904478578744463,
+ "learning_rate": 1.1948027183458182e-05,
+ "loss": 0.4601830244064331,
+ "step": 1416
+ },
+ {
+ "epoch": 0.7434417628541448,
+ "grad_norm": 0.1530755605235123,
+ "learning_rate": 1.1921247609953747e-05,
+ "loss": 0.4552589952945709,
+ "step": 1417
+ },
+ {
+ "epoch": 0.7439664218258132,
+ "grad_norm": 0.15375751072258462,
+ "learning_rate": 1.1894510360432106e-05,
+ "loss": 0.42965373396873474,
+ "step": 1418
+ },
+ {
+ "epoch": 0.7444910807974816,
+ "grad_norm": 0.15678686023937333,
+ "learning_rate": 1.1867815507532377e-05,
+ "loss": 0.44684016704559326,
+ "step": 1419
+ },
+ {
+ "epoch": 0.74501573976915,
+ "grad_norm": 0.15486952983783606,
+ "learning_rate": 1.1841163123778502e-05,
+ "loss": 0.44415023922920227,
+ "step": 1420
+ },
+ {
+ "epoch": 0.7455403987408185,
+ "grad_norm": 0.15342611490545593,
+ "learning_rate": 1.1814553281579034e-05,
+ "loss": 0.4501304626464844,
+ "step": 1421
+ },
+ {
+ "epoch": 0.7460650577124869,
+ "grad_norm": 0.15374260012788166,
+ "learning_rate": 1.1787986053226954e-05,
+ "loss": 0.4521749019622803,
+ "step": 1422
+ },
+ {
+ "epoch": 0.7465897166841553,
+ "grad_norm": 0.1543553258370361,
+ "learning_rate": 1.176146151089949e-05,
+ "loss": 0.44753748178482056,
+ "step": 1423
+ },
+ {
+ "epoch": 0.7471143756558237,
+ "grad_norm": 0.15921670193848306,
+ "learning_rate": 1.1734979726657878e-05,
+ "loss": 0.4473114311695099,
+ "step": 1424
+ },
+ {
+ "epoch": 0.7476390346274921,
+ "grad_norm": 0.15023498686186612,
+ "learning_rate": 1.17085407724472e-05,
+ "loss": 0.4568778872489929,
+ "step": 1425
+ },
+ {
+ "epoch": 0.7481636935991606,
+ "grad_norm": 0.14601156920274583,
+ "learning_rate": 1.1682144720096168e-05,
+ "loss": 0.4389464855194092,
+ "step": 1426
+ },
+ {
+ "epoch": 0.748688352570829,
+ "grad_norm": 0.1520548095149398,
+ "learning_rate": 1.1655791641316954e-05,
+ "loss": 0.440733402967453,
+ "step": 1427
+ },
+ {
+ "epoch": 0.7492130115424974,
+ "grad_norm": 0.1591915163097993,
+ "learning_rate": 1.1629481607704963e-05,
+ "loss": 0.46245887875556946,
+ "step": 1428
+ },
+ {
+ "epoch": 0.7497376705141658,
+ "grad_norm": 0.14867295658447263,
+ "learning_rate": 1.1603214690738683e-05,
+ "loss": 0.4332764744758606,
+ "step": 1429
+ },
+ {
+ "epoch": 0.7502623294858342,
+ "grad_norm": 0.1498957762903134,
+ "learning_rate": 1.1576990961779447e-05,
+ "loss": 0.45215466618537903,
+ "step": 1430
+ },
+ {
+ "epoch": 0.7507869884575026,
+ "grad_norm": 0.15430555803491824,
+ "learning_rate": 1.1550810492071243e-05,
+ "loss": 0.4454396963119507,
+ "step": 1431
+ },
+ {
+ "epoch": 0.751311647429171,
+ "grad_norm": 0.1517461021936533,
+ "learning_rate": 1.1524673352740543e-05,
+ "loss": 0.43869835138320923,
+ "step": 1432
+ },
+ {
+ "epoch": 0.7518363064008394,
+ "grad_norm": 0.14541657994692936,
+ "learning_rate": 1.1498579614796121e-05,
+ "loss": 0.4373054504394531,
+ "step": 1433
+ },
+ {
+ "epoch": 0.7523609653725079,
+ "grad_norm": 0.15539513763786977,
+ "learning_rate": 1.1472529349128812e-05,
+ "loss": 0.45451128482818604,
+ "step": 1434
+ },
+ {
+ "epoch": 0.7528856243441763,
+ "grad_norm": 0.1482043113690946,
+ "learning_rate": 1.144652262651136e-05,
+ "loss": 0.43644773960113525,
+ "step": 1435
+ },
+ {
+ "epoch": 0.7534102833158447,
+ "grad_norm": 0.16293628094190815,
+ "learning_rate": 1.1420559517598205e-05,
+ "loss": 0.4687269330024719,
+ "step": 1436
+ },
+ {
+ "epoch": 0.7539349422875131,
+ "grad_norm": 0.1553393974390287,
+ "learning_rate": 1.1394640092925302e-05,
+ "loss": 0.4501398205757141,
+ "step": 1437
+ },
+ {
+ "epoch": 0.7544596012591815,
+ "grad_norm": 0.16214837694535802,
+ "learning_rate": 1.1368764422909919e-05,
+ "loss": 0.4542577862739563,
+ "step": 1438
+ },
+ {
+ "epoch": 0.75498426023085,
+ "grad_norm": 0.15838878372428297,
+ "learning_rate": 1.1342932577850473e-05,
+ "loss": 0.46228161454200745,
+ "step": 1439
+ },
+ {
+ "epoch": 0.7555089192025184,
+ "grad_norm": 0.15090419497122687,
+ "learning_rate": 1.13171446279263e-05,
+ "loss": 0.4495932161808014,
+ "step": 1440
+ },
+ {
+ "epoch": 0.7560335781741868,
+ "grad_norm": 0.1543376786458394,
+ "learning_rate": 1.1291400643197484e-05,
+ "loss": 0.451630175113678,
+ "step": 1441
+ },
+ {
+ "epoch": 0.7565582371458552,
+ "grad_norm": 0.1539470751610696,
+ "learning_rate": 1.1265700693604668e-05,
+ "loss": 0.447547972202301,
+ "step": 1442
+ },
+ {
+ "epoch": 0.7570828961175237,
+ "grad_norm": 0.15761047703305445,
+ "learning_rate": 1.1240044848968856e-05,
+ "loss": 0.4516204595565796,
+ "step": 1443
+ },
+ {
+ "epoch": 0.757607555089192,
+ "grad_norm": 0.15675279683704735,
+ "learning_rate": 1.1214433178991258e-05,
+ "loss": 0.4597971737384796,
+ "step": 1444
+ },
+ {
+ "epoch": 0.7581322140608604,
+ "grad_norm": 0.15592110994904113,
+ "learning_rate": 1.1188865753253022e-05,
+ "loss": 0.443898469209671,
+ "step": 1445
+ },
+ {
+ "epoch": 0.7586568730325288,
+ "grad_norm": 0.15597501691030619,
+ "learning_rate": 1.116334264121513e-05,
+ "loss": 0.4581306278705597,
+ "step": 1446
+ },
+ {
+ "epoch": 0.7591815320041972,
+ "grad_norm": 0.15552304728589958,
+ "learning_rate": 1.1137863912218154e-05,
+ "loss": 0.4475254416465759,
+ "step": 1447
+ },
+ {
+ "epoch": 0.7597061909758657,
+ "grad_norm": 0.15759990123898035,
+ "learning_rate": 1.1112429635482118e-05,
+ "loss": 0.43585264682769775,
+ "step": 1448
+ },
+ {
+ "epoch": 0.7602308499475341,
+ "grad_norm": 0.15964657264676635,
+ "learning_rate": 1.1087039880106247e-05,
+ "loss": 0.46786290407180786,
+ "step": 1449
+ },
+ {
+ "epoch": 0.7607555089192025,
+ "grad_norm": 0.15155804051218852,
+ "learning_rate": 1.106169471506882e-05,
+ "loss": 0.42292726039886475,
+ "step": 1450
+ },
+ {
+ "epoch": 0.7612801678908709,
+ "grad_norm": 0.15519270966930884,
+ "learning_rate": 1.103639420922698e-05,
+ "loss": 0.4248972237110138,
+ "step": 1451
+ },
+ {
+ "epoch": 0.7618048268625394,
+ "grad_norm": 0.1526330511128821,
+ "learning_rate": 1.1011138431316537e-05,
+ "loss": 0.4317784309387207,
+ "step": 1452
+ },
+ {
+ "epoch": 0.7623294858342078,
+ "grad_norm": 0.15797682979272226,
+ "learning_rate": 1.0985927449951779e-05,
+ "loss": 0.44506698846817017,
+ "step": 1453
+ },
+ {
+ "epoch": 0.7628541448058762,
+ "grad_norm": 0.15479979838505462,
+ "learning_rate": 1.096076133362531e-05,
+ "loss": 0.4471631348133087,
+ "step": 1454
+ },
+ {
+ "epoch": 0.7633788037775446,
+ "grad_norm": 0.15712267274447342,
+ "learning_rate": 1.093564015070783e-05,
+ "loss": 0.4321915805339813,
+ "step": 1455
+ },
+ {
+ "epoch": 0.7639034627492131,
+ "grad_norm": 0.15857431494180016,
+ "learning_rate": 1.0910563969447969e-05,
+ "loss": 0.44455766677856445,
+ "step": 1456
+ },
+ {
+ "epoch": 0.7644281217208815,
+ "grad_norm": 0.15377057080997789,
+ "learning_rate": 1.0885532857972097e-05,
+ "loss": 0.4567924439907074,
+ "step": 1457
+ },
+ {
+ "epoch": 0.7649527806925498,
+ "grad_norm": 0.15544844005009104,
+ "learning_rate": 1.0860546884284143e-05,
+ "loss": 0.4491572976112366,
+ "step": 1458
+ },
+ {
+ "epoch": 0.7654774396642182,
+ "grad_norm": 0.16210735566479478,
+ "learning_rate": 1.08356061162654e-05,
+ "loss": 0.43823331594467163,
+ "step": 1459
+ },
+ {
+ "epoch": 0.7660020986358866,
+ "grad_norm": 0.15213559486379452,
+ "learning_rate": 1.0810710621674355e-05,
+ "loss": 0.45375770330429077,
+ "step": 1460
+ },
+ {
+ "epoch": 0.7665267576075551,
+ "grad_norm": 0.1727832306402992,
+ "learning_rate": 1.0785860468146494e-05,
+ "loss": 0.4639572501182556,
+ "step": 1461
+ },
+ {
+ "epoch": 0.7670514165792235,
+ "grad_norm": 0.1637555902013663,
+ "learning_rate": 1.076105572319412e-05,
+ "loss": 0.4355923533439636,
+ "step": 1462
+ },
+ {
+ "epoch": 0.7675760755508919,
+ "grad_norm": 0.16646936030470866,
+ "learning_rate": 1.0736296454206166e-05,
+ "loss": 0.4650861322879791,
+ "step": 1463
+ },
+ {
+ "epoch": 0.7681007345225603,
+ "grad_norm": 0.14987620615431385,
+ "learning_rate": 1.0711582728448033e-05,
+ "loss": 0.45445430278778076,
+ "step": 1464
+ },
+ {
+ "epoch": 0.7686253934942288,
+ "grad_norm": 0.16428618403547932,
+ "learning_rate": 1.0686914613061377e-05,
+ "loss": 0.46050935983657837,
+ "step": 1465
+ },
+ {
+ "epoch": 0.7691500524658972,
+ "grad_norm": 0.15052999053128632,
+ "learning_rate": 1.0662292175063949e-05,
+ "loss": 0.43385741114616394,
+ "step": 1466
+ },
+ {
+ "epoch": 0.7696747114375656,
+ "grad_norm": 0.1654949402740152,
+ "learning_rate": 1.0637715481349397e-05,
+ "loss": 0.4434410333633423,
+ "step": 1467
+ },
+ {
+ "epoch": 0.770199370409234,
+ "grad_norm": 0.15923836508490036,
+ "learning_rate": 1.0613184598687088e-05,
+ "loss": 0.4410185217857361,
+ "step": 1468
+ },
+ {
+ "epoch": 0.7707240293809025,
+ "grad_norm": 0.15813007111170696,
+ "learning_rate": 1.0588699593721957e-05,
+ "loss": 0.43588417768478394,
+ "step": 1469
+ },
+ {
+ "epoch": 0.7712486883525709,
+ "grad_norm": 0.16978817898431764,
+ "learning_rate": 1.0564260532974274e-05,
+ "loss": 0.4514843225479126,
+ "step": 1470
+ },
+ {
+ "epoch": 0.7717733473242392,
+ "grad_norm": 0.1613844852697828,
+ "learning_rate": 1.0539867482839499e-05,
+ "loss": 0.4554520845413208,
+ "step": 1471
+ },
+ {
+ "epoch": 0.7722980062959076,
+ "grad_norm": 0.1587788181984633,
+ "learning_rate": 1.0515520509588075e-05,
+ "loss": 0.43200671672821045,
+ "step": 1472
+ },
+ {
+ "epoch": 0.772822665267576,
+ "grad_norm": 0.1539930824827353,
+ "learning_rate": 1.0491219679365283e-05,
+ "loss": 0.4343116283416748,
+ "step": 1473
+ },
+ {
+ "epoch": 0.7733473242392445,
+ "grad_norm": 0.1561612537773709,
+ "learning_rate": 1.0466965058191049e-05,
+ "loss": 0.4477815330028534,
+ "step": 1474
+ },
+ {
+ "epoch": 0.7738719832109129,
+ "grad_norm": 0.169990777691657,
+ "learning_rate": 1.0442756711959745e-05,
+ "loss": 0.4476611614227295,
+ "step": 1475
+ },
+ {
+ "epoch": 0.7743966421825813,
+ "grad_norm": 0.1624313049193793,
+ "learning_rate": 1.041859470644002e-05,
+ "loss": 0.47136399149894714,
+ "step": 1476
+ },
+ {
+ "epoch": 0.7749213011542497,
+ "grad_norm": 0.15993315458994328,
+ "learning_rate": 1.0394479107274645e-05,
+ "loss": 0.4672670364379883,
+ "step": 1477
+ },
+ {
+ "epoch": 0.7754459601259182,
+ "grad_norm": 0.16755349272347217,
+ "learning_rate": 1.0370409979980294e-05,
+ "loss": 0.45135706663131714,
+ "step": 1478
+ },
+ {
+ "epoch": 0.7759706190975866,
+ "grad_norm": 0.15665059815830362,
+ "learning_rate": 1.0346387389947407e-05,
+ "loss": 0.45178359746932983,
+ "step": 1479
+ },
+ {
+ "epoch": 0.776495278069255,
+ "grad_norm": 0.1554612731452571,
+ "learning_rate": 1.0322411402439983e-05,
+ "loss": 0.4412590265274048,
+ "step": 1480
+ },
+ {
+ "epoch": 0.7770199370409234,
+ "grad_norm": 0.1532496869312065,
+ "learning_rate": 1.0298482082595408e-05,
+ "loss": 0.44104987382888794,
+ "step": 1481
+ },
+ {
+ "epoch": 0.7775445960125918,
+ "grad_norm": 0.15065868316960426,
+ "learning_rate": 1.0274599495424286e-05,
+ "loss": 0.4336603283882141,
+ "step": 1482
+ },
+ {
+ "epoch": 0.7780692549842603,
+ "grad_norm": 0.1548373981760097,
+ "learning_rate": 1.0250763705810267e-05,
+ "loss": 0.4658375680446625,
+ "step": 1483
+ },
+ {
+ "epoch": 0.7785939139559287,
+ "grad_norm": 0.15470181175866582,
+ "learning_rate": 1.0226974778509846e-05,
+ "loss": 0.45232075452804565,
+ "step": 1484
+ },
+ {
+ "epoch": 0.779118572927597,
+ "grad_norm": 0.15507371187727106,
+ "learning_rate": 1.0203232778152228e-05,
+ "loss": 0.43640458583831787,
+ "step": 1485
+ },
+ {
+ "epoch": 0.7796432318992654,
+ "grad_norm": 0.15269440736544085,
+ "learning_rate": 1.01795377692391e-05,
+ "loss": 0.4400939345359802,
+ "step": 1486
+ },
+ {
+ "epoch": 0.7801678908709339,
+ "grad_norm": 0.15610899550395085,
+ "learning_rate": 1.0155889816144495e-05,
+ "loss": 0.4546602666378021,
+ "step": 1487
+ },
+ {
+ "epoch": 0.7806925498426023,
+ "grad_norm": 0.1559328242601668,
+ "learning_rate": 1.0132288983114611e-05,
+ "loss": 0.45510709285736084,
+ "step": 1488
+ },
+ {
+ "epoch": 0.7812172088142707,
+ "grad_norm": 0.1537255194600787,
+ "learning_rate": 1.0108735334267639e-05,
+ "loss": 0.4475342631340027,
+ "step": 1489
+ },
+ {
+ "epoch": 0.7817418677859391,
+ "grad_norm": 0.14616575061664402,
+ "learning_rate": 1.0085228933593558e-05,
+ "loss": 0.4378190040588379,
+ "step": 1490
+ },
+ {
+ "epoch": 0.7822665267576075,
+ "grad_norm": 0.1624360967746598,
+ "learning_rate": 1.0061769844953999e-05,
+ "loss": 0.4439539909362793,
+ "step": 1491
+ },
+ {
+ "epoch": 0.782791185729276,
+ "grad_norm": 0.15407336125298413,
+ "learning_rate": 1.003835813208206e-05,
+ "loss": 0.44074907898902893,
+ "step": 1492
+ },
+ {
+ "epoch": 0.7833158447009444,
+ "grad_norm": 0.15786980710136125,
+ "learning_rate": 1.001499385858211e-05,
+ "loss": 0.4491667151451111,
+ "step": 1493
+ },
+ {
+ "epoch": 0.7838405036726128,
+ "grad_norm": 0.15338836738460088,
+ "learning_rate": 9.991677087929665e-06,
+ "loss": 0.446003794670105,
+ "step": 1494
+ },
+ {
+ "epoch": 0.7843651626442812,
+ "grad_norm": 0.156710894892,
+ "learning_rate": 9.968407883471165e-06,
+ "loss": 0.43604689836502075,
+ "step": 1495
+ },
+ {
+ "epoch": 0.7848898216159497,
+ "grad_norm": 0.16100539063191666,
+ "learning_rate": 9.945186308423832e-06,
+ "loss": 0.4431147873401642,
+ "step": 1496
+ },
+ {
+ "epoch": 0.7854144805876181,
+ "grad_norm": 0.15776018851122997,
+ "learning_rate": 9.92201242587548e-06,
+ "loss": 0.4532504975795746,
+ "step": 1497
+ },
+ {
+ "epoch": 0.7859391395592865,
+ "grad_norm": 0.1540333423540934,
+ "learning_rate": 9.898886298784368e-06,
+ "loss": 0.43605154752731323,
+ "step": 1498
+ },
+ {
+ "epoch": 0.7864637985309548,
+ "grad_norm": 0.16009602938972295,
+ "learning_rate": 9.875807989979005e-06,
+ "loss": 0.4605238437652588,
+ "step": 1499
+ },
+ {
+ "epoch": 0.7869884575026233,
+ "grad_norm": 0.15639368804171708,
+ "learning_rate": 9.852777562157983e-06,
+ "loss": 0.4509708881378174,
+ "step": 1500
+ },
+ {
+ "epoch": 0.7875131164742917,
+ "grad_norm": 0.15074011188741518,
+ "learning_rate": 9.829795077889828e-06,
+ "loss": 0.4527304172515869,
+ "step": 1501
+ },
+ {
+ "epoch": 0.7880377754459601,
+ "grad_norm": 0.15155083996490956,
+ "learning_rate": 9.806860599612799e-06,
+ "loss": 0.4306413531303406,
+ "step": 1502
+ },
+ {
+ "epoch": 0.7885624344176285,
+ "grad_norm": 0.1478956630795445,
+ "learning_rate": 9.783974189634734e-06,
+ "loss": 0.4434714913368225,
+ "step": 1503
+ },
+ {
+ "epoch": 0.789087093389297,
+ "grad_norm": 0.15327668434806432,
+ "learning_rate": 9.761135910132902e-06,
+ "loss": 0.44431227445602417,
+ "step": 1504
+ },
+ {
+ "epoch": 0.7896117523609654,
+ "grad_norm": 0.2243177585497212,
+ "learning_rate": 9.738345823153783e-06,
+ "loss": 0.4541007876396179,
+ "step": 1505
+ },
+ {
+ "epoch": 0.7901364113326338,
+ "grad_norm": 0.1600239865812282,
+ "learning_rate": 9.715603990612945e-06,
+ "loss": 0.4456910490989685,
+ "step": 1506
+ },
+ {
+ "epoch": 0.7906610703043022,
+ "grad_norm": 0.15304634075332066,
+ "learning_rate": 9.692910474294858e-06,
+ "loss": 0.43965545296669006,
+ "step": 1507
+ },
+ {
+ "epoch": 0.7911857292759706,
+ "grad_norm": 0.15657752062053032,
+ "learning_rate": 9.670265335852725e-06,
+ "loss": 0.43358081579208374,
+ "step": 1508
+ },
+ {
+ "epoch": 0.7917103882476391,
+ "grad_norm": 0.16110057783374115,
+ "learning_rate": 9.647668636808304e-06,
+ "loss": 0.45711368322372437,
+ "step": 1509
+ },
+ {
+ "epoch": 0.7922350472193075,
+ "grad_norm": 0.1551133690299623,
+ "learning_rate": 9.625120438551791e-06,
+ "loss": 0.4613519608974457,
+ "step": 1510
+ },
+ {
+ "epoch": 0.7927597061909759,
+ "grad_norm": 0.15690897380725946,
+ "learning_rate": 9.602620802341581e-06,
+ "loss": 0.4472261071205139,
+ "step": 1511
+ },
+ {
+ "epoch": 0.7932843651626443,
+ "grad_norm": 0.147796599322911,
+ "learning_rate": 9.580169789304152e-06,
+ "loss": 0.4447203576564789,
+ "step": 1512
+ },
+ {
+ "epoch": 0.7938090241343126,
+ "grad_norm": 0.152019559199633,
+ "learning_rate": 9.557767460433878e-06,
+ "loss": 0.445905864238739,
+ "step": 1513
+ },
+ {
+ "epoch": 0.7943336831059811,
+ "grad_norm": 0.1574763924872037,
+ "learning_rate": 9.535413876592872e-06,
+ "loss": 0.4502032995223999,
+ "step": 1514
+ },
+ {
+ "epoch": 0.7948583420776495,
+ "grad_norm": 0.1524218315499056,
+ "learning_rate": 9.513109098510816e-06,
+ "loss": 0.43753111362457275,
+ "step": 1515
+ },
+ {
+ "epoch": 0.7953830010493179,
+ "grad_norm": 0.15589147437937434,
+ "learning_rate": 9.490853186784801e-06,
+ "loss": 0.44524312019348145,
+ "step": 1516
+ },
+ {
+ "epoch": 0.7959076600209863,
+ "grad_norm": 0.15604009714081407,
+ "learning_rate": 9.468646201879156e-06,
+ "loss": 0.44141292572021484,
+ "step": 1517
+ },
+ {
+ "epoch": 0.7964323189926548,
+ "grad_norm": 0.1562432554004435,
+ "learning_rate": 9.44648820412529e-06,
+ "loss": 0.461681067943573,
+ "step": 1518
+ },
+ {
+ "epoch": 0.7969569779643232,
+ "grad_norm": 0.15233588531495665,
+ "learning_rate": 9.424379253721513e-06,
+ "loss": 0.4529497027397156,
+ "step": 1519
+ },
+ {
+ "epoch": 0.7974816369359916,
+ "grad_norm": 0.16856991685290923,
+ "learning_rate": 9.40231941073291e-06,
+ "loss": 0.4483596086502075,
+ "step": 1520
+ },
+ {
+ "epoch": 0.79800629590766,
+ "grad_norm": 0.1546876712981077,
+ "learning_rate": 9.380308735091125e-06,
+ "loss": 0.43846625089645386,
+ "step": 1521
+ },
+ {
+ "epoch": 0.7985309548793285,
+ "grad_norm": 0.15198637556890449,
+ "learning_rate": 9.35834728659424e-06,
+ "loss": 0.4625553488731384,
+ "step": 1522
+ },
+ {
+ "epoch": 0.7990556138509969,
+ "grad_norm": 0.1519777007198701,
+ "learning_rate": 9.336435124906592e-06,
+ "loss": 0.44183820486068726,
+ "step": 1523
+ },
+ {
+ "epoch": 0.7995802728226653,
+ "grad_norm": 0.16125823074979284,
+ "learning_rate": 9.314572309558616e-06,
+ "loss": 0.46057307720184326,
+ "step": 1524
+ },
+ {
+ "epoch": 0.8001049317943337,
+ "grad_norm": 0.16283323624848034,
+ "learning_rate": 9.2927588999467e-06,
+ "loss": 0.4530155062675476,
+ "step": 1525
+ },
+ {
+ "epoch": 0.8006295907660022,
+ "grad_norm": 0.14998654489027738,
+ "learning_rate": 9.270994955332984e-06,
+ "loss": 0.43155500292778015,
+ "step": 1526
+ },
+ {
+ "epoch": 0.8011542497376705,
+ "grad_norm": 0.15683095030848093,
+ "learning_rate": 9.249280534845241e-06,
+ "loss": 0.45097050070762634,
+ "step": 1527
+ },
+ {
+ "epoch": 0.8016789087093389,
+ "grad_norm": 0.1561578254492378,
+ "learning_rate": 9.227615697476674e-06,
+ "loss": 0.42797762155532837,
+ "step": 1528
+ },
+ {
+ "epoch": 0.8022035676810073,
+ "grad_norm": 0.15642385261469652,
+ "learning_rate": 9.206000502085813e-06,
+ "loss": 0.4560159742832184,
+ "step": 1529
+ },
+ {
+ "epoch": 0.8027282266526757,
+ "grad_norm": 0.1623126279097499,
+ "learning_rate": 9.1844350073963e-06,
+ "loss": 0.4704165458679199,
+ "step": 1530
+ },
+ {
+ "epoch": 0.8032528856243442,
+ "grad_norm": 0.15090934786253773,
+ "learning_rate": 9.162919271996755e-06,
+ "loss": 0.44642555713653564,
+ "step": 1531
+ },
+ {
+ "epoch": 0.8037775445960126,
+ "grad_norm": 0.15305216978534084,
+ "learning_rate": 9.14145335434062e-06,
+ "loss": 0.4634283781051636,
+ "step": 1532
+ },
+ {
+ "epoch": 0.804302203567681,
+ "grad_norm": 0.14653219861345698,
+ "learning_rate": 9.120037312745979e-06,
+ "loss": 0.44544070959091187,
+ "step": 1533
+ },
+ {
+ "epoch": 0.8048268625393494,
+ "grad_norm": 0.16247397938502464,
+ "learning_rate": 9.098671205395422e-06,
+ "loss": 0.4552367925643921,
+ "step": 1534
+ },
+ {
+ "epoch": 0.8053515215110179,
+ "grad_norm": 0.14963174165902007,
+ "learning_rate": 9.077355090335892e-06,
+ "loss": 0.43211787939071655,
+ "step": 1535
+ },
+ {
+ "epoch": 0.8058761804826863,
+ "grad_norm": 0.1554853846772283,
+ "learning_rate": 9.056089025478495e-06,
+ "loss": 0.4319286644458771,
+ "step": 1536
+ },
+ {
+ "epoch": 0.8064008394543547,
+ "grad_norm": 0.15840895821310472,
+ "learning_rate": 9.03487306859837e-06,
+ "loss": 0.4494994878768921,
+ "step": 1537
+ },
+ {
+ "epoch": 0.8069254984260231,
+ "grad_norm": 0.1543496716800982,
+ "learning_rate": 9.013707277334515e-06,
+ "loss": 0.45500510931015015,
+ "step": 1538
+ },
+ {
+ "epoch": 0.8074501573976915,
+ "grad_norm": 0.14797282318550045,
+ "learning_rate": 8.992591709189651e-06,
+ "loss": 0.4385896325111389,
+ "step": 1539
+ },
+ {
+ "epoch": 0.8079748163693599,
+ "grad_norm": 0.17045570682787758,
+ "learning_rate": 8.971526421530066e-06,
+ "loss": 0.4585988521575928,
+ "step": 1540
+ },
+ {
+ "epoch": 0.8084994753410283,
+ "grad_norm": 0.14758220713044884,
+ "learning_rate": 8.950511471585407e-06,
+ "loss": 0.4295082092285156,
+ "step": 1541
+ },
+ {
+ "epoch": 0.8090241343126967,
+ "grad_norm": 0.16005752068175036,
+ "learning_rate": 8.929546916448597e-06,
+ "loss": 0.4507679343223572,
+ "step": 1542
+ },
+ {
+ "epoch": 0.8095487932843651,
+ "grad_norm": 0.14884684083313074,
+ "learning_rate": 8.908632813075641e-06,
+ "loss": 0.43547600507736206,
+ "step": 1543
+ },
+ {
+ "epoch": 0.8100734522560336,
+ "grad_norm": 0.14860107595788136,
+ "learning_rate": 8.88776921828547e-06,
+ "loss": 0.42700642347335815,
+ "step": 1544
+ },
+ {
+ "epoch": 0.810598111227702,
+ "grad_norm": 0.1472630611843838,
+ "learning_rate": 8.86695618875981e-06,
+ "loss": 0.4473339319229126,
+ "step": 1545
+ },
+ {
+ "epoch": 0.8111227701993704,
+ "grad_norm": 0.15105523788317893,
+ "learning_rate": 8.846193781043e-06,
+ "loss": 0.44923150539398193,
+ "step": 1546
+ },
+ {
+ "epoch": 0.8116474291710388,
+ "grad_norm": 0.16121409015263036,
+ "learning_rate": 8.82548205154185e-06,
+ "loss": 0.4653576612472534,
+ "step": 1547
+ },
+ {
+ "epoch": 0.8121720881427072,
+ "grad_norm": 0.1595195150959421,
+ "learning_rate": 8.804821056525497e-06,
+ "loss": 0.4522424638271332,
+ "step": 1548
+ },
+ {
+ "epoch": 0.8126967471143757,
+ "grad_norm": 0.15163510106316194,
+ "learning_rate": 8.784210852125232e-06,
+ "loss": 0.43657630681991577,
+ "step": 1549
+ },
+ {
+ "epoch": 0.8132214060860441,
+ "grad_norm": 0.15155109347013793,
+ "learning_rate": 8.763651494334377e-06,
+ "loss": 0.43183624744415283,
+ "step": 1550
+ },
+ {
+ "epoch": 0.8137460650577125,
+ "grad_norm": 0.14930024958763324,
+ "learning_rate": 8.7431430390081e-06,
+ "loss": 0.43563947081565857,
+ "step": 1551
+ },
+ {
+ "epoch": 0.8142707240293809,
+ "grad_norm": 0.15361200153430712,
+ "learning_rate": 8.722685541863283e-06,
+ "loss": 0.4387405514717102,
+ "step": 1552
+ },
+ {
+ "epoch": 0.8147953830010494,
+ "grad_norm": 0.15549528796193493,
+ "learning_rate": 8.702279058478366e-06,
+ "loss": 0.44347938895225525,
+ "step": 1553
+ },
+ {
+ "epoch": 0.8153200419727177,
+ "grad_norm": 0.16243042867861712,
+ "learning_rate": 8.681923644293194e-06,
+ "loss": 0.44852834939956665,
+ "step": 1554
+ },
+ {
+ "epoch": 0.8158447009443861,
+ "grad_norm": 0.15706129652944026,
+ "learning_rate": 8.661619354608873e-06,
+ "loss": 0.45419150590896606,
+ "step": 1555
+ },
+ {
+ "epoch": 0.8163693599160545,
+ "grad_norm": 0.14854258573134127,
+ "learning_rate": 8.641366244587611e-06,
+ "loss": 0.44866859912872314,
+ "step": 1556
+ },
+ {
+ "epoch": 0.816894018887723,
+ "grad_norm": 0.14896937678017533,
+ "learning_rate": 8.62116436925258e-06,
+ "loss": 0.44596076011657715,
+ "step": 1557
+ },
+ {
+ "epoch": 0.8174186778593914,
+ "grad_norm": 0.15614508495620508,
+ "learning_rate": 8.601013783487746e-06,
+ "loss": 0.45645207166671753,
+ "step": 1558
+ },
+ {
+ "epoch": 0.8179433368310598,
+ "grad_norm": 0.1486538543050606,
+ "learning_rate": 8.580914542037734e-06,
+ "loss": 0.4592300355434418,
+ "step": 1559
+ },
+ {
+ "epoch": 0.8184679958027282,
+ "grad_norm": 0.15475320756830518,
+ "learning_rate": 8.560866699507702e-06,
+ "loss": 0.4605134129524231,
+ "step": 1560
+ },
+ {
+ "epoch": 0.8189926547743966,
+ "grad_norm": 0.14754406728057914,
+ "learning_rate": 8.540870310363137e-06,
+ "loss": 0.43651673197746277,
+ "step": 1561
+ },
+ {
+ "epoch": 0.8195173137460651,
+ "grad_norm": 0.14871990701077825,
+ "learning_rate": 8.520925428929755e-06,
+ "loss": 0.44830837845802307,
+ "step": 1562
+ },
+ {
+ "epoch": 0.8200419727177335,
+ "grad_norm": 0.15358241317712973,
+ "learning_rate": 8.501032109393334e-06,
+ "loss": 0.4546644687652588,
+ "step": 1563
+ },
+ {
+ "epoch": 0.8205666316894019,
+ "grad_norm": 0.15650856948126138,
+ "learning_rate": 8.481190405799569e-06,
+ "loss": 0.4423891305923462,
+ "step": 1564
+ },
+ {
+ "epoch": 0.8210912906610703,
+ "grad_norm": 0.15474996746242287,
+ "learning_rate": 8.461400372053923e-06,
+ "loss": 0.44242313504219055,
+ "step": 1565
+ },
+ {
+ "epoch": 0.8216159496327388,
+ "grad_norm": 0.16544926591962897,
+ "learning_rate": 8.441662061921497e-06,
+ "loss": 0.4487020969390869,
+ "step": 1566
+ },
+ {
+ "epoch": 0.8221406086044072,
+ "grad_norm": 0.1489733586650647,
+ "learning_rate": 8.421975529026855e-06,
+ "loss": 0.4371356964111328,
+ "step": 1567
+ },
+ {
+ "epoch": 0.8226652675760755,
+ "grad_norm": 0.15995455652906124,
+ "learning_rate": 8.402340826853907e-06,
+ "loss": 0.4566021263599396,
+ "step": 1568
+ },
+ {
+ "epoch": 0.8231899265477439,
+ "grad_norm": 0.1500716152622102,
+ "learning_rate": 8.382758008745726e-06,
+ "loss": 0.43888187408447266,
+ "step": 1569
+ },
+ {
+ "epoch": 0.8237145855194123,
+ "grad_norm": 0.15554904485712012,
+ "learning_rate": 8.363227127904464e-06,
+ "loss": 0.4584001898765564,
+ "step": 1570
+ },
+ {
+ "epoch": 0.8242392444910808,
+ "grad_norm": 0.14820917815884277,
+ "learning_rate": 8.343748237391144e-06,
+ "loss": 0.43289071321487427,
+ "step": 1571
+ },
+ {
+ "epoch": 0.8247639034627492,
+ "grad_norm": 0.15801200747859714,
+ "learning_rate": 8.324321390125547e-06,
+ "loss": 0.44196009635925293,
+ "step": 1572
+ },
+ {
+ "epoch": 0.8252885624344176,
+ "grad_norm": 0.15323277975380797,
+ "learning_rate": 8.304946638886078e-06,
+ "loss": 0.43244463205337524,
+ "step": 1573
+ },
+ {
+ "epoch": 0.825813221406086,
+ "grad_norm": 0.15076940176344772,
+ "learning_rate": 8.285624036309583e-06,
+ "loss": 0.44401073455810547,
+ "step": 1574
+ },
+ {
+ "epoch": 0.8263378803777545,
+ "grad_norm": 0.1544958300475393,
+ "learning_rate": 8.266353634891266e-06,
+ "loss": 0.45198434591293335,
+ "step": 1575
+ },
+ {
+ "epoch": 0.8268625393494229,
+ "grad_norm": 0.15093528292068448,
+ "learning_rate": 8.247135486984486e-06,
+ "loss": 0.4403747618198395,
+ "step": 1576
+ },
+ {
+ "epoch": 0.8273871983210913,
+ "grad_norm": 0.14680662784017148,
+ "learning_rate": 8.227969644800648e-06,
+ "loss": 0.4375043213367462,
+ "step": 1577
+ },
+ {
+ "epoch": 0.8279118572927597,
+ "grad_norm": 0.14791704009291914,
+ "learning_rate": 8.208856160409055e-06,
+ "loss": 0.43811577558517456,
+ "step": 1578
+ },
+ {
+ "epoch": 0.8284365162644282,
+ "grad_norm": 0.14924185710518617,
+ "learning_rate": 8.189795085736765e-06,
+ "loss": 0.43853428959846497,
+ "step": 1579
+ },
+ {
+ "epoch": 0.8289611752360966,
+ "grad_norm": 0.15420576936040198,
+ "learning_rate": 8.170786472568451e-06,
+ "loss": 0.445467084646225,
+ "step": 1580
+ },
+ {
+ "epoch": 0.829485834207765,
+ "grad_norm": 0.15577901772605834,
+ "learning_rate": 8.151830372546258e-06,
+ "loss": 0.4529452621936798,
+ "step": 1581
+ },
+ {
+ "epoch": 0.8300104931794333,
+ "grad_norm": 0.15387840488797552,
+ "learning_rate": 8.132926837169676e-06,
+ "loss": 0.46009817719459534,
+ "step": 1582
+ },
+ {
+ "epoch": 0.8305351521511017,
+ "grad_norm": 0.1527913309957562,
+ "learning_rate": 8.11407591779537e-06,
+ "loss": 0.4513983726501465,
+ "step": 1583
+ },
+ {
+ "epoch": 0.8310598111227702,
+ "grad_norm": 0.1537090286922398,
+ "learning_rate": 8.095277665637061e-06,
+ "loss": 0.4537739157676697,
+ "step": 1584
+ },
+ {
+ "epoch": 0.8315844700944386,
+ "grad_norm": 0.15722962107654076,
+ "learning_rate": 8.076532131765408e-06,
+ "loss": 0.45232969522476196,
+ "step": 1585
+ },
+ {
+ "epoch": 0.832109129066107,
+ "grad_norm": 0.1450142396611787,
+ "learning_rate": 8.057839367107817e-06,
+ "loss": 0.4270634651184082,
+ "step": 1586
+ },
+ {
+ "epoch": 0.8326337880377754,
+ "grad_norm": 0.150264001570592,
+ "learning_rate": 8.039199422448355e-06,
+ "loss": 0.43563878536224365,
+ "step": 1587
+ },
+ {
+ "epoch": 0.8331584470094439,
+ "grad_norm": 0.14904943307590535,
+ "learning_rate": 8.020612348427572e-06,
+ "loss": 0.4401509165763855,
+ "step": 1588
+ },
+ {
+ "epoch": 0.8336831059811123,
+ "grad_norm": 0.15453821662860034,
+ "learning_rate": 8.002078195542393e-06,
+ "loss": 0.44545161724090576,
+ "step": 1589
+ },
+ {
+ "epoch": 0.8342077649527807,
+ "grad_norm": 0.1570809678830383,
+ "learning_rate": 7.983597014145951e-06,
+ "loss": 0.45228466391563416,
+ "step": 1590
+ },
+ {
+ "epoch": 0.8347324239244491,
+ "grad_norm": 0.14741240769629818,
+ "learning_rate": 7.965168854447493e-06,
+ "loss": 0.4485381245613098,
+ "step": 1591
+ },
+ {
+ "epoch": 0.8352570828961176,
+ "grad_norm": 0.15362267095533516,
+ "learning_rate": 7.946793766512195e-06,
+ "loss": 0.4358059763908386,
+ "step": 1592
+ },
+ {
+ "epoch": 0.835781741867786,
+ "grad_norm": 0.1544754586495605,
+ "learning_rate": 7.92847180026106e-06,
+ "loss": 0.43327778577804565,
+ "step": 1593
+ },
+ {
+ "epoch": 0.8363064008394544,
+ "grad_norm": 0.15023225026164963,
+ "learning_rate": 7.91020300547077e-06,
+ "loss": 0.44359731674194336,
+ "step": 1594
+ },
+ {
+ "epoch": 0.8368310598111228,
+ "grad_norm": 0.15665530392095345,
+ "learning_rate": 7.891987431773545e-06,
+ "loss": 0.44503387808799744,
+ "step": 1595
+ },
+ {
+ "epoch": 0.8373557187827911,
+ "grad_norm": 0.14717614926392292,
+ "learning_rate": 7.873825128657033e-06,
+ "loss": 0.45314061641693115,
+ "step": 1596
+ },
+ {
+ "epoch": 0.8378803777544596,
+ "grad_norm": 0.15207464934030612,
+ "learning_rate": 7.855716145464131e-06,
+ "loss": 0.4526442885398865,
+ "step": 1597
+ },
+ {
+ "epoch": 0.838405036726128,
+ "grad_norm": 0.15129881541794873,
+ "learning_rate": 7.837660531392904e-06,
+ "loss": 0.45127853751182556,
+ "step": 1598
+ },
+ {
+ "epoch": 0.8389296956977964,
+ "grad_norm": 0.15623362715417285,
+ "learning_rate": 7.819658335496405e-06,
+ "loss": 0.46432793140411377,
+ "step": 1599
+ },
+ {
+ "epoch": 0.8394543546694648,
+ "grad_norm": 0.1529219879642125,
+ "learning_rate": 7.801709606682573e-06,
+ "loss": 0.4475358724594116,
+ "step": 1600
+ },
+ {
+ "epoch": 0.8399790136411333,
+ "grad_norm": 0.14649139854087698,
+ "learning_rate": 7.78381439371409e-06,
+ "loss": 0.4363391101360321,
+ "step": 1601
+ },
+ {
+ "epoch": 0.8405036726128017,
+ "grad_norm": 0.14571233251948087,
+ "learning_rate": 7.765972745208243e-06,
+ "loss": 0.44419488310813904,
+ "step": 1602
+ },
+ {
+ "epoch": 0.8410283315844701,
+ "grad_norm": 0.15168460066406128,
+ "learning_rate": 7.748184709636791e-06,
+ "loss": 0.4433790445327759,
+ "step": 1603
+ },
+ {
+ "epoch": 0.8415529905561385,
+ "grad_norm": 0.14852860876513757,
+ "learning_rate": 7.730450335325853e-06,
+ "loss": 0.4493539333343506,
+ "step": 1604
+ },
+ {
+ "epoch": 0.842077649527807,
+ "grad_norm": 0.1505615491087377,
+ "learning_rate": 7.712769670455741e-06,
+ "loss": 0.4646318554878235,
+ "step": 1605
+ },
+ {
+ "epoch": 0.8426023084994754,
+ "grad_norm": 0.15407610486660114,
+ "learning_rate": 7.695142763060878e-06,
+ "loss": 0.4484049677848816,
+ "step": 1606
+ },
+ {
+ "epoch": 0.8431269674711438,
+ "grad_norm": 0.1521449709584124,
+ "learning_rate": 7.67756966102962e-06,
+ "loss": 0.43366891145706177,
+ "step": 1607
+ },
+ {
+ "epoch": 0.8436516264428122,
+ "grad_norm": 0.14747762647596166,
+ "learning_rate": 7.660050412104153e-06,
+ "loss": 0.43950092792510986,
+ "step": 1608
+ },
+ {
+ "epoch": 0.8441762854144806,
+ "grad_norm": 0.15512610809240027,
+ "learning_rate": 7.642585063880351e-06,
+ "loss": 0.46813055872917175,
+ "step": 1609
+ },
+ {
+ "epoch": 0.844700944386149,
+ "grad_norm": 0.15825480008734216,
+ "learning_rate": 7.625173663807655e-06,
+ "loss": 0.4348755478858948,
+ "step": 1610
+ },
+ {
+ "epoch": 0.8452256033578174,
+ "grad_norm": 0.1573968981854278,
+ "learning_rate": 7.6078162591889434e-06,
+ "loss": 0.45048123598098755,
+ "step": 1611
+ },
+ {
+ "epoch": 0.8457502623294858,
+ "grad_norm": 0.1517038960096417,
+ "learning_rate": 7.590512897180399e-06,
+ "loss": 0.43563610315322876,
+ "step": 1612
+ },
+ {
+ "epoch": 0.8462749213011542,
+ "grad_norm": 0.15503576044724665,
+ "learning_rate": 7.573263624791382e-06,
+ "loss": 0.4467726945877075,
+ "step": 1613
+ },
+ {
+ "epoch": 0.8467995802728226,
+ "grad_norm": 0.1560549927192375,
+ "learning_rate": 7.556068488884304e-06,
+ "loss": 0.4320477545261383,
+ "step": 1614
+ },
+ {
+ "epoch": 0.8473242392444911,
+ "grad_norm": 0.1679830356844654,
+ "learning_rate": 7.538927536174498e-06,
+ "loss": 0.4476034343242645,
+ "step": 1615
+ },
+ {
+ "epoch": 0.8478488982161595,
+ "grad_norm": 0.15169257383162213,
+ "learning_rate": 7.521840813230098e-06,
+ "loss": 0.4325854778289795,
+ "step": 1616
+ },
+ {
+ "epoch": 0.8483735571878279,
+ "grad_norm": 0.15980226324329827,
+ "learning_rate": 7.504808366471909e-06,
+ "loss": 0.4472721815109253,
+ "step": 1617
+ },
+ {
+ "epoch": 0.8488982161594963,
+ "grad_norm": 0.16932727986736365,
+ "learning_rate": 7.48783024217327e-06,
+ "loss": 0.45428287982940674,
+ "step": 1618
+ },
+ {
+ "epoch": 0.8494228751311648,
+ "grad_norm": 0.15065676306311376,
+ "learning_rate": 7.470906486459946e-06,
+ "loss": 0.44418883323669434,
+ "step": 1619
+ },
+ {
+ "epoch": 0.8499475341028332,
+ "grad_norm": 0.1533992998314183,
+ "learning_rate": 7.454037145309986e-06,
+ "loss": 0.43552401661872864,
+ "step": 1620
+ },
+ {
+ "epoch": 0.8504721930745016,
+ "grad_norm": 0.15479480515120017,
+ "learning_rate": 7.437222264553628e-06,
+ "loss": 0.444465696811676,
+ "step": 1621
+ },
+ {
+ "epoch": 0.85099685204617,
+ "grad_norm": 0.15845678337942112,
+ "learning_rate": 7.420461889873129e-06,
+ "loss": 0.44896578788757324,
+ "step": 1622
+ },
+ {
+ "epoch": 0.8515215110178384,
+ "grad_norm": 0.15038940533473208,
+ "learning_rate": 7.4037560668026804e-06,
+ "loss": 0.43062180280685425,
+ "step": 1623
+ },
+ {
+ "epoch": 0.8520461699895068,
+ "grad_norm": 0.1480017115109342,
+ "learning_rate": 7.387104840728259e-06,
+ "loss": 0.4630432724952698,
+ "step": 1624
+ },
+ {
+ "epoch": 0.8525708289611752,
+ "grad_norm": 0.15321305032824092,
+ "learning_rate": 7.370508256887514e-06,
+ "loss": 0.4419487714767456,
+ "step": 1625
+ },
+ {
+ "epoch": 0.8530954879328436,
+ "grad_norm": 0.1568246735699272,
+ "learning_rate": 7.353966360369659e-06,
+ "loss": 0.4483609199523926,
+ "step": 1626
+ },
+ {
+ "epoch": 0.853620146904512,
+ "grad_norm": 0.15539536583814875,
+ "learning_rate": 7.337479196115317e-06,
+ "loss": 0.4567753076553345,
+ "step": 1627
+ },
+ {
+ "epoch": 0.8541448058761805,
+ "grad_norm": 0.15528695146633797,
+ "learning_rate": 7.321046808916426e-06,
+ "loss": 0.4560334086418152,
+ "step": 1628
+ },
+ {
+ "epoch": 0.8546694648478489,
+ "grad_norm": 0.15005728554864303,
+ "learning_rate": 7.304669243416102e-06,
+ "loss": 0.43833765387535095,
+ "step": 1629
+ },
+ {
+ "epoch": 0.8551941238195173,
+ "grad_norm": 0.15959100645189242,
+ "learning_rate": 7.288346544108518e-06,
+ "loss": 0.4713831841945648,
+ "step": 1630
+ },
+ {
+ "epoch": 0.8557187827911857,
+ "grad_norm": 0.14898121463872632,
+ "learning_rate": 7.272078755338807e-06,
+ "loss": 0.44003745913505554,
+ "step": 1631
+ },
+ {
+ "epoch": 0.8562434417628542,
+ "grad_norm": 0.15083110802343333,
+ "learning_rate": 7.255865921302903e-06,
+ "loss": 0.429004043340683,
+ "step": 1632
+ },
+ {
+ "epoch": 0.8567681007345226,
+ "grad_norm": 0.1522757086796616,
+ "learning_rate": 7.239708086047445e-06,
+ "loss": 0.4403839707374573,
+ "step": 1633
+ },
+ {
+ "epoch": 0.857292759706191,
+ "grad_norm": 0.15719824838921245,
+ "learning_rate": 7.223605293469654e-06,
+ "loss": 0.4377821087837219,
+ "step": 1634
+ },
+ {
+ "epoch": 0.8578174186778594,
+ "grad_norm": 0.1450616202708373,
+ "learning_rate": 7.207557587317216e-06,
+ "loss": 0.4357878267765045,
+ "step": 1635
+ },
+ {
+ "epoch": 0.8583420776495279,
+ "grad_norm": 0.15703162220646427,
+ "learning_rate": 7.191565011188149e-06,
+ "loss": 0.44412243366241455,
+ "step": 1636
+ },
+ {
+ "epoch": 0.8588667366211962,
+ "grad_norm": 0.150175313334118,
+ "learning_rate": 7.175627608530716e-06,
+ "loss": 0.4543853998184204,
+ "step": 1637
+ },
+ {
+ "epoch": 0.8593913955928646,
+ "grad_norm": 0.15091711492716975,
+ "learning_rate": 7.159745422643261e-06,
+ "loss": 0.4268660545349121,
+ "step": 1638
+ },
+ {
+ "epoch": 0.859916054564533,
+ "grad_norm": 0.15449835654964011,
+ "learning_rate": 7.1439184966741295e-06,
+ "loss": 0.4507150650024414,
+ "step": 1639
+ },
+ {
+ "epoch": 0.8604407135362014,
+ "grad_norm": 0.15689612593442026,
+ "learning_rate": 7.128146873621533e-06,
+ "loss": 0.45520636439323425,
+ "step": 1640
+ },
+ {
+ "epoch": 0.8609653725078699,
+ "grad_norm": 0.1565226785234465,
+ "learning_rate": 7.11243059633345e-06,
+ "loss": 0.45102012157440186,
+ "step": 1641
+ },
+ {
+ "epoch": 0.8614900314795383,
+ "grad_norm": 0.15500877922230677,
+ "learning_rate": 7.096769707507481e-06,
+ "loss": 0.447115421295166,
+ "step": 1642
+ },
+ {
+ "epoch": 0.8620146904512067,
+ "grad_norm": 0.1595248428000243,
+ "learning_rate": 7.081164249690761e-06,
+ "loss": 0.4627235233783722,
+ "step": 1643
+ },
+ {
+ "epoch": 0.8625393494228751,
+ "grad_norm": 0.15958177510163488,
+ "learning_rate": 7.065614265279818e-06,
+ "loss": 0.4386388063430786,
+ "step": 1644
+ },
+ {
+ "epoch": 0.8630640083945436,
+ "grad_norm": 0.15152834825396286,
+ "learning_rate": 7.050119796520482e-06,
+ "loss": 0.4531296491622925,
+ "step": 1645
+ },
+ {
+ "epoch": 0.863588667366212,
+ "grad_norm": 0.15635733905345023,
+ "learning_rate": 7.034680885507752e-06,
+ "loss": 0.45573747158050537,
+ "step": 1646
+ },
+ {
+ "epoch": 0.8641133263378804,
+ "grad_norm": 0.15666918852633857,
+ "learning_rate": 7.019297574185703e-06,
+ "loss": 0.4499792158603668,
+ "step": 1647
+ },
+ {
+ "epoch": 0.8646379853095488,
+ "grad_norm": 0.156444627477174,
+ "learning_rate": 7.0039699043473405e-06,
+ "loss": 0.4545009136199951,
+ "step": 1648
+ },
+ {
+ "epoch": 0.8651626442812173,
+ "grad_norm": 0.14638989347446527,
+ "learning_rate": 6.988697917634515e-06,
+ "loss": 0.4397326409816742,
+ "step": 1649
+ },
+ {
+ "epoch": 0.8656873032528857,
+ "grad_norm": 0.15492813701343372,
+ "learning_rate": 6.973481655537793e-06,
+ "loss": 0.4402945041656494,
+ "step": 1650
+ },
+ {
+ "epoch": 0.866211962224554,
+ "grad_norm": 0.14466615385057716,
+ "learning_rate": 6.958321159396352e-06,
+ "loss": 0.4479280114173889,
+ "step": 1651
+ },
+ {
+ "epoch": 0.8667366211962224,
+ "grad_norm": 0.15865895330952856,
+ "learning_rate": 6.943216470397871e-06,
+ "loss": 0.4553218483924866,
+ "step": 1652
+ },
+ {
+ "epoch": 0.8672612801678908,
+ "grad_norm": 0.15293327582467067,
+ "learning_rate": 6.928167629578407e-06,
+ "loss": 0.44509196281433105,
+ "step": 1653
+ },
+ {
+ "epoch": 0.8677859391395593,
+ "grad_norm": 0.1515507946195896,
+ "learning_rate": 6.91317467782229e-06,
+ "loss": 0.4498351216316223,
+ "step": 1654
+ },
+ {
+ "epoch": 0.8683105981112277,
+ "grad_norm": 0.15313372799407085,
+ "learning_rate": 6.898237655862007e-06,
+ "loss": 0.4445887804031372,
+ "step": 1655
+ },
+ {
+ "epoch": 0.8688352570828961,
+ "grad_norm": 0.1491970050887588,
+ "learning_rate": 6.883356604278115e-06,
+ "loss": 0.4323808252811432,
+ "step": 1656
+ },
+ {
+ "epoch": 0.8693599160545645,
+ "grad_norm": 0.14449269685472518,
+ "learning_rate": 6.868531563499096e-06,
+ "loss": 0.42508652806282043,
+ "step": 1657
+ },
+ {
+ "epoch": 0.869884575026233,
+ "grad_norm": 0.14997328847830618,
+ "learning_rate": 6.853762573801264e-06,
+ "loss": 0.43655943870544434,
+ "step": 1658
+ },
+ {
+ "epoch": 0.8704092339979014,
+ "grad_norm": 0.15288208370561934,
+ "learning_rate": 6.839049675308662e-06,
+ "loss": 0.4505232274532318,
+ "step": 1659
+ },
+ {
+ "epoch": 0.8709338929695698,
+ "grad_norm": 0.14416201723141706,
+ "learning_rate": 6.824392907992938e-06,
+ "loss": 0.42444318532943726,
+ "step": 1660
+ },
+ {
+ "epoch": 0.8714585519412382,
+ "grad_norm": 0.14468507187223545,
+ "learning_rate": 6.809792311673244e-06,
+ "loss": 0.4251434803009033,
+ "step": 1661
+ },
+ {
+ "epoch": 0.8719832109129066,
+ "grad_norm": 0.15406295659798216,
+ "learning_rate": 6.795247926016142e-06,
+ "loss": 0.45068371295928955,
+ "step": 1662
+ },
+ {
+ "epoch": 0.8725078698845751,
+ "grad_norm": 0.1516721215663676,
+ "learning_rate": 6.780759790535465e-06,
+ "loss": 0.441638320684433,
+ "step": 1663
+ },
+ {
+ "epoch": 0.8730325288562435,
+ "grad_norm": 0.1537661139082594,
+ "learning_rate": 6.766327944592239e-06,
+ "loss": 0.44044923782348633,
+ "step": 1664
+ },
+ {
+ "epoch": 0.8735571878279118,
+ "grad_norm": 0.15175839438218855,
+ "learning_rate": 6.751952427394553e-06,
+ "loss": 0.44454342126846313,
+ "step": 1665
+ },
+ {
+ "epoch": 0.8740818467995802,
+ "grad_norm": 0.1531195282211392,
+ "learning_rate": 6.737633277997471e-06,
+ "loss": 0.4540765881538391,
+ "step": 1666
+ },
+ {
+ "epoch": 0.8746065057712487,
+ "grad_norm": 0.14961744003585584,
+ "learning_rate": 6.723370535302915e-06,
+ "loss": 0.44223541021347046,
+ "step": 1667
+ },
+ {
+ "epoch": 0.8751311647429171,
+ "grad_norm": 0.14816820299503658,
+ "learning_rate": 6.709164238059568e-06,
+ "loss": 0.43701833486557007,
+ "step": 1668
+ },
+ {
+ "epoch": 0.8756558237145855,
+ "grad_norm": 0.14622402294164918,
+ "learning_rate": 6.695014424862756e-06,
+ "loss": 0.42922335863113403,
+ "step": 1669
+ },
+ {
+ "epoch": 0.8761804826862539,
+ "grad_norm": 0.1547291062309004,
+ "learning_rate": 6.680921134154355e-06,
+ "loss": 0.4233800768852234,
+ "step": 1670
+ },
+ {
+ "epoch": 0.8767051416579223,
+ "grad_norm": 0.15378903523440587,
+ "learning_rate": 6.666884404222678e-06,
+ "loss": 0.4432950019836426,
+ "step": 1671
+ },
+ {
+ "epoch": 0.8772298006295908,
+ "grad_norm": 0.15973037860999773,
+ "learning_rate": 6.652904273202384e-06,
+ "loss": 0.4560699462890625,
+ "step": 1672
+ },
+ {
+ "epoch": 0.8777544596012592,
+ "grad_norm": 0.15430233088028492,
+ "learning_rate": 6.638980779074358e-06,
+ "loss": 0.45776498317718506,
+ "step": 1673
+ },
+ {
+ "epoch": 0.8782791185729276,
+ "grad_norm": 0.1587664730992995,
+ "learning_rate": 6.625113959665614e-06,
+ "loss": 0.43579810857772827,
+ "step": 1674
+ },
+ {
+ "epoch": 0.878803777544596,
+ "grad_norm": 0.14370919039693034,
+ "learning_rate": 6.6113038526491946e-06,
+ "loss": 0.4426189064979553,
+ "step": 1675
+ },
+ {
+ "epoch": 0.8793284365162645,
+ "grad_norm": 0.14933009355373217,
+ "learning_rate": 6.59755049554407e-06,
+ "loss": 0.44747287034988403,
+ "step": 1676
+ },
+ {
+ "epoch": 0.8798530954879329,
+ "grad_norm": 0.15713697891738065,
+ "learning_rate": 6.583853925715039e-06,
+ "loss": 0.46320077776908875,
+ "step": 1677
+ },
+ {
+ "epoch": 0.8803777544596013,
+ "grad_norm": 0.1542096833927892,
+ "learning_rate": 6.570214180372609e-06,
+ "loss": 0.4473888874053955,
+ "step": 1678
+ },
+ {
+ "epoch": 0.8809024134312696,
+ "grad_norm": 0.16187259968193946,
+ "learning_rate": 6.556631296572922e-06,
+ "loss": 0.4421869218349457,
+ "step": 1679
+ },
+ {
+ "epoch": 0.881427072402938,
+ "grad_norm": 0.15978032996583555,
+ "learning_rate": 6.54310531121762e-06,
+ "loss": 0.4305853843688965,
+ "step": 1680
+ },
+ {
+ "epoch": 0.8819517313746065,
+ "grad_norm": 0.15631580579775123,
+ "learning_rate": 6.529636261053784e-06,
+ "loss": 0.4340512454509735,
+ "step": 1681
+ },
+ {
+ "epoch": 0.8824763903462749,
+ "grad_norm": 0.150084161390103,
+ "learning_rate": 6.516224182673812e-06,
+ "loss": 0.4383493661880493,
+ "step": 1682
+ },
+ {
+ "epoch": 0.8830010493179433,
+ "grad_norm": 0.15527917215838966,
+ "learning_rate": 6.502869112515314e-06,
+ "loss": 0.45441877841949463,
+ "step": 1683
+ },
+ {
+ "epoch": 0.8835257082896117,
+ "grad_norm": 0.15164994594035522,
+ "learning_rate": 6.489571086861027e-06,
+ "loss": 0.43363988399505615,
+ "step": 1684
+ },
+ {
+ "epoch": 0.8840503672612802,
+ "grad_norm": 0.1467230535452683,
+ "learning_rate": 6.47633014183871e-06,
+ "loss": 0.4263266623020172,
+ "step": 1685
+ },
+ {
+ "epoch": 0.8845750262329486,
+ "grad_norm": 0.14597001769185472,
+ "learning_rate": 6.463146313421044e-06,
+ "loss": 0.4223391115665436,
+ "step": 1686
+ },
+ {
+ "epoch": 0.885099685204617,
+ "grad_norm": 0.15050446002714588,
+ "learning_rate": 6.4500196374255435e-06,
+ "loss": 0.43097931146621704,
+ "step": 1687
+ },
+ {
+ "epoch": 0.8856243441762854,
+ "grad_norm": 0.14918033614543294,
+ "learning_rate": 6.436950149514448e-06,
+ "loss": 0.4587080478668213,
+ "step": 1688
+ },
+ {
+ "epoch": 0.8861490031479539,
+ "grad_norm": 0.15469492243631364,
+ "learning_rate": 6.423937885194631e-06,
+ "loss": 0.45661866664886475,
+ "step": 1689
+ },
+ {
+ "epoch": 0.8866736621196223,
+ "grad_norm": 0.15552798640177423,
+ "learning_rate": 6.410982879817498e-06,
+ "loss": 0.44370236992836,
+ "step": 1690
+ },
+ {
+ "epoch": 0.8871983210912907,
+ "grad_norm": 0.15595593365088142,
+ "learning_rate": 6.398085168578901e-06,
+ "loss": 0.4291878342628479,
+ "step": 1691
+ },
+ {
+ "epoch": 0.887722980062959,
+ "grad_norm": 0.14708046948755318,
+ "learning_rate": 6.385244786519036e-06,
+ "loss": 0.4422011375427246,
+ "step": 1692
+ },
+ {
+ "epoch": 0.8882476390346274,
+ "grad_norm": 0.1499253382866992,
+ "learning_rate": 6.3724617685223404e-06,
+ "loss": 0.4595015048980713,
+ "step": 1693
+ },
+ {
+ "epoch": 0.8887722980062959,
+ "grad_norm": 0.14702868569252941,
+ "learning_rate": 6.3597361493174205e-06,
+ "loss": 0.4347558319568634,
+ "step": 1694
+ },
+ {
+ "epoch": 0.8892969569779643,
+ "grad_norm": 0.14977289247327386,
+ "learning_rate": 6.347067963476928e-06,
+ "loss": 0.43877455592155457,
+ "step": 1695
+ },
+ {
+ "epoch": 0.8898216159496327,
+ "grad_norm": 0.1453902077891248,
+ "learning_rate": 6.334457245417491e-06,
+ "loss": 0.4282415807247162,
+ "step": 1696
+ },
+ {
+ "epoch": 0.8903462749213011,
+ "grad_norm": 0.1545933676895574,
+ "learning_rate": 6.321904029399607e-06,
+ "loss": 0.4429236054420471,
+ "step": 1697
+ },
+ {
+ "epoch": 0.8908709338929696,
+ "grad_norm": 0.15625844051167467,
+ "learning_rate": 6.3094083495275565e-06,
+ "loss": 0.4284421503543854,
+ "step": 1698
+ },
+ {
+ "epoch": 0.891395592864638,
+ "grad_norm": 0.1484963883381567,
+ "learning_rate": 6.2969702397493e-06,
+ "loss": 0.4545463025569916,
+ "step": 1699
+ },
+ {
+ "epoch": 0.8919202518363064,
+ "grad_norm": 0.15122402806869387,
+ "learning_rate": 6.284589733856403e-06,
+ "loss": 0.43840622901916504,
+ "step": 1700
+ },
+ {
+ "epoch": 0.8924449108079748,
+ "grad_norm": 0.14889313167657334,
+ "learning_rate": 6.27226686548392e-06,
+ "loss": 0.4451153874397278,
+ "step": 1701
+ },
+ {
+ "epoch": 0.8929695697796433,
+ "grad_norm": 0.14724940941036385,
+ "learning_rate": 6.260001668110331e-06,
+ "loss": 0.434165358543396,
+ "step": 1702
+ },
+ {
+ "epoch": 0.8934942287513117,
+ "grad_norm": 0.15520838093295836,
+ "learning_rate": 6.247794175057433e-06,
+ "loss": 0.44641873240470886,
+ "step": 1703
+ },
+ {
+ "epoch": 0.8940188877229801,
+ "grad_norm": 0.159167904797364,
+ "learning_rate": 6.235644419490253e-06,
+ "loss": 0.44591957330703735,
+ "step": 1704
+ },
+ {
+ "epoch": 0.8945435466946485,
+ "grad_norm": 0.14843235881370814,
+ "learning_rate": 6.223552434416951e-06,
+ "loss": 0.43010181188583374,
+ "step": 1705
+ },
+ {
+ "epoch": 0.8950682056663168,
+ "grad_norm": 0.1590998153483183,
+ "learning_rate": 6.211518252688749e-06,
+ "loss": 0.4602118134498596,
+ "step": 1706
+ },
+ {
+ "epoch": 0.8955928646379853,
+ "grad_norm": 0.15257025539137767,
+ "learning_rate": 6.19954190699982e-06,
+ "loss": 0.45429927110671997,
+ "step": 1707
+ },
+ {
+ "epoch": 0.8961175236096537,
+ "grad_norm": 0.15304035006408193,
+ "learning_rate": 6.18762342988722e-06,
+ "loss": 0.4405020773410797,
+ "step": 1708
+ },
+ {
+ "epoch": 0.8966421825813221,
+ "grad_norm": 0.15956303033928781,
+ "learning_rate": 6.175762853730777e-06,
+ "loss": 0.4455980062484741,
+ "step": 1709
+ },
+ {
+ "epoch": 0.8971668415529905,
+ "grad_norm": 0.14973831573751925,
+ "learning_rate": 6.163960210753025e-06,
+ "loss": 0.4331505298614502,
+ "step": 1710
+ },
+ {
+ "epoch": 0.897691500524659,
+ "grad_norm": 0.14573687980573624,
+ "learning_rate": 6.1522155330190985e-06,
+ "loss": 0.43307584524154663,
+ "step": 1711
+ },
+ {
+ "epoch": 0.8982161594963274,
+ "grad_norm": 0.1556523345686865,
+ "learning_rate": 6.140528852436665e-06,
+ "loss": 0.4531600773334503,
+ "step": 1712
+ },
+ {
+ "epoch": 0.8987408184679958,
+ "grad_norm": 0.14262442470666953,
+ "learning_rate": 6.128900200755817e-06,
+ "loss": 0.4350634813308716,
+ "step": 1713
+ },
+ {
+ "epoch": 0.8992654774396642,
+ "grad_norm": 0.14709392307082428,
+ "learning_rate": 6.117329609568998e-06,
+ "loss": 0.43465620279312134,
+ "step": 1714
+ },
+ {
+ "epoch": 0.8997901364113327,
+ "grad_norm": 0.15648040137428026,
+ "learning_rate": 6.105817110310909e-06,
+ "loss": 0.4344560503959656,
+ "step": 1715
+ },
+ {
+ "epoch": 0.9003147953830011,
+ "grad_norm": 0.15758204487742147,
+ "learning_rate": 6.094362734258438e-06,
+ "loss": 0.4409951865673065,
+ "step": 1716
+ },
+ {
+ "epoch": 0.9008394543546695,
+ "grad_norm": 0.15755547382189733,
+ "learning_rate": 6.082966512530558e-06,
+ "loss": 0.44050097465515137,
+ "step": 1717
+ },
+ {
+ "epoch": 0.9013641133263379,
+ "grad_norm": 0.1536165045521179,
+ "learning_rate": 6.071628476088257e-06,
+ "loss": 0.43460220098495483,
+ "step": 1718
+ },
+ {
+ "epoch": 0.9018887722980063,
+ "grad_norm": 0.16253540769141986,
+ "learning_rate": 6.060348655734439e-06,
+ "loss": 0.4490512013435364,
+ "step": 1719
+ },
+ {
+ "epoch": 0.9024134312696747,
+ "grad_norm": 0.15003242262029914,
+ "learning_rate": 6.049127082113853e-06,
+ "loss": 0.44427937269210815,
+ "step": 1720
+ },
+ {
+ "epoch": 0.9029380902413431,
+ "grad_norm": 0.1579908758627471,
+ "learning_rate": 6.037963785712995e-06,
+ "loss": 0.4421137571334839,
+ "step": 1721
+ },
+ {
+ "epoch": 0.9034627492130115,
+ "grad_norm": 0.15111957361250286,
+ "learning_rate": 6.026858796860048e-06,
+ "loss": 0.44011324644088745,
+ "step": 1722
+ },
+ {
+ "epoch": 0.9039874081846799,
+ "grad_norm": 0.1446470746259905,
+ "learning_rate": 6.0158121457247845e-06,
+ "loss": 0.43798986077308655,
+ "step": 1723
+ },
+ {
+ "epoch": 0.9045120671563484,
+ "grad_norm": 0.14882152623490427,
+ "learning_rate": 6.004823862318474e-06,
+ "loss": 0.4265901446342468,
+ "step": 1724
+ },
+ {
+ "epoch": 0.9050367261280168,
+ "grad_norm": 0.1506889859372045,
+ "learning_rate": 5.993893976493831e-06,
+ "loss": 0.4503232538700104,
+ "step": 1725
+ },
+ {
+ "epoch": 0.9055613850996852,
+ "grad_norm": 0.16757194686045146,
+ "learning_rate": 5.983022517944901e-06,
+ "loss": 0.4509217441082001,
+ "step": 1726
+ },
+ {
+ "epoch": 0.9060860440713536,
+ "grad_norm": 0.14447292841509515,
+ "learning_rate": 5.972209516207007e-06,
+ "loss": 0.4476190507411957,
+ "step": 1727
+ },
+ {
+ "epoch": 0.906610703043022,
+ "grad_norm": 0.1536564780391058,
+ "learning_rate": 5.961455000656656e-06,
+ "loss": 0.45448166131973267,
+ "step": 1728
+ },
+ {
+ "epoch": 0.9071353620146905,
+ "grad_norm": 0.15436323862710438,
+ "learning_rate": 5.950759000511455e-06,
+ "loss": 0.45128491520881653,
+ "step": 1729
+ },
+ {
+ "epoch": 0.9076600209863589,
+ "grad_norm": 0.15379817460186834,
+ "learning_rate": 5.940121544830047e-06,
+ "loss": 0.4490806460380554,
+ "step": 1730
+ },
+ {
+ "epoch": 0.9081846799580273,
+ "grad_norm": 0.1470796082413195,
+ "learning_rate": 5.929542662512013e-06,
+ "loss": 0.4371582865715027,
+ "step": 1731
+ },
+ {
+ "epoch": 0.9087093389296957,
+ "grad_norm": 0.15107600076433048,
+ "learning_rate": 5.919022382297809e-06,
+ "loss": 0.45175206661224365,
+ "step": 1732
+ },
+ {
+ "epoch": 0.9092339979013642,
+ "grad_norm": 0.14972329317276084,
+ "learning_rate": 5.908560732768684e-06,
+ "loss": 0.43177497386932373,
+ "step": 1733
+ },
+ {
+ "epoch": 0.9097586568730325,
+ "grad_norm": 0.14437715003340176,
+ "learning_rate": 5.898157742346601e-06,
+ "loss": 0.4243622422218323,
+ "step": 1734
+ },
+ {
+ "epoch": 0.9102833158447009,
+ "grad_norm": 0.15200146823831148,
+ "learning_rate": 5.8878134392941465e-06,
+ "loss": 0.47819408774375916,
+ "step": 1735
+ },
+ {
+ "epoch": 0.9108079748163693,
+ "grad_norm": 0.156864616244505,
+ "learning_rate": 5.877527851714479e-06,
+ "loss": 0.44744759798049927,
+ "step": 1736
+ },
+ {
+ "epoch": 0.9113326337880377,
+ "grad_norm": 0.14489667207174767,
+ "learning_rate": 5.867301007551239e-06,
+ "loss": 0.4390789866447449,
+ "step": 1737
+ },
+ {
+ "epoch": 0.9118572927597062,
+ "grad_norm": 0.1528011587102748,
+ "learning_rate": 5.857132934588479e-06,
+ "loss": 0.4338226616382599,
+ "step": 1738
+ },
+ {
+ "epoch": 0.9123819517313746,
+ "grad_norm": 0.14716603285971375,
+ "learning_rate": 5.847023660450568e-06,
+ "loss": 0.43975532054901123,
+ "step": 1739
+ },
+ {
+ "epoch": 0.912906610703043,
+ "grad_norm": 0.15266609903803277,
+ "learning_rate": 5.836973212602143e-06,
+ "loss": 0.449812650680542,
+ "step": 1740
+ },
+ {
+ "epoch": 0.9134312696747114,
+ "grad_norm": 0.1466770441701569,
+ "learning_rate": 5.826981618348025e-06,
+ "loss": 0.43771812319755554,
+ "step": 1741
+ },
+ {
+ "epoch": 0.9139559286463799,
+ "grad_norm": 0.1428112225146131,
+ "learning_rate": 5.8170489048331305e-06,
+ "loss": 0.42403268814086914,
+ "step": 1742
+ },
+ {
+ "epoch": 0.9144805876180483,
+ "grad_norm": 0.1550829416480759,
+ "learning_rate": 5.8071750990424275e-06,
+ "loss": 0.45707154273986816,
+ "step": 1743
+ },
+ {
+ "epoch": 0.9150052465897167,
+ "grad_norm": 0.14687341315819044,
+ "learning_rate": 5.797360227800832e-06,
+ "loss": 0.4558846354484558,
+ "step": 1744
+ },
+ {
+ "epoch": 0.9155299055613851,
+ "grad_norm": 0.15660848922413237,
+ "learning_rate": 5.787604317773154e-06,
+ "loss": 0.44760435819625854,
+ "step": 1745
+ },
+ {
+ "epoch": 0.9160545645330536,
+ "grad_norm": 0.14891742746821762,
+ "learning_rate": 5.777907395464015e-06,
+ "loss": 0.4266620874404907,
+ "step": 1746
+ },
+ {
+ "epoch": 0.916579223504722,
+ "grad_norm": 0.14501110957005472,
+ "learning_rate": 5.768269487217777e-06,
+ "loss": 0.4388393759727478,
+ "step": 1747
+ },
+ {
+ "epoch": 0.9171038824763903,
+ "grad_norm": 0.14651438673799927,
+ "learning_rate": 5.758690619218493e-06,
+ "loss": 0.4416884183883667,
+ "step": 1748
+ },
+ {
+ "epoch": 0.9176285414480587,
+ "grad_norm": 0.14785104880890787,
+ "learning_rate": 5.749170817489792e-06,
+ "loss": 0.4368356466293335,
+ "step": 1749
+ },
+ {
+ "epoch": 0.9181532004197271,
+ "grad_norm": 0.16586567883267808,
+ "learning_rate": 5.739710107894847e-06,
+ "loss": 0.44294649362564087,
+ "step": 1750
+ },
+ {
+ "epoch": 0.9186778593913956,
+ "grad_norm": 0.14377347738227095,
+ "learning_rate": 5.730308516136288e-06,
+ "loss": 0.4405091106891632,
+ "step": 1751
+ },
+ {
+ "epoch": 0.919202518363064,
+ "grad_norm": 0.14703077660050376,
+ "learning_rate": 5.72096606775613e-06,
+ "loss": 0.44809678196907043,
+ "step": 1752
+ },
+ {
+ "epoch": 0.9197271773347324,
+ "grad_norm": 0.15073590419140934,
+ "learning_rate": 5.711682788135726e-06,
+ "loss": 0.43815505504608154,
+ "step": 1753
+ },
+ {
+ "epoch": 0.9202518363064008,
+ "grad_norm": 0.14774775313473323,
+ "learning_rate": 5.702458702495658e-06,
+ "loss": 0.4417206048965454,
+ "step": 1754
+ },
+ {
+ "epoch": 0.9207764952780693,
+ "grad_norm": 0.14289335805130654,
+ "learning_rate": 5.693293835895706e-06,
+ "loss": 0.42886218428611755,
+ "step": 1755
+ },
+ {
+ "epoch": 0.9213011542497377,
+ "grad_norm": 0.15571224276102627,
+ "learning_rate": 5.684188213234763e-06,
+ "loss": 0.44488853216171265,
+ "step": 1756
+ },
+ {
+ "epoch": 0.9218258132214061,
+ "grad_norm": 0.147115948090654,
+ "learning_rate": 5.675141859250763e-06,
+ "loss": 0.44682830572128296,
+ "step": 1757
+ },
+ {
+ "epoch": 0.9223504721930745,
+ "grad_norm": 0.15349533122985531,
+ "learning_rate": 5.666154798520627e-06,
+ "loss": 0.45550864934921265,
+ "step": 1758
+ },
+ {
+ "epoch": 0.922875131164743,
+ "grad_norm": 0.16960595806748266,
+ "learning_rate": 5.657227055460184e-06,
+ "loss": 0.4605559706687927,
+ "step": 1759
+ },
+ {
+ "epoch": 0.9233997901364114,
+ "grad_norm": 0.14665490922506455,
+ "learning_rate": 5.648358654324119e-06,
+ "loss": 0.45442628860473633,
+ "step": 1760
+ },
+ {
+ "epoch": 0.9239244491080797,
+ "grad_norm": 0.14780665809126886,
+ "learning_rate": 5.6395496192058905e-06,
+ "loss": 0.45085081458091736,
+ "step": 1761
+ },
+ {
+ "epoch": 0.9244491080797481,
+ "grad_norm": 0.15599036700169416,
+ "learning_rate": 5.6307999740376714e-06,
+ "loss": 0.4581005573272705,
+ "step": 1762
+ },
+ {
+ "epoch": 0.9249737670514165,
+ "grad_norm": 0.14383027603716222,
+ "learning_rate": 5.622109742590294e-06,
+ "loss": 0.43986010551452637,
+ "step": 1763
+ },
+ {
+ "epoch": 0.925498426023085,
+ "grad_norm": 0.1510369649046927,
+ "learning_rate": 5.613478948473166e-06,
+ "loss": 0.4317719340324402,
+ "step": 1764
+ },
+ {
+ "epoch": 0.9260230849947534,
+ "grad_norm": 0.16608560055258376,
+ "learning_rate": 5.60490761513423e-06,
+ "loss": 0.4500439167022705,
+ "step": 1765
+ },
+ {
+ "epoch": 0.9265477439664218,
+ "grad_norm": 0.15364415343465948,
+ "learning_rate": 5.596395765859875e-06,
+ "loss": 0.43703922629356384,
+ "step": 1766
+ },
+ {
+ "epoch": 0.9270724029380902,
+ "grad_norm": 0.145164767782082,
+ "learning_rate": 5.587943423774886e-06,
+ "loss": 0.4389477074146271,
+ "step": 1767
+ },
+ {
+ "epoch": 0.9275970619097587,
+ "grad_norm": 0.15044099308887526,
+ "learning_rate": 5.579550611842395e-06,
+ "loss": 0.43600398302078247,
+ "step": 1768
+ },
+ {
+ "epoch": 0.9281217208814271,
+ "grad_norm": 0.14694381874423756,
+ "learning_rate": 5.571217352863786e-06,
+ "loss": 0.4525601863861084,
+ "step": 1769
+ },
+ {
+ "epoch": 0.9286463798530955,
+ "grad_norm": 0.15979616589190854,
+ "learning_rate": 5.562943669478661e-06,
+ "loss": 0.44730353355407715,
+ "step": 1770
+ },
+ {
+ "epoch": 0.9291710388247639,
+ "grad_norm": 0.1545596231710442,
+ "learning_rate": 5.554729584164762e-06,
+ "loss": 0.43867969512939453,
+ "step": 1771
+ },
+ {
+ "epoch": 0.9296956977964324,
+ "grad_norm": 0.15923930406000852,
+ "learning_rate": 5.5465751192379235e-06,
+ "loss": 0.4648996889591217,
+ "step": 1772
+ },
+ {
+ "epoch": 0.9302203567681008,
+ "grad_norm": 0.14001241391446362,
+ "learning_rate": 5.538480296852001e-06,
+ "loss": 0.41449663043022156,
+ "step": 1773
+ },
+ {
+ "epoch": 0.9307450157397692,
+ "grad_norm": 0.16538765001279418,
+ "learning_rate": 5.530445138998815e-06,
+ "loss": 0.4503074288368225,
+ "step": 1774
+ },
+ {
+ "epoch": 0.9312696747114375,
+ "grad_norm": 0.15007886784403335,
+ "learning_rate": 5.5224696675080904e-06,
+ "loss": 0.44717323780059814,
+ "step": 1775
+ },
+ {
+ "epoch": 0.9317943336831059,
+ "grad_norm": 0.17834856046133146,
+ "learning_rate": 5.514553904047395e-06,
+ "loss": 0.4374881386756897,
+ "step": 1776
+ },
+ {
+ "epoch": 0.9323189926547744,
+ "grad_norm": 0.1494478926649096,
+ "learning_rate": 5.5066978701220865e-06,
+ "loss": 0.4464038610458374,
+ "step": 1777
+ },
+ {
+ "epoch": 0.9328436516264428,
+ "grad_norm": 0.14887448793632826,
+ "learning_rate": 5.498901587075253e-06,
+ "loss": 0.443928062915802,
+ "step": 1778
+ },
+ {
+ "epoch": 0.9333683105981112,
+ "grad_norm": 0.14192884484464632,
+ "learning_rate": 5.491165076087652e-06,
+ "loss": 0.42002278566360474,
+ "step": 1779
+ },
+ {
+ "epoch": 0.9338929695697796,
+ "grad_norm": 0.14783356724791524,
+ "learning_rate": 5.483488358177647e-06,
+ "loss": 0.43647563457489014,
+ "step": 1780
+ },
+ {
+ "epoch": 0.934417628541448,
+ "grad_norm": 0.15004597148119703,
+ "learning_rate": 5.475871454201163e-06,
+ "loss": 0.4396110773086548,
+ "step": 1781
+ },
+ {
+ "epoch": 0.9349422875131165,
+ "grad_norm": 0.15668918073856763,
+ "learning_rate": 5.468314384851626e-06,
+ "loss": 0.42848116159439087,
+ "step": 1782
+ },
+ {
+ "epoch": 0.9354669464847849,
+ "grad_norm": 0.1449937199040528,
+ "learning_rate": 5.460817170659901e-06,
+ "loss": 0.44262629747390747,
+ "step": 1783
+ },
+ {
+ "epoch": 0.9359916054564533,
+ "grad_norm": 0.15868260486048544,
+ "learning_rate": 5.4533798319942375e-06,
+ "loss": 0.44874119758605957,
+ "step": 1784
+ },
+ {
+ "epoch": 0.9365162644281217,
+ "grad_norm": 0.15004870783258292,
+ "learning_rate": 5.446002389060224e-06,
+ "loss": 0.4575575590133667,
+ "step": 1785
+ },
+ {
+ "epoch": 0.9370409233997902,
+ "grad_norm": 0.14830248639010987,
+ "learning_rate": 5.438684861900721e-06,
+ "loss": 0.44039857387542725,
+ "step": 1786
+ },
+ {
+ "epoch": 0.9375655823714586,
+ "grad_norm": 0.15429570419507257,
+ "learning_rate": 5.431427270395808e-06,
+ "loss": 0.43363744020462036,
+ "step": 1787
+ },
+ {
+ "epoch": 0.938090241343127,
+ "grad_norm": 0.15982575395123055,
+ "learning_rate": 5.4242296342627354e-06,
+ "loss": 0.4484976530075073,
+ "step": 1788
+ },
+ {
+ "epoch": 0.9386149003147953,
+ "grad_norm": 0.1466121447285136,
+ "learning_rate": 5.417091973055878e-06,
+ "loss": 0.44059592485427856,
+ "step": 1789
+ },
+ {
+ "epoch": 0.9391395592864638,
+ "grad_norm": 0.15353848724097463,
+ "learning_rate": 5.410014306166657e-06,
+ "loss": 0.4600589871406555,
+ "step": 1790
+ },
+ {
+ "epoch": 0.9396642182581322,
+ "grad_norm": 0.14968071650065867,
+ "learning_rate": 5.4029966528235065e-06,
+ "loss": 0.4415352940559387,
+ "step": 1791
+ },
+ {
+ "epoch": 0.9401888772298006,
+ "grad_norm": 0.14819069534768417,
+ "learning_rate": 5.396039032091826e-06,
+ "loss": 0.4436836242675781,
+ "step": 1792
+ },
+ {
+ "epoch": 0.940713536201469,
+ "grad_norm": 0.14683654708689367,
+ "learning_rate": 5.389141462873908e-06,
+ "loss": 0.4243587851524353,
+ "step": 1793
+ },
+ {
+ "epoch": 0.9412381951731374,
+ "grad_norm": 0.15431293103561666,
+ "learning_rate": 5.382303963908908e-06,
+ "loss": 0.450438916683197,
+ "step": 1794
+ },
+ {
+ "epoch": 0.9417628541448059,
+ "grad_norm": 0.15125346717879914,
+ "learning_rate": 5.3755265537727775e-06,
+ "loss": 0.4469110369682312,
+ "step": 1795
+ },
+ {
+ "epoch": 0.9422875131164743,
+ "grad_norm": 0.15430864502191044,
+ "learning_rate": 5.368809250878227e-06,
+ "loss": 0.45182669162750244,
+ "step": 1796
+ },
+ {
+ "epoch": 0.9428121720881427,
+ "grad_norm": 0.1515431753176237,
+ "learning_rate": 5.362152073474657e-06,
+ "loss": 0.43683797121047974,
+ "step": 1797
+ },
+ {
+ "epoch": 0.9433368310598111,
+ "grad_norm": 0.1524915522484605,
+ "learning_rate": 5.355555039648133e-06,
+ "loss": 0.44641250371932983,
+ "step": 1798
+ },
+ {
+ "epoch": 0.9438614900314796,
+ "grad_norm": 0.15722684273308743,
+ "learning_rate": 5.3490181673213206e-06,
+ "loss": 0.43558835983276367,
+ "step": 1799
+ },
+ {
+ "epoch": 0.944386149003148,
+ "grad_norm": 0.15527272713567492,
+ "learning_rate": 5.342541474253438e-06,
+ "loss": 0.4423788785934448,
+ "step": 1800
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 4.76536688869376e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1800/video_preprocessor_config.json b/checkpoints/checkpoint-1800/video_preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..a933f0ec6941222c33aeba7980bd5acf8dd98e91
--- /dev/null
+++ b/checkpoints/checkpoint-1800/video_preprocessor_config.json
@@ -0,0 +1,30 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "do_sample_frames": true,
+ "fps": 2,
+ "image_mean": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "image_std": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "max_frames": 768,
+ "merge_size": 2,
+ "min_frames": 4,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 786432,
+ "shortest_edge": 131072
+ },
+ "temporal_patch_size": 2,
+ "video_processor_type": "Qwen3VLVideoProcessor"
+}
diff --git a/checkpoints/checkpoint-1850/chat_template.jinja b/checkpoints/checkpoint-1850/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1850/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1850/config.json b/checkpoints/checkpoint-1850/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1850/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1850/expert_prune_metadata.json b/checkpoints/checkpoint-1850/expert_prune_metadata.json
new file mode 100644
index 0000000000000000000000000000000000000000..16e01c8a10636331e334732fa9629579c68df44e
--- /dev/null
+++ b/checkpoints/checkpoint-1850/expert_prune_metadata.json
@@ -0,0 +1,14 @@
+{
+ "source_model": "/mnt/data/wuwei/model_list/Qwen3.6-35B-A3B/models--Qwen--Qwen3.6-35B-A3B/snapshots/995ad96eacd98c81ed38be0c5b274b04031597b0",
+ "output_model": "/mnt/data/wuwei/model_list/qwen3-19b-a3b",
+ "total_parameters": 18990568816,
+ "parameter_billions_decimal": 18.990568816,
+ "directory_size_label": "19B",
+ "original_num_experts": 256,
+ "kept_experts_per_layer": 128,
+ "num_profiled_layers": 40,
+ "mtp_weights": "removed because Qwen3_5MoeForConditionalGeneration does not instantiate or use the source checkpoint's mtp.* training-only tensors",
+ "expert_order": "ascending original expert index; router rows and expert tensors use the same remapping",
+ "router_statistics": "router_stats.json and router_stats.pt",
+ "selection_score": "top-8 router hit frequency * mean(abs(actual expert output after down_proj and before routing-weight multiplication))"
+}
diff --git a/checkpoints/checkpoint-1850/generation_config.json b/checkpoints/checkpoint-1850/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1850/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1850/model.safetensors b/checkpoints/checkpoint-1850/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..f2619d07ebeaa22ee6993af274fdfca6be14721f
--- /dev/null
+++ b/checkpoints/checkpoint-1850/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a719ad2b8c605b81d12f24e43d2463f50fba9c23322e922cd5e08bcb62adccd
+size 37983421000
diff --git a/checkpoints/checkpoint-1850/preprocessor_config.json b/checkpoints/checkpoint-1850/preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ec43d1f859ecadf800ebae92fda4a812d34975d3
--- /dev/null
+++ b/checkpoints/checkpoint-1850/preprocessor_config.json
@@ -0,0 +1,26 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "image_mean": [
+ 0.48145466,
+ 0.4578275,
+ 0.40821073
+ ],
+ "image_processor_type": "Qwen2VLImageProcessorFast",
+ "image_std": [
+ 0.26862954,
+ 0.26130258,
+ 0.27577711
+ ],
+ "merge_size": 2,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 1003520,
+ "shortest_edge": 3136
+ },
+ "temporal_patch_size": 2
+}
diff --git a/checkpoints/checkpoint-1850/tokenizer.json b/checkpoints/checkpoint-1850/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1850/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1850/tokenizer_config.json b/checkpoints/checkpoint-1850/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1850/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1850/trainer_state.json b/checkpoints/checkpoint-1850/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..547b536bd04d0299f79cc9aa10cedcafd8cd4b16
--- /dev/null
+++ b/checkpoints/checkpoint-1850/trainer_state.json
@@ -0,0 +1,12984 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.9706190975865687,
+ "eval_steps": 500,
+ "global_step": 1850,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ },
+ {
+ "epoch": 0.5776495278069255,
+ "grad_norm": 0.1862053032854098,
+ "learning_rate": 2.2065553964033126e-05,
+ "loss": 0.4816875457763672,
+ "step": 1101
+ },
+ {
+ "epoch": 0.5781741867785939,
+ "grad_norm": 0.16996015894363117,
+ "learning_rate": 2.20295734045532e-05,
+ "loss": 0.4471197724342346,
+ "step": 1102
+ },
+ {
+ "epoch": 0.5786988457502623,
+ "grad_norm": 0.16889227008390692,
+ "learning_rate": 2.1993607706997725e-05,
+ "loss": 0.468391478061676,
+ "step": 1103
+ },
+ {
+ "epoch": 0.5792235047219307,
+ "grad_norm": 0.16670931681562373,
+ "learning_rate": 2.195765696907745e-05,
+ "loss": 0.44934162497520447,
+ "step": 1104
+ },
+ {
+ "epoch": 0.5797481636935992,
+ "grad_norm": 0.16328418741223716,
+ "learning_rate": 2.1921721288462482e-05,
+ "loss": 0.4558270275592804,
+ "step": 1105
+ },
+ {
+ "epoch": 0.5802728226652676,
+ "grad_norm": 0.17364660408703936,
+ "learning_rate": 2.188580076278203e-05,
+ "loss": 0.4597751796245575,
+ "step": 1106
+ },
+ {
+ "epoch": 0.580797481636936,
+ "grad_norm": 0.1628618453800601,
+ "learning_rate": 2.18498954896241e-05,
+ "loss": 0.4521859288215637,
+ "step": 1107
+ },
+ {
+ "epoch": 0.5813221406086044,
+ "grad_norm": 0.16813817951871482,
+ "learning_rate": 2.1814005566535302e-05,
+ "loss": 0.4581005573272705,
+ "step": 1108
+ },
+ {
+ "epoch": 0.5818467995802729,
+ "grad_norm": 0.16825923629726025,
+ "learning_rate": 2.1778131091020516e-05,
+ "loss": 0.4433916211128235,
+ "step": 1109
+ },
+ {
+ "epoch": 0.5823714585519413,
+ "grad_norm": 0.17194743467883944,
+ "learning_rate": 2.1742272160542654e-05,
+ "loss": 0.47296425700187683,
+ "step": 1110
+ },
+ {
+ "epoch": 0.5828961175236097,
+ "grad_norm": 0.16947730546039969,
+ "learning_rate": 2.170642887252241e-05,
+ "loss": 0.4815937876701355,
+ "step": 1111
+ },
+ {
+ "epoch": 0.583420776495278,
+ "grad_norm": 0.1657449928668473,
+ "learning_rate": 2.167060132433798e-05,
+ "loss": 0.4479164779186249,
+ "step": 1112
+ },
+ {
+ "epoch": 0.5839454354669464,
+ "grad_norm": 0.1658952516607364,
+ "learning_rate": 2.1634789613324778e-05,
+ "loss": 0.4637463688850403,
+ "step": 1113
+ },
+ {
+ "epoch": 0.5844700944386149,
+ "grad_norm": 0.16777002188998236,
+ "learning_rate": 2.1598993836775218e-05,
+ "loss": 0.457524836063385,
+ "step": 1114
+ },
+ {
+ "epoch": 0.5849947534102833,
+ "grad_norm": 0.15837566447341322,
+ "learning_rate": 2.1563214091938405e-05,
+ "loss": 0.44938087463378906,
+ "step": 1115
+ },
+ {
+ "epoch": 0.5855194123819517,
+ "grad_norm": 0.17162206617035314,
+ "learning_rate": 2.1527450476019896e-05,
+ "loss": 0.4393450617790222,
+ "step": 1116
+ },
+ {
+ "epoch": 0.5860440713536201,
+ "grad_norm": 0.16224101921433814,
+ "learning_rate": 2.1491703086181426e-05,
+ "loss": 0.451913982629776,
+ "step": 1117
+ },
+ {
+ "epoch": 0.5865687303252886,
+ "grad_norm": 0.1736253504575144,
+ "learning_rate": 2.145597201954066e-05,
+ "loss": 0.4611290395259857,
+ "step": 1118
+ },
+ {
+ "epoch": 0.587093389296957,
+ "grad_norm": 0.16308408333737234,
+ "learning_rate": 2.1420257373170903e-05,
+ "loss": 0.45898568630218506,
+ "step": 1119
+ },
+ {
+ "epoch": 0.5876180482686254,
+ "grad_norm": 0.16948634794175524,
+ "learning_rate": 2.1384559244100848e-05,
+ "loss": 0.45982882380485535,
+ "step": 1120
+ },
+ {
+ "epoch": 0.5881427072402938,
+ "grad_norm": 0.16676366538565252,
+ "learning_rate": 2.1348877729314336e-05,
+ "loss": 0.44826576113700867,
+ "step": 1121
+ },
+ {
+ "epoch": 0.5886673662119623,
+ "grad_norm": 0.16431125630712295,
+ "learning_rate": 2.1313212925750037e-05,
+ "loss": 0.4616445302963257,
+ "step": 1122
+ },
+ {
+ "epoch": 0.5891920251836307,
+ "grad_norm": 0.1631659061616955,
+ "learning_rate": 2.1277564930301252e-05,
+ "loss": 0.45651188492774963,
+ "step": 1123
+ },
+ {
+ "epoch": 0.5897166841552991,
+ "grad_norm": 0.15805950502689237,
+ "learning_rate": 2.1241933839815604e-05,
+ "loss": 0.4486045837402344,
+ "step": 1124
+ },
+ {
+ "epoch": 0.5902413431269675,
+ "grad_norm": 0.16530288038603502,
+ "learning_rate": 2.1206319751094773e-05,
+ "loss": 0.4485066533088684,
+ "step": 1125
+ },
+ {
+ "epoch": 0.5907660020986358,
+ "grad_norm": 0.16061903194239047,
+ "learning_rate": 2.1170722760894295e-05,
+ "loss": 0.4565330743789673,
+ "step": 1126
+ },
+ {
+ "epoch": 0.5912906610703043,
+ "grad_norm": 0.1643346999353627,
+ "learning_rate": 2.1135142965923195e-05,
+ "loss": 0.45560717582702637,
+ "step": 1127
+ },
+ {
+ "epoch": 0.5918153200419727,
+ "grad_norm": 0.17484644315215753,
+ "learning_rate": 2.109958046284382e-05,
+ "loss": 0.4795891046524048,
+ "step": 1128
+ },
+ {
+ "epoch": 0.5923399790136411,
+ "grad_norm": 0.1623762678622651,
+ "learning_rate": 2.1064035348271528e-05,
+ "loss": 0.47395604848861694,
+ "step": 1129
+ },
+ {
+ "epoch": 0.5928646379853095,
+ "grad_norm": 0.16888824241669015,
+ "learning_rate": 2.1028507718774456e-05,
+ "loss": 0.4601455330848694,
+ "step": 1130
+ },
+ {
+ "epoch": 0.593389296956978,
+ "grad_norm": 0.16291744568806302,
+ "learning_rate": 2.0992997670873193e-05,
+ "loss": 0.46248236298561096,
+ "step": 1131
+ },
+ {
+ "epoch": 0.5939139559286464,
+ "grad_norm": 0.16185598691236988,
+ "learning_rate": 2.0957505301040603e-05,
+ "loss": 0.45085301995277405,
+ "step": 1132
+ },
+ {
+ "epoch": 0.5944386149003148,
+ "grad_norm": 0.1678000806117814,
+ "learning_rate": 2.0922030705701518e-05,
+ "loss": 0.44939014315605164,
+ "step": 1133
+ },
+ {
+ "epoch": 0.5949632738719832,
+ "grad_norm": 0.16104474282288994,
+ "learning_rate": 2.088657398123246e-05,
+ "loss": 0.4313976764678955,
+ "step": 1134
+ },
+ {
+ "epoch": 0.5954879328436516,
+ "grad_norm": 0.1631128968199669,
+ "learning_rate": 2.085113522396141e-05,
+ "loss": 0.4508278965950012,
+ "step": 1135
+ },
+ {
+ "epoch": 0.5960125918153201,
+ "grad_norm": 0.16307688195844533,
+ "learning_rate": 2.0815714530167557e-05,
+ "loss": 0.4467576742172241,
+ "step": 1136
+ },
+ {
+ "epoch": 0.5965372507869885,
+ "grad_norm": 0.16456832304532207,
+ "learning_rate": 2.0780311996080976e-05,
+ "loss": 0.4621703624725342,
+ "step": 1137
+ },
+ {
+ "epoch": 0.5970619097586569,
+ "grad_norm": 0.16346217748325367,
+ "learning_rate": 2.0744927717882435e-05,
+ "loss": 0.4589449465274811,
+ "step": 1138
+ },
+ {
+ "epoch": 0.5975865687303253,
+ "grad_norm": 0.16180691501076994,
+ "learning_rate": 2.0709561791703097e-05,
+ "loss": 0.43698400259017944,
+ "step": 1139
+ },
+ {
+ "epoch": 0.5981112277019937,
+ "grad_norm": 0.15695459976898835,
+ "learning_rate": 2.067421431362427e-05,
+ "loss": 0.452271044254303,
+ "step": 1140
+ },
+ {
+ "epoch": 0.5986358866736621,
+ "grad_norm": 0.15930893553721054,
+ "learning_rate": 2.063888537967713e-05,
+ "loss": 0.44868963956832886,
+ "step": 1141
+ },
+ {
+ "epoch": 0.5991605456453305,
+ "grad_norm": 0.15963802284017928,
+ "learning_rate": 2.0603575085842485e-05,
+ "loss": 0.4376864731311798,
+ "step": 1142
+ },
+ {
+ "epoch": 0.5996852046169989,
+ "grad_norm": 0.16610266900507178,
+ "learning_rate": 2.05682835280505e-05,
+ "loss": 0.44676727056503296,
+ "step": 1143
+ },
+ {
+ "epoch": 0.6002098635886673,
+ "grad_norm": 0.15978253236669612,
+ "learning_rate": 2.053301080218044e-05,
+ "loss": 0.434773713350296,
+ "step": 1144
+ },
+ {
+ "epoch": 0.6007345225603358,
+ "grad_norm": 0.16305141803526324,
+ "learning_rate": 2.0497757004060403e-05,
+ "loss": 0.4529780149459839,
+ "step": 1145
+ },
+ {
+ "epoch": 0.6012591815320042,
+ "grad_norm": 0.1649737864101215,
+ "learning_rate": 2.046252222946707e-05,
+ "loss": 0.4571434259414673,
+ "step": 1146
+ },
+ {
+ "epoch": 0.6017838405036726,
+ "grad_norm": 0.15738824968326598,
+ "learning_rate": 2.0427306574125427e-05,
+ "loss": 0.4446289539337158,
+ "step": 1147
+ },
+ {
+ "epoch": 0.602308499475341,
+ "grad_norm": 0.1731139535185187,
+ "learning_rate": 2.039211013370853e-05,
+ "loss": 0.4551090598106384,
+ "step": 1148
+ },
+ {
+ "epoch": 0.6028331584470095,
+ "grad_norm": 0.15857251900043037,
+ "learning_rate": 2.0356933003837237e-05,
+ "loss": 0.4644010663032532,
+ "step": 1149
+ },
+ {
+ "epoch": 0.6033578174186779,
+ "grad_norm": 0.1589582861134211,
+ "learning_rate": 2.032177528007992e-05,
+ "loss": 0.45741650462150574,
+ "step": 1150
+ },
+ {
+ "epoch": 0.6038824763903463,
+ "grad_norm": 0.15639092706575122,
+ "learning_rate": 2.0286637057952257e-05,
+ "loss": 0.4375547766685486,
+ "step": 1151
+ },
+ {
+ "epoch": 0.6044071353620147,
+ "grad_norm": 0.16335605581337187,
+ "learning_rate": 2.0251518432916933e-05,
+ "loss": 0.4636140465736389,
+ "step": 1152
+ },
+ {
+ "epoch": 0.6049317943336832,
+ "grad_norm": 0.15437437113359356,
+ "learning_rate": 2.0216419500383372e-05,
+ "loss": 0.44632142782211304,
+ "step": 1153
+ },
+ {
+ "epoch": 0.6054564533053515,
+ "grad_norm": 0.15522039198698664,
+ "learning_rate": 2.0181340355707545e-05,
+ "loss": 0.449049174785614,
+ "step": 1154
+ },
+ {
+ "epoch": 0.6059811122770199,
+ "grad_norm": 0.1599951917909323,
+ "learning_rate": 2.0146281094191604e-05,
+ "loss": 0.44395962357521057,
+ "step": 1155
+ },
+ {
+ "epoch": 0.6065057712486883,
+ "grad_norm": 0.16866621620527716,
+ "learning_rate": 2.011124181108373e-05,
+ "loss": 0.45693716406822205,
+ "step": 1156
+ },
+ {
+ "epoch": 0.6070304302203567,
+ "grad_norm": 0.15508935650135075,
+ "learning_rate": 2.0076222601577815e-05,
+ "loss": 0.4525151252746582,
+ "step": 1157
+ },
+ {
+ "epoch": 0.6075550891920252,
+ "grad_norm": 0.16508998118005971,
+ "learning_rate": 2.004122356081321e-05,
+ "loss": 0.466006875038147,
+ "step": 1158
+ },
+ {
+ "epoch": 0.6080797481636936,
+ "grad_norm": 0.15817418864022237,
+ "learning_rate": 2.0006244783874466e-05,
+ "loss": 0.4496110677719116,
+ "step": 1159
+ },
+ {
+ "epoch": 0.608604407135362,
+ "grad_norm": 0.16127270259749824,
+ "learning_rate": 1.997128636579109e-05,
+ "loss": 0.45918628573417664,
+ "step": 1160
+ },
+ {
+ "epoch": 0.6091290661070304,
+ "grad_norm": 0.1702599924159677,
+ "learning_rate": 1.9936348401537292e-05,
+ "loss": 0.47218555212020874,
+ "step": 1161
+ },
+ {
+ "epoch": 0.6096537250786989,
+ "grad_norm": 0.15690176248846127,
+ "learning_rate": 1.9901430986031695e-05,
+ "loss": 0.4516097903251648,
+ "step": 1162
+ },
+ {
+ "epoch": 0.6101783840503673,
+ "grad_norm": 0.1713614405614144,
+ "learning_rate": 1.9866534214137085e-05,
+ "loss": 0.45030856132507324,
+ "step": 1163
+ },
+ {
+ "epoch": 0.6107030430220357,
+ "grad_norm": 0.1670865563026229,
+ "learning_rate": 1.9831658180660195e-05,
+ "loss": 0.45603489875793457,
+ "step": 1164
+ },
+ {
+ "epoch": 0.6112277019937041,
+ "grad_norm": 0.16897539113401952,
+ "learning_rate": 1.9796802980351392e-05,
+ "loss": 0.46314537525177,
+ "step": 1165
+ },
+ {
+ "epoch": 0.6117523609653726,
+ "grad_norm": 0.15832868870697644,
+ "learning_rate": 1.976196870790446e-05,
+ "loss": 0.44784611463546753,
+ "step": 1166
+ },
+ {
+ "epoch": 0.612277019937041,
+ "grad_norm": 0.16595365551891114,
+ "learning_rate": 1.9727155457956315e-05,
+ "loss": 0.4409624934196472,
+ "step": 1167
+ },
+ {
+ "epoch": 0.6128016789087093,
+ "grad_norm": 0.1642630411212017,
+ "learning_rate": 1.969236332508677e-05,
+ "loss": 0.46120351552963257,
+ "step": 1168
+ },
+ {
+ "epoch": 0.6133263378803777,
+ "grad_norm": 0.16481796135868812,
+ "learning_rate": 1.9657592403818253e-05,
+ "loss": 0.45544904470443726,
+ "step": 1169
+ },
+ {
+ "epoch": 0.6138509968520461,
+ "grad_norm": 0.1629832779852364,
+ "learning_rate": 1.962284278861557e-05,
+ "loss": 0.4578462839126587,
+ "step": 1170
+ },
+ {
+ "epoch": 0.6143756558237146,
+ "grad_norm": 0.15916219613603688,
+ "learning_rate": 1.9588114573885664e-05,
+ "loss": 0.4546983540058136,
+ "step": 1171
+ },
+ {
+ "epoch": 0.614900314795383,
+ "grad_norm": 0.1660520165166075,
+ "learning_rate": 1.95534078539773e-05,
+ "loss": 0.4601942300796509,
+ "step": 1172
+ },
+ {
+ "epoch": 0.6154249737670514,
+ "grad_norm": 0.16924794578577768,
+ "learning_rate": 1.9518722723180877e-05,
+ "loss": 0.45997846126556396,
+ "step": 1173
+ },
+ {
+ "epoch": 0.6159496327387198,
+ "grad_norm": 0.16597646133045382,
+ "learning_rate": 1.9484059275728127e-05,
+ "loss": 0.4594605565071106,
+ "step": 1174
+ },
+ {
+ "epoch": 0.6164742917103883,
+ "grad_norm": 0.16964243792519748,
+ "learning_rate": 1.9449417605791876e-05,
+ "loss": 0.4638690650463104,
+ "step": 1175
+ },
+ {
+ "epoch": 0.6169989506820567,
+ "grad_norm": 0.1758591426791696,
+ "learning_rate": 1.9414797807485796e-05,
+ "loss": 0.4666447639465332,
+ "step": 1176
+ },
+ {
+ "epoch": 0.6175236096537251,
+ "grad_norm": 0.16217661227891753,
+ "learning_rate": 1.9380199974864126e-05,
+ "loss": 0.44252535700798035,
+ "step": 1177
+ },
+ {
+ "epoch": 0.6180482686253935,
+ "grad_norm": 0.1576446184499019,
+ "learning_rate": 1.934562420192142e-05,
+ "loss": 0.4499641954898834,
+ "step": 1178
+ },
+ {
+ "epoch": 0.618572927597062,
+ "grad_norm": 0.16231008092636912,
+ "learning_rate": 1.9311070582592327e-05,
+ "loss": 0.4561542570590973,
+ "step": 1179
+ },
+ {
+ "epoch": 0.6190975865687304,
+ "grad_norm": 0.16790382249577043,
+ "learning_rate": 1.9276539210751292e-05,
+ "loss": 0.4484325647354126,
+ "step": 1180
+ },
+ {
+ "epoch": 0.6196222455403987,
+ "grad_norm": 0.18794729932605855,
+ "learning_rate": 1.9242030180212323e-05,
+ "loss": 0.44169026613235474,
+ "step": 1181
+ },
+ {
+ "epoch": 0.6201469045120671,
+ "grad_norm": 0.16948051243889814,
+ "learning_rate": 1.920754358472872e-05,
+ "loss": 0.4592876434326172,
+ "step": 1182
+ },
+ {
+ "epoch": 0.6206715634837355,
+ "grad_norm": 0.16762558710029374,
+ "learning_rate": 1.917307951799286e-05,
+ "loss": 0.4402095079421997,
+ "step": 1183
+ },
+ {
+ "epoch": 0.621196222455404,
+ "grad_norm": 0.18170280116387003,
+ "learning_rate": 1.9138638073635885e-05,
+ "loss": 0.4633026123046875,
+ "step": 1184
+ },
+ {
+ "epoch": 0.6217208814270724,
+ "grad_norm": 0.1718051709136108,
+ "learning_rate": 1.9104219345227486e-05,
+ "loss": 0.4510066509246826,
+ "step": 1185
+ },
+ {
+ "epoch": 0.6222455403987408,
+ "grad_norm": 0.16829066326345576,
+ "learning_rate": 1.906982342627565e-05,
+ "loss": 0.4632919430732727,
+ "step": 1186
+ },
+ {
+ "epoch": 0.6227701993704092,
+ "grad_norm": 0.16669141808548274,
+ "learning_rate": 1.9035450410226395e-05,
+ "loss": 0.455544650554657,
+ "step": 1187
+ },
+ {
+ "epoch": 0.6232948583420777,
+ "grad_norm": 0.16679996559722232,
+ "learning_rate": 1.90011003904635e-05,
+ "loss": 0.459994375705719,
+ "step": 1188
+ },
+ {
+ "epoch": 0.6238195173137461,
+ "grad_norm": 0.17379754713278164,
+ "learning_rate": 1.8966773460308284e-05,
+ "loss": 0.4576749801635742,
+ "step": 1189
+ },
+ {
+ "epoch": 0.6243441762854145,
+ "grad_norm": 0.17088539733389246,
+ "learning_rate": 1.8932469713019335e-05,
+ "loss": 0.44853052496910095,
+ "step": 1190
+ },
+ {
+ "epoch": 0.6248688352570829,
+ "grad_norm": 0.15556883889201015,
+ "learning_rate": 1.8898189241792248e-05,
+ "loss": 0.44060447812080383,
+ "step": 1191
+ },
+ {
+ "epoch": 0.6253934942287513,
+ "grad_norm": 0.1640093917439215,
+ "learning_rate": 1.8863932139759388e-05,
+ "loss": 0.4531911611557007,
+ "step": 1192
+ },
+ {
+ "epoch": 0.6259181532004198,
+ "grad_norm": 0.16444284826591576,
+ "learning_rate": 1.882969849998964e-05,
+ "loss": 0.44176363945007324,
+ "step": 1193
+ },
+ {
+ "epoch": 0.6264428121720882,
+ "grad_norm": 0.1696462520995642,
+ "learning_rate": 1.8795488415488143e-05,
+ "loss": 0.44516652822494507,
+ "step": 1194
+ },
+ {
+ "epoch": 0.6269674711437565,
+ "grad_norm": 0.16810724705191796,
+ "learning_rate": 1.8761301979196032e-05,
+ "loss": 0.4542563855648041,
+ "step": 1195
+ },
+ {
+ "epoch": 0.6274921301154249,
+ "grad_norm": 0.17334813313855824,
+ "learning_rate": 1.8727139283990204e-05,
+ "loss": 0.47769248485565186,
+ "step": 1196
+ },
+ {
+ "epoch": 0.6280167890870934,
+ "grad_norm": 0.16308212035303235,
+ "learning_rate": 1.8693000422683043e-05,
+ "loss": 0.4407799243927002,
+ "step": 1197
+ },
+ {
+ "epoch": 0.6285414480587618,
+ "grad_norm": 0.16853319139928977,
+ "learning_rate": 1.8658885488022204e-05,
+ "loss": 0.4565497040748596,
+ "step": 1198
+ },
+ {
+ "epoch": 0.6290661070304302,
+ "grad_norm": 0.16694968781284703,
+ "learning_rate": 1.8624794572690328e-05,
+ "loss": 0.44522374868392944,
+ "step": 1199
+ },
+ {
+ "epoch": 0.6295907660020986,
+ "grad_norm": 0.16418579468339103,
+ "learning_rate": 1.8590727769304788e-05,
+ "loss": 0.4441048502922058,
+ "step": 1200
+ },
+ {
+ "epoch": 0.630115424973767,
+ "grad_norm": 0.1661930777124156,
+ "learning_rate": 1.8556685170417486e-05,
+ "loss": 0.457327663898468,
+ "step": 1201
+ },
+ {
+ "epoch": 0.6306400839454355,
+ "grad_norm": 0.16536630027039814,
+ "learning_rate": 1.852266686851452e-05,
+ "loss": 0.462494432926178,
+ "step": 1202
+ },
+ {
+ "epoch": 0.6311647429171039,
+ "grad_norm": 0.16611474088307454,
+ "learning_rate": 1.8488672956016013e-05,
+ "loss": 0.4557701647281647,
+ "step": 1203
+ },
+ {
+ "epoch": 0.6316894018887723,
+ "grad_norm": 0.16759343079749697,
+ "learning_rate": 1.8454703525275813e-05,
+ "loss": 0.45069995522499084,
+ "step": 1204
+ },
+ {
+ "epoch": 0.6322140608604407,
+ "grad_norm": 0.16186676294485336,
+ "learning_rate": 1.8420758668581255e-05,
+ "loss": 0.44954800605773926,
+ "step": 1205
+ },
+ {
+ "epoch": 0.6327387198321092,
+ "grad_norm": 0.15687159835368553,
+ "learning_rate": 1.8386838478152916e-05,
+ "loss": 0.4490429759025574,
+ "step": 1206
+ },
+ {
+ "epoch": 0.6332633788037776,
+ "grad_norm": 0.16242951976115577,
+ "learning_rate": 1.8352943046144372e-05,
+ "loss": 0.45819586515426636,
+ "step": 1207
+ },
+ {
+ "epoch": 0.633788037775446,
+ "grad_norm": 0.16516087266517232,
+ "learning_rate": 1.831907246464191e-05,
+ "loss": 0.44769883155822754,
+ "step": 1208
+ },
+ {
+ "epoch": 0.6343126967471143,
+ "grad_norm": 0.16709475871512644,
+ "learning_rate": 1.8285226825664332e-05,
+ "loss": 0.4699128568172455,
+ "step": 1209
+ },
+ {
+ "epoch": 0.6348373557187827,
+ "grad_norm": 0.16288623150595774,
+ "learning_rate": 1.8251406221162636e-05,
+ "loss": 0.45322558283805847,
+ "step": 1210
+ },
+ {
+ "epoch": 0.6353620146904512,
+ "grad_norm": 0.1675722848054237,
+ "learning_rate": 1.821761074301986e-05,
+ "loss": 0.4752347469329834,
+ "step": 1211
+ },
+ {
+ "epoch": 0.6358866736621196,
+ "grad_norm": 0.163285963514504,
+ "learning_rate": 1.818384048305074e-05,
+ "loss": 0.441958487033844,
+ "step": 1212
+ },
+ {
+ "epoch": 0.636411332633788,
+ "grad_norm": 0.17078102287557764,
+ "learning_rate": 1.8150095533001517e-05,
+ "loss": 0.4551333487033844,
+ "step": 1213
+ },
+ {
+ "epoch": 0.6369359916054564,
+ "grad_norm": 0.16188540746097138,
+ "learning_rate": 1.811637598454966e-05,
+ "loss": 0.4470779597759247,
+ "step": 1214
+ },
+ {
+ "epoch": 0.6374606505771249,
+ "grad_norm": 0.15235628218958286,
+ "learning_rate": 1.808268192930364e-05,
+ "loss": 0.43604153394699097,
+ "step": 1215
+ },
+ {
+ "epoch": 0.6379853095487933,
+ "grad_norm": 0.1604601949959104,
+ "learning_rate": 1.8049013458802665e-05,
+ "loss": 0.4445713758468628,
+ "step": 1216
+ },
+ {
+ "epoch": 0.6385099685204617,
+ "grad_norm": 0.15511797131215901,
+ "learning_rate": 1.8015370664516427e-05,
+ "loss": 0.4455301761627197,
+ "step": 1217
+ },
+ {
+ "epoch": 0.6390346274921301,
+ "grad_norm": 0.16815669546944095,
+ "learning_rate": 1.798175363784487e-05,
+ "loss": 0.46135014295578003,
+ "step": 1218
+ },
+ {
+ "epoch": 0.6395592864637986,
+ "grad_norm": 0.161558177688922,
+ "learning_rate": 1.7948162470117917e-05,
+ "loss": 0.4519304037094116,
+ "step": 1219
+ },
+ {
+ "epoch": 0.640083945435467,
+ "grad_norm": 0.16677842130048232,
+ "learning_rate": 1.7914597252595265e-05,
+ "loss": 0.4550550878047943,
+ "step": 1220
+ },
+ {
+ "epoch": 0.6406086044071354,
+ "grad_norm": 0.1597331480423138,
+ "learning_rate": 1.7881058076466095e-05,
+ "loss": 0.441120445728302,
+ "step": 1221
+ },
+ {
+ "epoch": 0.6411332633788038,
+ "grad_norm": 0.15951106545246824,
+ "learning_rate": 1.7847545032848844e-05,
+ "loss": 0.4471382200717926,
+ "step": 1222
+ },
+ {
+ "epoch": 0.6416579223504721,
+ "grad_norm": 0.16157787674667565,
+ "learning_rate": 1.7814058212790947e-05,
+ "loss": 0.4550648331642151,
+ "step": 1223
+ },
+ {
+ "epoch": 0.6421825813221406,
+ "grad_norm": 0.16989186763352415,
+ "learning_rate": 1.7780597707268583e-05,
+ "loss": 0.4515089988708496,
+ "step": 1224
+ },
+ {
+ "epoch": 0.642707240293809,
+ "grad_norm": 0.15952153223099508,
+ "learning_rate": 1.7747163607186468e-05,
+ "loss": 0.4658638834953308,
+ "step": 1225
+ },
+ {
+ "epoch": 0.6432318992654774,
+ "grad_norm": 0.15205926699647218,
+ "learning_rate": 1.7713756003377567e-05,
+ "loss": 0.4330868422985077,
+ "step": 1226
+ },
+ {
+ "epoch": 0.6437565582371458,
+ "grad_norm": 0.15784806480936794,
+ "learning_rate": 1.7680374986602865e-05,
+ "loss": 0.4506068527698517,
+ "step": 1227
+ },
+ {
+ "epoch": 0.6442812172088143,
+ "grad_norm": 0.16006709280686687,
+ "learning_rate": 1.7647020647551105e-05,
+ "loss": 0.446941077709198,
+ "step": 1228
+ },
+ {
+ "epoch": 0.6448058761804827,
+ "grad_norm": 0.16064631158222062,
+ "learning_rate": 1.761369307683856e-05,
+ "loss": 0.4418525695800781,
+ "step": 1229
+ },
+ {
+ "epoch": 0.6453305351521511,
+ "grad_norm": 0.16454136480444392,
+ "learning_rate": 1.7580392365008767e-05,
+ "loss": 0.46268391609191895,
+ "step": 1230
+ },
+ {
+ "epoch": 0.6458551941238195,
+ "grad_norm": 0.15401246535771462,
+ "learning_rate": 1.754711860253233e-05,
+ "loss": 0.4372989535331726,
+ "step": 1231
+ },
+ {
+ "epoch": 0.646379853095488,
+ "grad_norm": 0.1544567975028383,
+ "learning_rate": 1.7513871879806603e-05,
+ "loss": 0.44806575775146484,
+ "step": 1232
+ },
+ {
+ "epoch": 0.6469045120671564,
+ "grad_norm": 0.16411225112730965,
+ "learning_rate": 1.748065228715549e-05,
+ "loss": 0.43782058358192444,
+ "step": 1233
+ },
+ {
+ "epoch": 0.6474291710388248,
+ "grad_norm": 0.1552767832972093,
+ "learning_rate": 1.7447459914829183e-05,
+ "loss": 0.44531378149986267,
+ "step": 1234
+ },
+ {
+ "epoch": 0.6479538300104932,
+ "grad_norm": 0.16214695257123513,
+ "learning_rate": 1.741429485300392e-05,
+ "loss": 0.47399842739105225,
+ "step": 1235
+ },
+ {
+ "epoch": 0.6484784889821616,
+ "grad_norm": 0.16684861111484112,
+ "learning_rate": 1.7381157191781766e-05,
+ "loss": 0.4654282331466675,
+ "step": 1236
+ },
+ {
+ "epoch": 0.64900314795383,
+ "grad_norm": 0.15901864178630273,
+ "learning_rate": 1.7348047021190334e-05,
+ "loss": 0.45128414034843445,
+ "step": 1237
+ },
+ {
+ "epoch": 0.6495278069254984,
+ "grad_norm": 0.15360991961742707,
+ "learning_rate": 1.7314964431182523e-05,
+ "loss": 0.4303375482559204,
+ "step": 1238
+ },
+ {
+ "epoch": 0.6500524658971668,
+ "grad_norm": 0.1633151574385741,
+ "learning_rate": 1.728190951163633e-05,
+ "loss": 0.45568403601646423,
+ "step": 1239
+ },
+ {
+ "epoch": 0.6505771248688352,
+ "grad_norm": 0.16377853571058645,
+ "learning_rate": 1.724888235235457e-05,
+ "loss": 0.46503543853759766,
+ "step": 1240
+ },
+ {
+ "epoch": 0.6511017838405037,
+ "grad_norm": 0.1595023801275368,
+ "learning_rate": 1.7215883043064655e-05,
+ "loss": 0.46337294578552246,
+ "step": 1241
+ },
+ {
+ "epoch": 0.6516264428121721,
+ "grad_norm": 0.16324796182212292,
+ "learning_rate": 1.7182911673418313e-05,
+ "loss": 0.46197354793548584,
+ "step": 1242
+ },
+ {
+ "epoch": 0.6521511017838405,
+ "grad_norm": 0.16353734830879765,
+ "learning_rate": 1.714996833299138e-05,
+ "loss": 0.45741671323776245,
+ "step": 1243
+ },
+ {
+ "epoch": 0.6526757607555089,
+ "grad_norm": 0.15941587678263386,
+ "learning_rate": 1.7117053111283528e-05,
+ "loss": 0.4514503479003906,
+ "step": 1244
+ },
+ {
+ "epoch": 0.6532004197271774,
+ "grad_norm": 0.1642653230145008,
+ "learning_rate": 1.708416609771804e-05,
+ "loss": 0.4437655210494995,
+ "step": 1245
+ },
+ {
+ "epoch": 0.6537250786988458,
+ "grad_norm": 0.15814155519130152,
+ "learning_rate": 1.7051307381641587e-05,
+ "loss": 0.4314163327217102,
+ "step": 1246
+ },
+ {
+ "epoch": 0.6542497376705142,
+ "grad_norm": 0.16086078370768056,
+ "learning_rate": 1.701847705232394e-05,
+ "loss": 0.4526430666446686,
+ "step": 1247
+ },
+ {
+ "epoch": 0.6547743966421826,
+ "grad_norm": 0.16389360190715183,
+ "learning_rate": 1.698567519895775e-05,
+ "loss": 0.4278966188430786,
+ "step": 1248
+ },
+ {
+ "epoch": 0.655299055613851,
+ "grad_norm": 0.16303142287804262,
+ "learning_rate": 1.6952901910658316e-05,
+ "loss": 0.45778220891952515,
+ "step": 1249
+ },
+ {
+ "epoch": 0.6558237145855194,
+ "grad_norm": 0.1675209956904048,
+ "learning_rate": 1.6920157276463318e-05,
+ "loss": 0.4489403963088989,
+ "step": 1250
+ },
+ {
+ "epoch": 0.6563483735571878,
+ "grad_norm": 0.16228311551074323,
+ "learning_rate": 1.6887441385332604e-05,
+ "loss": 0.4534342885017395,
+ "step": 1251
+ },
+ {
+ "epoch": 0.6568730325288562,
+ "grad_norm": 0.16408957851046913,
+ "learning_rate": 1.6854754326147916e-05,
+ "loss": 0.4549208879470825,
+ "step": 1252
+ },
+ {
+ "epoch": 0.6573976915005246,
+ "grad_norm": 0.15967922111432845,
+ "learning_rate": 1.682209618771268e-05,
+ "loss": 0.4451141357421875,
+ "step": 1253
+ },
+ {
+ "epoch": 0.657922350472193,
+ "grad_norm": 0.16440698722204208,
+ "learning_rate": 1.678946705875176e-05,
+ "loss": 0.454200804233551,
+ "step": 1254
+ },
+ {
+ "epoch": 0.6584470094438615,
+ "grad_norm": 0.16251120184938497,
+ "learning_rate": 1.675686702791117e-05,
+ "loss": 0.4580027461051941,
+ "step": 1255
+ },
+ {
+ "epoch": 0.6589716684155299,
+ "grad_norm": 0.1643717386150457,
+ "learning_rate": 1.6724296183757924e-05,
+ "loss": 0.4422311782836914,
+ "step": 1256
+ },
+ {
+ "epoch": 0.6594963273871983,
+ "grad_norm": 0.1644586414378722,
+ "learning_rate": 1.66917546147797e-05,
+ "loss": 0.4593179225921631,
+ "step": 1257
+ },
+ {
+ "epoch": 0.6600209863588667,
+ "grad_norm": 0.16751074186967685,
+ "learning_rate": 1.6659242409384657e-05,
+ "loss": 0.457930326461792,
+ "step": 1258
+ },
+ {
+ "epoch": 0.6605456453305352,
+ "grad_norm": 0.15398418609155334,
+ "learning_rate": 1.6626759655901182e-05,
+ "loss": 0.44002899527549744,
+ "step": 1259
+ },
+ {
+ "epoch": 0.6610703043022036,
+ "grad_norm": 0.16328113175790226,
+ "learning_rate": 1.659430644257763e-05,
+ "loss": 0.4500438868999481,
+ "step": 1260
+ },
+ {
+ "epoch": 0.661594963273872,
+ "grad_norm": 0.1694244340615106,
+ "learning_rate": 1.656188285758214e-05,
+ "loss": 0.4818427264690399,
+ "step": 1261
+ },
+ {
+ "epoch": 0.6621196222455404,
+ "grad_norm": 0.1578320701754368,
+ "learning_rate": 1.652948898900232e-05,
+ "loss": 0.4492436647415161,
+ "step": 1262
+ },
+ {
+ "epoch": 0.6626442812172089,
+ "grad_norm": 0.16538865917828469,
+ "learning_rate": 1.6497124924845064e-05,
+ "loss": 0.45070338249206543,
+ "step": 1263
+ },
+ {
+ "epoch": 0.6631689401888772,
+ "grad_norm": 0.15670817848256852,
+ "learning_rate": 1.6464790753036296e-05,
+ "loss": 0.45439285039901733,
+ "step": 1264
+ },
+ {
+ "epoch": 0.6636935991605456,
+ "grad_norm": 0.16184786585965885,
+ "learning_rate": 1.64324865614207e-05,
+ "loss": 0.446522980928421,
+ "step": 1265
+ },
+ {
+ "epoch": 0.664218258132214,
+ "grad_norm": 0.15719726462251984,
+ "learning_rate": 1.6400212437761554e-05,
+ "loss": 0.4495280981063843,
+ "step": 1266
+ },
+ {
+ "epoch": 0.6647429171038824,
+ "grad_norm": 0.1680804414134405,
+ "learning_rate": 1.6367968469740416e-05,
+ "loss": 0.4652596712112427,
+ "step": 1267
+ },
+ {
+ "epoch": 0.6652675760755509,
+ "grad_norm": 0.15919541846683566,
+ "learning_rate": 1.6335754744956937e-05,
+ "loss": 0.43842798471450806,
+ "step": 1268
+ },
+ {
+ "epoch": 0.6657922350472193,
+ "grad_norm": 0.16322634187139418,
+ "learning_rate": 1.630357135092859e-05,
+ "loss": 0.45242664217948914,
+ "step": 1269
+ },
+ {
+ "epoch": 0.6663168940188877,
+ "grad_norm": 0.15644867734734783,
+ "learning_rate": 1.6271418375090445e-05,
+ "loss": 0.44801655411720276,
+ "step": 1270
+ },
+ {
+ "epoch": 0.6668415529905561,
+ "grad_norm": 0.1514717416124423,
+ "learning_rate": 1.623929590479496e-05,
+ "loss": 0.4506194293498993,
+ "step": 1271
+ },
+ {
+ "epoch": 0.6673662119622246,
+ "grad_norm": 0.1740345085495889,
+ "learning_rate": 1.6207204027311683e-05,
+ "loss": 0.4818177819252014,
+ "step": 1272
+ },
+ {
+ "epoch": 0.667890870933893,
+ "grad_norm": 0.15797701900765457,
+ "learning_rate": 1.6175142829827067e-05,
+ "loss": 0.4639382064342499,
+ "step": 1273
+ },
+ {
+ "epoch": 0.6684155299055614,
+ "grad_norm": 0.15663812316653875,
+ "learning_rate": 1.6143112399444198e-05,
+ "loss": 0.46217501163482666,
+ "step": 1274
+ },
+ {
+ "epoch": 0.6689401888772298,
+ "grad_norm": 0.15474018977043758,
+ "learning_rate": 1.6111112823182592e-05,
+ "loss": 0.43691688776016235,
+ "step": 1275
+ },
+ {
+ "epoch": 0.6694648478488983,
+ "grad_norm": 0.1615634942529431,
+ "learning_rate": 1.6079144187977923e-05,
+ "loss": 0.4412601590156555,
+ "step": 1276
+ },
+ {
+ "epoch": 0.6699895068205667,
+ "grad_norm": 0.15440022842754444,
+ "learning_rate": 1.604720658068184e-05,
+ "loss": 0.4523273706436157,
+ "step": 1277
+ },
+ {
+ "epoch": 0.670514165792235,
+ "grad_norm": 0.1634985665776242,
+ "learning_rate": 1.6015300088061664e-05,
+ "loss": 0.46264398097991943,
+ "step": 1278
+ },
+ {
+ "epoch": 0.6710388247639034,
+ "grad_norm": 0.15037886370689527,
+ "learning_rate": 1.598342479680018e-05,
+ "loss": 0.451465904712677,
+ "step": 1279
+ },
+ {
+ "epoch": 0.6715634837355718,
+ "grad_norm": 0.16895242437746613,
+ "learning_rate": 1.595158079349542e-05,
+ "loss": 0.4466145634651184,
+ "step": 1280
+ },
+ {
+ "epoch": 0.6720881427072403,
+ "grad_norm": 0.1629548326999642,
+ "learning_rate": 1.5919768164660425e-05,
+ "loss": 0.45415782928466797,
+ "step": 1281
+ },
+ {
+ "epoch": 0.6726128016789087,
+ "grad_norm": 0.1563872102268991,
+ "learning_rate": 1.588798699672298e-05,
+ "loss": 0.4386737048625946,
+ "step": 1282
+ },
+ {
+ "epoch": 0.6731374606505771,
+ "grad_norm": 0.16038172773344717,
+ "learning_rate": 1.5856237376025403e-05,
+ "loss": 0.446260929107666,
+ "step": 1283
+ },
+ {
+ "epoch": 0.6736621196222455,
+ "grad_norm": 0.16189780714407623,
+ "learning_rate": 1.582451938882431e-05,
+ "loss": 0.45458072423934937,
+ "step": 1284
+ },
+ {
+ "epoch": 0.674186778593914,
+ "grad_norm": 0.15204521785049285,
+ "learning_rate": 1.5792833121290367e-05,
+ "loss": 0.4425617754459381,
+ "step": 1285
+ },
+ {
+ "epoch": 0.6747114375655824,
+ "grad_norm": 0.16237376704792778,
+ "learning_rate": 1.5761178659508063e-05,
+ "loss": 0.44631800055503845,
+ "step": 1286
+ },
+ {
+ "epoch": 0.6752360965372508,
+ "grad_norm": 0.1582063618466069,
+ "learning_rate": 1.57295560894755e-05,
+ "loss": 0.44280534982681274,
+ "step": 1287
+ },
+ {
+ "epoch": 0.6757607555089192,
+ "grad_norm": 0.16587032349847394,
+ "learning_rate": 1.5697965497104116e-05,
+ "loss": 0.4495997130870819,
+ "step": 1288
+ },
+ {
+ "epoch": 0.6762854144805877,
+ "grad_norm": 0.1557269262755563,
+ "learning_rate": 1.566640696821848e-05,
+ "loss": 0.4493359625339508,
+ "step": 1289
+ },
+ {
+ "epoch": 0.6768100734522561,
+ "grad_norm": 0.16038490357012256,
+ "learning_rate": 1.5634880588556043e-05,
+ "loss": 0.4447544813156128,
+ "step": 1290
+ },
+ {
+ "epoch": 0.6773347324239245,
+ "grad_norm": 0.15355201908546406,
+ "learning_rate": 1.560338644376692e-05,
+ "loss": 0.4468432664871216,
+ "step": 1291
+ },
+ {
+ "epoch": 0.6778593913955928,
+ "grad_norm": 0.15077009515589776,
+ "learning_rate": 1.5571924619413668e-05,
+ "loss": 0.44959044456481934,
+ "step": 1292
+ },
+ {
+ "epoch": 0.6783840503672612,
+ "grad_norm": 0.1583038146613965,
+ "learning_rate": 1.5540495200971012e-05,
+ "loss": 0.42450425028800964,
+ "step": 1293
+ },
+ {
+ "epoch": 0.6789087093389297,
+ "grad_norm": 0.15950687190598087,
+ "learning_rate": 1.5509098273825635e-05,
+ "loss": 0.4527238607406616,
+ "step": 1294
+ },
+ {
+ "epoch": 0.6794333683105981,
+ "grad_norm": 0.1526262208847808,
+ "learning_rate": 1.5477733923275968e-05,
+ "loss": 0.4379724860191345,
+ "step": 1295
+ },
+ {
+ "epoch": 0.6799580272822665,
+ "grad_norm": 0.15541579650021778,
+ "learning_rate": 1.5446402234531934e-05,
+ "loss": 0.4462166130542755,
+ "step": 1296
+ },
+ {
+ "epoch": 0.6804826862539349,
+ "grad_norm": 0.15483930734313786,
+ "learning_rate": 1.5415103292714718e-05,
+ "loss": 0.4390607178211212,
+ "step": 1297
+ },
+ {
+ "epoch": 0.6810073452256034,
+ "grad_norm": 0.1578005731149154,
+ "learning_rate": 1.5383837182856532e-05,
+ "loss": 0.44462254643440247,
+ "step": 1298
+ },
+ {
+ "epoch": 0.6815320041972718,
+ "grad_norm": 0.15830183884958823,
+ "learning_rate": 1.5352603989900406e-05,
+ "loss": 0.4563537538051605,
+ "step": 1299
+ },
+ {
+ "epoch": 0.6820566631689402,
+ "grad_norm": 0.15605472163445522,
+ "learning_rate": 1.5321403798699923e-05,
+ "loss": 0.43852412700653076,
+ "step": 1300
+ },
+ {
+ "epoch": 0.6825813221406086,
+ "grad_norm": 0.15770470227141278,
+ "learning_rate": 1.5290236694019017e-05,
+ "loss": 0.4438081979751587,
+ "step": 1301
+ },
+ {
+ "epoch": 0.683105981112277,
+ "grad_norm": 0.16175339845012077,
+ "learning_rate": 1.5259102760531747e-05,
+ "loss": 0.46670734882354736,
+ "step": 1302
+ },
+ {
+ "epoch": 0.6836306400839455,
+ "grad_norm": 0.16359504716154477,
+ "learning_rate": 1.5228002082822036e-05,
+ "loss": 0.4590187668800354,
+ "step": 1303
+ },
+ {
+ "epoch": 0.6841552990556139,
+ "grad_norm": 0.15422879500254713,
+ "learning_rate": 1.5196934745383457e-05,
+ "loss": 0.45287445187568665,
+ "step": 1304
+ },
+ {
+ "epoch": 0.6846799580272823,
+ "grad_norm": 0.15929902143574254,
+ "learning_rate": 1.5165900832619015e-05,
+ "loss": 0.44286859035491943,
+ "step": 1305
+ },
+ {
+ "epoch": 0.6852046169989506,
+ "grad_norm": 0.14838868714016124,
+ "learning_rate": 1.5134900428840904e-05,
+ "loss": 0.45155346393585205,
+ "step": 1306
+ },
+ {
+ "epoch": 0.6857292759706191,
+ "grad_norm": 0.15494288640397672,
+ "learning_rate": 1.5103933618270283e-05,
+ "loss": 0.44010141491889954,
+ "step": 1307
+ },
+ {
+ "epoch": 0.6862539349422875,
+ "grad_norm": 0.1686262360598338,
+ "learning_rate": 1.5073000485037042e-05,
+ "loss": 0.446555495262146,
+ "step": 1308
+ },
+ {
+ "epoch": 0.6867785939139559,
+ "grad_norm": 0.16122110823982602,
+ "learning_rate": 1.5042101113179579e-05,
+ "loss": 0.44640374183654785,
+ "step": 1309
+ },
+ {
+ "epoch": 0.6873032528856243,
+ "grad_norm": 0.16849937797919867,
+ "learning_rate": 1.5011235586644572e-05,
+ "loss": 0.4465785026550293,
+ "step": 1310
+ },
+ {
+ "epoch": 0.6878279118572928,
+ "grad_norm": 0.16062971262800754,
+ "learning_rate": 1.498040398928674e-05,
+ "loss": 0.445321649312973,
+ "step": 1311
+ },
+ {
+ "epoch": 0.6883525708289612,
+ "grad_norm": 0.1608953185993034,
+ "learning_rate": 1.4949606404868638e-05,
+ "loss": 0.46518227458000183,
+ "step": 1312
+ },
+ {
+ "epoch": 0.6888772298006296,
+ "grad_norm": 0.15880255672830915,
+ "learning_rate": 1.491884291706041e-05,
+ "loss": 0.44582217931747437,
+ "step": 1313
+ },
+ {
+ "epoch": 0.689401888772298,
+ "grad_norm": 0.16715145200789822,
+ "learning_rate": 1.4888113609439563e-05,
+ "loss": 0.4435983896255493,
+ "step": 1314
+ },
+ {
+ "epoch": 0.6899265477439664,
+ "grad_norm": 0.15263740340694454,
+ "learning_rate": 1.485741856549075e-05,
+ "loss": 0.4405066967010498,
+ "step": 1315
+ },
+ {
+ "epoch": 0.6904512067156349,
+ "grad_norm": 0.16247034518813966,
+ "learning_rate": 1.4826757868605517e-05,
+ "loss": 0.4544215500354767,
+ "step": 1316
+ },
+ {
+ "epoch": 0.6909758656873033,
+ "grad_norm": 0.15649146705978162,
+ "learning_rate": 1.4796131602082142e-05,
+ "loss": 0.4454413056373596,
+ "step": 1317
+ },
+ {
+ "epoch": 0.6915005246589717,
+ "grad_norm": 0.16286718636058178,
+ "learning_rate": 1.4765539849125318e-05,
+ "loss": 0.4497183561325073,
+ "step": 1318
+ },
+ {
+ "epoch": 0.6920251836306401,
+ "grad_norm": 0.15812342085005582,
+ "learning_rate": 1.473498269284599e-05,
+ "loss": 0.445032000541687,
+ "step": 1319
+ },
+ {
+ "epoch": 0.6925498426023085,
+ "grad_norm": 0.1558101290511678,
+ "learning_rate": 1.4704460216261123e-05,
+ "loss": 0.45769619941711426,
+ "step": 1320
+ },
+ {
+ "epoch": 0.6930745015739769,
+ "grad_norm": 0.15642889059495207,
+ "learning_rate": 1.4673972502293428e-05,
+ "loss": 0.44788360595703125,
+ "step": 1321
+ },
+ {
+ "epoch": 0.6935991605456453,
+ "grad_norm": 0.15722513518113074,
+ "learning_rate": 1.4643519633771222e-05,
+ "loss": 0.449631929397583,
+ "step": 1322
+ },
+ {
+ "epoch": 0.6941238195173137,
+ "grad_norm": 0.15587633748081542,
+ "learning_rate": 1.4613101693428125e-05,
+ "loss": 0.4450187087059021,
+ "step": 1323
+ },
+ {
+ "epoch": 0.6946484784889821,
+ "grad_norm": 0.1577107074882479,
+ "learning_rate": 1.458271876390287e-05,
+ "loss": 0.45108890533447266,
+ "step": 1324
+ },
+ {
+ "epoch": 0.6951731374606506,
+ "grad_norm": 0.15649839486990696,
+ "learning_rate": 1.4552370927739073e-05,
+ "loss": 0.46430301666259766,
+ "step": 1325
+ },
+ {
+ "epoch": 0.695697796432319,
+ "grad_norm": 0.18573600420690636,
+ "learning_rate": 1.4522058267385008e-05,
+ "loss": 0.4701972007751465,
+ "step": 1326
+ },
+ {
+ "epoch": 0.6962224554039874,
+ "grad_norm": 0.16296497084037542,
+ "learning_rate": 1.4491780865193402e-05,
+ "loss": 0.4607710838317871,
+ "step": 1327
+ },
+ {
+ "epoch": 0.6967471143756558,
+ "grad_norm": 0.161124007217343,
+ "learning_rate": 1.446153880342117e-05,
+ "loss": 0.45095115900039673,
+ "step": 1328
+ },
+ {
+ "epoch": 0.6972717733473243,
+ "grad_norm": 0.1537884039166766,
+ "learning_rate": 1.4431332164229222e-05,
+ "loss": 0.424068808555603,
+ "step": 1329
+ },
+ {
+ "epoch": 0.6977964323189927,
+ "grad_norm": 0.16588834503456087,
+ "learning_rate": 1.4401161029682237e-05,
+ "loss": 0.4622839093208313,
+ "step": 1330
+ },
+ {
+ "epoch": 0.6983210912906611,
+ "grad_norm": 0.16100510604015544,
+ "learning_rate": 1.4371025481748435e-05,
+ "loss": 0.4660705626010895,
+ "step": 1331
+ },
+ {
+ "epoch": 0.6988457502623295,
+ "grad_norm": 0.15559888940590075,
+ "learning_rate": 1.4340925602299337e-05,
+ "loss": 0.4484976530075073,
+ "step": 1332
+ },
+ {
+ "epoch": 0.6993704092339978,
+ "grad_norm": 0.15503821966213488,
+ "learning_rate": 1.4310861473109616e-05,
+ "loss": 0.45636600255966187,
+ "step": 1333
+ },
+ {
+ "epoch": 0.6998950682056663,
+ "grad_norm": 0.15539074896109334,
+ "learning_rate": 1.4280833175856743e-05,
+ "loss": 0.45863068103790283,
+ "step": 1334
+ },
+ {
+ "epoch": 0.7004197271773347,
+ "grad_norm": 0.15664821009154872,
+ "learning_rate": 1.4250840792120896e-05,
+ "loss": 0.44043463468551636,
+ "step": 1335
+ },
+ {
+ "epoch": 0.7009443861490031,
+ "grad_norm": 0.15825478808520285,
+ "learning_rate": 1.4220884403384655e-05,
+ "loss": 0.46522051095962524,
+ "step": 1336
+ },
+ {
+ "epoch": 0.7014690451206715,
+ "grad_norm": 0.16004268568141544,
+ "learning_rate": 1.4190964091032843e-05,
+ "loss": 0.46114686131477356,
+ "step": 1337
+ },
+ {
+ "epoch": 0.70199370409234,
+ "grad_norm": 0.16413139949945635,
+ "learning_rate": 1.4161079936352234e-05,
+ "loss": 0.4457515478134155,
+ "step": 1338
+ },
+ {
+ "epoch": 0.7025183630640084,
+ "grad_norm": 0.1538890577972702,
+ "learning_rate": 1.4131232020531398e-05,
+ "loss": 0.4616093337535858,
+ "step": 1339
+ },
+ {
+ "epoch": 0.7030430220356768,
+ "grad_norm": 0.16240737854323284,
+ "learning_rate": 1.4101420424660434e-05,
+ "loss": 0.44631537795066833,
+ "step": 1340
+ },
+ {
+ "epoch": 0.7035676810073452,
+ "grad_norm": 0.15984708603406073,
+ "learning_rate": 1.4071645229730762e-05,
+ "loss": 0.4366717040538788,
+ "step": 1341
+ },
+ {
+ "epoch": 0.7040923399790137,
+ "grad_norm": 0.15149695695635904,
+ "learning_rate": 1.4041906516634946e-05,
+ "loss": 0.44129428267478943,
+ "step": 1342
+ },
+ {
+ "epoch": 0.7046169989506821,
+ "grad_norm": 0.15041167501768415,
+ "learning_rate": 1.4012204366166399e-05,
+ "loss": 0.43654775619506836,
+ "step": 1343
+ },
+ {
+ "epoch": 0.7051416579223505,
+ "grad_norm": 0.1596097627963756,
+ "learning_rate": 1.3982538859019217e-05,
+ "loss": 0.4652445614337921,
+ "step": 1344
+ },
+ {
+ "epoch": 0.7056663168940189,
+ "grad_norm": 0.15621704205027884,
+ "learning_rate": 1.3952910075787941e-05,
+ "loss": 0.44577041268348694,
+ "step": 1345
+ },
+ {
+ "epoch": 0.7061909758656874,
+ "grad_norm": 0.16073041098008783,
+ "learning_rate": 1.3923318096967341e-05,
+ "loss": 0.4391976594924927,
+ "step": 1346
+ },
+ {
+ "epoch": 0.7067156348373557,
+ "grad_norm": 0.14793853946372448,
+ "learning_rate": 1.3893763002952204e-05,
+ "loss": 0.4463799297809601,
+ "step": 1347
+ },
+ {
+ "epoch": 0.7072402938090241,
+ "grad_norm": 0.16176080358867354,
+ "learning_rate": 1.3864244874037097e-05,
+ "loss": 0.45183271169662476,
+ "step": 1348
+ },
+ {
+ "epoch": 0.7077649527806925,
+ "grad_norm": 0.16372771179026924,
+ "learning_rate": 1.3834763790416171e-05,
+ "loss": 0.44205576181411743,
+ "step": 1349
+ },
+ {
+ "epoch": 0.7082896117523609,
+ "grad_norm": 0.15135222516127125,
+ "learning_rate": 1.3805319832182937e-05,
+ "loss": 0.4390261173248291,
+ "step": 1350
+ },
+ {
+ "epoch": 0.7088142707240294,
+ "grad_norm": 0.15493183225792365,
+ "learning_rate": 1.377591307933002e-05,
+ "loss": 0.45837974548339844,
+ "step": 1351
+ },
+ {
+ "epoch": 0.7093389296956978,
+ "grad_norm": 0.16405906906516504,
+ "learning_rate": 1.3746543611749008e-05,
+ "loss": 0.45032668113708496,
+ "step": 1352
+ },
+ {
+ "epoch": 0.7098635886673662,
+ "grad_norm": 0.15725677976524505,
+ "learning_rate": 1.3717211509230155e-05,
+ "loss": 0.4486224949359894,
+ "step": 1353
+ },
+ {
+ "epoch": 0.7103882476390346,
+ "grad_norm": 0.15781469066983164,
+ "learning_rate": 1.3687916851462227e-05,
+ "loss": 0.4502638578414917,
+ "step": 1354
+ },
+ {
+ "epoch": 0.710912906610703,
+ "grad_norm": 0.16115007729054298,
+ "learning_rate": 1.3658659718032241e-05,
+ "loss": 0.4508914351463318,
+ "step": 1355
+ },
+ {
+ "epoch": 0.7114375655823715,
+ "grad_norm": 0.16755060475968744,
+ "learning_rate": 1.3629440188425286e-05,
+ "loss": 0.45653945207595825,
+ "step": 1356
+ },
+ {
+ "epoch": 0.7119622245540399,
+ "grad_norm": 0.1574994294756123,
+ "learning_rate": 1.3600258342024268e-05,
+ "loss": 0.4349849820137024,
+ "step": 1357
+ },
+ {
+ "epoch": 0.7124868835257083,
+ "grad_norm": 0.16104440922657715,
+ "learning_rate": 1.3571114258109758e-05,
+ "loss": 0.4501233696937561,
+ "step": 1358
+ },
+ {
+ "epoch": 0.7130115424973767,
+ "grad_norm": 0.15163184290939022,
+ "learning_rate": 1.3542008015859691e-05,
+ "loss": 0.44171035289764404,
+ "step": 1359
+ },
+ {
+ "epoch": 0.7135362014690452,
+ "grad_norm": 0.15955896674456582,
+ "learning_rate": 1.3512939694349211e-05,
+ "loss": 0.4558671712875366,
+ "step": 1360
+ },
+ {
+ "epoch": 0.7140608604407135,
+ "grad_norm": 0.1589376751536802,
+ "learning_rate": 1.3483909372550446e-05,
+ "loss": 0.4430835247039795,
+ "step": 1361
+ },
+ {
+ "epoch": 0.7145855194123819,
+ "grad_norm": 0.15979088397822155,
+ "learning_rate": 1.3454917129332279e-05,
+ "loss": 0.4426245391368866,
+ "step": 1362
+ },
+ {
+ "epoch": 0.7151101783840503,
+ "grad_norm": 0.15493179714072167,
+ "learning_rate": 1.3425963043460143e-05,
+ "loss": 0.4550474286079407,
+ "step": 1363
+ },
+ {
+ "epoch": 0.7156348373557188,
+ "grad_norm": 0.15396498968995032,
+ "learning_rate": 1.3397047193595808e-05,
+ "loss": 0.44220536947250366,
+ "step": 1364
+ },
+ {
+ "epoch": 0.7161594963273872,
+ "grad_norm": 0.1489346822667273,
+ "learning_rate": 1.336816965829717e-05,
+ "loss": 0.4437323212623596,
+ "step": 1365
+ },
+ {
+ "epoch": 0.7166841552990556,
+ "grad_norm": 0.16311964931734477,
+ "learning_rate": 1.333933051601802e-05,
+ "loss": 0.45308777689933777,
+ "step": 1366
+ },
+ {
+ "epoch": 0.717208814270724,
+ "grad_norm": 0.1534354619021179,
+ "learning_rate": 1.331052984510785e-05,
+ "loss": 0.4334886372089386,
+ "step": 1367
+ },
+ {
+ "epoch": 0.7177334732423925,
+ "grad_norm": 0.15541089758065404,
+ "learning_rate": 1.328176772381165e-05,
+ "loss": 0.45248323678970337,
+ "step": 1368
+ },
+ {
+ "epoch": 0.7182581322140609,
+ "grad_norm": 0.1617921305700596,
+ "learning_rate": 1.3253044230269651e-05,
+ "loss": 0.44092947244644165,
+ "step": 1369
+ },
+ {
+ "epoch": 0.7187827911857293,
+ "grad_norm": 0.15047353793760151,
+ "learning_rate": 1.3224359442517165e-05,
+ "loss": 0.4424414336681366,
+ "step": 1370
+ },
+ {
+ "epoch": 0.7193074501573977,
+ "grad_norm": 0.15958089102258743,
+ "learning_rate": 1.319571343848433e-05,
+ "loss": 0.4458580017089844,
+ "step": 1371
+ },
+ {
+ "epoch": 0.7198321091290661,
+ "grad_norm": 0.15349258306968733,
+ "learning_rate": 1.3167106295995924e-05,
+ "loss": 0.4285386800765991,
+ "step": 1372
+ },
+ {
+ "epoch": 0.7203567681007346,
+ "grad_norm": 0.16042529563759886,
+ "learning_rate": 1.3138538092771164e-05,
+ "loss": 0.4497690200805664,
+ "step": 1373
+ },
+ {
+ "epoch": 0.720881427072403,
+ "grad_norm": 0.1565113736588754,
+ "learning_rate": 1.3110008906423454e-05,
+ "loss": 0.43562617897987366,
+ "step": 1374
+ },
+ {
+ "epoch": 0.7214060860440713,
+ "grad_norm": 0.15177291086871506,
+ "learning_rate": 1.3081518814460215e-05,
+ "loss": 0.44456908106803894,
+ "step": 1375
+ },
+ {
+ "epoch": 0.7219307450157397,
+ "grad_norm": 0.15928490480439225,
+ "learning_rate": 1.3053067894282622e-05,
+ "loss": 0.46170780062675476,
+ "step": 1376
+ },
+ {
+ "epoch": 0.7224554039874082,
+ "grad_norm": 0.1554814469217892,
+ "learning_rate": 1.302465622318549e-05,
+ "loss": 0.4422169327735901,
+ "step": 1377
+ },
+ {
+ "epoch": 0.7229800629590766,
+ "grad_norm": 0.15375370214966386,
+ "learning_rate": 1.2996283878356947e-05,
+ "loss": 0.4564763903617859,
+ "step": 1378
+ },
+ {
+ "epoch": 0.723504721930745,
+ "grad_norm": 0.15942841969395802,
+ "learning_rate": 1.2967950936878311e-05,
+ "loss": 0.45121875405311584,
+ "step": 1379
+ },
+ {
+ "epoch": 0.7240293809024134,
+ "grad_norm": 0.15731308716774903,
+ "learning_rate": 1.2939657475723844e-05,
+ "loss": 0.4285027086734772,
+ "step": 1380
+ },
+ {
+ "epoch": 0.7245540398740818,
+ "grad_norm": 0.15997753613932147,
+ "learning_rate": 1.291140357176054e-05,
+ "loss": 0.4566386342048645,
+ "step": 1381
+ },
+ {
+ "epoch": 0.7250786988457503,
+ "grad_norm": 0.15333715267473727,
+ "learning_rate": 1.2883189301747927e-05,
+ "loss": 0.45093846321105957,
+ "step": 1382
+ },
+ {
+ "epoch": 0.7256033578174187,
+ "grad_norm": 0.1514510304709061,
+ "learning_rate": 1.2855014742337874e-05,
+ "loss": 0.44256967306137085,
+ "step": 1383
+ },
+ {
+ "epoch": 0.7261280167890871,
+ "grad_norm": 0.15624301298995746,
+ "learning_rate": 1.2826879970074346e-05,
+ "loss": 0.4361875653266907,
+ "step": 1384
+ },
+ {
+ "epoch": 0.7266526757607555,
+ "grad_norm": 0.15767850122006694,
+ "learning_rate": 1.279878506139322e-05,
+ "loss": 0.4477671980857849,
+ "step": 1385
+ },
+ {
+ "epoch": 0.727177334732424,
+ "grad_norm": 0.15537676969694064,
+ "learning_rate": 1.2770730092622072e-05,
+ "loss": 0.4339832663536072,
+ "step": 1386
+ },
+ {
+ "epoch": 0.7277019937040924,
+ "grad_norm": 0.15944113633281293,
+ "learning_rate": 1.2742715139979965e-05,
+ "loss": 0.45407935976982117,
+ "step": 1387
+ },
+ {
+ "epoch": 0.7282266526757608,
+ "grad_norm": 0.16763519518606218,
+ "learning_rate": 1.271474027957727e-05,
+ "loss": 0.46442392468452454,
+ "step": 1388
+ },
+ {
+ "epoch": 0.7287513116474291,
+ "grad_norm": 0.1564572774601902,
+ "learning_rate": 1.2686805587415423e-05,
+ "loss": 0.4463968276977539,
+ "step": 1389
+ },
+ {
+ "epoch": 0.7292759706190975,
+ "grad_norm": 0.16384626314986128,
+ "learning_rate": 1.2658911139386709e-05,
+ "loss": 0.45519885420799255,
+ "step": 1390
+ },
+ {
+ "epoch": 0.729800629590766,
+ "grad_norm": 0.1562023802820964,
+ "learning_rate": 1.2631057011274106e-05,
+ "loss": 0.4365222454071045,
+ "step": 1391
+ },
+ {
+ "epoch": 0.7303252885624344,
+ "grad_norm": 0.15597287232940926,
+ "learning_rate": 1.2603243278751042e-05,
+ "loss": 0.4328266680240631,
+ "step": 1392
+ },
+ {
+ "epoch": 0.7308499475341028,
+ "grad_norm": 0.14762833706132167,
+ "learning_rate": 1.2575470017381214e-05,
+ "loss": 0.43902426958084106,
+ "step": 1393
+ },
+ {
+ "epoch": 0.7313746065057712,
+ "grad_norm": 0.1688371515856315,
+ "learning_rate": 1.2547737302618351e-05,
+ "loss": 0.4649963676929474,
+ "step": 1394
+ },
+ {
+ "epoch": 0.7318992654774397,
+ "grad_norm": 0.15501514509586015,
+ "learning_rate": 1.2520045209806024e-05,
+ "loss": 0.4501558542251587,
+ "step": 1395
+ },
+ {
+ "epoch": 0.7324239244491081,
+ "grad_norm": 0.15516748917693537,
+ "learning_rate": 1.2492393814177457e-05,
+ "loss": 0.4405498206615448,
+ "step": 1396
+ },
+ {
+ "epoch": 0.7329485834207765,
+ "grad_norm": 0.17115400468199243,
+ "learning_rate": 1.2464783190855295e-05,
+ "loss": 0.4445379674434662,
+ "step": 1397
+ },
+ {
+ "epoch": 0.7334732423924449,
+ "grad_norm": 0.15966386341604186,
+ "learning_rate": 1.243721341485143e-05,
+ "loss": 0.4769889712333679,
+ "step": 1398
+ },
+ {
+ "epoch": 0.7339979013641134,
+ "grad_norm": 0.15614706359781472,
+ "learning_rate": 1.2409684561066766e-05,
+ "loss": 0.43643918633461,
+ "step": 1399
+ },
+ {
+ "epoch": 0.7345225603357818,
+ "grad_norm": 0.17467968960064226,
+ "learning_rate": 1.238219670429104e-05,
+ "loss": 0.45413005352020264,
+ "step": 1400
+ },
+ {
+ "epoch": 0.7350472193074502,
+ "grad_norm": 0.1544517430810819,
+ "learning_rate": 1.2354749919202598e-05,
+ "loss": 0.45218425989151,
+ "step": 1401
+ },
+ {
+ "epoch": 0.7355718782791185,
+ "grad_norm": 0.16551083673374484,
+ "learning_rate": 1.2327344280368217e-05,
+ "loss": 0.4459628462791443,
+ "step": 1402
+ },
+ {
+ "epoch": 0.7360965372507869,
+ "grad_norm": 0.17614490459613263,
+ "learning_rate": 1.2299979862242882e-05,
+ "loss": 0.4406280815601349,
+ "step": 1403
+ },
+ {
+ "epoch": 0.7366211962224554,
+ "grad_norm": 0.15554925394435032,
+ "learning_rate": 1.2272656739169588e-05,
+ "loss": 0.4380159378051758,
+ "step": 1404
+ },
+ {
+ "epoch": 0.7371458551941238,
+ "grad_norm": 0.16475466799870916,
+ "learning_rate": 1.2245374985379147e-05,
+ "loss": 0.4442812204360962,
+ "step": 1405
+ },
+ {
+ "epoch": 0.7376705141657922,
+ "grad_norm": 0.16877301168875541,
+ "learning_rate": 1.2218134674989978e-05,
+ "loss": 0.4589429199695587,
+ "step": 1406
+ },
+ {
+ "epoch": 0.7381951731374606,
+ "grad_norm": 0.1583278589457438,
+ "learning_rate": 1.2190935882007894e-05,
+ "loss": 0.44896745681762695,
+ "step": 1407
+ },
+ {
+ "epoch": 0.7387198321091291,
+ "grad_norm": 0.1582022541656192,
+ "learning_rate": 1.2163778680325947e-05,
+ "loss": 0.4367997646331787,
+ "step": 1408
+ },
+ {
+ "epoch": 0.7392444910807975,
+ "grad_norm": 0.1669573982300068,
+ "learning_rate": 1.2136663143724163e-05,
+ "loss": 0.4503386616706848,
+ "step": 1409
+ },
+ {
+ "epoch": 0.7397691500524659,
+ "grad_norm": 0.1637877523924539,
+ "learning_rate": 1.2109589345869387e-05,
+ "loss": 0.4660584330558777,
+ "step": 1410
+ },
+ {
+ "epoch": 0.7402938090241343,
+ "grad_norm": 0.1640136159943701,
+ "learning_rate": 1.2082557360315065e-05,
+ "loss": 0.45206528902053833,
+ "step": 1411
+ },
+ {
+ "epoch": 0.7408184679958028,
+ "grad_norm": 0.1596580168682554,
+ "learning_rate": 1.2055567260501054e-05,
+ "loss": 0.4405035376548767,
+ "step": 1412
+ },
+ {
+ "epoch": 0.7413431269674712,
+ "grad_norm": 0.15224511741945293,
+ "learning_rate": 1.2028619119753402e-05,
+ "loss": 0.42810291051864624,
+ "step": 1413
+ },
+ {
+ "epoch": 0.7418677859391396,
+ "grad_norm": 0.15631381931817098,
+ "learning_rate": 1.2001713011284193e-05,
+ "loss": 0.4421626925468445,
+ "step": 1414
+ },
+ {
+ "epoch": 0.742392444910808,
+ "grad_norm": 0.15721819981265503,
+ "learning_rate": 1.1974849008191288e-05,
+ "loss": 0.4540696144104004,
+ "step": 1415
+ },
+ {
+ "epoch": 0.7429171038824763,
+ "grad_norm": 0.15904478578744463,
+ "learning_rate": 1.1948027183458182e-05,
+ "loss": 0.4601830244064331,
+ "step": 1416
+ },
+ {
+ "epoch": 0.7434417628541448,
+ "grad_norm": 0.1530755605235123,
+ "learning_rate": 1.1921247609953747e-05,
+ "loss": 0.4552589952945709,
+ "step": 1417
+ },
+ {
+ "epoch": 0.7439664218258132,
+ "grad_norm": 0.15375751072258462,
+ "learning_rate": 1.1894510360432106e-05,
+ "loss": 0.42965373396873474,
+ "step": 1418
+ },
+ {
+ "epoch": 0.7444910807974816,
+ "grad_norm": 0.15678686023937333,
+ "learning_rate": 1.1867815507532377e-05,
+ "loss": 0.44684016704559326,
+ "step": 1419
+ },
+ {
+ "epoch": 0.74501573976915,
+ "grad_norm": 0.15486952983783606,
+ "learning_rate": 1.1841163123778502e-05,
+ "loss": 0.44415023922920227,
+ "step": 1420
+ },
+ {
+ "epoch": 0.7455403987408185,
+ "grad_norm": 0.15342611490545593,
+ "learning_rate": 1.1814553281579034e-05,
+ "loss": 0.4501304626464844,
+ "step": 1421
+ },
+ {
+ "epoch": 0.7460650577124869,
+ "grad_norm": 0.15374260012788166,
+ "learning_rate": 1.1787986053226954e-05,
+ "loss": 0.4521749019622803,
+ "step": 1422
+ },
+ {
+ "epoch": 0.7465897166841553,
+ "grad_norm": 0.1543553258370361,
+ "learning_rate": 1.176146151089949e-05,
+ "loss": 0.44753748178482056,
+ "step": 1423
+ },
+ {
+ "epoch": 0.7471143756558237,
+ "grad_norm": 0.15921670193848306,
+ "learning_rate": 1.1734979726657878e-05,
+ "loss": 0.4473114311695099,
+ "step": 1424
+ },
+ {
+ "epoch": 0.7476390346274921,
+ "grad_norm": 0.15023498686186612,
+ "learning_rate": 1.17085407724472e-05,
+ "loss": 0.4568778872489929,
+ "step": 1425
+ },
+ {
+ "epoch": 0.7481636935991606,
+ "grad_norm": 0.14601156920274583,
+ "learning_rate": 1.1682144720096168e-05,
+ "loss": 0.4389464855194092,
+ "step": 1426
+ },
+ {
+ "epoch": 0.748688352570829,
+ "grad_norm": 0.1520548095149398,
+ "learning_rate": 1.1655791641316954e-05,
+ "loss": 0.440733402967453,
+ "step": 1427
+ },
+ {
+ "epoch": 0.7492130115424974,
+ "grad_norm": 0.1591915163097993,
+ "learning_rate": 1.1629481607704963e-05,
+ "loss": 0.46245887875556946,
+ "step": 1428
+ },
+ {
+ "epoch": 0.7497376705141658,
+ "grad_norm": 0.14867295658447263,
+ "learning_rate": 1.1603214690738683e-05,
+ "loss": 0.4332764744758606,
+ "step": 1429
+ },
+ {
+ "epoch": 0.7502623294858342,
+ "grad_norm": 0.1498957762903134,
+ "learning_rate": 1.1576990961779447e-05,
+ "loss": 0.45215466618537903,
+ "step": 1430
+ },
+ {
+ "epoch": 0.7507869884575026,
+ "grad_norm": 0.15430555803491824,
+ "learning_rate": 1.1550810492071243e-05,
+ "loss": 0.4454396963119507,
+ "step": 1431
+ },
+ {
+ "epoch": 0.751311647429171,
+ "grad_norm": 0.1517461021936533,
+ "learning_rate": 1.1524673352740543e-05,
+ "loss": 0.43869835138320923,
+ "step": 1432
+ },
+ {
+ "epoch": 0.7518363064008394,
+ "grad_norm": 0.14541657994692936,
+ "learning_rate": 1.1498579614796121e-05,
+ "loss": 0.4373054504394531,
+ "step": 1433
+ },
+ {
+ "epoch": 0.7523609653725079,
+ "grad_norm": 0.15539513763786977,
+ "learning_rate": 1.1472529349128812e-05,
+ "loss": 0.45451128482818604,
+ "step": 1434
+ },
+ {
+ "epoch": 0.7528856243441763,
+ "grad_norm": 0.1482043113690946,
+ "learning_rate": 1.144652262651136e-05,
+ "loss": 0.43644773960113525,
+ "step": 1435
+ },
+ {
+ "epoch": 0.7534102833158447,
+ "grad_norm": 0.16293628094190815,
+ "learning_rate": 1.1420559517598205e-05,
+ "loss": 0.4687269330024719,
+ "step": 1436
+ },
+ {
+ "epoch": 0.7539349422875131,
+ "grad_norm": 0.1553393974390287,
+ "learning_rate": 1.1394640092925302e-05,
+ "loss": 0.4501398205757141,
+ "step": 1437
+ },
+ {
+ "epoch": 0.7544596012591815,
+ "grad_norm": 0.16214837694535802,
+ "learning_rate": 1.1368764422909919e-05,
+ "loss": 0.4542577862739563,
+ "step": 1438
+ },
+ {
+ "epoch": 0.75498426023085,
+ "grad_norm": 0.15838878372428297,
+ "learning_rate": 1.1342932577850473e-05,
+ "loss": 0.46228161454200745,
+ "step": 1439
+ },
+ {
+ "epoch": 0.7555089192025184,
+ "grad_norm": 0.15090419497122687,
+ "learning_rate": 1.13171446279263e-05,
+ "loss": 0.4495932161808014,
+ "step": 1440
+ },
+ {
+ "epoch": 0.7560335781741868,
+ "grad_norm": 0.1543376786458394,
+ "learning_rate": 1.1291400643197484e-05,
+ "loss": 0.451630175113678,
+ "step": 1441
+ },
+ {
+ "epoch": 0.7565582371458552,
+ "grad_norm": 0.1539470751610696,
+ "learning_rate": 1.1265700693604668e-05,
+ "loss": 0.447547972202301,
+ "step": 1442
+ },
+ {
+ "epoch": 0.7570828961175237,
+ "grad_norm": 0.15761047703305445,
+ "learning_rate": 1.1240044848968856e-05,
+ "loss": 0.4516204595565796,
+ "step": 1443
+ },
+ {
+ "epoch": 0.757607555089192,
+ "grad_norm": 0.15675279683704735,
+ "learning_rate": 1.1214433178991258e-05,
+ "loss": 0.4597971737384796,
+ "step": 1444
+ },
+ {
+ "epoch": 0.7581322140608604,
+ "grad_norm": 0.15592110994904113,
+ "learning_rate": 1.1188865753253022e-05,
+ "loss": 0.443898469209671,
+ "step": 1445
+ },
+ {
+ "epoch": 0.7586568730325288,
+ "grad_norm": 0.15597501691030619,
+ "learning_rate": 1.116334264121513e-05,
+ "loss": 0.4581306278705597,
+ "step": 1446
+ },
+ {
+ "epoch": 0.7591815320041972,
+ "grad_norm": 0.15552304728589958,
+ "learning_rate": 1.1137863912218154e-05,
+ "loss": 0.4475254416465759,
+ "step": 1447
+ },
+ {
+ "epoch": 0.7597061909758657,
+ "grad_norm": 0.15759990123898035,
+ "learning_rate": 1.1112429635482118e-05,
+ "loss": 0.43585264682769775,
+ "step": 1448
+ },
+ {
+ "epoch": 0.7602308499475341,
+ "grad_norm": 0.15964657264676635,
+ "learning_rate": 1.1087039880106247e-05,
+ "loss": 0.46786290407180786,
+ "step": 1449
+ },
+ {
+ "epoch": 0.7607555089192025,
+ "grad_norm": 0.15155804051218852,
+ "learning_rate": 1.106169471506882e-05,
+ "loss": 0.42292726039886475,
+ "step": 1450
+ },
+ {
+ "epoch": 0.7612801678908709,
+ "grad_norm": 0.15519270966930884,
+ "learning_rate": 1.103639420922698e-05,
+ "loss": 0.4248972237110138,
+ "step": 1451
+ },
+ {
+ "epoch": 0.7618048268625394,
+ "grad_norm": 0.1526330511128821,
+ "learning_rate": 1.1011138431316537e-05,
+ "loss": 0.4317784309387207,
+ "step": 1452
+ },
+ {
+ "epoch": 0.7623294858342078,
+ "grad_norm": 0.15797682979272226,
+ "learning_rate": 1.0985927449951779e-05,
+ "loss": 0.44506698846817017,
+ "step": 1453
+ },
+ {
+ "epoch": 0.7628541448058762,
+ "grad_norm": 0.15479979838505462,
+ "learning_rate": 1.096076133362531e-05,
+ "loss": 0.4471631348133087,
+ "step": 1454
+ },
+ {
+ "epoch": 0.7633788037775446,
+ "grad_norm": 0.15712267274447342,
+ "learning_rate": 1.093564015070783e-05,
+ "loss": 0.4321915805339813,
+ "step": 1455
+ },
+ {
+ "epoch": 0.7639034627492131,
+ "grad_norm": 0.15857431494180016,
+ "learning_rate": 1.0910563969447969e-05,
+ "loss": 0.44455766677856445,
+ "step": 1456
+ },
+ {
+ "epoch": 0.7644281217208815,
+ "grad_norm": 0.15377057080997789,
+ "learning_rate": 1.0885532857972097e-05,
+ "loss": 0.4567924439907074,
+ "step": 1457
+ },
+ {
+ "epoch": 0.7649527806925498,
+ "grad_norm": 0.15544844005009104,
+ "learning_rate": 1.0860546884284143e-05,
+ "loss": 0.4491572976112366,
+ "step": 1458
+ },
+ {
+ "epoch": 0.7654774396642182,
+ "grad_norm": 0.16210735566479478,
+ "learning_rate": 1.08356061162654e-05,
+ "loss": 0.43823331594467163,
+ "step": 1459
+ },
+ {
+ "epoch": 0.7660020986358866,
+ "grad_norm": 0.15213559486379452,
+ "learning_rate": 1.0810710621674355e-05,
+ "loss": 0.45375770330429077,
+ "step": 1460
+ },
+ {
+ "epoch": 0.7665267576075551,
+ "grad_norm": 0.1727832306402992,
+ "learning_rate": 1.0785860468146494e-05,
+ "loss": 0.4639572501182556,
+ "step": 1461
+ },
+ {
+ "epoch": 0.7670514165792235,
+ "grad_norm": 0.1637555902013663,
+ "learning_rate": 1.076105572319412e-05,
+ "loss": 0.4355923533439636,
+ "step": 1462
+ },
+ {
+ "epoch": 0.7675760755508919,
+ "grad_norm": 0.16646936030470866,
+ "learning_rate": 1.0736296454206166e-05,
+ "loss": 0.4650861322879791,
+ "step": 1463
+ },
+ {
+ "epoch": 0.7681007345225603,
+ "grad_norm": 0.14987620615431385,
+ "learning_rate": 1.0711582728448033e-05,
+ "loss": 0.45445430278778076,
+ "step": 1464
+ },
+ {
+ "epoch": 0.7686253934942288,
+ "grad_norm": 0.16428618403547932,
+ "learning_rate": 1.0686914613061377e-05,
+ "loss": 0.46050935983657837,
+ "step": 1465
+ },
+ {
+ "epoch": 0.7691500524658972,
+ "grad_norm": 0.15052999053128632,
+ "learning_rate": 1.0662292175063949e-05,
+ "loss": 0.43385741114616394,
+ "step": 1466
+ },
+ {
+ "epoch": 0.7696747114375656,
+ "grad_norm": 0.1654949402740152,
+ "learning_rate": 1.0637715481349397e-05,
+ "loss": 0.4434410333633423,
+ "step": 1467
+ },
+ {
+ "epoch": 0.770199370409234,
+ "grad_norm": 0.15923836508490036,
+ "learning_rate": 1.0613184598687088e-05,
+ "loss": 0.4410185217857361,
+ "step": 1468
+ },
+ {
+ "epoch": 0.7707240293809025,
+ "grad_norm": 0.15813007111170696,
+ "learning_rate": 1.0588699593721957e-05,
+ "loss": 0.43588417768478394,
+ "step": 1469
+ },
+ {
+ "epoch": 0.7712486883525709,
+ "grad_norm": 0.16978817898431764,
+ "learning_rate": 1.0564260532974274e-05,
+ "loss": 0.4514843225479126,
+ "step": 1470
+ },
+ {
+ "epoch": 0.7717733473242392,
+ "grad_norm": 0.1613844852697828,
+ "learning_rate": 1.0539867482839499e-05,
+ "loss": 0.4554520845413208,
+ "step": 1471
+ },
+ {
+ "epoch": 0.7722980062959076,
+ "grad_norm": 0.1587788181984633,
+ "learning_rate": 1.0515520509588075e-05,
+ "loss": 0.43200671672821045,
+ "step": 1472
+ },
+ {
+ "epoch": 0.772822665267576,
+ "grad_norm": 0.1539930824827353,
+ "learning_rate": 1.0491219679365283e-05,
+ "loss": 0.4343116283416748,
+ "step": 1473
+ },
+ {
+ "epoch": 0.7733473242392445,
+ "grad_norm": 0.1561612537773709,
+ "learning_rate": 1.0466965058191049e-05,
+ "loss": 0.4477815330028534,
+ "step": 1474
+ },
+ {
+ "epoch": 0.7738719832109129,
+ "grad_norm": 0.169990777691657,
+ "learning_rate": 1.0442756711959745e-05,
+ "loss": 0.4476611614227295,
+ "step": 1475
+ },
+ {
+ "epoch": 0.7743966421825813,
+ "grad_norm": 0.1624313049193793,
+ "learning_rate": 1.041859470644002e-05,
+ "loss": 0.47136399149894714,
+ "step": 1476
+ },
+ {
+ "epoch": 0.7749213011542497,
+ "grad_norm": 0.15993315458994328,
+ "learning_rate": 1.0394479107274645e-05,
+ "loss": 0.4672670364379883,
+ "step": 1477
+ },
+ {
+ "epoch": 0.7754459601259182,
+ "grad_norm": 0.16755349272347217,
+ "learning_rate": 1.0370409979980294e-05,
+ "loss": 0.45135706663131714,
+ "step": 1478
+ },
+ {
+ "epoch": 0.7759706190975866,
+ "grad_norm": 0.15665059815830362,
+ "learning_rate": 1.0346387389947407e-05,
+ "loss": 0.45178359746932983,
+ "step": 1479
+ },
+ {
+ "epoch": 0.776495278069255,
+ "grad_norm": 0.1554612731452571,
+ "learning_rate": 1.0322411402439983e-05,
+ "loss": 0.4412590265274048,
+ "step": 1480
+ },
+ {
+ "epoch": 0.7770199370409234,
+ "grad_norm": 0.1532496869312065,
+ "learning_rate": 1.0298482082595408e-05,
+ "loss": 0.44104987382888794,
+ "step": 1481
+ },
+ {
+ "epoch": 0.7775445960125918,
+ "grad_norm": 0.15065868316960426,
+ "learning_rate": 1.0274599495424286e-05,
+ "loss": 0.4336603283882141,
+ "step": 1482
+ },
+ {
+ "epoch": 0.7780692549842603,
+ "grad_norm": 0.1548373981760097,
+ "learning_rate": 1.0250763705810267e-05,
+ "loss": 0.4658375680446625,
+ "step": 1483
+ },
+ {
+ "epoch": 0.7785939139559287,
+ "grad_norm": 0.15470181175866582,
+ "learning_rate": 1.0226974778509846e-05,
+ "loss": 0.45232075452804565,
+ "step": 1484
+ },
+ {
+ "epoch": 0.779118572927597,
+ "grad_norm": 0.15507371187727106,
+ "learning_rate": 1.0203232778152228e-05,
+ "loss": 0.43640458583831787,
+ "step": 1485
+ },
+ {
+ "epoch": 0.7796432318992654,
+ "grad_norm": 0.15269440736544085,
+ "learning_rate": 1.01795377692391e-05,
+ "loss": 0.4400939345359802,
+ "step": 1486
+ },
+ {
+ "epoch": 0.7801678908709339,
+ "grad_norm": 0.15610899550395085,
+ "learning_rate": 1.0155889816144495e-05,
+ "loss": 0.4546602666378021,
+ "step": 1487
+ },
+ {
+ "epoch": 0.7806925498426023,
+ "grad_norm": 0.1559328242601668,
+ "learning_rate": 1.0132288983114611e-05,
+ "loss": 0.45510709285736084,
+ "step": 1488
+ },
+ {
+ "epoch": 0.7812172088142707,
+ "grad_norm": 0.1537255194600787,
+ "learning_rate": 1.0108735334267639e-05,
+ "loss": 0.4475342631340027,
+ "step": 1489
+ },
+ {
+ "epoch": 0.7817418677859391,
+ "grad_norm": 0.14616575061664402,
+ "learning_rate": 1.0085228933593558e-05,
+ "loss": 0.4378190040588379,
+ "step": 1490
+ },
+ {
+ "epoch": 0.7822665267576075,
+ "grad_norm": 0.1624360967746598,
+ "learning_rate": 1.0061769844953999e-05,
+ "loss": 0.4439539909362793,
+ "step": 1491
+ },
+ {
+ "epoch": 0.782791185729276,
+ "grad_norm": 0.15407336125298413,
+ "learning_rate": 1.003835813208206e-05,
+ "loss": 0.44074907898902893,
+ "step": 1492
+ },
+ {
+ "epoch": 0.7833158447009444,
+ "grad_norm": 0.15786980710136125,
+ "learning_rate": 1.001499385858211e-05,
+ "loss": 0.4491667151451111,
+ "step": 1493
+ },
+ {
+ "epoch": 0.7838405036726128,
+ "grad_norm": 0.15338836738460088,
+ "learning_rate": 9.991677087929665e-06,
+ "loss": 0.446003794670105,
+ "step": 1494
+ },
+ {
+ "epoch": 0.7843651626442812,
+ "grad_norm": 0.156710894892,
+ "learning_rate": 9.968407883471165e-06,
+ "loss": 0.43604689836502075,
+ "step": 1495
+ },
+ {
+ "epoch": 0.7848898216159497,
+ "grad_norm": 0.16100539063191666,
+ "learning_rate": 9.945186308423832e-06,
+ "loss": 0.4431147873401642,
+ "step": 1496
+ },
+ {
+ "epoch": 0.7854144805876181,
+ "grad_norm": 0.15776018851122997,
+ "learning_rate": 9.92201242587548e-06,
+ "loss": 0.4532504975795746,
+ "step": 1497
+ },
+ {
+ "epoch": 0.7859391395592865,
+ "grad_norm": 0.1540333423540934,
+ "learning_rate": 9.898886298784368e-06,
+ "loss": 0.43605154752731323,
+ "step": 1498
+ },
+ {
+ "epoch": 0.7864637985309548,
+ "grad_norm": 0.16009602938972295,
+ "learning_rate": 9.875807989979005e-06,
+ "loss": 0.4605238437652588,
+ "step": 1499
+ },
+ {
+ "epoch": 0.7869884575026233,
+ "grad_norm": 0.15639368804171708,
+ "learning_rate": 9.852777562157983e-06,
+ "loss": 0.4509708881378174,
+ "step": 1500
+ },
+ {
+ "epoch": 0.7875131164742917,
+ "grad_norm": 0.15074011188741518,
+ "learning_rate": 9.829795077889828e-06,
+ "loss": 0.4527304172515869,
+ "step": 1501
+ },
+ {
+ "epoch": 0.7880377754459601,
+ "grad_norm": 0.15155083996490956,
+ "learning_rate": 9.806860599612799e-06,
+ "loss": 0.4306413531303406,
+ "step": 1502
+ },
+ {
+ "epoch": 0.7885624344176285,
+ "grad_norm": 0.1478956630795445,
+ "learning_rate": 9.783974189634734e-06,
+ "loss": 0.4434714913368225,
+ "step": 1503
+ },
+ {
+ "epoch": 0.789087093389297,
+ "grad_norm": 0.15327668434806432,
+ "learning_rate": 9.761135910132902e-06,
+ "loss": 0.44431227445602417,
+ "step": 1504
+ },
+ {
+ "epoch": 0.7896117523609654,
+ "grad_norm": 0.2243177585497212,
+ "learning_rate": 9.738345823153783e-06,
+ "loss": 0.4541007876396179,
+ "step": 1505
+ },
+ {
+ "epoch": 0.7901364113326338,
+ "grad_norm": 0.1600239865812282,
+ "learning_rate": 9.715603990612945e-06,
+ "loss": 0.4456910490989685,
+ "step": 1506
+ },
+ {
+ "epoch": 0.7906610703043022,
+ "grad_norm": 0.15304634075332066,
+ "learning_rate": 9.692910474294858e-06,
+ "loss": 0.43965545296669006,
+ "step": 1507
+ },
+ {
+ "epoch": 0.7911857292759706,
+ "grad_norm": 0.15657752062053032,
+ "learning_rate": 9.670265335852725e-06,
+ "loss": 0.43358081579208374,
+ "step": 1508
+ },
+ {
+ "epoch": 0.7917103882476391,
+ "grad_norm": 0.16110057783374115,
+ "learning_rate": 9.647668636808304e-06,
+ "loss": 0.45711368322372437,
+ "step": 1509
+ },
+ {
+ "epoch": 0.7922350472193075,
+ "grad_norm": 0.1551133690299623,
+ "learning_rate": 9.625120438551791e-06,
+ "loss": 0.4613519608974457,
+ "step": 1510
+ },
+ {
+ "epoch": 0.7927597061909759,
+ "grad_norm": 0.15690897380725946,
+ "learning_rate": 9.602620802341581e-06,
+ "loss": 0.4472261071205139,
+ "step": 1511
+ },
+ {
+ "epoch": 0.7932843651626443,
+ "grad_norm": 0.147796599322911,
+ "learning_rate": 9.580169789304152e-06,
+ "loss": 0.4447203576564789,
+ "step": 1512
+ },
+ {
+ "epoch": 0.7938090241343126,
+ "grad_norm": 0.152019559199633,
+ "learning_rate": 9.557767460433878e-06,
+ "loss": 0.445905864238739,
+ "step": 1513
+ },
+ {
+ "epoch": 0.7943336831059811,
+ "grad_norm": 0.1574763924872037,
+ "learning_rate": 9.535413876592872e-06,
+ "loss": 0.4502032995223999,
+ "step": 1514
+ },
+ {
+ "epoch": 0.7948583420776495,
+ "grad_norm": 0.1524218315499056,
+ "learning_rate": 9.513109098510816e-06,
+ "loss": 0.43753111362457275,
+ "step": 1515
+ },
+ {
+ "epoch": 0.7953830010493179,
+ "grad_norm": 0.15589147437937434,
+ "learning_rate": 9.490853186784801e-06,
+ "loss": 0.44524312019348145,
+ "step": 1516
+ },
+ {
+ "epoch": 0.7959076600209863,
+ "grad_norm": 0.15604009714081407,
+ "learning_rate": 9.468646201879156e-06,
+ "loss": 0.44141292572021484,
+ "step": 1517
+ },
+ {
+ "epoch": 0.7964323189926548,
+ "grad_norm": 0.1562432554004435,
+ "learning_rate": 9.44648820412529e-06,
+ "loss": 0.461681067943573,
+ "step": 1518
+ },
+ {
+ "epoch": 0.7969569779643232,
+ "grad_norm": 0.15233588531495665,
+ "learning_rate": 9.424379253721513e-06,
+ "loss": 0.4529497027397156,
+ "step": 1519
+ },
+ {
+ "epoch": 0.7974816369359916,
+ "grad_norm": 0.16856991685290923,
+ "learning_rate": 9.40231941073291e-06,
+ "loss": 0.4483596086502075,
+ "step": 1520
+ },
+ {
+ "epoch": 0.79800629590766,
+ "grad_norm": 0.1546876712981077,
+ "learning_rate": 9.380308735091125e-06,
+ "loss": 0.43846625089645386,
+ "step": 1521
+ },
+ {
+ "epoch": 0.7985309548793285,
+ "grad_norm": 0.15198637556890449,
+ "learning_rate": 9.35834728659424e-06,
+ "loss": 0.4625553488731384,
+ "step": 1522
+ },
+ {
+ "epoch": 0.7990556138509969,
+ "grad_norm": 0.1519777007198701,
+ "learning_rate": 9.336435124906592e-06,
+ "loss": 0.44183820486068726,
+ "step": 1523
+ },
+ {
+ "epoch": 0.7995802728226653,
+ "grad_norm": 0.16125823074979284,
+ "learning_rate": 9.314572309558616e-06,
+ "loss": 0.46057307720184326,
+ "step": 1524
+ },
+ {
+ "epoch": 0.8001049317943337,
+ "grad_norm": 0.16283323624848034,
+ "learning_rate": 9.2927588999467e-06,
+ "loss": 0.4530155062675476,
+ "step": 1525
+ },
+ {
+ "epoch": 0.8006295907660022,
+ "grad_norm": 0.14998654489027738,
+ "learning_rate": 9.270994955332984e-06,
+ "loss": 0.43155500292778015,
+ "step": 1526
+ },
+ {
+ "epoch": 0.8011542497376705,
+ "grad_norm": 0.15683095030848093,
+ "learning_rate": 9.249280534845241e-06,
+ "loss": 0.45097050070762634,
+ "step": 1527
+ },
+ {
+ "epoch": 0.8016789087093389,
+ "grad_norm": 0.1561578254492378,
+ "learning_rate": 9.227615697476674e-06,
+ "loss": 0.42797762155532837,
+ "step": 1528
+ },
+ {
+ "epoch": 0.8022035676810073,
+ "grad_norm": 0.15642385261469652,
+ "learning_rate": 9.206000502085813e-06,
+ "loss": 0.4560159742832184,
+ "step": 1529
+ },
+ {
+ "epoch": 0.8027282266526757,
+ "grad_norm": 0.1623126279097499,
+ "learning_rate": 9.1844350073963e-06,
+ "loss": 0.4704165458679199,
+ "step": 1530
+ },
+ {
+ "epoch": 0.8032528856243442,
+ "grad_norm": 0.15090934786253773,
+ "learning_rate": 9.162919271996755e-06,
+ "loss": 0.44642555713653564,
+ "step": 1531
+ },
+ {
+ "epoch": 0.8037775445960126,
+ "grad_norm": 0.15305216978534084,
+ "learning_rate": 9.14145335434062e-06,
+ "loss": 0.4634283781051636,
+ "step": 1532
+ },
+ {
+ "epoch": 0.804302203567681,
+ "grad_norm": 0.14653219861345698,
+ "learning_rate": 9.120037312745979e-06,
+ "loss": 0.44544070959091187,
+ "step": 1533
+ },
+ {
+ "epoch": 0.8048268625393494,
+ "grad_norm": 0.16247397938502464,
+ "learning_rate": 9.098671205395422e-06,
+ "loss": 0.4552367925643921,
+ "step": 1534
+ },
+ {
+ "epoch": 0.8053515215110179,
+ "grad_norm": 0.14963174165902007,
+ "learning_rate": 9.077355090335892e-06,
+ "loss": 0.43211787939071655,
+ "step": 1535
+ },
+ {
+ "epoch": 0.8058761804826863,
+ "grad_norm": 0.1554853846772283,
+ "learning_rate": 9.056089025478495e-06,
+ "loss": 0.4319286644458771,
+ "step": 1536
+ },
+ {
+ "epoch": 0.8064008394543547,
+ "grad_norm": 0.15840895821310472,
+ "learning_rate": 9.03487306859837e-06,
+ "loss": 0.4494994878768921,
+ "step": 1537
+ },
+ {
+ "epoch": 0.8069254984260231,
+ "grad_norm": 0.1543496716800982,
+ "learning_rate": 9.013707277334515e-06,
+ "loss": 0.45500510931015015,
+ "step": 1538
+ },
+ {
+ "epoch": 0.8074501573976915,
+ "grad_norm": 0.14797282318550045,
+ "learning_rate": 8.992591709189651e-06,
+ "loss": 0.4385896325111389,
+ "step": 1539
+ },
+ {
+ "epoch": 0.8079748163693599,
+ "grad_norm": 0.17045570682787758,
+ "learning_rate": 8.971526421530066e-06,
+ "loss": 0.4585988521575928,
+ "step": 1540
+ },
+ {
+ "epoch": 0.8084994753410283,
+ "grad_norm": 0.14758220713044884,
+ "learning_rate": 8.950511471585407e-06,
+ "loss": 0.4295082092285156,
+ "step": 1541
+ },
+ {
+ "epoch": 0.8090241343126967,
+ "grad_norm": 0.16005752068175036,
+ "learning_rate": 8.929546916448597e-06,
+ "loss": 0.4507679343223572,
+ "step": 1542
+ },
+ {
+ "epoch": 0.8095487932843651,
+ "grad_norm": 0.14884684083313074,
+ "learning_rate": 8.908632813075641e-06,
+ "loss": 0.43547600507736206,
+ "step": 1543
+ },
+ {
+ "epoch": 0.8100734522560336,
+ "grad_norm": 0.14860107595788136,
+ "learning_rate": 8.88776921828547e-06,
+ "loss": 0.42700642347335815,
+ "step": 1544
+ },
+ {
+ "epoch": 0.810598111227702,
+ "grad_norm": 0.1472630611843838,
+ "learning_rate": 8.86695618875981e-06,
+ "loss": 0.4473339319229126,
+ "step": 1545
+ },
+ {
+ "epoch": 0.8111227701993704,
+ "grad_norm": 0.15105523788317893,
+ "learning_rate": 8.846193781043e-06,
+ "loss": 0.44923150539398193,
+ "step": 1546
+ },
+ {
+ "epoch": 0.8116474291710388,
+ "grad_norm": 0.16121409015263036,
+ "learning_rate": 8.82548205154185e-06,
+ "loss": 0.4653576612472534,
+ "step": 1547
+ },
+ {
+ "epoch": 0.8121720881427072,
+ "grad_norm": 0.1595195150959421,
+ "learning_rate": 8.804821056525497e-06,
+ "loss": 0.4522424638271332,
+ "step": 1548
+ },
+ {
+ "epoch": 0.8126967471143757,
+ "grad_norm": 0.15163510106316194,
+ "learning_rate": 8.784210852125232e-06,
+ "loss": 0.43657630681991577,
+ "step": 1549
+ },
+ {
+ "epoch": 0.8132214060860441,
+ "grad_norm": 0.15155109347013793,
+ "learning_rate": 8.763651494334377e-06,
+ "loss": 0.43183624744415283,
+ "step": 1550
+ },
+ {
+ "epoch": 0.8137460650577125,
+ "grad_norm": 0.14930024958763324,
+ "learning_rate": 8.7431430390081e-06,
+ "loss": 0.43563947081565857,
+ "step": 1551
+ },
+ {
+ "epoch": 0.8142707240293809,
+ "grad_norm": 0.15361200153430712,
+ "learning_rate": 8.722685541863283e-06,
+ "loss": 0.4387405514717102,
+ "step": 1552
+ },
+ {
+ "epoch": 0.8147953830010494,
+ "grad_norm": 0.15549528796193493,
+ "learning_rate": 8.702279058478366e-06,
+ "loss": 0.44347938895225525,
+ "step": 1553
+ },
+ {
+ "epoch": 0.8153200419727177,
+ "grad_norm": 0.16243042867861712,
+ "learning_rate": 8.681923644293194e-06,
+ "loss": 0.44852834939956665,
+ "step": 1554
+ },
+ {
+ "epoch": 0.8158447009443861,
+ "grad_norm": 0.15706129652944026,
+ "learning_rate": 8.661619354608873e-06,
+ "loss": 0.45419150590896606,
+ "step": 1555
+ },
+ {
+ "epoch": 0.8163693599160545,
+ "grad_norm": 0.14854258573134127,
+ "learning_rate": 8.641366244587611e-06,
+ "loss": 0.44866859912872314,
+ "step": 1556
+ },
+ {
+ "epoch": 0.816894018887723,
+ "grad_norm": 0.14896937678017533,
+ "learning_rate": 8.62116436925258e-06,
+ "loss": 0.44596076011657715,
+ "step": 1557
+ },
+ {
+ "epoch": 0.8174186778593914,
+ "grad_norm": 0.15614508495620508,
+ "learning_rate": 8.601013783487746e-06,
+ "loss": 0.45645207166671753,
+ "step": 1558
+ },
+ {
+ "epoch": 0.8179433368310598,
+ "grad_norm": 0.1486538543050606,
+ "learning_rate": 8.580914542037734e-06,
+ "loss": 0.4592300355434418,
+ "step": 1559
+ },
+ {
+ "epoch": 0.8184679958027282,
+ "grad_norm": 0.15475320756830518,
+ "learning_rate": 8.560866699507702e-06,
+ "loss": 0.4605134129524231,
+ "step": 1560
+ },
+ {
+ "epoch": 0.8189926547743966,
+ "grad_norm": 0.14754406728057914,
+ "learning_rate": 8.540870310363137e-06,
+ "loss": 0.43651673197746277,
+ "step": 1561
+ },
+ {
+ "epoch": 0.8195173137460651,
+ "grad_norm": 0.14871990701077825,
+ "learning_rate": 8.520925428929755e-06,
+ "loss": 0.44830837845802307,
+ "step": 1562
+ },
+ {
+ "epoch": 0.8200419727177335,
+ "grad_norm": 0.15358241317712973,
+ "learning_rate": 8.501032109393334e-06,
+ "loss": 0.4546644687652588,
+ "step": 1563
+ },
+ {
+ "epoch": 0.8205666316894019,
+ "grad_norm": 0.15650856948126138,
+ "learning_rate": 8.481190405799569e-06,
+ "loss": 0.4423891305923462,
+ "step": 1564
+ },
+ {
+ "epoch": 0.8210912906610703,
+ "grad_norm": 0.15474996746242287,
+ "learning_rate": 8.461400372053923e-06,
+ "loss": 0.44242313504219055,
+ "step": 1565
+ },
+ {
+ "epoch": 0.8216159496327388,
+ "grad_norm": 0.16544926591962897,
+ "learning_rate": 8.441662061921497e-06,
+ "loss": 0.4487020969390869,
+ "step": 1566
+ },
+ {
+ "epoch": 0.8221406086044072,
+ "grad_norm": 0.1489733586650647,
+ "learning_rate": 8.421975529026855e-06,
+ "loss": 0.4371356964111328,
+ "step": 1567
+ },
+ {
+ "epoch": 0.8226652675760755,
+ "grad_norm": 0.15995455652906124,
+ "learning_rate": 8.402340826853907e-06,
+ "loss": 0.4566021263599396,
+ "step": 1568
+ },
+ {
+ "epoch": 0.8231899265477439,
+ "grad_norm": 0.1500716152622102,
+ "learning_rate": 8.382758008745726e-06,
+ "loss": 0.43888187408447266,
+ "step": 1569
+ },
+ {
+ "epoch": 0.8237145855194123,
+ "grad_norm": 0.15554904485712012,
+ "learning_rate": 8.363227127904464e-06,
+ "loss": 0.4584001898765564,
+ "step": 1570
+ },
+ {
+ "epoch": 0.8242392444910808,
+ "grad_norm": 0.14820917815884277,
+ "learning_rate": 8.343748237391144e-06,
+ "loss": 0.43289071321487427,
+ "step": 1571
+ },
+ {
+ "epoch": 0.8247639034627492,
+ "grad_norm": 0.15801200747859714,
+ "learning_rate": 8.324321390125547e-06,
+ "loss": 0.44196009635925293,
+ "step": 1572
+ },
+ {
+ "epoch": 0.8252885624344176,
+ "grad_norm": 0.15323277975380797,
+ "learning_rate": 8.304946638886078e-06,
+ "loss": 0.43244463205337524,
+ "step": 1573
+ },
+ {
+ "epoch": 0.825813221406086,
+ "grad_norm": 0.15076940176344772,
+ "learning_rate": 8.285624036309583e-06,
+ "loss": 0.44401073455810547,
+ "step": 1574
+ },
+ {
+ "epoch": 0.8263378803777545,
+ "grad_norm": 0.1544958300475393,
+ "learning_rate": 8.266353634891266e-06,
+ "loss": 0.45198434591293335,
+ "step": 1575
+ },
+ {
+ "epoch": 0.8268625393494229,
+ "grad_norm": 0.15093528292068448,
+ "learning_rate": 8.247135486984486e-06,
+ "loss": 0.4403747618198395,
+ "step": 1576
+ },
+ {
+ "epoch": 0.8273871983210913,
+ "grad_norm": 0.14680662784017148,
+ "learning_rate": 8.227969644800648e-06,
+ "loss": 0.4375043213367462,
+ "step": 1577
+ },
+ {
+ "epoch": 0.8279118572927597,
+ "grad_norm": 0.14791704009291914,
+ "learning_rate": 8.208856160409055e-06,
+ "loss": 0.43811577558517456,
+ "step": 1578
+ },
+ {
+ "epoch": 0.8284365162644282,
+ "grad_norm": 0.14924185710518617,
+ "learning_rate": 8.189795085736765e-06,
+ "loss": 0.43853428959846497,
+ "step": 1579
+ },
+ {
+ "epoch": 0.8289611752360966,
+ "grad_norm": 0.15420576936040198,
+ "learning_rate": 8.170786472568451e-06,
+ "loss": 0.445467084646225,
+ "step": 1580
+ },
+ {
+ "epoch": 0.829485834207765,
+ "grad_norm": 0.15577901772605834,
+ "learning_rate": 8.151830372546258e-06,
+ "loss": 0.4529452621936798,
+ "step": 1581
+ },
+ {
+ "epoch": 0.8300104931794333,
+ "grad_norm": 0.15387840488797552,
+ "learning_rate": 8.132926837169676e-06,
+ "loss": 0.46009817719459534,
+ "step": 1582
+ },
+ {
+ "epoch": 0.8305351521511017,
+ "grad_norm": 0.1527913309957562,
+ "learning_rate": 8.11407591779537e-06,
+ "loss": 0.4513983726501465,
+ "step": 1583
+ },
+ {
+ "epoch": 0.8310598111227702,
+ "grad_norm": 0.1537090286922398,
+ "learning_rate": 8.095277665637061e-06,
+ "loss": 0.4537739157676697,
+ "step": 1584
+ },
+ {
+ "epoch": 0.8315844700944386,
+ "grad_norm": 0.15722962107654076,
+ "learning_rate": 8.076532131765408e-06,
+ "loss": 0.45232969522476196,
+ "step": 1585
+ },
+ {
+ "epoch": 0.832109129066107,
+ "grad_norm": 0.1450142396611787,
+ "learning_rate": 8.057839367107817e-06,
+ "loss": 0.4270634651184082,
+ "step": 1586
+ },
+ {
+ "epoch": 0.8326337880377754,
+ "grad_norm": 0.150264001570592,
+ "learning_rate": 8.039199422448355e-06,
+ "loss": 0.43563878536224365,
+ "step": 1587
+ },
+ {
+ "epoch": 0.8331584470094439,
+ "grad_norm": 0.14904943307590535,
+ "learning_rate": 8.020612348427572e-06,
+ "loss": 0.4401509165763855,
+ "step": 1588
+ },
+ {
+ "epoch": 0.8336831059811123,
+ "grad_norm": 0.15453821662860034,
+ "learning_rate": 8.002078195542393e-06,
+ "loss": 0.44545161724090576,
+ "step": 1589
+ },
+ {
+ "epoch": 0.8342077649527807,
+ "grad_norm": 0.1570809678830383,
+ "learning_rate": 7.983597014145951e-06,
+ "loss": 0.45228466391563416,
+ "step": 1590
+ },
+ {
+ "epoch": 0.8347324239244491,
+ "grad_norm": 0.14741240769629818,
+ "learning_rate": 7.965168854447493e-06,
+ "loss": 0.4485381245613098,
+ "step": 1591
+ },
+ {
+ "epoch": 0.8352570828961176,
+ "grad_norm": 0.15362267095533516,
+ "learning_rate": 7.946793766512195e-06,
+ "loss": 0.4358059763908386,
+ "step": 1592
+ },
+ {
+ "epoch": 0.835781741867786,
+ "grad_norm": 0.1544754586495605,
+ "learning_rate": 7.92847180026106e-06,
+ "loss": 0.43327778577804565,
+ "step": 1593
+ },
+ {
+ "epoch": 0.8363064008394544,
+ "grad_norm": 0.15023225026164963,
+ "learning_rate": 7.91020300547077e-06,
+ "loss": 0.44359731674194336,
+ "step": 1594
+ },
+ {
+ "epoch": 0.8368310598111228,
+ "grad_norm": 0.15665530392095345,
+ "learning_rate": 7.891987431773545e-06,
+ "loss": 0.44503387808799744,
+ "step": 1595
+ },
+ {
+ "epoch": 0.8373557187827911,
+ "grad_norm": 0.14717614926392292,
+ "learning_rate": 7.873825128657033e-06,
+ "loss": 0.45314061641693115,
+ "step": 1596
+ },
+ {
+ "epoch": 0.8378803777544596,
+ "grad_norm": 0.15207464934030612,
+ "learning_rate": 7.855716145464131e-06,
+ "loss": 0.4526442885398865,
+ "step": 1597
+ },
+ {
+ "epoch": 0.838405036726128,
+ "grad_norm": 0.15129881541794873,
+ "learning_rate": 7.837660531392904e-06,
+ "loss": 0.45127853751182556,
+ "step": 1598
+ },
+ {
+ "epoch": 0.8389296956977964,
+ "grad_norm": 0.15623362715417285,
+ "learning_rate": 7.819658335496405e-06,
+ "loss": 0.46432793140411377,
+ "step": 1599
+ },
+ {
+ "epoch": 0.8394543546694648,
+ "grad_norm": 0.1529219879642125,
+ "learning_rate": 7.801709606682573e-06,
+ "loss": 0.4475358724594116,
+ "step": 1600
+ },
+ {
+ "epoch": 0.8399790136411333,
+ "grad_norm": 0.14649139854087698,
+ "learning_rate": 7.78381439371409e-06,
+ "loss": 0.4363391101360321,
+ "step": 1601
+ },
+ {
+ "epoch": 0.8405036726128017,
+ "grad_norm": 0.14571233251948087,
+ "learning_rate": 7.765972745208243e-06,
+ "loss": 0.44419488310813904,
+ "step": 1602
+ },
+ {
+ "epoch": 0.8410283315844701,
+ "grad_norm": 0.15168460066406128,
+ "learning_rate": 7.748184709636791e-06,
+ "loss": 0.4433790445327759,
+ "step": 1603
+ },
+ {
+ "epoch": 0.8415529905561385,
+ "grad_norm": 0.14852860876513757,
+ "learning_rate": 7.730450335325853e-06,
+ "loss": 0.4493539333343506,
+ "step": 1604
+ },
+ {
+ "epoch": 0.842077649527807,
+ "grad_norm": 0.1505615491087377,
+ "learning_rate": 7.712769670455741e-06,
+ "loss": 0.4646318554878235,
+ "step": 1605
+ },
+ {
+ "epoch": 0.8426023084994754,
+ "grad_norm": 0.15407610486660114,
+ "learning_rate": 7.695142763060878e-06,
+ "loss": 0.4484049677848816,
+ "step": 1606
+ },
+ {
+ "epoch": 0.8431269674711438,
+ "grad_norm": 0.1521449709584124,
+ "learning_rate": 7.67756966102962e-06,
+ "loss": 0.43366891145706177,
+ "step": 1607
+ },
+ {
+ "epoch": 0.8436516264428122,
+ "grad_norm": 0.14747762647596166,
+ "learning_rate": 7.660050412104153e-06,
+ "loss": 0.43950092792510986,
+ "step": 1608
+ },
+ {
+ "epoch": 0.8441762854144806,
+ "grad_norm": 0.15512610809240027,
+ "learning_rate": 7.642585063880351e-06,
+ "loss": 0.46813055872917175,
+ "step": 1609
+ },
+ {
+ "epoch": 0.844700944386149,
+ "grad_norm": 0.15825480008734216,
+ "learning_rate": 7.625173663807655e-06,
+ "loss": 0.4348755478858948,
+ "step": 1610
+ },
+ {
+ "epoch": 0.8452256033578174,
+ "grad_norm": 0.1573968981854278,
+ "learning_rate": 7.6078162591889434e-06,
+ "loss": 0.45048123598098755,
+ "step": 1611
+ },
+ {
+ "epoch": 0.8457502623294858,
+ "grad_norm": 0.1517038960096417,
+ "learning_rate": 7.590512897180399e-06,
+ "loss": 0.43563610315322876,
+ "step": 1612
+ },
+ {
+ "epoch": 0.8462749213011542,
+ "grad_norm": 0.15503576044724665,
+ "learning_rate": 7.573263624791382e-06,
+ "loss": 0.4467726945877075,
+ "step": 1613
+ },
+ {
+ "epoch": 0.8467995802728226,
+ "grad_norm": 0.1560549927192375,
+ "learning_rate": 7.556068488884304e-06,
+ "loss": 0.4320477545261383,
+ "step": 1614
+ },
+ {
+ "epoch": 0.8473242392444911,
+ "grad_norm": 0.1679830356844654,
+ "learning_rate": 7.538927536174498e-06,
+ "loss": 0.4476034343242645,
+ "step": 1615
+ },
+ {
+ "epoch": 0.8478488982161595,
+ "grad_norm": 0.15169257383162213,
+ "learning_rate": 7.521840813230098e-06,
+ "loss": 0.4325854778289795,
+ "step": 1616
+ },
+ {
+ "epoch": 0.8483735571878279,
+ "grad_norm": 0.15980226324329827,
+ "learning_rate": 7.504808366471909e-06,
+ "loss": 0.4472721815109253,
+ "step": 1617
+ },
+ {
+ "epoch": 0.8488982161594963,
+ "grad_norm": 0.16932727986736365,
+ "learning_rate": 7.48783024217327e-06,
+ "loss": 0.45428287982940674,
+ "step": 1618
+ },
+ {
+ "epoch": 0.8494228751311648,
+ "grad_norm": 0.15065676306311376,
+ "learning_rate": 7.470906486459946e-06,
+ "loss": 0.44418883323669434,
+ "step": 1619
+ },
+ {
+ "epoch": 0.8499475341028332,
+ "grad_norm": 0.1533992998314183,
+ "learning_rate": 7.454037145309986e-06,
+ "loss": 0.43552401661872864,
+ "step": 1620
+ },
+ {
+ "epoch": 0.8504721930745016,
+ "grad_norm": 0.15479480515120017,
+ "learning_rate": 7.437222264553628e-06,
+ "loss": 0.444465696811676,
+ "step": 1621
+ },
+ {
+ "epoch": 0.85099685204617,
+ "grad_norm": 0.15845678337942112,
+ "learning_rate": 7.420461889873129e-06,
+ "loss": 0.44896578788757324,
+ "step": 1622
+ },
+ {
+ "epoch": 0.8515215110178384,
+ "grad_norm": 0.15038940533473208,
+ "learning_rate": 7.4037560668026804e-06,
+ "loss": 0.43062180280685425,
+ "step": 1623
+ },
+ {
+ "epoch": 0.8520461699895068,
+ "grad_norm": 0.1480017115109342,
+ "learning_rate": 7.387104840728259e-06,
+ "loss": 0.4630432724952698,
+ "step": 1624
+ },
+ {
+ "epoch": 0.8525708289611752,
+ "grad_norm": 0.15321305032824092,
+ "learning_rate": 7.370508256887514e-06,
+ "loss": 0.4419487714767456,
+ "step": 1625
+ },
+ {
+ "epoch": 0.8530954879328436,
+ "grad_norm": 0.1568246735699272,
+ "learning_rate": 7.353966360369659e-06,
+ "loss": 0.4483609199523926,
+ "step": 1626
+ },
+ {
+ "epoch": 0.853620146904512,
+ "grad_norm": 0.15539536583814875,
+ "learning_rate": 7.337479196115317e-06,
+ "loss": 0.4567753076553345,
+ "step": 1627
+ },
+ {
+ "epoch": 0.8541448058761805,
+ "grad_norm": 0.15528695146633797,
+ "learning_rate": 7.321046808916426e-06,
+ "loss": 0.4560334086418152,
+ "step": 1628
+ },
+ {
+ "epoch": 0.8546694648478489,
+ "grad_norm": 0.15005728554864303,
+ "learning_rate": 7.304669243416102e-06,
+ "loss": 0.43833765387535095,
+ "step": 1629
+ },
+ {
+ "epoch": 0.8551941238195173,
+ "grad_norm": 0.15959100645189242,
+ "learning_rate": 7.288346544108518e-06,
+ "loss": 0.4713831841945648,
+ "step": 1630
+ },
+ {
+ "epoch": 0.8557187827911857,
+ "grad_norm": 0.14898121463872632,
+ "learning_rate": 7.272078755338807e-06,
+ "loss": 0.44003745913505554,
+ "step": 1631
+ },
+ {
+ "epoch": 0.8562434417628542,
+ "grad_norm": 0.15083110802343333,
+ "learning_rate": 7.255865921302903e-06,
+ "loss": 0.429004043340683,
+ "step": 1632
+ },
+ {
+ "epoch": 0.8567681007345226,
+ "grad_norm": 0.1522757086796616,
+ "learning_rate": 7.239708086047445e-06,
+ "loss": 0.4403839707374573,
+ "step": 1633
+ },
+ {
+ "epoch": 0.857292759706191,
+ "grad_norm": 0.15719824838921245,
+ "learning_rate": 7.223605293469654e-06,
+ "loss": 0.4377821087837219,
+ "step": 1634
+ },
+ {
+ "epoch": 0.8578174186778594,
+ "grad_norm": 0.1450616202708373,
+ "learning_rate": 7.207557587317216e-06,
+ "loss": 0.4357878267765045,
+ "step": 1635
+ },
+ {
+ "epoch": 0.8583420776495279,
+ "grad_norm": 0.15703162220646427,
+ "learning_rate": 7.191565011188149e-06,
+ "loss": 0.44412243366241455,
+ "step": 1636
+ },
+ {
+ "epoch": 0.8588667366211962,
+ "grad_norm": 0.150175313334118,
+ "learning_rate": 7.175627608530716e-06,
+ "loss": 0.4543853998184204,
+ "step": 1637
+ },
+ {
+ "epoch": 0.8593913955928646,
+ "grad_norm": 0.15091711492716975,
+ "learning_rate": 7.159745422643261e-06,
+ "loss": 0.4268660545349121,
+ "step": 1638
+ },
+ {
+ "epoch": 0.859916054564533,
+ "grad_norm": 0.15449835654964011,
+ "learning_rate": 7.1439184966741295e-06,
+ "loss": 0.4507150650024414,
+ "step": 1639
+ },
+ {
+ "epoch": 0.8604407135362014,
+ "grad_norm": 0.15689612593442026,
+ "learning_rate": 7.128146873621533e-06,
+ "loss": 0.45520636439323425,
+ "step": 1640
+ },
+ {
+ "epoch": 0.8609653725078699,
+ "grad_norm": 0.1565226785234465,
+ "learning_rate": 7.11243059633345e-06,
+ "loss": 0.45102012157440186,
+ "step": 1641
+ },
+ {
+ "epoch": 0.8614900314795383,
+ "grad_norm": 0.15500877922230677,
+ "learning_rate": 7.096769707507481e-06,
+ "loss": 0.447115421295166,
+ "step": 1642
+ },
+ {
+ "epoch": 0.8620146904512067,
+ "grad_norm": 0.1595248428000243,
+ "learning_rate": 7.081164249690761e-06,
+ "loss": 0.4627235233783722,
+ "step": 1643
+ },
+ {
+ "epoch": 0.8625393494228751,
+ "grad_norm": 0.15958177510163488,
+ "learning_rate": 7.065614265279818e-06,
+ "loss": 0.4386388063430786,
+ "step": 1644
+ },
+ {
+ "epoch": 0.8630640083945436,
+ "grad_norm": 0.15152834825396286,
+ "learning_rate": 7.050119796520482e-06,
+ "loss": 0.4531296491622925,
+ "step": 1645
+ },
+ {
+ "epoch": 0.863588667366212,
+ "grad_norm": 0.15635733905345023,
+ "learning_rate": 7.034680885507752e-06,
+ "loss": 0.45573747158050537,
+ "step": 1646
+ },
+ {
+ "epoch": 0.8641133263378804,
+ "grad_norm": 0.15666918852633857,
+ "learning_rate": 7.019297574185703e-06,
+ "loss": 0.4499792158603668,
+ "step": 1647
+ },
+ {
+ "epoch": 0.8646379853095488,
+ "grad_norm": 0.156444627477174,
+ "learning_rate": 7.0039699043473405e-06,
+ "loss": 0.4545009136199951,
+ "step": 1648
+ },
+ {
+ "epoch": 0.8651626442812173,
+ "grad_norm": 0.14638989347446527,
+ "learning_rate": 6.988697917634515e-06,
+ "loss": 0.4397326409816742,
+ "step": 1649
+ },
+ {
+ "epoch": 0.8656873032528857,
+ "grad_norm": 0.15492813701343372,
+ "learning_rate": 6.973481655537793e-06,
+ "loss": 0.4402945041656494,
+ "step": 1650
+ },
+ {
+ "epoch": 0.866211962224554,
+ "grad_norm": 0.14466615385057716,
+ "learning_rate": 6.958321159396352e-06,
+ "loss": 0.4479280114173889,
+ "step": 1651
+ },
+ {
+ "epoch": 0.8667366211962224,
+ "grad_norm": 0.15865895330952856,
+ "learning_rate": 6.943216470397871e-06,
+ "loss": 0.4553218483924866,
+ "step": 1652
+ },
+ {
+ "epoch": 0.8672612801678908,
+ "grad_norm": 0.15293327582467067,
+ "learning_rate": 6.928167629578407e-06,
+ "loss": 0.44509196281433105,
+ "step": 1653
+ },
+ {
+ "epoch": 0.8677859391395593,
+ "grad_norm": 0.1515507946195896,
+ "learning_rate": 6.91317467782229e-06,
+ "loss": 0.4498351216316223,
+ "step": 1654
+ },
+ {
+ "epoch": 0.8683105981112277,
+ "grad_norm": 0.15313372799407085,
+ "learning_rate": 6.898237655862007e-06,
+ "loss": 0.4445887804031372,
+ "step": 1655
+ },
+ {
+ "epoch": 0.8688352570828961,
+ "grad_norm": 0.1491970050887588,
+ "learning_rate": 6.883356604278115e-06,
+ "loss": 0.4323808252811432,
+ "step": 1656
+ },
+ {
+ "epoch": 0.8693599160545645,
+ "grad_norm": 0.14449269685472518,
+ "learning_rate": 6.868531563499096e-06,
+ "loss": 0.42508652806282043,
+ "step": 1657
+ },
+ {
+ "epoch": 0.869884575026233,
+ "grad_norm": 0.14997328847830618,
+ "learning_rate": 6.853762573801264e-06,
+ "loss": 0.43655943870544434,
+ "step": 1658
+ },
+ {
+ "epoch": 0.8704092339979014,
+ "grad_norm": 0.15288208370561934,
+ "learning_rate": 6.839049675308662e-06,
+ "loss": 0.4505232274532318,
+ "step": 1659
+ },
+ {
+ "epoch": 0.8709338929695698,
+ "grad_norm": 0.14416201723141706,
+ "learning_rate": 6.824392907992938e-06,
+ "loss": 0.42444318532943726,
+ "step": 1660
+ },
+ {
+ "epoch": 0.8714585519412382,
+ "grad_norm": 0.14468507187223545,
+ "learning_rate": 6.809792311673244e-06,
+ "loss": 0.4251434803009033,
+ "step": 1661
+ },
+ {
+ "epoch": 0.8719832109129066,
+ "grad_norm": 0.15406295659798216,
+ "learning_rate": 6.795247926016142e-06,
+ "loss": 0.45068371295928955,
+ "step": 1662
+ },
+ {
+ "epoch": 0.8725078698845751,
+ "grad_norm": 0.1516721215663676,
+ "learning_rate": 6.780759790535465e-06,
+ "loss": 0.441638320684433,
+ "step": 1663
+ },
+ {
+ "epoch": 0.8730325288562435,
+ "grad_norm": 0.1537661139082594,
+ "learning_rate": 6.766327944592239e-06,
+ "loss": 0.44044923782348633,
+ "step": 1664
+ },
+ {
+ "epoch": 0.8735571878279118,
+ "grad_norm": 0.15175839438218855,
+ "learning_rate": 6.751952427394553e-06,
+ "loss": 0.44454342126846313,
+ "step": 1665
+ },
+ {
+ "epoch": 0.8740818467995802,
+ "grad_norm": 0.1531195282211392,
+ "learning_rate": 6.737633277997471e-06,
+ "loss": 0.4540765881538391,
+ "step": 1666
+ },
+ {
+ "epoch": 0.8746065057712487,
+ "grad_norm": 0.14961744003585584,
+ "learning_rate": 6.723370535302915e-06,
+ "loss": 0.44223541021347046,
+ "step": 1667
+ },
+ {
+ "epoch": 0.8751311647429171,
+ "grad_norm": 0.14816820299503658,
+ "learning_rate": 6.709164238059568e-06,
+ "loss": 0.43701833486557007,
+ "step": 1668
+ },
+ {
+ "epoch": 0.8756558237145855,
+ "grad_norm": 0.14622402294164918,
+ "learning_rate": 6.695014424862756e-06,
+ "loss": 0.42922335863113403,
+ "step": 1669
+ },
+ {
+ "epoch": 0.8761804826862539,
+ "grad_norm": 0.1547291062309004,
+ "learning_rate": 6.680921134154355e-06,
+ "loss": 0.4233800768852234,
+ "step": 1670
+ },
+ {
+ "epoch": 0.8767051416579223,
+ "grad_norm": 0.15378903523440587,
+ "learning_rate": 6.666884404222678e-06,
+ "loss": 0.4432950019836426,
+ "step": 1671
+ },
+ {
+ "epoch": 0.8772298006295908,
+ "grad_norm": 0.15973037860999773,
+ "learning_rate": 6.652904273202384e-06,
+ "loss": 0.4560699462890625,
+ "step": 1672
+ },
+ {
+ "epoch": 0.8777544596012592,
+ "grad_norm": 0.15430233088028492,
+ "learning_rate": 6.638980779074358e-06,
+ "loss": 0.45776498317718506,
+ "step": 1673
+ },
+ {
+ "epoch": 0.8782791185729276,
+ "grad_norm": 0.1587664730992995,
+ "learning_rate": 6.625113959665614e-06,
+ "loss": 0.43579810857772827,
+ "step": 1674
+ },
+ {
+ "epoch": 0.878803777544596,
+ "grad_norm": 0.14370919039693034,
+ "learning_rate": 6.6113038526491946e-06,
+ "loss": 0.4426189064979553,
+ "step": 1675
+ },
+ {
+ "epoch": 0.8793284365162645,
+ "grad_norm": 0.14933009355373217,
+ "learning_rate": 6.59755049554407e-06,
+ "loss": 0.44747287034988403,
+ "step": 1676
+ },
+ {
+ "epoch": 0.8798530954879329,
+ "grad_norm": 0.15713697891738065,
+ "learning_rate": 6.583853925715039e-06,
+ "loss": 0.46320077776908875,
+ "step": 1677
+ },
+ {
+ "epoch": 0.8803777544596013,
+ "grad_norm": 0.1542096833927892,
+ "learning_rate": 6.570214180372609e-06,
+ "loss": 0.4473888874053955,
+ "step": 1678
+ },
+ {
+ "epoch": 0.8809024134312696,
+ "grad_norm": 0.16187259968193946,
+ "learning_rate": 6.556631296572922e-06,
+ "loss": 0.4421869218349457,
+ "step": 1679
+ },
+ {
+ "epoch": 0.881427072402938,
+ "grad_norm": 0.15978032996583555,
+ "learning_rate": 6.54310531121762e-06,
+ "loss": 0.4305853843688965,
+ "step": 1680
+ },
+ {
+ "epoch": 0.8819517313746065,
+ "grad_norm": 0.15631580579775123,
+ "learning_rate": 6.529636261053784e-06,
+ "loss": 0.4340512454509735,
+ "step": 1681
+ },
+ {
+ "epoch": 0.8824763903462749,
+ "grad_norm": 0.150084161390103,
+ "learning_rate": 6.516224182673812e-06,
+ "loss": 0.4383493661880493,
+ "step": 1682
+ },
+ {
+ "epoch": 0.8830010493179433,
+ "grad_norm": 0.15527917215838966,
+ "learning_rate": 6.502869112515314e-06,
+ "loss": 0.45441877841949463,
+ "step": 1683
+ },
+ {
+ "epoch": 0.8835257082896117,
+ "grad_norm": 0.15164994594035522,
+ "learning_rate": 6.489571086861027e-06,
+ "loss": 0.43363988399505615,
+ "step": 1684
+ },
+ {
+ "epoch": 0.8840503672612802,
+ "grad_norm": 0.1467230535452683,
+ "learning_rate": 6.47633014183871e-06,
+ "loss": 0.4263266623020172,
+ "step": 1685
+ },
+ {
+ "epoch": 0.8845750262329486,
+ "grad_norm": 0.14597001769185472,
+ "learning_rate": 6.463146313421044e-06,
+ "loss": 0.4223391115665436,
+ "step": 1686
+ },
+ {
+ "epoch": 0.885099685204617,
+ "grad_norm": 0.15050446002714588,
+ "learning_rate": 6.4500196374255435e-06,
+ "loss": 0.43097931146621704,
+ "step": 1687
+ },
+ {
+ "epoch": 0.8856243441762854,
+ "grad_norm": 0.14918033614543294,
+ "learning_rate": 6.436950149514448e-06,
+ "loss": 0.4587080478668213,
+ "step": 1688
+ },
+ {
+ "epoch": 0.8861490031479539,
+ "grad_norm": 0.15469492243631364,
+ "learning_rate": 6.423937885194631e-06,
+ "loss": 0.45661866664886475,
+ "step": 1689
+ },
+ {
+ "epoch": 0.8866736621196223,
+ "grad_norm": 0.15552798640177423,
+ "learning_rate": 6.410982879817498e-06,
+ "loss": 0.44370236992836,
+ "step": 1690
+ },
+ {
+ "epoch": 0.8871983210912907,
+ "grad_norm": 0.15595593365088142,
+ "learning_rate": 6.398085168578901e-06,
+ "loss": 0.4291878342628479,
+ "step": 1691
+ },
+ {
+ "epoch": 0.887722980062959,
+ "grad_norm": 0.14708046948755318,
+ "learning_rate": 6.385244786519036e-06,
+ "loss": 0.4422011375427246,
+ "step": 1692
+ },
+ {
+ "epoch": 0.8882476390346274,
+ "grad_norm": 0.1499253382866992,
+ "learning_rate": 6.3724617685223404e-06,
+ "loss": 0.4595015048980713,
+ "step": 1693
+ },
+ {
+ "epoch": 0.8887722980062959,
+ "grad_norm": 0.14702868569252941,
+ "learning_rate": 6.3597361493174205e-06,
+ "loss": 0.4347558319568634,
+ "step": 1694
+ },
+ {
+ "epoch": 0.8892969569779643,
+ "grad_norm": 0.14977289247327386,
+ "learning_rate": 6.347067963476928e-06,
+ "loss": 0.43877455592155457,
+ "step": 1695
+ },
+ {
+ "epoch": 0.8898216159496327,
+ "grad_norm": 0.1453902077891248,
+ "learning_rate": 6.334457245417491e-06,
+ "loss": 0.4282415807247162,
+ "step": 1696
+ },
+ {
+ "epoch": 0.8903462749213011,
+ "grad_norm": 0.1545933676895574,
+ "learning_rate": 6.321904029399607e-06,
+ "loss": 0.4429236054420471,
+ "step": 1697
+ },
+ {
+ "epoch": 0.8908709338929696,
+ "grad_norm": 0.15625844051167467,
+ "learning_rate": 6.3094083495275565e-06,
+ "loss": 0.4284421503543854,
+ "step": 1698
+ },
+ {
+ "epoch": 0.891395592864638,
+ "grad_norm": 0.1484963883381567,
+ "learning_rate": 6.2969702397493e-06,
+ "loss": 0.4545463025569916,
+ "step": 1699
+ },
+ {
+ "epoch": 0.8919202518363064,
+ "grad_norm": 0.15122402806869387,
+ "learning_rate": 6.284589733856403e-06,
+ "loss": 0.43840622901916504,
+ "step": 1700
+ },
+ {
+ "epoch": 0.8924449108079748,
+ "grad_norm": 0.14889313167657334,
+ "learning_rate": 6.27226686548392e-06,
+ "loss": 0.4451153874397278,
+ "step": 1701
+ },
+ {
+ "epoch": 0.8929695697796433,
+ "grad_norm": 0.14724940941036385,
+ "learning_rate": 6.260001668110331e-06,
+ "loss": 0.434165358543396,
+ "step": 1702
+ },
+ {
+ "epoch": 0.8934942287513117,
+ "grad_norm": 0.15520838093295836,
+ "learning_rate": 6.247794175057433e-06,
+ "loss": 0.44641873240470886,
+ "step": 1703
+ },
+ {
+ "epoch": 0.8940188877229801,
+ "grad_norm": 0.159167904797364,
+ "learning_rate": 6.235644419490253e-06,
+ "loss": 0.44591957330703735,
+ "step": 1704
+ },
+ {
+ "epoch": 0.8945435466946485,
+ "grad_norm": 0.14843235881370814,
+ "learning_rate": 6.223552434416951e-06,
+ "loss": 0.43010181188583374,
+ "step": 1705
+ },
+ {
+ "epoch": 0.8950682056663168,
+ "grad_norm": 0.1590998153483183,
+ "learning_rate": 6.211518252688749e-06,
+ "loss": 0.4602118134498596,
+ "step": 1706
+ },
+ {
+ "epoch": 0.8955928646379853,
+ "grad_norm": 0.15257025539137767,
+ "learning_rate": 6.19954190699982e-06,
+ "loss": 0.45429927110671997,
+ "step": 1707
+ },
+ {
+ "epoch": 0.8961175236096537,
+ "grad_norm": 0.15304035006408193,
+ "learning_rate": 6.18762342988722e-06,
+ "loss": 0.4405020773410797,
+ "step": 1708
+ },
+ {
+ "epoch": 0.8966421825813221,
+ "grad_norm": 0.15956303033928781,
+ "learning_rate": 6.175762853730777e-06,
+ "loss": 0.4455980062484741,
+ "step": 1709
+ },
+ {
+ "epoch": 0.8971668415529905,
+ "grad_norm": 0.14973831573751925,
+ "learning_rate": 6.163960210753025e-06,
+ "loss": 0.4331505298614502,
+ "step": 1710
+ },
+ {
+ "epoch": 0.897691500524659,
+ "grad_norm": 0.14573687980573624,
+ "learning_rate": 6.1522155330190985e-06,
+ "loss": 0.43307584524154663,
+ "step": 1711
+ },
+ {
+ "epoch": 0.8982161594963274,
+ "grad_norm": 0.1556523345686865,
+ "learning_rate": 6.140528852436665e-06,
+ "loss": 0.4531600773334503,
+ "step": 1712
+ },
+ {
+ "epoch": 0.8987408184679958,
+ "grad_norm": 0.14262442470666953,
+ "learning_rate": 6.128900200755817e-06,
+ "loss": 0.4350634813308716,
+ "step": 1713
+ },
+ {
+ "epoch": 0.8992654774396642,
+ "grad_norm": 0.14709392307082428,
+ "learning_rate": 6.117329609568998e-06,
+ "loss": 0.43465620279312134,
+ "step": 1714
+ },
+ {
+ "epoch": 0.8997901364113327,
+ "grad_norm": 0.15648040137428026,
+ "learning_rate": 6.105817110310909e-06,
+ "loss": 0.4344560503959656,
+ "step": 1715
+ },
+ {
+ "epoch": 0.9003147953830011,
+ "grad_norm": 0.15758204487742147,
+ "learning_rate": 6.094362734258438e-06,
+ "loss": 0.4409951865673065,
+ "step": 1716
+ },
+ {
+ "epoch": 0.9008394543546695,
+ "grad_norm": 0.15755547382189733,
+ "learning_rate": 6.082966512530558e-06,
+ "loss": 0.44050097465515137,
+ "step": 1717
+ },
+ {
+ "epoch": 0.9013641133263379,
+ "grad_norm": 0.1536165045521179,
+ "learning_rate": 6.071628476088257e-06,
+ "loss": 0.43460220098495483,
+ "step": 1718
+ },
+ {
+ "epoch": 0.9018887722980063,
+ "grad_norm": 0.16253540769141986,
+ "learning_rate": 6.060348655734439e-06,
+ "loss": 0.4490512013435364,
+ "step": 1719
+ },
+ {
+ "epoch": 0.9024134312696747,
+ "grad_norm": 0.15003242262029914,
+ "learning_rate": 6.049127082113853e-06,
+ "loss": 0.44427937269210815,
+ "step": 1720
+ },
+ {
+ "epoch": 0.9029380902413431,
+ "grad_norm": 0.1579908758627471,
+ "learning_rate": 6.037963785712995e-06,
+ "loss": 0.4421137571334839,
+ "step": 1721
+ },
+ {
+ "epoch": 0.9034627492130115,
+ "grad_norm": 0.15111957361250286,
+ "learning_rate": 6.026858796860048e-06,
+ "loss": 0.44011324644088745,
+ "step": 1722
+ },
+ {
+ "epoch": 0.9039874081846799,
+ "grad_norm": 0.1446470746259905,
+ "learning_rate": 6.0158121457247845e-06,
+ "loss": 0.43798986077308655,
+ "step": 1723
+ },
+ {
+ "epoch": 0.9045120671563484,
+ "grad_norm": 0.14882152623490427,
+ "learning_rate": 6.004823862318474e-06,
+ "loss": 0.4265901446342468,
+ "step": 1724
+ },
+ {
+ "epoch": 0.9050367261280168,
+ "grad_norm": 0.1506889859372045,
+ "learning_rate": 5.993893976493831e-06,
+ "loss": 0.4503232538700104,
+ "step": 1725
+ },
+ {
+ "epoch": 0.9055613850996852,
+ "grad_norm": 0.16757194686045146,
+ "learning_rate": 5.983022517944901e-06,
+ "loss": 0.4509217441082001,
+ "step": 1726
+ },
+ {
+ "epoch": 0.9060860440713536,
+ "grad_norm": 0.14447292841509515,
+ "learning_rate": 5.972209516207007e-06,
+ "loss": 0.4476190507411957,
+ "step": 1727
+ },
+ {
+ "epoch": 0.906610703043022,
+ "grad_norm": 0.1536564780391058,
+ "learning_rate": 5.961455000656656e-06,
+ "loss": 0.45448166131973267,
+ "step": 1728
+ },
+ {
+ "epoch": 0.9071353620146905,
+ "grad_norm": 0.15436323862710438,
+ "learning_rate": 5.950759000511455e-06,
+ "loss": 0.45128491520881653,
+ "step": 1729
+ },
+ {
+ "epoch": 0.9076600209863589,
+ "grad_norm": 0.15379817460186834,
+ "learning_rate": 5.940121544830047e-06,
+ "loss": 0.4490806460380554,
+ "step": 1730
+ },
+ {
+ "epoch": 0.9081846799580273,
+ "grad_norm": 0.1470796082413195,
+ "learning_rate": 5.929542662512013e-06,
+ "loss": 0.4371582865715027,
+ "step": 1731
+ },
+ {
+ "epoch": 0.9087093389296957,
+ "grad_norm": 0.15107600076433048,
+ "learning_rate": 5.919022382297809e-06,
+ "loss": 0.45175206661224365,
+ "step": 1732
+ },
+ {
+ "epoch": 0.9092339979013642,
+ "grad_norm": 0.14972329317276084,
+ "learning_rate": 5.908560732768684e-06,
+ "loss": 0.43177497386932373,
+ "step": 1733
+ },
+ {
+ "epoch": 0.9097586568730325,
+ "grad_norm": 0.14437715003340176,
+ "learning_rate": 5.898157742346601e-06,
+ "loss": 0.4243622422218323,
+ "step": 1734
+ },
+ {
+ "epoch": 0.9102833158447009,
+ "grad_norm": 0.15200146823831148,
+ "learning_rate": 5.8878134392941465e-06,
+ "loss": 0.47819408774375916,
+ "step": 1735
+ },
+ {
+ "epoch": 0.9108079748163693,
+ "grad_norm": 0.156864616244505,
+ "learning_rate": 5.877527851714479e-06,
+ "loss": 0.44744759798049927,
+ "step": 1736
+ },
+ {
+ "epoch": 0.9113326337880377,
+ "grad_norm": 0.14489667207174767,
+ "learning_rate": 5.867301007551239e-06,
+ "loss": 0.4390789866447449,
+ "step": 1737
+ },
+ {
+ "epoch": 0.9118572927597062,
+ "grad_norm": 0.1528011587102748,
+ "learning_rate": 5.857132934588479e-06,
+ "loss": 0.4338226616382599,
+ "step": 1738
+ },
+ {
+ "epoch": 0.9123819517313746,
+ "grad_norm": 0.14716603285971375,
+ "learning_rate": 5.847023660450568e-06,
+ "loss": 0.43975532054901123,
+ "step": 1739
+ },
+ {
+ "epoch": 0.912906610703043,
+ "grad_norm": 0.15266609903803277,
+ "learning_rate": 5.836973212602143e-06,
+ "loss": 0.449812650680542,
+ "step": 1740
+ },
+ {
+ "epoch": 0.9134312696747114,
+ "grad_norm": 0.1466770441701569,
+ "learning_rate": 5.826981618348025e-06,
+ "loss": 0.43771812319755554,
+ "step": 1741
+ },
+ {
+ "epoch": 0.9139559286463799,
+ "grad_norm": 0.1428112225146131,
+ "learning_rate": 5.8170489048331305e-06,
+ "loss": 0.42403268814086914,
+ "step": 1742
+ },
+ {
+ "epoch": 0.9144805876180483,
+ "grad_norm": 0.1550829416480759,
+ "learning_rate": 5.8071750990424275e-06,
+ "loss": 0.45707154273986816,
+ "step": 1743
+ },
+ {
+ "epoch": 0.9150052465897167,
+ "grad_norm": 0.14687341315819044,
+ "learning_rate": 5.797360227800832e-06,
+ "loss": 0.4558846354484558,
+ "step": 1744
+ },
+ {
+ "epoch": 0.9155299055613851,
+ "grad_norm": 0.15660848922413237,
+ "learning_rate": 5.787604317773154e-06,
+ "loss": 0.44760435819625854,
+ "step": 1745
+ },
+ {
+ "epoch": 0.9160545645330536,
+ "grad_norm": 0.14891742746821762,
+ "learning_rate": 5.777907395464015e-06,
+ "loss": 0.4266620874404907,
+ "step": 1746
+ },
+ {
+ "epoch": 0.916579223504722,
+ "grad_norm": 0.14501110957005472,
+ "learning_rate": 5.768269487217777e-06,
+ "loss": 0.4388393759727478,
+ "step": 1747
+ },
+ {
+ "epoch": 0.9171038824763903,
+ "grad_norm": 0.14651438673799927,
+ "learning_rate": 5.758690619218493e-06,
+ "loss": 0.4416884183883667,
+ "step": 1748
+ },
+ {
+ "epoch": 0.9176285414480587,
+ "grad_norm": 0.14785104880890787,
+ "learning_rate": 5.749170817489792e-06,
+ "loss": 0.4368356466293335,
+ "step": 1749
+ },
+ {
+ "epoch": 0.9181532004197271,
+ "grad_norm": 0.16586567883267808,
+ "learning_rate": 5.739710107894847e-06,
+ "loss": 0.44294649362564087,
+ "step": 1750
+ },
+ {
+ "epoch": 0.9186778593913956,
+ "grad_norm": 0.14377347738227095,
+ "learning_rate": 5.730308516136288e-06,
+ "loss": 0.4405091106891632,
+ "step": 1751
+ },
+ {
+ "epoch": 0.919202518363064,
+ "grad_norm": 0.14703077660050376,
+ "learning_rate": 5.72096606775613e-06,
+ "loss": 0.44809678196907043,
+ "step": 1752
+ },
+ {
+ "epoch": 0.9197271773347324,
+ "grad_norm": 0.15073590419140934,
+ "learning_rate": 5.711682788135726e-06,
+ "loss": 0.43815505504608154,
+ "step": 1753
+ },
+ {
+ "epoch": 0.9202518363064008,
+ "grad_norm": 0.14774775313473323,
+ "learning_rate": 5.702458702495658e-06,
+ "loss": 0.4417206048965454,
+ "step": 1754
+ },
+ {
+ "epoch": 0.9207764952780693,
+ "grad_norm": 0.14289335805130654,
+ "learning_rate": 5.693293835895706e-06,
+ "loss": 0.42886218428611755,
+ "step": 1755
+ },
+ {
+ "epoch": 0.9213011542497377,
+ "grad_norm": 0.15571224276102627,
+ "learning_rate": 5.684188213234763e-06,
+ "loss": 0.44488853216171265,
+ "step": 1756
+ },
+ {
+ "epoch": 0.9218258132214061,
+ "grad_norm": 0.147115948090654,
+ "learning_rate": 5.675141859250763e-06,
+ "loss": 0.44682830572128296,
+ "step": 1757
+ },
+ {
+ "epoch": 0.9223504721930745,
+ "grad_norm": 0.15349533122985531,
+ "learning_rate": 5.666154798520627e-06,
+ "loss": 0.45550864934921265,
+ "step": 1758
+ },
+ {
+ "epoch": 0.922875131164743,
+ "grad_norm": 0.16960595806748266,
+ "learning_rate": 5.657227055460184e-06,
+ "loss": 0.4605559706687927,
+ "step": 1759
+ },
+ {
+ "epoch": 0.9233997901364114,
+ "grad_norm": 0.14665490922506455,
+ "learning_rate": 5.648358654324119e-06,
+ "loss": 0.45442628860473633,
+ "step": 1760
+ },
+ {
+ "epoch": 0.9239244491080797,
+ "grad_norm": 0.14780665809126886,
+ "learning_rate": 5.6395496192058905e-06,
+ "loss": 0.45085081458091736,
+ "step": 1761
+ },
+ {
+ "epoch": 0.9244491080797481,
+ "grad_norm": 0.15599036700169416,
+ "learning_rate": 5.6307999740376714e-06,
+ "loss": 0.4581005573272705,
+ "step": 1762
+ },
+ {
+ "epoch": 0.9249737670514165,
+ "grad_norm": 0.14383027603716222,
+ "learning_rate": 5.622109742590294e-06,
+ "loss": 0.43986010551452637,
+ "step": 1763
+ },
+ {
+ "epoch": 0.925498426023085,
+ "grad_norm": 0.1510369649046927,
+ "learning_rate": 5.613478948473166e-06,
+ "loss": 0.4317719340324402,
+ "step": 1764
+ },
+ {
+ "epoch": 0.9260230849947534,
+ "grad_norm": 0.16608560055258376,
+ "learning_rate": 5.60490761513423e-06,
+ "loss": 0.4500439167022705,
+ "step": 1765
+ },
+ {
+ "epoch": 0.9265477439664218,
+ "grad_norm": 0.15364415343465948,
+ "learning_rate": 5.596395765859875e-06,
+ "loss": 0.43703922629356384,
+ "step": 1766
+ },
+ {
+ "epoch": 0.9270724029380902,
+ "grad_norm": 0.145164767782082,
+ "learning_rate": 5.587943423774886e-06,
+ "loss": 0.4389477074146271,
+ "step": 1767
+ },
+ {
+ "epoch": 0.9275970619097587,
+ "grad_norm": 0.15044099308887526,
+ "learning_rate": 5.579550611842395e-06,
+ "loss": 0.43600398302078247,
+ "step": 1768
+ },
+ {
+ "epoch": 0.9281217208814271,
+ "grad_norm": 0.14694381874423756,
+ "learning_rate": 5.571217352863786e-06,
+ "loss": 0.4525601863861084,
+ "step": 1769
+ },
+ {
+ "epoch": 0.9286463798530955,
+ "grad_norm": 0.15979616589190854,
+ "learning_rate": 5.562943669478661e-06,
+ "loss": 0.44730353355407715,
+ "step": 1770
+ },
+ {
+ "epoch": 0.9291710388247639,
+ "grad_norm": 0.1545596231710442,
+ "learning_rate": 5.554729584164762e-06,
+ "loss": 0.43867969512939453,
+ "step": 1771
+ },
+ {
+ "epoch": 0.9296956977964324,
+ "grad_norm": 0.15923930406000852,
+ "learning_rate": 5.5465751192379235e-06,
+ "loss": 0.4648996889591217,
+ "step": 1772
+ },
+ {
+ "epoch": 0.9302203567681008,
+ "grad_norm": 0.14001241391446362,
+ "learning_rate": 5.538480296852001e-06,
+ "loss": 0.41449663043022156,
+ "step": 1773
+ },
+ {
+ "epoch": 0.9307450157397692,
+ "grad_norm": 0.16538765001279418,
+ "learning_rate": 5.530445138998815e-06,
+ "loss": 0.4503074288368225,
+ "step": 1774
+ },
+ {
+ "epoch": 0.9312696747114375,
+ "grad_norm": 0.15007886784403335,
+ "learning_rate": 5.5224696675080904e-06,
+ "loss": 0.44717323780059814,
+ "step": 1775
+ },
+ {
+ "epoch": 0.9317943336831059,
+ "grad_norm": 0.17834856046133146,
+ "learning_rate": 5.514553904047395e-06,
+ "loss": 0.4374881386756897,
+ "step": 1776
+ },
+ {
+ "epoch": 0.9323189926547744,
+ "grad_norm": 0.1494478926649096,
+ "learning_rate": 5.5066978701220865e-06,
+ "loss": 0.4464038610458374,
+ "step": 1777
+ },
+ {
+ "epoch": 0.9328436516264428,
+ "grad_norm": 0.14887448793632826,
+ "learning_rate": 5.498901587075253e-06,
+ "loss": 0.443928062915802,
+ "step": 1778
+ },
+ {
+ "epoch": 0.9333683105981112,
+ "grad_norm": 0.14192884484464632,
+ "learning_rate": 5.491165076087652e-06,
+ "loss": 0.42002278566360474,
+ "step": 1779
+ },
+ {
+ "epoch": 0.9338929695697796,
+ "grad_norm": 0.14783356724791524,
+ "learning_rate": 5.483488358177647e-06,
+ "loss": 0.43647563457489014,
+ "step": 1780
+ },
+ {
+ "epoch": 0.934417628541448,
+ "grad_norm": 0.15004597148119703,
+ "learning_rate": 5.475871454201163e-06,
+ "loss": 0.4396110773086548,
+ "step": 1781
+ },
+ {
+ "epoch": 0.9349422875131165,
+ "grad_norm": 0.15668918073856763,
+ "learning_rate": 5.468314384851626e-06,
+ "loss": 0.42848116159439087,
+ "step": 1782
+ },
+ {
+ "epoch": 0.9354669464847849,
+ "grad_norm": 0.1449937199040528,
+ "learning_rate": 5.460817170659901e-06,
+ "loss": 0.44262629747390747,
+ "step": 1783
+ },
+ {
+ "epoch": 0.9359916054564533,
+ "grad_norm": 0.15868260486048544,
+ "learning_rate": 5.4533798319942375e-06,
+ "loss": 0.44874119758605957,
+ "step": 1784
+ },
+ {
+ "epoch": 0.9365162644281217,
+ "grad_norm": 0.15004870783258292,
+ "learning_rate": 5.446002389060224e-06,
+ "loss": 0.4575575590133667,
+ "step": 1785
+ },
+ {
+ "epoch": 0.9370409233997902,
+ "grad_norm": 0.14830248639010987,
+ "learning_rate": 5.438684861900721e-06,
+ "loss": 0.44039857387542725,
+ "step": 1786
+ },
+ {
+ "epoch": 0.9375655823714586,
+ "grad_norm": 0.15429570419507257,
+ "learning_rate": 5.431427270395808e-06,
+ "loss": 0.43363744020462036,
+ "step": 1787
+ },
+ {
+ "epoch": 0.938090241343127,
+ "grad_norm": 0.15982575395123055,
+ "learning_rate": 5.4242296342627354e-06,
+ "loss": 0.4484976530075073,
+ "step": 1788
+ },
+ {
+ "epoch": 0.9386149003147953,
+ "grad_norm": 0.1466121447285136,
+ "learning_rate": 5.417091973055878e-06,
+ "loss": 0.44059592485427856,
+ "step": 1789
+ },
+ {
+ "epoch": 0.9391395592864638,
+ "grad_norm": 0.15353848724097463,
+ "learning_rate": 5.410014306166657e-06,
+ "loss": 0.4600589871406555,
+ "step": 1790
+ },
+ {
+ "epoch": 0.9396642182581322,
+ "grad_norm": 0.14968071650065867,
+ "learning_rate": 5.4029966528235065e-06,
+ "loss": 0.4415352940559387,
+ "step": 1791
+ },
+ {
+ "epoch": 0.9401888772298006,
+ "grad_norm": 0.14819069534768417,
+ "learning_rate": 5.396039032091826e-06,
+ "loss": 0.4436836242675781,
+ "step": 1792
+ },
+ {
+ "epoch": 0.940713536201469,
+ "grad_norm": 0.14683654708689367,
+ "learning_rate": 5.389141462873908e-06,
+ "loss": 0.4243587851524353,
+ "step": 1793
+ },
+ {
+ "epoch": 0.9412381951731374,
+ "grad_norm": 0.15431293103561666,
+ "learning_rate": 5.382303963908908e-06,
+ "loss": 0.450438916683197,
+ "step": 1794
+ },
+ {
+ "epoch": 0.9417628541448059,
+ "grad_norm": 0.15125346717879914,
+ "learning_rate": 5.3755265537727775e-06,
+ "loss": 0.4469110369682312,
+ "step": 1795
+ },
+ {
+ "epoch": 0.9422875131164743,
+ "grad_norm": 0.15430864502191044,
+ "learning_rate": 5.368809250878227e-06,
+ "loss": 0.45182669162750244,
+ "step": 1796
+ },
+ {
+ "epoch": 0.9428121720881427,
+ "grad_norm": 0.1515431753176237,
+ "learning_rate": 5.362152073474657e-06,
+ "loss": 0.43683797121047974,
+ "step": 1797
+ },
+ {
+ "epoch": 0.9433368310598111,
+ "grad_norm": 0.1524915522484605,
+ "learning_rate": 5.355555039648133e-06,
+ "loss": 0.44641250371932983,
+ "step": 1798
+ },
+ {
+ "epoch": 0.9438614900314796,
+ "grad_norm": 0.15722684273308743,
+ "learning_rate": 5.3490181673213206e-06,
+ "loss": 0.43558835983276367,
+ "step": 1799
+ },
+ {
+ "epoch": 0.944386149003148,
+ "grad_norm": 0.15527272713567492,
+ "learning_rate": 5.342541474253438e-06,
+ "loss": 0.4423788785934448,
+ "step": 1800
+ },
+ {
+ "epoch": 0.9449108079748164,
+ "grad_norm": 0.1578831929456481,
+ "learning_rate": 5.336124978040208e-06,
+ "loss": 0.4506484866142273,
+ "step": 1801
+ },
+ {
+ "epoch": 0.9454354669464848,
+ "grad_norm": 0.14812076004561775,
+ "learning_rate": 5.329768696113815e-06,
+ "loss": 0.4432883858680725,
+ "step": 1802
+ },
+ {
+ "epoch": 0.9459601259181531,
+ "grad_norm": 0.1511327211030653,
+ "learning_rate": 5.323472645742854e-06,
+ "loss": 0.4321002960205078,
+ "step": 1803
+ },
+ {
+ "epoch": 0.9464847848898216,
+ "grad_norm": 0.1531668263170348,
+ "learning_rate": 5.3172368440322854e-06,
+ "loss": 0.44666337966918945,
+ "step": 1804
+ },
+ {
+ "epoch": 0.94700944386149,
+ "grad_norm": 0.14571994558609092,
+ "learning_rate": 5.311061307923387e-06,
+ "loss": 0.4436221420764923,
+ "step": 1805
+ },
+ {
+ "epoch": 0.9475341028331584,
+ "grad_norm": 0.15158206242691336,
+ "learning_rate": 5.304946054193706e-06,
+ "loss": 0.45979544520378113,
+ "step": 1806
+ },
+ {
+ "epoch": 0.9480587618048268,
+ "grad_norm": 0.15374308675707662,
+ "learning_rate": 5.298891099457018e-06,
+ "loss": 0.44095706939697266,
+ "step": 1807
+ },
+ {
+ "epoch": 0.9485834207764953,
+ "grad_norm": 0.14777162806829072,
+ "learning_rate": 5.292896460163283e-06,
+ "loss": 0.44000717997550964,
+ "step": 1808
+ },
+ {
+ "epoch": 0.9491080797481637,
+ "grad_norm": 0.15416014526696098,
+ "learning_rate": 5.286962152598589e-06,
+ "loss": 0.4367929697036743,
+ "step": 1809
+ },
+ {
+ "epoch": 0.9496327387198321,
+ "grad_norm": 0.1586289813390648,
+ "learning_rate": 5.281088192885125e-06,
+ "loss": 0.44338831305503845,
+ "step": 1810
+ },
+ {
+ "epoch": 0.9501573976915005,
+ "grad_norm": 0.1480189398685719,
+ "learning_rate": 5.275274596981124e-06,
+ "loss": 0.44127213954925537,
+ "step": 1811
+ },
+ {
+ "epoch": 0.950682056663169,
+ "grad_norm": 0.15893037852856393,
+ "learning_rate": 5.269521380680823e-06,
+ "loss": 0.43447786569595337,
+ "step": 1812
+ },
+ {
+ "epoch": 0.9512067156348374,
+ "grad_norm": 0.15257675815194532,
+ "learning_rate": 5.2638285596144236e-06,
+ "loss": 0.4478892385959625,
+ "step": 1813
+ },
+ {
+ "epoch": 0.9517313746065058,
+ "grad_norm": 0.14985387736695718,
+ "learning_rate": 5.258196149248046e-06,
+ "loss": 0.4517829418182373,
+ "step": 1814
+ },
+ {
+ "epoch": 0.9522560335781742,
+ "grad_norm": 0.14441329680781126,
+ "learning_rate": 5.252624164883689e-06,
+ "loss": 0.4337430000305176,
+ "step": 1815
+ },
+ {
+ "epoch": 0.9527806925498427,
+ "grad_norm": 0.15113217761739892,
+ "learning_rate": 5.247112621659186e-06,
+ "loss": 0.44544464349746704,
+ "step": 1816
+ },
+ {
+ "epoch": 0.953305351521511,
+ "grad_norm": 0.15511548841769227,
+ "learning_rate": 5.241661534548169e-06,
+ "loss": 0.4420471787452698,
+ "step": 1817
+ },
+ {
+ "epoch": 0.9538300104931794,
+ "grad_norm": 0.141474135090229,
+ "learning_rate": 5.236270918360016e-06,
+ "loss": 0.4323197603225708,
+ "step": 1818
+ },
+ {
+ "epoch": 0.9543546694648478,
+ "grad_norm": 0.14741159689952743,
+ "learning_rate": 5.23094078773983e-06,
+ "loss": 0.4486733675003052,
+ "step": 1819
+ },
+ {
+ "epoch": 0.9548793284365162,
+ "grad_norm": 0.14938242936939558,
+ "learning_rate": 5.225671157168377e-06,
+ "loss": 0.4469614624977112,
+ "step": 1820
+ },
+ {
+ "epoch": 0.9554039874081847,
+ "grad_norm": 0.14960446458046617,
+ "learning_rate": 5.220462040962068e-06,
+ "loss": 0.4371514320373535,
+ "step": 1821
+ },
+ {
+ "epoch": 0.9559286463798531,
+ "grad_norm": 0.1453826830908917,
+ "learning_rate": 5.215313453272906e-06,
+ "loss": 0.43045082688331604,
+ "step": 1822
+ },
+ {
+ "epoch": 0.9564533053515215,
+ "grad_norm": 0.14561549586732817,
+ "learning_rate": 5.210225408088447e-06,
+ "loss": 0.44164514541625977,
+ "step": 1823
+ },
+ {
+ "epoch": 0.9569779643231899,
+ "grad_norm": 0.15381074610209994,
+ "learning_rate": 5.205197919231773e-06,
+ "loss": 0.45496129989624023,
+ "step": 1824
+ },
+ {
+ "epoch": 0.9575026232948584,
+ "grad_norm": 0.14729572792188594,
+ "learning_rate": 5.200231000361448e-06,
+ "loss": 0.4330272078514099,
+ "step": 1825
+ },
+ {
+ "epoch": 0.9580272822665268,
+ "grad_norm": 0.1549156051160634,
+ "learning_rate": 5.195324664971476e-06,
+ "loss": 0.44893333315849304,
+ "step": 1826
+ },
+ {
+ "epoch": 0.9585519412381952,
+ "grad_norm": 0.1544667437543616,
+ "learning_rate": 5.190478926391271e-06,
+ "loss": 0.4463026821613312,
+ "step": 1827
+ },
+ {
+ "epoch": 0.9590766002098636,
+ "grad_norm": 0.1482681888796153,
+ "learning_rate": 5.1856937977856235e-06,
+ "loss": 0.42915546894073486,
+ "step": 1828
+ },
+ {
+ "epoch": 0.959601259181532,
+ "grad_norm": 0.15393225997035367,
+ "learning_rate": 5.180969292154653e-06,
+ "loss": 0.45655858516693115,
+ "step": 1829
+ },
+ {
+ "epoch": 0.9601259181532005,
+ "grad_norm": 0.15402682858559275,
+ "learning_rate": 5.176305422333788e-06,
+ "loss": 0.45586466789245605,
+ "step": 1830
+ },
+ {
+ "epoch": 0.9606505771248688,
+ "grad_norm": 0.15136206757704734,
+ "learning_rate": 5.171702200993716e-06,
+ "loss": 0.4525001049041748,
+ "step": 1831
+ },
+ {
+ "epoch": 0.9611752360965372,
+ "grad_norm": 0.15443301434325082,
+ "learning_rate": 5.167159640640357e-06,
+ "loss": 0.4500652551651001,
+ "step": 1832
+ },
+ {
+ "epoch": 0.9616998950682056,
+ "grad_norm": 0.1553996888683437,
+ "learning_rate": 5.162677753614834e-06,
+ "loss": 0.4426611363887787,
+ "step": 1833
+ },
+ {
+ "epoch": 0.9622245540398741,
+ "grad_norm": 0.14719352293992807,
+ "learning_rate": 5.158256552093428e-06,
+ "loss": 0.43979009985923767,
+ "step": 1834
+ },
+ {
+ "epoch": 0.9627492130115425,
+ "grad_norm": 0.14547547543617104,
+ "learning_rate": 5.153896048087557e-06,
+ "loss": 0.443497896194458,
+ "step": 1835
+ },
+ {
+ "epoch": 0.9632738719832109,
+ "grad_norm": 0.15665555976593595,
+ "learning_rate": 5.149596253443734e-06,
+ "loss": 0.441650390625,
+ "step": 1836
+ },
+ {
+ "epoch": 0.9637985309548793,
+ "grad_norm": 0.1552107125031637,
+ "learning_rate": 5.145357179843536e-06,
+ "loss": 0.44342389702796936,
+ "step": 1837
+ },
+ {
+ "epoch": 0.9643231899265478,
+ "grad_norm": 0.1450717817395646,
+ "learning_rate": 5.141178838803574e-06,
+ "loss": 0.4459199011325836,
+ "step": 1838
+ },
+ {
+ "epoch": 0.9648478488982162,
+ "grad_norm": 0.15194030785499474,
+ "learning_rate": 5.137061241675472e-06,
+ "loss": 0.43879520893096924,
+ "step": 1839
+ },
+ {
+ "epoch": 0.9653725078698846,
+ "grad_norm": 0.15876627179026034,
+ "learning_rate": 5.133004399645817e-06,
+ "loss": 0.4629112184047699,
+ "step": 1840
+ },
+ {
+ "epoch": 0.965897166841553,
+ "grad_norm": 0.15025078854799365,
+ "learning_rate": 5.129008323736142e-06,
+ "loss": 0.4596017599105835,
+ "step": 1841
+ },
+ {
+ "epoch": 0.9664218258132214,
+ "grad_norm": 0.15108123279475658,
+ "learning_rate": 5.125073024802887e-06,
+ "loss": 0.4394477605819702,
+ "step": 1842
+ },
+ {
+ "epoch": 0.9669464847848899,
+ "grad_norm": 0.15040459550862542,
+ "learning_rate": 5.121198513537382e-06,
+ "loss": 0.44406938552856445,
+ "step": 1843
+ },
+ {
+ "epoch": 0.9674711437565582,
+ "grad_norm": 0.14829125607751426,
+ "learning_rate": 5.1173848004658065e-06,
+ "loss": 0.4245413541793823,
+ "step": 1844
+ },
+ {
+ "epoch": 0.9679958027282266,
+ "grad_norm": 0.15182555299407333,
+ "learning_rate": 5.11363189594916e-06,
+ "loss": 0.44984906911849976,
+ "step": 1845
+ },
+ {
+ "epoch": 0.968520461699895,
+ "grad_norm": 0.1455263692654042,
+ "learning_rate": 5.109939810183251e-06,
+ "loss": 0.41511911153793335,
+ "step": 1846
+ },
+ {
+ "epoch": 0.9690451206715635,
+ "grad_norm": 0.14474034020024001,
+ "learning_rate": 5.106308553198644e-06,
+ "loss": 0.42551544308662415,
+ "step": 1847
+ },
+ {
+ "epoch": 0.9695697796432319,
+ "grad_norm": 0.14650897119054226,
+ "learning_rate": 5.102738134860656e-06,
+ "loss": 0.4347913861274719,
+ "step": 1848
+ },
+ {
+ "epoch": 0.9700944386149003,
+ "grad_norm": 0.15017963706984508,
+ "learning_rate": 5.099228564869313e-06,
+ "loss": 0.442004919052124,
+ "step": 1849
+ },
+ {
+ "epoch": 0.9706190975865687,
+ "grad_norm": 0.14854765100160125,
+ "learning_rate": 5.09577985275933e-06,
+ "loss": 0.4416872262954712,
+ "step": 1850
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 4.899408023283302e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1850/video_preprocessor_config.json b/checkpoints/checkpoint-1850/video_preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..a933f0ec6941222c33aeba7980bd5acf8dd98e91
--- /dev/null
+++ b/checkpoints/checkpoint-1850/video_preprocessor_config.json
@@ -0,0 +1,30 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "do_sample_frames": true,
+ "fps": 2,
+ "image_mean": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "image_std": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "max_frames": 768,
+ "merge_size": 2,
+ "min_frames": 4,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 786432,
+ "shortest_edge": 131072
+ },
+ "temporal_patch_size": 2,
+ "video_processor_type": "Qwen3VLVideoProcessor"
+}
diff --git a/checkpoints/checkpoint-1900/chat_template.jinja b/checkpoints/checkpoint-1900/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1900/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1900/config.json b/checkpoints/checkpoint-1900/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1900/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1900/expert_prune_metadata.json b/checkpoints/checkpoint-1900/expert_prune_metadata.json
new file mode 100644
index 0000000000000000000000000000000000000000..16e01c8a10636331e334732fa9629579c68df44e
--- /dev/null
+++ b/checkpoints/checkpoint-1900/expert_prune_metadata.json
@@ -0,0 +1,14 @@
+{
+ "source_model": "/mnt/data/wuwei/model_list/Qwen3.6-35B-A3B/models--Qwen--Qwen3.6-35B-A3B/snapshots/995ad96eacd98c81ed38be0c5b274b04031597b0",
+ "output_model": "/mnt/data/wuwei/model_list/qwen3-19b-a3b",
+ "total_parameters": 18990568816,
+ "parameter_billions_decimal": 18.990568816,
+ "directory_size_label": "19B",
+ "original_num_experts": 256,
+ "kept_experts_per_layer": 128,
+ "num_profiled_layers": 40,
+ "mtp_weights": "removed because Qwen3_5MoeForConditionalGeneration does not instantiate or use the source checkpoint's mtp.* training-only tensors",
+ "expert_order": "ascending original expert index; router rows and expert tensors use the same remapping",
+ "router_statistics": "router_stats.json and router_stats.pt",
+ "selection_score": "top-8 router hit frequency * mean(abs(actual expert output after down_proj and before routing-weight multiplication))"
+}
diff --git a/checkpoints/checkpoint-1900/generation_config.json b/checkpoints/checkpoint-1900/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1900/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1900/model.safetensors b/checkpoints/checkpoint-1900/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..a8f602690e4c070f35fed8c70fd6c4e63fb96bfe
--- /dev/null
+++ b/checkpoints/checkpoint-1900/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ff577a39e88f9ec9891f57a7f7887842a024b6df6e5d4307c3f8c2ea37b3f79
+size 37983421000
diff --git a/checkpoints/checkpoint-1900/preprocessor_config.json b/checkpoints/checkpoint-1900/preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ec43d1f859ecadf800ebae92fda4a812d34975d3
--- /dev/null
+++ b/checkpoints/checkpoint-1900/preprocessor_config.json
@@ -0,0 +1,26 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "image_mean": [
+ 0.48145466,
+ 0.4578275,
+ 0.40821073
+ ],
+ "image_processor_type": "Qwen2VLImageProcessorFast",
+ "image_std": [
+ 0.26862954,
+ 0.26130258,
+ 0.27577711
+ ],
+ "merge_size": 2,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 1003520,
+ "shortest_edge": 3136
+ },
+ "temporal_patch_size": 2
+}
diff --git a/checkpoints/checkpoint-1900/tokenizer.json b/checkpoints/checkpoint-1900/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1900/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1900/tokenizer_config.json b/checkpoints/checkpoint-1900/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1900/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1900/trainer_state.json b/checkpoints/checkpoint-1900/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..b207e76626b7b6c334354714ef412a38dcf2dfc6
--- /dev/null
+++ b/checkpoints/checkpoint-1900/trainer_state.json
@@ -0,0 +1,13334 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.9968520461699895,
+ "eval_steps": 500,
+ "global_step": 1900,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ },
+ {
+ "epoch": 0.5776495278069255,
+ "grad_norm": 0.1862053032854098,
+ "learning_rate": 2.2065553964033126e-05,
+ "loss": 0.4816875457763672,
+ "step": 1101
+ },
+ {
+ "epoch": 0.5781741867785939,
+ "grad_norm": 0.16996015894363117,
+ "learning_rate": 2.20295734045532e-05,
+ "loss": 0.4471197724342346,
+ "step": 1102
+ },
+ {
+ "epoch": 0.5786988457502623,
+ "grad_norm": 0.16889227008390692,
+ "learning_rate": 2.1993607706997725e-05,
+ "loss": 0.468391478061676,
+ "step": 1103
+ },
+ {
+ "epoch": 0.5792235047219307,
+ "grad_norm": 0.16670931681562373,
+ "learning_rate": 2.195765696907745e-05,
+ "loss": 0.44934162497520447,
+ "step": 1104
+ },
+ {
+ "epoch": 0.5797481636935992,
+ "grad_norm": 0.16328418741223716,
+ "learning_rate": 2.1921721288462482e-05,
+ "loss": 0.4558270275592804,
+ "step": 1105
+ },
+ {
+ "epoch": 0.5802728226652676,
+ "grad_norm": 0.17364660408703936,
+ "learning_rate": 2.188580076278203e-05,
+ "loss": 0.4597751796245575,
+ "step": 1106
+ },
+ {
+ "epoch": 0.580797481636936,
+ "grad_norm": 0.1628618453800601,
+ "learning_rate": 2.18498954896241e-05,
+ "loss": 0.4521859288215637,
+ "step": 1107
+ },
+ {
+ "epoch": 0.5813221406086044,
+ "grad_norm": 0.16813817951871482,
+ "learning_rate": 2.1814005566535302e-05,
+ "loss": 0.4581005573272705,
+ "step": 1108
+ },
+ {
+ "epoch": 0.5818467995802729,
+ "grad_norm": 0.16825923629726025,
+ "learning_rate": 2.1778131091020516e-05,
+ "loss": 0.4433916211128235,
+ "step": 1109
+ },
+ {
+ "epoch": 0.5823714585519413,
+ "grad_norm": 0.17194743467883944,
+ "learning_rate": 2.1742272160542654e-05,
+ "loss": 0.47296425700187683,
+ "step": 1110
+ },
+ {
+ "epoch": 0.5828961175236097,
+ "grad_norm": 0.16947730546039969,
+ "learning_rate": 2.170642887252241e-05,
+ "loss": 0.4815937876701355,
+ "step": 1111
+ },
+ {
+ "epoch": 0.583420776495278,
+ "grad_norm": 0.1657449928668473,
+ "learning_rate": 2.167060132433798e-05,
+ "loss": 0.4479164779186249,
+ "step": 1112
+ },
+ {
+ "epoch": 0.5839454354669464,
+ "grad_norm": 0.1658952516607364,
+ "learning_rate": 2.1634789613324778e-05,
+ "loss": 0.4637463688850403,
+ "step": 1113
+ },
+ {
+ "epoch": 0.5844700944386149,
+ "grad_norm": 0.16777002188998236,
+ "learning_rate": 2.1598993836775218e-05,
+ "loss": 0.457524836063385,
+ "step": 1114
+ },
+ {
+ "epoch": 0.5849947534102833,
+ "grad_norm": 0.15837566447341322,
+ "learning_rate": 2.1563214091938405e-05,
+ "loss": 0.44938087463378906,
+ "step": 1115
+ },
+ {
+ "epoch": 0.5855194123819517,
+ "grad_norm": 0.17162206617035314,
+ "learning_rate": 2.1527450476019896e-05,
+ "loss": 0.4393450617790222,
+ "step": 1116
+ },
+ {
+ "epoch": 0.5860440713536201,
+ "grad_norm": 0.16224101921433814,
+ "learning_rate": 2.1491703086181426e-05,
+ "loss": 0.451913982629776,
+ "step": 1117
+ },
+ {
+ "epoch": 0.5865687303252886,
+ "grad_norm": 0.1736253504575144,
+ "learning_rate": 2.145597201954066e-05,
+ "loss": 0.4611290395259857,
+ "step": 1118
+ },
+ {
+ "epoch": 0.587093389296957,
+ "grad_norm": 0.16308408333737234,
+ "learning_rate": 2.1420257373170903e-05,
+ "loss": 0.45898568630218506,
+ "step": 1119
+ },
+ {
+ "epoch": 0.5876180482686254,
+ "grad_norm": 0.16948634794175524,
+ "learning_rate": 2.1384559244100848e-05,
+ "loss": 0.45982882380485535,
+ "step": 1120
+ },
+ {
+ "epoch": 0.5881427072402938,
+ "grad_norm": 0.16676366538565252,
+ "learning_rate": 2.1348877729314336e-05,
+ "loss": 0.44826576113700867,
+ "step": 1121
+ },
+ {
+ "epoch": 0.5886673662119623,
+ "grad_norm": 0.16431125630712295,
+ "learning_rate": 2.1313212925750037e-05,
+ "loss": 0.4616445302963257,
+ "step": 1122
+ },
+ {
+ "epoch": 0.5891920251836307,
+ "grad_norm": 0.1631659061616955,
+ "learning_rate": 2.1277564930301252e-05,
+ "loss": 0.45651188492774963,
+ "step": 1123
+ },
+ {
+ "epoch": 0.5897166841552991,
+ "grad_norm": 0.15805950502689237,
+ "learning_rate": 2.1241933839815604e-05,
+ "loss": 0.4486045837402344,
+ "step": 1124
+ },
+ {
+ "epoch": 0.5902413431269675,
+ "grad_norm": 0.16530288038603502,
+ "learning_rate": 2.1206319751094773e-05,
+ "loss": 0.4485066533088684,
+ "step": 1125
+ },
+ {
+ "epoch": 0.5907660020986358,
+ "grad_norm": 0.16061903194239047,
+ "learning_rate": 2.1170722760894295e-05,
+ "loss": 0.4565330743789673,
+ "step": 1126
+ },
+ {
+ "epoch": 0.5912906610703043,
+ "grad_norm": 0.1643346999353627,
+ "learning_rate": 2.1135142965923195e-05,
+ "loss": 0.45560717582702637,
+ "step": 1127
+ },
+ {
+ "epoch": 0.5918153200419727,
+ "grad_norm": 0.17484644315215753,
+ "learning_rate": 2.109958046284382e-05,
+ "loss": 0.4795891046524048,
+ "step": 1128
+ },
+ {
+ "epoch": 0.5923399790136411,
+ "grad_norm": 0.1623762678622651,
+ "learning_rate": 2.1064035348271528e-05,
+ "loss": 0.47395604848861694,
+ "step": 1129
+ },
+ {
+ "epoch": 0.5928646379853095,
+ "grad_norm": 0.16888824241669015,
+ "learning_rate": 2.1028507718774456e-05,
+ "loss": 0.4601455330848694,
+ "step": 1130
+ },
+ {
+ "epoch": 0.593389296956978,
+ "grad_norm": 0.16291744568806302,
+ "learning_rate": 2.0992997670873193e-05,
+ "loss": 0.46248236298561096,
+ "step": 1131
+ },
+ {
+ "epoch": 0.5939139559286464,
+ "grad_norm": 0.16185598691236988,
+ "learning_rate": 2.0957505301040603e-05,
+ "loss": 0.45085301995277405,
+ "step": 1132
+ },
+ {
+ "epoch": 0.5944386149003148,
+ "grad_norm": 0.1678000806117814,
+ "learning_rate": 2.0922030705701518e-05,
+ "loss": 0.44939014315605164,
+ "step": 1133
+ },
+ {
+ "epoch": 0.5949632738719832,
+ "grad_norm": 0.16104474282288994,
+ "learning_rate": 2.088657398123246e-05,
+ "loss": 0.4313976764678955,
+ "step": 1134
+ },
+ {
+ "epoch": 0.5954879328436516,
+ "grad_norm": 0.1631128968199669,
+ "learning_rate": 2.085113522396141e-05,
+ "loss": 0.4508278965950012,
+ "step": 1135
+ },
+ {
+ "epoch": 0.5960125918153201,
+ "grad_norm": 0.16307688195844533,
+ "learning_rate": 2.0815714530167557e-05,
+ "loss": 0.4467576742172241,
+ "step": 1136
+ },
+ {
+ "epoch": 0.5965372507869885,
+ "grad_norm": 0.16456832304532207,
+ "learning_rate": 2.0780311996080976e-05,
+ "loss": 0.4621703624725342,
+ "step": 1137
+ },
+ {
+ "epoch": 0.5970619097586569,
+ "grad_norm": 0.16346217748325367,
+ "learning_rate": 2.0744927717882435e-05,
+ "loss": 0.4589449465274811,
+ "step": 1138
+ },
+ {
+ "epoch": 0.5975865687303253,
+ "grad_norm": 0.16180691501076994,
+ "learning_rate": 2.0709561791703097e-05,
+ "loss": 0.43698400259017944,
+ "step": 1139
+ },
+ {
+ "epoch": 0.5981112277019937,
+ "grad_norm": 0.15695459976898835,
+ "learning_rate": 2.067421431362427e-05,
+ "loss": 0.452271044254303,
+ "step": 1140
+ },
+ {
+ "epoch": 0.5986358866736621,
+ "grad_norm": 0.15930893553721054,
+ "learning_rate": 2.063888537967713e-05,
+ "loss": 0.44868963956832886,
+ "step": 1141
+ },
+ {
+ "epoch": 0.5991605456453305,
+ "grad_norm": 0.15963802284017928,
+ "learning_rate": 2.0603575085842485e-05,
+ "loss": 0.4376864731311798,
+ "step": 1142
+ },
+ {
+ "epoch": 0.5996852046169989,
+ "grad_norm": 0.16610266900507178,
+ "learning_rate": 2.05682835280505e-05,
+ "loss": 0.44676727056503296,
+ "step": 1143
+ },
+ {
+ "epoch": 0.6002098635886673,
+ "grad_norm": 0.15978253236669612,
+ "learning_rate": 2.053301080218044e-05,
+ "loss": 0.434773713350296,
+ "step": 1144
+ },
+ {
+ "epoch": 0.6007345225603358,
+ "grad_norm": 0.16305141803526324,
+ "learning_rate": 2.0497757004060403e-05,
+ "loss": 0.4529780149459839,
+ "step": 1145
+ },
+ {
+ "epoch": 0.6012591815320042,
+ "grad_norm": 0.1649737864101215,
+ "learning_rate": 2.046252222946707e-05,
+ "loss": 0.4571434259414673,
+ "step": 1146
+ },
+ {
+ "epoch": 0.6017838405036726,
+ "grad_norm": 0.15738824968326598,
+ "learning_rate": 2.0427306574125427e-05,
+ "loss": 0.4446289539337158,
+ "step": 1147
+ },
+ {
+ "epoch": 0.602308499475341,
+ "grad_norm": 0.1731139535185187,
+ "learning_rate": 2.039211013370853e-05,
+ "loss": 0.4551090598106384,
+ "step": 1148
+ },
+ {
+ "epoch": 0.6028331584470095,
+ "grad_norm": 0.15857251900043037,
+ "learning_rate": 2.0356933003837237e-05,
+ "loss": 0.4644010663032532,
+ "step": 1149
+ },
+ {
+ "epoch": 0.6033578174186779,
+ "grad_norm": 0.1589582861134211,
+ "learning_rate": 2.032177528007992e-05,
+ "loss": 0.45741650462150574,
+ "step": 1150
+ },
+ {
+ "epoch": 0.6038824763903463,
+ "grad_norm": 0.15639092706575122,
+ "learning_rate": 2.0286637057952257e-05,
+ "loss": 0.4375547766685486,
+ "step": 1151
+ },
+ {
+ "epoch": 0.6044071353620147,
+ "grad_norm": 0.16335605581337187,
+ "learning_rate": 2.0251518432916933e-05,
+ "loss": 0.4636140465736389,
+ "step": 1152
+ },
+ {
+ "epoch": 0.6049317943336832,
+ "grad_norm": 0.15437437113359356,
+ "learning_rate": 2.0216419500383372e-05,
+ "loss": 0.44632142782211304,
+ "step": 1153
+ },
+ {
+ "epoch": 0.6054564533053515,
+ "grad_norm": 0.15522039198698664,
+ "learning_rate": 2.0181340355707545e-05,
+ "loss": 0.449049174785614,
+ "step": 1154
+ },
+ {
+ "epoch": 0.6059811122770199,
+ "grad_norm": 0.1599951917909323,
+ "learning_rate": 2.0146281094191604e-05,
+ "loss": 0.44395962357521057,
+ "step": 1155
+ },
+ {
+ "epoch": 0.6065057712486883,
+ "grad_norm": 0.16866621620527716,
+ "learning_rate": 2.011124181108373e-05,
+ "loss": 0.45693716406822205,
+ "step": 1156
+ },
+ {
+ "epoch": 0.6070304302203567,
+ "grad_norm": 0.15508935650135075,
+ "learning_rate": 2.0076222601577815e-05,
+ "loss": 0.4525151252746582,
+ "step": 1157
+ },
+ {
+ "epoch": 0.6075550891920252,
+ "grad_norm": 0.16508998118005971,
+ "learning_rate": 2.004122356081321e-05,
+ "loss": 0.466006875038147,
+ "step": 1158
+ },
+ {
+ "epoch": 0.6080797481636936,
+ "grad_norm": 0.15817418864022237,
+ "learning_rate": 2.0006244783874466e-05,
+ "loss": 0.4496110677719116,
+ "step": 1159
+ },
+ {
+ "epoch": 0.608604407135362,
+ "grad_norm": 0.16127270259749824,
+ "learning_rate": 1.997128636579109e-05,
+ "loss": 0.45918628573417664,
+ "step": 1160
+ },
+ {
+ "epoch": 0.6091290661070304,
+ "grad_norm": 0.1702599924159677,
+ "learning_rate": 1.9936348401537292e-05,
+ "loss": 0.47218555212020874,
+ "step": 1161
+ },
+ {
+ "epoch": 0.6096537250786989,
+ "grad_norm": 0.15690176248846127,
+ "learning_rate": 1.9901430986031695e-05,
+ "loss": 0.4516097903251648,
+ "step": 1162
+ },
+ {
+ "epoch": 0.6101783840503673,
+ "grad_norm": 0.1713614405614144,
+ "learning_rate": 1.9866534214137085e-05,
+ "loss": 0.45030856132507324,
+ "step": 1163
+ },
+ {
+ "epoch": 0.6107030430220357,
+ "grad_norm": 0.1670865563026229,
+ "learning_rate": 1.9831658180660195e-05,
+ "loss": 0.45603489875793457,
+ "step": 1164
+ },
+ {
+ "epoch": 0.6112277019937041,
+ "grad_norm": 0.16897539113401952,
+ "learning_rate": 1.9796802980351392e-05,
+ "loss": 0.46314537525177,
+ "step": 1165
+ },
+ {
+ "epoch": 0.6117523609653726,
+ "grad_norm": 0.15832868870697644,
+ "learning_rate": 1.976196870790446e-05,
+ "loss": 0.44784611463546753,
+ "step": 1166
+ },
+ {
+ "epoch": 0.612277019937041,
+ "grad_norm": 0.16595365551891114,
+ "learning_rate": 1.9727155457956315e-05,
+ "loss": 0.4409624934196472,
+ "step": 1167
+ },
+ {
+ "epoch": 0.6128016789087093,
+ "grad_norm": 0.1642630411212017,
+ "learning_rate": 1.969236332508677e-05,
+ "loss": 0.46120351552963257,
+ "step": 1168
+ },
+ {
+ "epoch": 0.6133263378803777,
+ "grad_norm": 0.16481796135868812,
+ "learning_rate": 1.9657592403818253e-05,
+ "loss": 0.45544904470443726,
+ "step": 1169
+ },
+ {
+ "epoch": 0.6138509968520461,
+ "grad_norm": 0.1629832779852364,
+ "learning_rate": 1.962284278861557e-05,
+ "loss": 0.4578462839126587,
+ "step": 1170
+ },
+ {
+ "epoch": 0.6143756558237146,
+ "grad_norm": 0.15916219613603688,
+ "learning_rate": 1.9588114573885664e-05,
+ "loss": 0.4546983540058136,
+ "step": 1171
+ },
+ {
+ "epoch": 0.614900314795383,
+ "grad_norm": 0.1660520165166075,
+ "learning_rate": 1.95534078539773e-05,
+ "loss": 0.4601942300796509,
+ "step": 1172
+ },
+ {
+ "epoch": 0.6154249737670514,
+ "grad_norm": 0.16924794578577768,
+ "learning_rate": 1.9518722723180877e-05,
+ "loss": 0.45997846126556396,
+ "step": 1173
+ },
+ {
+ "epoch": 0.6159496327387198,
+ "grad_norm": 0.16597646133045382,
+ "learning_rate": 1.9484059275728127e-05,
+ "loss": 0.4594605565071106,
+ "step": 1174
+ },
+ {
+ "epoch": 0.6164742917103883,
+ "grad_norm": 0.16964243792519748,
+ "learning_rate": 1.9449417605791876e-05,
+ "loss": 0.4638690650463104,
+ "step": 1175
+ },
+ {
+ "epoch": 0.6169989506820567,
+ "grad_norm": 0.1758591426791696,
+ "learning_rate": 1.9414797807485796e-05,
+ "loss": 0.4666447639465332,
+ "step": 1176
+ },
+ {
+ "epoch": 0.6175236096537251,
+ "grad_norm": 0.16217661227891753,
+ "learning_rate": 1.9380199974864126e-05,
+ "loss": 0.44252535700798035,
+ "step": 1177
+ },
+ {
+ "epoch": 0.6180482686253935,
+ "grad_norm": 0.1576446184499019,
+ "learning_rate": 1.934562420192142e-05,
+ "loss": 0.4499641954898834,
+ "step": 1178
+ },
+ {
+ "epoch": 0.618572927597062,
+ "grad_norm": 0.16231008092636912,
+ "learning_rate": 1.9311070582592327e-05,
+ "loss": 0.4561542570590973,
+ "step": 1179
+ },
+ {
+ "epoch": 0.6190975865687304,
+ "grad_norm": 0.16790382249577043,
+ "learning_rate": 1.9276539210751292e-05,
+ "loss": 0.4484325647354126,
+ "step": 1180
+ },
+ {
+ "epoch": 0.6196222455403987,
+ "grad_norm": 0.18794729932605855,
+ "learning_rate": 1.9242030180212323e-05,
+ "loss": 0.44169026613235474,
+ "step": 1181
+ },
+ {
+ "epoch": 0.6201469045120671,
+ "grad_norm": 0.16948051243889814,
+ "learning_rate": 1.920754358472872e-05,
+ "loss": 0.4592876434326172,
+ "step": 1182
+ },
+ {
+ "epoch": 0.6206715634837355,
+ "grad_norm": 0.16762558710029374,
+ "learning_rate": 1.917307951799286e-05,
+ "loss": 0.4402095079421997,
+ "step": 1183
+ },
+ {
+ "epoch": 0.621196222455404,
+ "grad_norm": 0.18170280116387003,
+ "learning_rate": 1.9138638073635885e-05,
+ "loss": 0.4633026123046875,
+ "step": 1184
+ },
+ {
+ "epoch": 0.6217208814270724,
+ "grad_norm": 0.1718051709136108,
+ "learning_rate": 1.9104219345227486e-05,
+ "loss": 0.4510066509246826,
+ "step": 1185
+ },
+ {
+ "epoch": 0.6222455403987408,
+ "grad_norm": 0.16829066326345576,
+ "learning_rate": 1.906982342627565e-05,
+ "loss": 0.4632919430732727,
+ "step": 1186
+ },
+ {
+ "epoch": 0.6227701993704092,
+ "grad_norm": 0.16669141808548274,
+ "learning_rate": 1.9035450410226395e-05,
+ "loss": 0.455544650554657,
+ "step": 1187
+ },
+ {
+ "epoch": 0.6232948583420777,
+ "grad_norm": 0.16679996559722232,
+ "learning_rate": 1.90011003904635e-05,
+ "loss": 0.459994375705719,
+ "step": 1188
+ },
+ {
+ "epoch": 0.6238195173137461,
+ "grad_norm": 0.17379754713278164,
+ "learning_rate": 1.8966773460308284e-05,
+ "loss": 0.4576749801635742,
+ "step": 1189
+ },
+ {
+ "epoch": 0.6243441762854145,
+ "grad_norm": 0.17088539733389246,
+ "learning_rate": 1.8932469713019335e-05,
+ "loss": 0.44853052496910095,
+ "step": 1190
+ },
+ {
+ "epoch": 0.6248688352570829,
+ "grad_norm": 0.15556883889201015,
+ "learning_rate": 1.8898189241792248e-05,
+ "loss": 0.44060447812080383,
+ "step": 1191
+ },
+ {
+ "epoch": 0.6253934942287513,
+ "grad_norm": 0.1640093917439215,
+ "learning_rate": 1.8863932139759388e-05,
+ "loss": 0.4531911611557007,
+ "step": 1192
+ },
+ {
+ "epoch": 0.6259181532004198,
+ "grad_norm": 0.16444284826591576,
+ "learning_rate": 1.882969849998964e-05,
+ "loss": 0.44176363945007324,
+ "step": 1193
+ },
+ {
+ "epoch": 0.6264428121720882,
+ "grad_norm": 0.1696462520995642,
+ "learning_rate": 1.8795488415488143e-05,
+ "loss": 0.44516652822494507,
+ "step": 1194
+ },
+ {
+ "epoch": 0.6269674711437565,
+ "grad_norm": 0.16810724705191796,
+ "learning_rate": 1.8761301979196032e-05,
+ "loss": 0.4542563855648041,
+ "step": 1195
+ },
+ {
+ "epoch": 0.6274921301154249,
+ "grad_norm": 0.17334813313855824,
+ "learning_rate": 1.8727139283990204e-05,
+ "loss": 0.47769248485565186,
+ "step": 1196
+ },
+ {
+ "epoch": 0.6280167890870934,
+ "grad_norm": 0.16308212035303235,
+ "learning_rate": 1.8693000422683043e-05,
+ "loss": 0.4407799243927002,
+ "step": 1197
+ },
+ {
+ "epoch": 0.6285414480587618,
+ "grad_norm": 0.16853319139928977,
+ "learning_rate": 1.8658885488022204e-05,
+ "loss": 0.4565497040748596,
+ "step": 1198
+ },
+ {
+ "epoch": 0.6290661070304302,
+ "grad_norm": 0.16694968781284703,
+ "learning_rate": 1.8624794572690328e-05,
+ "loss": 0.44522374868392944,
+ "step": 1199
+ },
+ {
+ "epoch": 0.6295907660020986,
+ "grad_norm": 0.16418579468339103,
+ "learning_rate": 1.8590727769304788e-05,
+ "loss": 0.4441048502922058,
+ "step": 1200
+ },
+ {
+ "epoch": 0.630115424973767,
+ "grad_norm": 0.1661930777124156,
+ "learning_rate": 1.8556685170417486e-05,
+ "loss": 0.457327663898468,
+ "step": 1201
+ },
+ {
+ "epoch": 0.6306400839454355,
+ "grad_norm": 0.16536630027039814,
+ "learning_rate": 1.852266686851452e-05,
+ "loss": 0.462494432926178,
+ "step": 1202
+ },
+ {
+ "epoch": 0.6311647429171039,
+ "grad_norm": 0.16611474088307454,
+ "learning_rate": 1.8488672956016013e-05,
+ "loss": 0.4557701647281647,
+ "step": 1203
+ },
+ {
+ "epoch": 0.6316894018887723,
+ "grad_norm": 0.16759343079749697,
+ "learning_rate": 1.8454703525275813e-05,
+ "loss": 0.45069995522499084,
+ "step": 1204
+ },
+ {
+ "epoch": 0.6322140608604407,
+ "grad_norm": 0.16186676294485336,
+ "learning_rate": 1.8420758668581255e-05,
+ "loss": 0.44954800605773926,
+ "step": 1205
+ },
+ {
+ "epoch": 0.6327387198321092,
+ "grad_norm": 0.15687159835368553,
+ "learning_rate": 1.8386838478152916e-05,
+ "loss": 0.4490429759025574,
+ "step": 1206
+ },
+ {
+ "epoch": 0.6332633788037776,
+ "grad_norm": 0.16242951976115577,
+ "learning_rate": 1.8352943046144372e-05,
+ "loss": 0.45819586515426636,
+ "step": 1207
+ },
+ {
+ "epoch": 0.633788037775446,
+ "grad_norm": 0.16516087266517232,
+ "learning_rate": 1.831907246464191e-05,
+ "loss": 0.44769883155822754,
+ "step": 1208
+ },
+ {
+ "epoch": 0.6343126967471143,
+ "grad_norm": 0.16709475871512644,
+ "learning_rate": 1.8285226825664332e-05,
+ "loss": 0.4699128568172455,
+ "step": 1209
+ },
+ {
+ "epoch": 0.6348373557187827,
+ "grad_norm": 0.16288623150595774,
+ "learning_rate": 1.8251406221162636e-05,
+ "loss": 0.45322558283805847,
+ "step": 1210
+ },
+ {
+ "epoch": 0.6353620146904512,
+ "grad_norm": 0.1675722848054237,
+ "learning_rate": 1.821761074301986e-05,
+ "loss": 0.4752347469329834,
+ "step": 1211
+ },
+ {
+ "epoch": 0.6358866736621196,
+ "grad_norm": 0.163285963514504,
+ "learning_rate": 1.818384048305074e-05,
+ "loss": 0.441958487033844,
+ "step": 1212
+ },
+ {
+ "epoch": 0.636411332633788,
+ "grad_norm": 0.17078102287557764,
+ "learning_rate": 1.8150095533001517e-05,
+ "loss": 0.4551333487033844,
+ "step": 1213
+ },
+ {
+ "epoch": 0.6369359916054564,
+ "grad_norm": 0.16188540746097138,
+ "learning_rate": 1.811637598454966e-05,
+ "loss": 0.4470779597759247,
+ "step": 1214
+ },
+ {
+ "epoch": 0.6374606505771249,
+ "grad_norm": 0.15235628218958286,
+ "learning_rate": 1.808268192930364e-05,
+ "loss": 0.43604153394699097,
+ "step": 1215
+ },
+ {
+ "epoch": 0.6379853095487933,
+ "grad_norm": 0.1604601949959104,
+ "learning_rate": 1.8049013458802665e-05,
+ "loss": 0.4445713758468628,
+ "step": 1216
+ },
+ {
+ "epoch": 0.6385099685204617,
+ "grad_norm": 0.15511797131215901,
+ "learning_rate": 1.8015370664516427e-05,
+ "loss": 0.4455301761627197,
+ "step": 1217
+ },
+ {
+ "epoch": 0.6390346274921301,
+ "grad_norm": 0.16815669546944095,
+ "learning_rate": 1.798175363784487e-05,
+ "loss": 0.46135014295578003,
+ "step": 1218
+ },
+ {
+ "epoch": 0.6395592864637986,
+ "grad_norm": 0.161558177688922,
+ "learning_rate": 1.7948162470117917e-05,
+ "loss": 0.4519304037094116,
+ "step": 1219
+ },
+ {
+ "epoch": 0.640083945435467,
+ "grad_norm": 0.16677842130048232,
+ "learning_rate": 1.7914597252595265e-05,
+ "loss": 0.4550550878047943,
+ "step": 1220
+ },
+ {
+ "epoch": 0.6406086044071354,
+ "grad_norm": 0.1597331480423138,
+ "learning_rate": 1.7881058076466095e-05,
+ "loss": 0.441120445728302,
+ "step": 1221
+ },
+ {
+ "epoch": 0.6411332633788038,
+ "grad_norm": 0.15951106545246824,
+ "learning_rate": 1.7847545032848844e-05,
+ "loss": 0.4471382200717926,
+ "step": 1222
+ },
+ {
+ "epoch": 0.6416579223504721,
+ "grad_norm": 0.16157787674667565,
+ "learning_rate": 1.7814058212790947e-05,
+ "loss": 0.4550648331642151,
+ "step": 1223
+ },
+ {
+ "epoch": 0.6421825813221406,
+ "grad_norm": 0.16989186763352415,
+ "learning_rate": 1.7780597707268583e-05,
+ "loss": 0.4515089988708496,
+ "step": 1224
+ },
+ {
+ "epoch": 0.642707240293809,
+ "grad_norm": 0.15952153223099508,
+ "learning_rate": 1.7747163607186468e-05,
+ "loss": 0.4658638834953308,
+ "step": 1225
+ },
+ {
+ "epoch": 0.6432318992654774,
+ "grad_norm": 0.15205926699647218,
+ "learning_rate": 1.7713756003377567e-05,
+ "loss": 0.4330868422985077,
+ "step": 1226
+ },
+ {
+ "epoch": 0.6437565582371458,
+ "grad_norm": 0.15784806480936794,
+ "learning_rate": 1.7680374986602865e-05,
+ "loss": 0.4506068527698517,
+ "step": 1227
+ },
+ {
+ "epoch": 0.6442812172088143,
+ "grad_norm": 0.16006709280686687,
+ "learning_rate": 1.7647020647551105e-05,
+ "loss": 0.446941077709198,
+ "step": 1228
+ },
+ {
+ "epoch": 0.6448058761804827,
+ "grad_norm": 0.16064631158222062,
+ "learning_rate": 1.761369307683856e-05,
+ "loss": 0.4418525695800781,
+ "step": 1229
+ },
+ {
+ "epoch": 0.6453305351521511,
+ "grad_norm": 0.16454136480444392,
+ "learning_rate": 1.7580392365008767e-05,
+ "loss": 0.46268391609191895,
+ "step": 1230
+ },
+ {
+ "epoch": 0.6458551941238195,
+ "grad_norm": 0.15401246535771462,
+ "learning_rate": 1.754711860253233e-05,
+ "loss": 0.4372989535331726,
+ "step": 1231
+ },
+ {
+ "epoch": 0.646379853095488,
+ "grad_norm": 0.1544567975028383,
+ "learning_rate": 1.7513871879806603e-05,
+ "loss": 0.44806575775146484,
+ "step": 1232
+ },
+ {
+ "epoch": 0.6469045120671564,
+ "grad_norm": 0.16411225112730965,
+ "learning_rate": 1.748065228715549e-05,
+ "loss": 0.43782058358192444,
+ "step": 1233
+ },
+ {
+ "epoch": 0.6474291710388248,
+ "grad_norm": 0.1552767832972093,
+ "learning_rate": 1.7447459914829183e-05,
+ "loss": 0.44531378149986267,
+ "step": 1234
+ },
+ {
+ "epoch": 0.6479538300104932,
+ "grad_norm": 0.16214695257123513,
+ "learning_rate": 1.741429485300392e-05,
+ "loss": 0.47399842739105225,
+ "step": 1235
+ },
+ {
+ "epoch": 0.6484784889821616,
+ "grad_norm": 0.16684861111484112,
+ "learning_rate": 1.7381157191781766e-05,
+ "loss": 0.4654282331466675,
+ "step": 1236
+ },
+ {
+ "epoch": 0.64900314795383,
+ "grad_norm": 0.15901864178630273,
+ "learning_rate": 1.7348047021190334e-05,
+ "loss": 0.45128414034843445,
+ "step": 1237
+ },
+ {
+ "epoch": 0.6495278069254984,
+ "grad_norm": 0.15360991961742707,
+ "learning_rate": 1.7314964431182523e-05,
+ "loss": 0.4303375482559204,
+ "step": 1238
+ },
+ {
+ "epoch": 0.6500524658971668,
+ "grad_norm": 0.1633151574385741,
+ "learning_rate": 1.728190951163633e-05,
+ "loss": 0.45568403601646423,
+ "step": 1239
+ },
+ {
+ "epoch": 0.6505771248688352,
+ "grad_norm": 0.16377853571058645,
+ "learning_rate": 1.724888235235457e-05,
+ "loss": 0.46503543853759766,
+ "step": 1240
+ },
+ {
+ "epoch": 0.6511017838405037,
+ "grad_norm": 0.1595023801275368,
+ "learning_rate": 1.7215883043064655e-05,
+ "loss": 0.46337294578552246,
+ "step": 1241
+ },
+ {
+ "epoch": 0.6516264428121721,
+ "grad_norm": 0.16324796182212292,
+ "learning_rate": 1.7182911673418313e-05,
+ "loss": 0.46197354793548584,
+ "step": 1242
+ },
+ {
+ "epoch": 0.6521511017838405,
+ "grad_norm": 0.16353734830879765,
+ "learning_rate": 1.714996833299138e-05,
+ "loss": 0.45741671323776245,
+ "step": 1243
+ },
+ {
+ "epoch": 0.6526757607555089,
+ "grad_norm": 0.15941587678263386,
+ "learning_rate": 1.7117053111283528e-05,
+ "loss": 0.4514503479003906,
+ "step": 1244
+ },
+ {
+ "epoch": 0.6532004197271774,
+ "grad_norm": 0.1642653230145008,
+ "learning_rate": 1.708416609771804e-05,
+ "loss": 0.4437655210494995,
+ "step": 1245
+ },
+ {
+ "epoch": 0.6537250786988458,
+ "grad_norm": 0.15814155519130152,
+ "learning_rate": 1.7051307381641587e-05,
+ "loss": 0.4314163327217102,
+ "step": 1246
+ },
+ {
+ "epoch": 0.6542497376705142,
+ "grad_norm": 0.16086078370768056,
+ "learning_rate": 1.701847705232394e-05,
+ "loss": 0.4526430666446686,
+ "step": 1247
+ },
+ {
+ "epoch": 0.6547743966421826,
+ "grad_norm": 0.16389360190715183,
+ "learning_rate": 1.698567519895775e-05,
+ "loss": 0.4278966188430786,
+ "step": 1248
+ },
+ {
+ "epoch": 0.655299055613851,
+ "grad_norm": 0.16303142287804262,
+ "learning_rate": 1.6952901910658316e-05,
+ "loss": 0.45778220891952515,
+ "step": 1249
+ },
+ {
+ "epoch": 0.6558237145855194,
+ "grad_norm": 0.1675209956904048,
+ "learning_rate": 1.6920157276463318e-05,
+ "loss": 0.4489403963088989,
+ "step": 1250
+ },
+ {
+ "epoch": 0.6563483735571878,
+ "grad_norm": 0.16228311551074323,
+ "learning_rate": 1.6887441385332604e-05,
+ "loss": 0.4534342885017395,
+ "step": 1251
+ },
+ {
+ "epoch": 0.6568730325288562,
+ "grad_norm": 0.16408957851046913,
+ "learning_rate": 1.6854754326147916e-05,
+ "loss": 0.4549208879470825,
+ "step": 1252
+ },
+ {
+ "epoch": 0.6573976915005246,
+ "grad_norm": 0.15967922111432845,
+ "learning_rate": 1.682209618771268e-05,
+ "loss": 0.4451141357421875,
+ "step": 1253
+ },
+ {
+ "epoch": 0.657922350472193,
+ "grad_norm": 0.16440698722204208,
+ "learning_rate": 1.678946705875176e-05,
+ "loss": 0.454200804233551,
+ "step": 1254
+ },
+ {
+ "epoch": 0.6584470094438615,
+ "grad_norm": 0.16251120184938497,
+ "learning_rate": 1.675686702791117e-05,
+ "loss": 0.4580027461051941,
+ "step": 1255
+ },
+ {
+ "epoch": 0.6589716684155299,
+ "grad_norm": 0.1643717386150457,
+ "learning_rate": 1.6724296183757924e-05,
+ "loss": 0.4422311782836914,
+ "step": 1256
+ },
+ {
+ "epoch": 0.6594963273871983,
+ "grad_norm": 0.1644586414378722,
+ "learning_rate": 1.66917546147797e-05,
+ "loss": 0.4593179225921631,
+ "step": 1257
+ },
+ {
+ "epoch": 0.6600209863588667,
+ "grad_norm": 0.16751074186967685,
+ "learning_rate": 1.6659242409384657e-05,
+ "loss": 0.457930326461792,
+ "step": 1258
+ },
+ {
+ "epoch": 0.6605456453305352,
+ "grad_norm": 0.15398418609155334,
+ "learning_rate": 1.6626759655901182e-05,
+ "loss": 0.44002899527549744,
+ "step": 1259
+ },
+ {
+ "epoch": 0.6610703043022036,
+ "grad_norm": 0.16328113175790226,
+ "learning_rate": 1.659430644257763e-05,
+ "loss": 0.4500438868999481,
+ "step": 1260
+ },
+ {
+ "epoch": 0.661594963273872,
+ "grad_norm": 0.1694244340615106,
+ "learning_rate": 1.656188285758214e-05,
+ "loss": 0.4818427264690399,
+ "step": 1261
+ },
+ {
+ "epoch": 0.6621196222455404,
+ "grad_norm": 0.1578320701754368,
+ "learning_rate": 1.652948898900232e-05,
+ "loss": 0.4492436647415161,
+ "step": 1262
+ },
+ {
+ "epoch": 0.6626442812172089,
+ "grad_norm": 0.16538865917828469,
+ "learning_rate": 1.6497124924845064e-05,
+ "loss": 0.45070338249206543,
+ "step": 1263
+ },
+ {
+ "epoch": 0.6631689401888772,
+ "grad_norm": 0.15670817848256852,
+ "learning_rate": 1.6464790753036296e-05,
+ "loss": 0.45439285039901733,
+ "step": 1264
+ },
+ {
+ "epoch": 0.6636935991605456,
+ "grad_norm": 0.16184786585965885,
+ "learning_rate": 1.64324865614207e-05,
+ "loss": 0.446522980928421,
+ "step": 1265
+ },
+ {
+ "epoch": 0.664218258132214,
+ "grad_norm": 0.15719726462251984,
+ "learning_rate": 1.6400212437761554e-05,
+ "loss": 0.4495280981063843,
+ "step": 1266
+ },
+ {
+ "epoch": 0.6647429171038824,
+ "grad_norm": 0.1680804414134405,
+ "learning_rate": 1.6367968469740416e-05,
+ "loss": 0.4652596712112427,
+ "step": 1267
+ },
+ {
+ "epoch": 0.6652675760755509,
+ "grad_norm": 0.15919541846683566,
+ "learning_rate": 1.6335754744956937e-05,
+ "loss": 0.43842798471450806,
+ "step": 1268
+ },
+ {
+ "epoch": 0.6657922350472193,
+ "grad_norm": 0.16322634187139418,
+ "learning_rate": 1.630357135092859e-05,
+ "loss": 0.45242664217948914,
+ "step": 1269
+ },
+ {
+ "epoch": 0.6663168940188877,
+ "grad_norm": 0.15644867734734783,
+ "learning_rate": 1.6271418375090445e-05,
+ "loss": 0.44801655411720276,
+ "step": 1270
+ },
+ {
+ "epoch": 0.6668415529905561,
+ "grad_norm": 0.1514717416124423,
+ "learning_rate": 1.623929590479496e-05,
+ "loss": 0.4506194293498993,
+ "step": 1271
+ },
+ {
+ "epoch": 0.6673662119622246,
+ "grad_norm": 0.1740345085495889,
+ "learning_rate": 1.6207204027311683e-05,
+ "loss": 0.4818177819252014,
+ "step": 1272
+ },
+ {
+ "epoch": 0.667890870933893,
+ "grad_norm": 0.15797701900765457,
+ "learning_rate": 1.6175142829827067e-05,
+ "loss": 0.4639382064342499,
+ "step": 1273
+ },
+ {
+ "epoch": 0.6684155299055614,
+ "grad_norm": 0.15663812316653875,
+ "learning_rate": 1.6143112399444198e-05,
+ "loss": 0.46217501163482666,
+ "step": 1274
+ },
+ {
+ "epoch": 0.6689401888772298,
+ "grad_norm": 0.15474018977043758,
+ "learning_rate": 1.6111112823182592e-05,
+ "loss": 0.43691688776016235,
+ "step": 1275
+ },
+ {
+ "epoch": 0.6694648478488983,
+ "grad_norm": 0.1615634942529431,
+ "learning_rate": 1.6079144187977923e-05,
+ "loss": 0.4412601590156555,
+ "step": 1276
+ },
+ {
+ "epoch": 0.6699895068205667,
+ "grad_norm": 0.15440022842754444,
+ "learning_rate": 1.604720658068184e-05,
+ "loss": 0.4523273706436157,
+ "step": 1277
+ },
+ {
+ "epoch": 0.670514165792235,
+ "grad_norm": 0.1634985665776242,
+ "learning_rate": 1.6015300088061664e-05,
+ "loss": 0.46264398097991943,
+ "step": 1278
+ },
+ {
+ "epoch": 0.6710388247639034,
+ "grad_norm": 0.15037886370689527,
+ "learning_rate": 1.598342479680018e-05,
+ "loss": 0.451465904712677,
+ "step": 1279
+ },
+ {
+ "epoch": 0.6715634837355718,
+ "grad_norm": 0.16895242437746613,
+ "learning_rate": 1.595158079349542e-05,
+ "loss": 0.4466145634651184,
+ "step": 1280
+ },
+ {
+ "epoch": 0.6720881427072403,
+ "grad_norm": 0.1629548326999642,
+ "learning_rate": 1.5919768164660425e-05,
+ "loss": 0.45415782928466797,
+ "step": 1281
+ },
+ {
+ "epoch": 0.6726128016789087,
+ "grad_norm": 0.1563872102268991,
+ "learning_rate": 1.588798699672298e-05,
+ "loss": 0.4386737048625946,
+ "step": 1282
+ },
+ {
+ "epoch": 0.6731374606505771,
+ "grad_norm": 0.16038172773344717,
+ "learning_rate": 1.5856237376025403e-05,
+ "loss": 0.446260929107666,
+ "step": 1283
+ },
+ {
+ "epoch": 0.6736621196222455,
+ "grad_norm": 0.16189780714407623,
+ "learning_rate": 1.582451938882431e-05,
+ "loss": 0.45458072423934937,
+ "step": 1284
+ },
+ {
+ "epoch": 0.674186778593914,
+ "grad_norm": 0.15204521785049285,
+ "learning_rate": 1.5792833121290367e-05,
+ "loss": 0.4425617754459381,
+ "step": 1285
+ },
+ {
+ "epoch": 0.6747114375655824,
+ "grad_norm": 0.16237376704792778,
+ "learning_rate": 1.5761178659508063e-05,
+ "loss": 0.44631800055503845,
+ "step": 1286
+ },
+ {
+ "epoch": 0.6752360965372508,
+ "grad_norm": 0.1582063618466069,
+ "learning_rate": 1.57295560894755e-05,
+ "loss": 0.44280534982681274,
+ "step": 1287
+ },
+ {
+ "epoch": 0.6757607555089192,
+ "grad_norm": 0.16587032349847394,
+ "learning_rate": 1.5697965497104116e-05,
+ "loss": 0.4495997130870819,
+ "step": 1288
+ },
+ {
+ "epoch": 0.6762854144805877,
+ "grad_norm": 0.1557269262755563,
+ "learning_rate": 1.566640696821848e-05,
+ "loss": 0.4493359625339508,
+ "step": 1289
+ },
+ {
+ "epoch": 0.6768100734522561,
+ "grad_norm": 0.16038490357012256,
+ "learning_rate": 1.5634880588556043e-05,
+ "loss": 0.4447544813156128,
+ "step": 1290
+ },
+ {
+ "epoch": 0.6773347324239245,
+ "grad_norm": 0.15355201908546406,
+ "learning_rate": 1.560338644376692e-05,
+ "loss": 0.4468432664871216,
+ "step": 1291
+ },
+ {
+ "epoch": 0.6778593913955928,
+ "grad_norm": 0.15077009515589776,
+ "learning_rate": 1.5571924619413668e-05,
+ "loss": 0.44959044456481934,
+ "step": 1292
+ },
+ {
+ "epoch": 0.6783840503672612,
+ "grad_norm": 0.1583038146613965,
+ "learning_rate": 1.5540495200971012e-05,
+ "loss": 0.42450425028800964,
+ "step": 1293
+ },
+ {
+ "epoch": 0.6789087093389297,
+ "grad_norm": 0.15950687190598087,
+ "learning_rate": 1.5509098273825635e-05,
+ "loss": 0.4527238607406616,
+ "step": 1294
+ },
+ {
+ "epoch": 0.6794333683105981,
+ "grad_norm": 0.1526262208847808,
+ "learning_rate": 1.5477733923275968e-05,
+ "loss": 0.4379724860191345,
+ "step": 1295
+ },
+ {
+ "epoch": 0.6799580272822665,
+ "grad_norm": 0.15541579650021778,
+ "learning_rate": 1.5446402234531934e-05,
+ "loss": 0.4462166130542755,
+ "step": 1296
+ },
+ {
+ "epoch": 0.6804826862539349,
+ "grad_norm": 0.15483930734313786,
+ "learning_rate": 1.5415103292714718e-05,
+ "loss": 0.4390607178211212,
+ "step": 1297
+ },
+ {
+ "epoch": 0.6810073452256034,
+ "grad_norm": 0.1578005731149154,
+ "learning_rate": 1.5383837182856532e-05,
+ "loss": 0.44462254643440247,
+ "step": 1298
+ },
+ {
+ "epoch": 0.6815320041972718,
+ "grad_norm": 0.15830183884958823,
+ "learning_rate": 1.5352603989900406e-05,
+ "loss": 0.4563537538051605,
+ "step": 1299
+ },
+ {
+ "epoch": 0.6820566631689402,
+ "grad_norm": 0.15605472163445522,
+ "learning_rate": 1.5321403798699923e-05,
+ "loss": 0.43852412700653076,
+ "step": 1300
+ },
+ {
+ "epoch": 0.6825813221406086,
+ "grad_norm": 0.15770470227141278,
+ "learning_rate": 1.5290236694019017e-05,
+ "loss": 0.4438081979751587,
+ "step": 1301
+ },
+ {
+ "epoch": 0.683105981112277,
+ "grad_norm": 0.16175339845012077,
+ "learning_rate": 1.5259102760531747e-05,
+ "loss": 0.46670734882354736,
+ "step": 1302
+ },
+ {
+ "epoch": 0.6836306400839455,
+ "grad_norm": 0.16359504716154477,
+ "learning_rate": 1.5228002082822036e-05,
+ "loss": 0.4590187668800354,
+ "step": 1303
+ },
+ {
+ "epoch": 0.6841552990556139,
+ "grad_norm": 0.15422879500254713,
+ "learning_rate": 1.5196934745383457e-05,
+ "loss": 0.45287445187568665,
+ "step": 1304
+ },
+ {
+ "epoch": 0.6846799580272823,
+ "grad_norm": 0.15929902143574254,
+ "learning_rate": 1.5165900832619015e-05,
+ "loss": 0.44286859035491943,
+ "step": 1305
+ },
+ {
+ "epoch": 0.6852046169989506,
+ "grad_norm": 0.14838868714016124,
+ "learning_rate": 1.5134900428840904e-05,
+ "loss": 0.45155346393585205,
+ "step": 1306
+ },
+ {
+ "epoch": 0.6857292759706191,
+ "grad_norm": 0.15494288640397672,
+ "learning_rate": 1.5103933618270283e-05,
+ "loss": 0.44010141491889954,
+ "step": 1307
+ },
+ {
+ "epoch": 0.6862539349422875,
+ "grad_norm": 0.1686262360598338,
+ "learning_rate": 1.5073000485037042e-05,
+ "loss": 0.446555495262146,
+ "step": 1308
+ },
+ {
+ "epoch": 0.6867785939139559,
+ "grad_norm": 0.16122110823982602,
+ "learning_rate": 1.5042101113179579e-05,
+ "loss": 0.44640374183654785,
+ "step": 1309
+ },
+ {
+ "epoch": 0.6873032528856243,
+ "grad_norm": 0.16849937797919867,
+ "learning_rate": 1.5011235586644572e-05,
+ "loss": 0.4465785026550293,
+ "step": 1310
+ },
+ {
+ "epoch": 0.6878279118572928,
+ "grad_norm": 0.16062971262800754,
+ "learning_rate": 1.498040398928674e-05,
+ "loss": 0.445321649312973,
+ "step": 1311
+ },
+ {
+ "epoch": 0.6883525708289612,
+ "grad_norm": 0.1608953185993034,
+ "learning_rate": 1.4949606404868638e-05,
+ "loss": 0.46518227458000183,
+ "step": 1312
+ },
+ {
+ "epoch": 0.6888772298006296,
+ "grad_norm": 0.15880255672830915,
+ "learning_rate": 1.491884291706041e-05,
+ "loss": 0.44582217931747437,
+ "step": 1313
+ },
+ {
+ "epoch": 0.689401888772298,
+ "grad_norm": 0.16715145200789822,
+ "learning_rate": 1.4888113609439563e-05,
+ "loss": 0.4435983896255493,
+ "step": 1314
+ },
+ {
+ "epoch": 0.6899265477439664,
+ "grad_norm": 0.15263740340694454,
+ "learning_rate": 1.485741856549075e-05,
+ "loss": 0.4405066967010498,
+ "step": 1315
+ },
+ {
+ "epoch": 0.6904512067156349,
+ "grad_norm": 0.16247034518813966,
+ "learning_rate": 1.4826757868605517e-05,
+ "loss": 0.4544215500354767,
+ "step": 1316
+ },
+ {
+ "epoch": 0.6909758656873033,
+ "grad_norm": 0.15649146705978162,
+ "learning_rate": 1.4796131602082142e-05,
+ "loss": 0.4454413056373596,
+ "step": 1317
+ },
+ {
+ "epoch": 0.6915005246589717,
+ "grad_norm": 0.16286718636058178,
+ "learning_rate": 1.4765539849125318e-05,
+ "loss": 0.4497183561325073,
+ "step": 1318
+ },
+ {
+ "epoch": 0.6920251836306401,
+ "grad_norm": 0.15812342085005582,
+ "learning_rate": 1.473498269284599e-05,
+ "loss": 0.445032000541687,
+ "step": 1319
+ },
+ {
+ "epoch": 0.6925498426023085,
+ "grad_norm": 0.1558101290511678,
+ "learning_rate": 1.4704460216261123e-05,
+ "loss": 0.45769619941711426,
+ "step": 1320
+ },
+ {
+ "epoch": 0.6930745015739769,
+ "grad_norm": 0.15642889059495207,
+ "learning_rate": 1.4673972502293428e-05,
+ "loss": 0.44788360595703125,
+ "step": 1321
+ },
+ {
+ "epoch": 0.6935991605456453,
+ "grad_norm": 0.15722513518113074,
+ "learning_rate": 1.4643519633771222e-05,
+ "loss": 0.449631929397583,
+ "step": 1322
+ },
+ {
+ "epoch": 0.6941238195173137,
+ "grad_norm": 0.15587633748081542,
+ "learning_rate": 1.4613101693428125e-05,
+ "loss": 0.4450187087059021,
+ "step": 1323
+ },
+ {
+ "epoch": 0.6946484784889821,
+ "grad_norm": 0.1577107074882479,
+ "learning_rate": 1.458271876390287e-05,
+ "loss": 0.45108890533447266,
+ "step": 1324
+ },
+ {
+ "epoch": 0.6951731374606506,
+ "grad_norm": 0.15649839486990696,
+ "learning_rate": 1.4552370927739073e-05,
+ "loss": 0.46430301666259766,
+ "step": 1325
+ },
+ {
+ "epoch": 0.695697796432319,
+ "grad_norm": 0.18573600420690636,
+ "learning_rate": 1.4522058267385008e-05,
+ "loss": 0.4701972007751465,
+ "step": 1326
+ },
+ {
+ "epoch": 0.6962224554039874,
+ "grad_norm": 0.16296497084037542,
+ "learning_rate": 1.4491780865193402e-05,
+ "loss": 0.4607710838317871,
+ "step": 1327
+ },
+ {
+ "epoch": 0.6967471143756558,
+ "grad_norm": 0.161124007217343,
+ "learning_rate": 1.446153880342117e-05,
+ "loss": 0.45095115900039673,
+ "step": 1328
+ },
+ {
+ "epoch": 0.6972717733473243,
+ "grad_norm": 0.1537884039166766,
+ "learning_rate": 1.4431332164229222e-05,
+ "loss": 0.424068808555603,
+ "step": 1329
+ },
+ {
+ "epoch": 0.6977964323189927,
+ "grad_norm": 0.16588834503456087,
+ "learning_rate": 1.4401161029682237e-05,
+ "loss": 0.4622839093208313,
+ "step": 1330
+ },
+ {
+ "epoch": 0.6983210912906611,
+ "grad_norm": 0.16100510604015544,
+ "learning_rate": 1.4371025481748435e-05,
+ "loss": 0.4660705626010895,
+ "step": 1331
+ },
+ {
+ "epoch": 0.6988457502623295,
+ "grad_norm": 0.15559888940590075,
+ "learning_rate": 1.4340925602299337e-05,
+ "loss": 0.4484976530075073,
+ "step": 1332
+ },
+ {
+ "epoch": 0.6993704092339978,
+ "grad_norm": 0.15503821966213488,
+ "learning_rate": 1.4310861473109616e-05,
+ "loss": 0.45636600255966187,
+ "step": 1333
+ },
+ {
+ "epoch": 0.6998950682056663,
+ "grad_norm": 0.15539074896109334,
+ "learning_rate": 1.4280833175856743e-05,
+ "loss": 0.45863068103790283,
+ "step": 1334
+ },
+ {
+ "epoch": 0.7004197271773347,
+ "grad_norm": 0.15664821009154872,
+ "learning_rate": 1.4250840792120896e-05,
+ "loss": 0.44043463468551636,
+ "step": 1335
+ },
+ {
+ "epoch": 0.7009443861490031,
+ "grad_norm": 0.15825478808520285,
+ "learning_rate": 1.4220884403384655e-05,
+ "loss": 0.46522051095962524,
+ "step": 1336
+ },
+ {
+ "epoch": 0.7014690451206715,
+ "grad_norm": 0.16004268568141544,
+ "learning_rate": 1.4190964091032843e-05,
+ "loss": 0.46114686131477356,
+ "step": 1337
+ },
+ {
+ "epoch": 0.70199370409234,
+ "grad_norm": 0.16413139949945635,
+ "learning_rate": 1.4161079936352234e-05,
+ "loss": 0.4457515478134155,
+ "step": 1338
+ },
+ {
+ "epoch": 0.7025183630640084,
+ "grad_norm": 0.1538890577972702,
+ "learning_rate": 1.4131232020531398e-05,
+ "loss": 0.4616093337535858,
+ "step": 1339
+ },
+ {
+ "epoch": 0.7030430220356768,
+ "grad_norm": 0.16240737854323284,
+ "learning_rate": 1.4101420424660434e-05,
+ "loss": 0.44631537795066833,
+ "step": 1340
+ },
+ {
+ "epoch": 0.7035676810073452,
+ "grad_norm": 0.15984708603406073,
+ "learning_rate": 1.4071645229730762e-05,
+ "loss": 0.4366717040538788,
+ "step": 1341
+ },
+ {
+ "epoch": 0.7040923399790137,
+ "grad_norm": 0.15149695695635904,
+ "learning_rate": 1.4041906516634946e-05,
+ "loss": 0.44129428267478943,
+ "step": 1342
+ },
+ {
+ "epoch": 0.7046169989506821,
+ "grad_norm": 0.15041167501768415,
+ "learning_rate": 1.4012204366166399e-05,
+ "loss": 0.43654775619506836,
+ "step": 1343
+ },
+ {
+ "epoch": 0.7051416579223505,
+ "grad_norm": 0.1596097627963756,
+ "learning_rate": 1.3982538859019217e-05,
+ "loss": 0.4652445614337921,
+ "step": 1344
+ },
+ {
+ "epoch": 0.7056663168940189,
+ "grad_norm": 0.15621704205027884,
+ "learning_rate": 1.3952910075787941e-05,
+ "loss": 0.44577041268348694,
+ "step": 1345
+ },
+ {
+ "epoch": 0.7061909758656874,
+ "grad_norm": 0.16073041098008783,
+ "learning_rate": 1.3923318096967341e-05,
+ "loss": 0.4391976594924927,
+ "step": 1346
+ },
+ {
+ "epoch": 0.7067156348373557,
+ "grad_norm": 0.14793853946372448,
+ "learning_rate": 1.3893763002952204e-05,
+ "loss": 0.4463799297809601,
+ "step": 1347
+ },
+ {
+ "epoch": 0.7072402938090241,
+ "grad_norm": 0.16176080358867354,
+ "learning_rate": 1.3864244874037097e-05,
+ "loss": 0.45183271169662476,
+ "step": 1348
+ },
+ {
+ "epoch": 0.7077649527806925,
+ "grad_norm": 0.16372771179026924,
+ "learning_rate": 1.3834763790416171e-05,
+ "loss": 0.44205576181411743,
+ "step": 1349
+ },
+ {
+ "epoch": 0.7082896117523609,
+ "grad_norm": 0.15135222516127125,
+ "learning_rate": 1.3805319832182937e-05,
+ "loss": 0.4390261173248291,
+ "step": 1350
+ },
+ {
+ "epoch": 0.7088142707240294,
+ "grad_norm": 0.15493183225792365,
+ "learning_rate": 1.377591307933002e-05,
+ "loss": 0.45837974548339844,
+ "step": 1351
+ },
+ {
+ "epoch": 0.7093389296956978,
+ "grad_norm": 0.16405906906516504,
+ "learning_rate": 1.3746543611749008e-05,
+ "loss": 0.45032668113708496,
+ "step": 1352
+ },
+ {
+ "epoch": 0.7098635886673662,
+ "grad_norm": 0.15725677976524505,
+ "learning_rate": 1.3717211509230155e-05,
+ "loss": 0.4486224949359894,
+ "step": 1353
+ },
+ {
+ "epoch": 0.7103882476390346,
+ "grad_norm": 0.15781469066983164,
+ "learning_rate": 1.3687916851462227e-05,
+ "loss": 0.4502638578414917,
+ "step": 1354
+ },
+ {
+ "epoch": 0.710912906610703,
+ "grad_norm": 0.16115007729054298,
+ "learning_rate": 1.3658659718032241e-05,
+ "loss": 0.4508914351463318,
+ "step": 1355
+ },
+ {
+ "epoch": 0.7114375655823715,
+ "grad_norm": 0.16755060475968744,
+ "learning_rate": 1.3629440188425286e-05,
+ "loss": 0.45653945207595825,
+ "step": 1356
+ },
+ {
+ "epoch": 0.7119622245540399,
+ "grad_norm": 0.1574994294756123,
+ "learning_rate": 1.3600258342024268e-05,
+ "loss": 0.4349849820137024,
+ "step": 1357
+ },
+ {
+ "epoch": 0.7124868835257083,
+ "grad_norm": 0.16104440922657715,
+ "learning_rate": 1.3571114258109758e-05,
+ "loss": 0.4501233696937561,
+ "step": 1358
+ },
+ {
+ "epoch": 0.7130115424973767,
+ "grad_norm": 0.15163184290939022,
+ "learning_rate": 1.3542008015859691e-05,
+ "loss": 0.44171035289764404,
+ "step": 1359
+ },
+ {
+ "epoch": 0.7135362014690452,
+ "grad_norm": 0.15955896674456582,
+ "learning_rate": 1.3512939694349211e-05,
+ "loss": 0.4558671712875366,
+ "step": 1360
+ },
+ {
+ "epoch": 0.7140608604407135,
+ "grad_norm": 0.1589376751536802,
+ "learning_rate": 1.3483909372550446e-05,
+ "loss": 0.4430835247039795,
+ "step": 1361
+ },
+ {
+ "epoch": 0.7145855194123819,
+ "grad_norm": 0.15979088397822155,
+ "learning_rate": 1.3454917129332279e-05,
+ "loss": 0.4426245391368866,
+ "step": 1362
+ },
+ {
+ "epoch": 0.7151101783840503,
+ "grad_norm": 0.15493179714072167,
+ "learning_rate": 1.3425963043460143e-05,
+ "loss": 0.4550474286079407,
+ "step": 1363
+ },
+ {
+ "epoch": 0.7156348373557188,
+ "grad_norm": 0.15396498968995032,
+ "learning_rate": 1.3397047193595808e-05,
+ "loss": 0.44220536947250366,
+ "step": 1364
+ },
+ {
+ "epoch": 0.7161594963273872,
+ "grad_norm": 0.1489346822667273,
+ "learning_rate": 1.336816965829717e-05,
+ "loss": 0.4437323212623596,
+ "step": 1365
+ },
+ {
+ "epoch": 0.7166841552990556,
+ "grad_norm": 0.16311964931734477,
+ "learning_rate": 1.333933051601802e-05,
+ "loss": 0.45308777689933777,
+ "step": 1366
+ },
+ {
+ "epoch": 0.717208814270724,
+ "grad_norm": 0.1534354619021179,
+ "learning_rate": 1.331052984510785e-05,
+ "loss": 0.4334886372089386,
+ "step": 1367
+ },
+ {
+ "epoch": 0.7177334732423925,
+ "grad_norm": 0.15541089758065404,
+ "learning_rate": 1.328176772381165e-05,
+ "loss": 0.45248323678970337,
+ "step": 1368
+ },
+ {
+ "epoch": 0.7182581322140609,
+ "grad_norm": 0.1617921305700596,
+ "learning_rate": 1.3253044230269651e-05,
+ "loss": 0.44092947244644165,
+ "step": 1369
+ },
+ {
+ "epoch": 0.7187827911857293,
+ "grad_norm": 0.15047353793760151,
+ "learning_rate": 1.3224359442517165e-05,
+ "loss": 0.4424414336681366,
+ "step": 1370
+ },
+ {
+ "epoch": 0.7193074501573977,
+ "grad_norm": 0.15958089102258743,
+ "learning_rate": 1.319571343848433e-05,
+ "loss": 0.4458580017089844,
+ "step": 1371
+ },
+ {
+ "epoch": 0.7198321091290661,
+ "grad_norm": 0.15349258306968733,
+ "learning_rate": 1.3167106295995924e-05,
+ "loss": 0.4285386800765991,
+ "step": 1372
+ },
+ {
+ "epoch": 0.7203567681007346,
+ "grad_norm": 0.16042529563759886,
+ "learning_rate": 1.3138538092771164e-05,
+ "loss": 0.4497690200805664,
+ "step": 1373
+ },
+ {
+ "epoch": 0.720881427072403,
+ "grad_norm": 0.1565113736588754,
+ "learning_rate": 1.3110008906423454e-05,
+ "loss": 0.43562617897987366,
+ "step": 1374
+ },
+ {
+ "epoch": 0.7214060860440713,
+ "grad_norm": 0.15177291086871506,
+ "learning_rate": 1.3081518814460215e-05,
+ "loss": 0.44456908106803894,
+ "step": 1375
+ },
+ {
+ "epoch": 0.7219307450157397,
+ "grad_norm": 0.15928490480439225,
+ "learning_rate": 1.3053067894282622e-05,
+ "loss": 0.46170780062675476,
+ "step": 1376
+ },
+ {
+ "epoch": 0.7224554039874082,
+ "grad_norm": 0.1554814469217892,
+ "learning_rate": 1.302465622318549e-05,
+ "loss": 0.4422169327735901,
+ "step": 1377
+ },
+ {
+ "epoch": 0.7229800629590766,
+ "grad_norm": 0.15375370214966386,
+ "learning_rate": 1.2996283878356947e-05,
+ "loss": 0.4564763903617859,
+ "step": 1378
+ },
+ {
+ "epoch": 0.723504721930745,
+ "grad_norm": 0.15942841969395802,
+ "learning_rate": 1.2967950936878311e-05,
+ "loss": 0.45121875405311584,
+ "step": 1379
+ },
+ {
+ "epoch": 0.7240293809024134,
+ "grad_norm": 0.15731308716774903,
+ "learning_rate": 1.2939657475723844e-05,
+ "loss": 0.4285027086734772,
+ "step": 1380
+ },
+ {
+ "epoch": 0.7245540398740818,
+ "grad_norm": 0.15997753613932147,
+ "learning_rate": 1.291140357176054e-05,
+ "loss": 0.4566386342048645,
+ "step": 1381
+ },
+ {
+ "epoch": 0.7250786988457503,
+ "grad_norm": 0.15333715267473727,
+ "learning_rate": 1.2883189301747927e-05,
+ "loss": 0.45093846321105957,
+ "step": 1382
+ },
+ {
+ "epoch": 0.7256033578174187,
+ "grad_norm": 0.1514510304709061,
+ "learning_rate": 1.2855014742337874e-05,
+ "loss": 0.44256967306137085,
+ "step": 1383
+ },
+ {
+ "epoch": 0.7261280167890871,
+ "grad_norm": 0.15624301298995746,
+ "learning_rate": 1.2826879970074346e-05,
+ "loss": 0.4361875653266907,
+ "step": 1384
+ },
+ {
+ "epoch": 0.7266526757607555,
+ "grad_norm": 0.15767850122006694,
+ "learning_rate": 1.279878506139322e-05,
+ "loss": 0.4477671980857849,
+ "step": 1385
+ },
+ {
+ "epoch": 0.727177334732424,
+ "grad_norm": 0.15537676969694064,
+ "learning_rate": 1.2770730092622072e-05,
+ "loss": 0.4339832663536072,
+ "step": 1386
+ },
+ {
+ "epoch": 0.7277019937040924,
+ "grad_norm": 0.15944113633281293,
+ "learning_rate": 1.2742715139979965e-05,
+ "loss": 0.45407935976982117,
+ "step": 1387
+ },
+ {
+ "epoch": 0.7282266526757608,
+ "grad_norm": 0.16763519518606218,
+ "learning_rate": 1.271474027957727e-05,
+ "loss": 0.46442392468452454,
+ "step": 1388
+ },
+ {
+ "epoch": 0.7287513116474291,
+ "grad_norm": 0.1564572774601902,
+ "learning_rate": 1.2686805587415423e-05,
+ "loss": 0.4463968276977539,
+ "step": 1389
+ },
+ {
+ "epoch": 0.7292759706190975,
+ "grad_norm": 0.16384626314986128,
+ "learning_rate": 1.2658911139386709e-05,
+ "loss": 0.45519885420799255,
+ "step": 1390
+ },
+ {
+ "epoch": 0.729800629590766,
+ "grad_norm": 0.1562023802820964,
+ "learning_rate": 1.2631057011274106e-05,
+ "loss": 0.4365222454071045,
+ "step": 1391
+ },
+ {
+ "epoch": 0.7303252885624344,
+ "grad_norm": 0.15597287232940926,
+ "learning_rate": 1.2603243278751042e-05,
+ "loss": 0.4328266680240631,
+ "step": 1392
+ },
+ {
+ "epoch": 0.7308499475341028,
+ "grad_norm": 0.14762833706132167,
+ "learning_rate": 1.2575470017381214e-05,
+ "loss": 0.43902426958084106,
+ "step": 1393
+ },
+ {
+ "epoch": 0.7313746065057712,
+ "grad_norm": 0.1688371515856315,
+ "learning_rate": 1.2547737302618351e-05,
+ "loss": 0.4649963676929474,
+ "step": 1394
+ },
+ {
+ "epoch": 0.7318992654774397,
+ "grad_norm": 0.15501514509586015,
+ "learning_rate": 1.2520045209806024e-05,
+ "loss": 0.4501558542251587,
+ "step": 1395
+ },
+ {
+ "epoch": 0.7324239244491081,
+ "grad_norm": 0.15516748917693537,
+ "learning_rate": 1.2492393814177457e-05,
+ "loss": 0.4405498206615448,
+ "step": 1396
+ },
+ {
+ "epoch": 0.7329485834207765,
+ "grad_norm": 0.17115400468199243,
+ "learning_rate": 1.2464783190855295e-05,
+ "loss": 0.4445379674434662,
+ "step": 1397
+ },
+ {
+ "epoch": 0.7334732423924449,
+ "grad_norm": 0.15966386341604186,
+ "learning_rate": 1.243721341485143e-05,
+ "loss": 0.4769889712333679,
+ "step": 1398
+ },
+ {
+ "epoch": 0.7339979013641134,
+ "grad_norm": 0.15614706359781472,
+ "learning_rate": 1.2409684561066766e-05,
+ "loss": 0.43643918633461,
+ "step": 1399
+ },
+ {
+ "epoch": 0.7345225603357818,
+ "grad_norm": 0.17467968960064226,
+ "learning_rate": 1.238219670429104e-05,
+ "loss": 0.45413005352020264,
+ "step": 1400
+ },
+ {
+ "epoch": 0.7350472193074502,
+ "grad_norm": 0.1544517430810819,
+ "learning_rate": 1.2354749919202598e-05,
+ "loss": 0.45218425989151,
+ "step": 1401
+ },
+ {
+ "epoch": 0.7355718782791185,
+ "grad_norm": 0.16551083673374484,
+ "learning_rate": 1.2327344280368217e-05,
+ "loss": 0.4459628462791443,
+ "step": 1402
+ },
+ {
+ "epoch": 0.7360965372507869,
+ "grad_norm": 0.17614490459613263,
+ "learning_rate": 1.2299979862242882e-05,
+ "loss": 0.4406280815601349,
+ "step": 1403
+ },
+ {
+ "epoch": 0.7366211962224554,
+ "grad_norm": 0.15554925394435032,
+ "learning_rate": 1.2272656739169588e-05,
+ "loss": 0.4380159378051758,
+ "step": 1404
+ },
+ {
+ "epoch": 0.7371458551941238,
+ "grad_norm": 0.16475466799870916,
+ "learning_rate": 1.2245374985379147e-05,
+ "loss": 0.4442812204360962,
+ "step": 1405
+ },
+ {
+ "epoch": 0.7376705141657922,
+ "grad_norm": 0.16877301168875541,
+ "learning_rate": 1.2218134674989978e-05,
+ "loss": 0.4589429199695587,
+ "step": 1406
+ },
+ {
+ "epoch": 0.7381951731374606,
+ "grad_norm": 0.1583278589457438,
+ "learning_rate": 1.2190935882007894e-05,
+ "loss": 0.44896745681762695,
+ "step": 1407
+ },
+ {
+ "epoch": 0.7387198321091291,
+ "grad_norm": 0.1582022541656192,
+ "learning_rate": 1.2163778680325947e-05,
+ "loss": 0.4367997646331787,
+ "step": 1408
+ },
+ {
+ "epoch": 0.7392444910807975,
+ "grad_norm": 0.1669573982300068,
+ "learning_rate": 1.2136663143724163e-05,
+ "loss": 0.4503386616706848,
+ "step": 1409
+ },
+ {
+ "epoch": 0.7397691500524659,
+ "grad_norm": 0.1637877523924539,
+ "learning_rate": 1.2109589345869387e-05,
+ "loss": 0.4660584330558777,
+ "step": 1410
+ },
+ {
+ "epoch": 0.7402938090241343,
+ "grad_norm": 0.1640136159943701,
+ "learning_rate": 1.2082557360315065e-05,
+ "loss": 0.45206528902053833,
+ "step": 1411
+ },
+ {
+ "epoch": 0.7408184679958028,
+ "grad_norm": 0.1596580168682554,
+ "learning_rate": 1.2055567260501054e-05,
+ "loss": 0.4405035376548767,
+ "step": 1412
+ },
+ {
+ "epoch": 0.7413431269674712,
+ "grad_norm": 0.15224511741945293,
+ "learning_rate": 1.2028619119753402e-05,
+ "loss": 0.42810291051864624,
+ "step": 1413
+ },
+ {
+ "epoch": 0.7418677859391396,
+ "grad_norm": 0.15631381931817098,
+ "learning_rate": 1.2001713011284193e-05,
+ "loss": 0.4421626925468445,
+ "step": 1414
+ },
+ {
+ "epoch": 0.742392444910808,
+ "grad_norm": 0.15721819981265503,
+ "learning_rate": 1.1974849008191288e-05,
+ "loss": 0.4540696144104004,
+ "step": 1415
+ },
+ {
+ "epoch": 0.7429171038824763,
+ "grad_norm": 0.15904478578744463,
+ "learning_rate": 1.1948027183458182e-05,
+ "loss": 0.4601830244064331,
+ "step": 1416
+ },
+ {
+ "epoch": 0.7434417628541448,
+ "grad_norm": 0.1530755605235123,
+ "learning_rate": 1.1921247609953747e-05,
+ "loss": 0.4552589952945709,
+ "step": 1417
+ },
+ {
+ "epoch": 0.7439664218258132,
+ "grad_norm": 0.15375751072258462,
+ "learning_rate": 1.1894510360432106e-05,
+ "loss": 0.42965373396873474,
+ "step": 1418
+ },
+ {
+ "epoch": 0.7444910807974816,
+ "grad_norm": 0.15678686023937333,
+ "learning_rate": 1.1867815507532377e-05,
+ "loss": 0.44684016704559326,
+ "step": 1419
+ },
+ {
+ "epoch": 0.74501573976915,
+ "grad_norm": 0.15486952983783606,
+ "learning_rate": 1.1841163123778502e-05,
+ "loss": 0.44415023922920227,
+ "step": 1420
+ },
+ {
+ "epoch": 0.7455403987408185,
+ "grad_norm": 0.15342611490545593,
+ "learning_rate": 1.1814553281579034e-05,
+ "loss": 0.4501304626464844,
+ "step": 1421
+ },
+ {
+ "epoch": 0.7460650577124869,
+ "grad_norm": 0.15374260012788166,
+ "learning_rate": 1.1787986053226954e-05,
+ "loss": 0.4521749019622803,
+ "step": 1422
+ },
+ {
+ "epoch": 0.7465897166841553,
+ "grad_norm": 0.1543553258370361,
+ "learning_rate": 1.176146151089949e-05,
+ "loss": 0.44753748178482056,
+ "step": 1423
+ },
+ {
+ "epoch": 0.7471143756558237,
+ "grad_norm": 0.15921670193848306,
+ "learning_rate": 1.1734979726657878e-05,
+ "loss": 0.4473114311695099,
+ "step": 1424
+ },
+ {
+ "epoch": 0.7476390346274921,
+ "grad_norm": 0.15023498686186612,
+ "learning_rate": 1.17085407724472e-05,
+ "loss": 0.4568778872489929,
+ "step": 1425
+ },
+ {
+ "epoch": 0.7481636935991606,
+ "grad_norm": 0.14601156920274583,
+ "learning_rate": 1.1682144720096168e-05,
+ "loss": 0.4389464855194092,
+ "step": 1426
+ },
+ {
+ "epoch": 0.748688352570829,
+ "grad_norm": 0.1520548095149398,
+ "learning_rate": 1.1655791641316954e-05,
+ "loss": 0.440733402967453,
+ "step": 1427
+ },
+ {
+ "epoch": 0.7492130115424974,
+ "grad_norm": 0.1591915163097993,
+ "learning_rate": 1.1629481607704963e-05,
+ "loss": 0.46245887875556946,
+ "step": 1428
+ },
+ {
+ "epoch": 0.7497376705141658,
+ "grad_norm": 0.14867295658447263,
+ "learning_rate": 1.1603214690738683e-05,
+ "loss": 0.4332764744758606,
+ "step": 1429
+ },
+ {
+ "epoch": 0.7502623294858342,
+ "grad_norm": 0.1498957762903134,
+ "learning_rate": 1.1576990961779447e-05,
+ "loss": 0.45215466618537903,
+ "step": 1430
+ },
+ {
+ "epoch": 0.7507869884575026,
+ "grad_norm": 0.15430555803491824,
+ "learning_rate": 1.1550810492071243e-05,
+ "loss": 0.4454396963119507,
+ "step": 1431
+ },
+ {
+ "epoch": 0.751311647429171,
+ "grad_norm": 0.1517461021936533,
+ "learning_rate": 1.1524673352740543e-05,
+ "loss": 0.43869835138320923,
+ "step": 1432
+ },
+ {
+ "epoch": 0.7518363064008394,
+ "grad_norm": 0.14541657994692936,
+ "learning_rate": 1.1498579614796121e-05,
+ "loss": 0.4373054504394531,
+ "step": 1433
+ },
+ {
+ "epoch": 0.7523609653725079,
+ "grad_norm": 0.15539513763786977,
+ "learning_rate": 1.1472529349128812e-05,
+ "loss": 0.45451128482818604,
+ "step": 1434
+ },
+ {
+ "epoch": 0.7528856243441763,
+ "grad_norm": 0.1482043113690946,
+ "learning_rate": 1.144652262651136e-05,
+ "loss": 0.43644773960113525,
+ "step": 1435
+ },
+ {
+ "epoch": 0.7534102833158447,
+ "grad_norm": 0.16293628094190815,
+ "learning_rate": 1.1420559517598205e-05,
+ "loss": 0.4687269330024719,
+ "step": 1436
+ },
+ {
+ "epoch": 0.7539349422875131,
+ "grad_norm": 0.1553393974390287,
+ "learning_rate": 1.1394640092925302e-05,
+ "loss": 0.4501398205757141,
+ "step": 1437
+ },
+ {
+ "epoch": 0.7544596012591815,
+ "grad_norm": 0.16214837694535802,
+ "learning_rate": 1.1368764422909919e-05,
+ "loss": 0.4542577862739563,
+ "step": 1438
+ },
+ {
+ "epoch": 0.75498426023085,
+ "grad_norm": 0.15838878372428297,
+ "learning_rate": 1.1342932577850473e-05,
+ "loss": 0.46228161454200745,
+ "step": 1439
+ },
+ {
+ "epoch": 0.7555089192025184,
+ "grad_norm": 0.15090419497122687,
+ "learning_rate": 1.13171446279263e-05,
+ "loss": 0.4495932161808014,
+ "step": 1440
+ },
+ {
+ "epoch": 0.7560335781741868,
+ "grad_norm": 0.1543376786458394,
+ "learning_rate": 1.1291400643197484e-05,
+ "loss": 0.451630175113678,
+ "step": 1441
+ },
+ {
+ "epoch": 0.7565582371458552,
+ "grad_norm": 0.1539470751610696,
+ "learning_rate": 1.1265700693604668e-05,
+ "loss": 0.447547972202301,
+ "step": 1442
+ },
+ {
+ "epoch": 0.7570828961175237,
+ "grad_norm": 0.15761047703305445,
+ "learning_rate": 1.1240044848968856e-05,
+ "loss": 0.4516204595565796,
+ "step": 1443
+ },
+ {
+ "epoch": 0.757607555089192,
+ "grad_norm": 0.15675279683704735,
+ "learning_rate": 1.1214433178991258e-05,
+ "loss": 0.4597971737384796,
+ "step": 1444
+ },
+ {
+ "epoch": 0.7581322140608604,
+ "grad_norm": 0.15592110994904113,
+ "learning_rate": 1.1188865753253022e-05,
+ "loss": 0.443898469209671,
+ "step": 1445
+ },
+ {
+ "epoch": 0.7586568730325288,
+ "grad_norm": 0.15597501691030619,
+ "learning_rate": 1.116334264121513e-05,
+ "loss": 0.4581306278705597,
+ "step": 1446
+ },
+ {
+ "epoch": 0.7591815320041972,
+ "grad_norm": 0.15552304728589958,
+ "learning_rate": 1.1137863912218154e-05,
+ "loss": 0.4475254416465759,
+ "step": 1447
+ },
+ {
+ "epoch": 0.7597061909758657,
+ "grad_norm": 0.15759990123898035,
+ "learning_rate": 1.1112429635482118e-05,
+ "loss": 0.43585264682769775,
+ "step": 1448
+ },
+ {
+ "epoch": 0.7602308499475341,
+ "grad_norm": 0.15964657264676635,
+ "learning_rate": 1.1087039880106247e-05,
+ "loss": 0.46786290407180786,
+ "step": 1449
+ },
+ {
+ "epoch": 0.7607555089192025,
+ "grad_norm": 0.15155804051218852,
+ "learning_rate": 1.106169471506882e-05,
+ "loss": 0.42292726039886475,
+ "step": 1450
+ },
+ {
+ "epoch": 0.7612801678908709,
+ "grad_norm": 0.15519270966930884,
+ "learning_rate": 1.103639420922698e-05,
+ "loss": 0.4248972237110138,
+ "step": 1451
+ },
+ {
+ "epoch": 0.7618048268625394,
+ "grad_norm": 0.1526330511128821,
+ "learning_rate": 1.1011138431316537e-05,
+ "loss": 0.4317784309387207,
+ "step": 1452
+ },
+ {
+ "epoch": 0.7623294858342078,
+ "grad_norm": 0.15797682979272226,
+ "learning_rate": 1.0985927449951779e-05,
+ "loss": 0.44506698846817017,
+ "step": 1453
+ },
+ {
+ "epoch": 0.7628541448058762,
+ "grad_norm": 0.15479979838505462,
+ "learning_rate": 1.096076133362531e-05,
+ "loss": 0.4471631348133087,
+ "step": 1454
+ },
+ {
+ "epoch": 0.7633788037775446,
+ "grad_norm": 0.15712267274447342,
+ "learning_rate": 1.093564015070783e-05,
+ "loss": 0.4321915805339813,
+ "step": 1455
+ },
+ {
+ "epoch": 0.7639034627492131,
+ "grad_norm": 0.15857431494180016,
+ "learning_rate": 1.0910563969447969e-05,
+ "loss": 0.44455766677856445,
+ "step": 1456
+ },
+ {
+ "epoch": 0.7644281217208815,
+ "grad_norm": 0.15377057080997789,
+ "learning_rate": 1.0885532857972097e-05,
+ "loss": 0.4567924439907074,
+ "step": 1457
+ },
+ {
+ "epoch": 0.7649527806925498,
+ "grad_norm": 0.15544844005009104,
+ "learning_rate": 1.0860546884284143e-05,
+ "loss": 0.4491572976112366,
+ "step": 1458
+ },
+ {
+ "epoch": 0.7654774396642182,
+ "grad_norm": 0.16210735566479478,
+ "learning_rate": 1.08356061162654e-05,
+ "loss": 0.43823331594467163,
+ "step": 1459
+ },
+ {
+ "epoch": 0.7660020986358866,
+ "grad_norm": 0.15213559486379452,
+ "learning_rate": 1.0810710621674355e-05,
+ "loss": 0.45375770330429077,
+ "step": 1460
+ },
+ {
+ "epoch": 0.7665267576075551,
+ "grad_norm": 0.1727832306402992,
+ "learning_rate": 1.0785860468146494e-05,
+ "loss": 0.4639572501182556,
+ "step": 1461
+ },
+ {
+ "epoch": 0.7670514165792235,
+ "grad_norm": 0.1637555902013663,
+ "learning_rate": 1.076105572319412e-05,
+ "loss": 0.4355923533439636,
+ "step": 1462
+ },
+ {
+ "epoch": 0.7675760755508919,
+ "grad_norm": 0.16646936030470866,
+ "learning_rate": 1.0736296454206166e-05,
+ "loss": 0.4650861322879791,
+ "step": 1463
+ },
+ {
+ "epoch": 0.7681007345225603,
+ "grad_norm": 0.14987620615431385,
+ "learning_rate": 1.0711582728448033e-05,
+ "loss": 0.45445430278778076,
+ "step": 1464
+ },
+ {
+ "epoch": 0.7686253934942288,
+ "grad_norm": 0.16428618403547932,
+ "learning_rate": 1.0686914613061377e-05,
+ "loss": 0.46050935983657837,
+ "step": 1465
+ },
+ {
+ "epoch": 0.7691500524658972,
+ "grad_norm": 0.15052999053128632,
+ "learning_rate": 1.0662292175063949e-05,
+ "loss": 0.43385741114616394,
+ "step": 1466
+ },
+ {
+ "epoch": 0.7696747114375656,
+ "grad_norm": 0.1654949402740152,
+ "learning_rate": 1.0637715481349397e-05,
+ "loss": 0.4434410333633423,
+ "step": 1467
+ },
+ {
+ "epoch": 0.770199370409234,
+ "grad_norm": 0.15923836508490036,
+ "learning_rate": 1.0613184598687088e-05,
+ "loss": 0.4410185217857361,
+ "step": 1468
+ },
+ {
+ "epoch": 0.7707240293809025,
+ "grad_norm": 0.15813007111170696,
+ "learning_rate": 1.0588699593721957e-05,
+ "loss": 0.43588417768478394,
+ "step": 1469
+ },
+ {
+ "epoch": 0.7712486883525709,
+ "grad_norm": 0.16978817898431764,
+ "learning_rate": 1.0564260532974274e-05,
+ "loss": 0.4514843225479126,
+ "step": 1470
+ },
+ {
+ "epoch": 0.7717733473242392,
+ "grad_norm": 0.1613844852697828,
+ "learning_rate": 1.0539867482839499e-05,
+ "loss": 0.4554520845413208,
+ "step": 1471
+ },
+ {
+ "epoch": 0.7722980062959076,
+ "grad_norm": 0.1587788181984633,
+ "learning_rate": 1.0515520509588075e-05,
+ "loss": 0.43200671672821045,
+ "step": 1472
+ },
+ {
+ "epoch": 0.772822665267576,
+ "grad_norm": 0.1539930824827353,
+ "learning_rate": 1.0491219679365283e-05,
+ "loss": 0.4343116283416748,
+ "step": 1473
+ },
+ {
+ "epoch": 0.7733473242392445,
+ "grad_norm": 0.1561612537773709,
+ "learning_rate": 1.0466965058191049e-05,
+ "loss": 0.4477815330028534,
+ "step": 1474
+ },
+ {
+ "epoch": 0.7738719832109129,
+ "grad_norm": 0.169990777691657,
+ "learning_rate": 1.0442756711959745e-05,
+ "loss": 0.4476611614227295,
+ "step": 1475
+ },
+ {
+ "epoch": 0.7743966421825813,
+ "grad_norm": 0.1624313049193793,
+ "learning_rate": 1.041859470644002e-05,
+ "loss": 0.47136399149894714,
+ "step": 1476
+ },
+ {
+ "epoch": 0.7749213011542497,
+ "grad_norm": 0.15993315458994328,
+ "learning_rate": 1.0394479107274645e-05,
+ "loss": 0.4672670364379883,
+ "step": 1477
+ },
+ {
+ "epoch": 0.7754459601259182,
+ "grad_norm": 0.16755349272347217,
+ "learning_rate": 1.0370409979980294e-05,
+ "loss": 0.45135706663131714,
+ "step": 1478
+ },
+ {
+ "epoch": 0.7759706190975866,
+ "grad_norm": 0.15665059815830362,
+ "learning_rate": 1.0346387389947407e-05,
+ "loss": 0.45178359746932983,
+ "step": 1479
+ },
+ {
+ "epoch": 0.776495278069255,
+ "grad_norm": 0.1554612731452571,
+ "learning_rate": 1.0322411402439983e-05,
+ "loss": 0.4412590265274048,
+ "step": 1480
+ },
+ {
+ "epoch": 0.7770199370409234,
+ "grad_norm": 0.1532496869312065,
+ "learning_rate": 1.0298482082595408e-05,
+ "loss": 0.44104987382888794,
+ "step": 1481
+ },
+ {
+ "epoch": 0.7775445960125918,
+ "grad_norm": 0.15065868316960426,
+ "learning_rate": 1.0274599495424286e-05,
+ "loss": 0.4336603283882141,
+ "step": 1482
+ },
+ {
+ "epoch": 0.7780692549842603,
+ "grad_norm": 0.1548373981760097,
+ "learning_rate": 1.0250763705810267e-05,
+ "loss": 0.4658375680446625,
+ "step": 1483
+ },
+ {
+ "epoch": 0.7785939139559287,
+ "grad_norm": 0.15470181175866582,
+ "learning_rate": 1.0226974778509846e-05,
+ "loss": 0.45232075452804565,
+ "step": 1484
+ },
+ {
+ "epoch": 0.779118572927597,
+ "grad_norm": 0.15507371187727106,
+ "learning_rate": 1.0203232778152228e-05,
+ "loss": 0.43640458583831787,
+ "step": 1485
+ },
+ {
+ "epoch": 0.7796432318992654,
+ "grad_norm": 0.15269440736544085,
+ "learning_rate": 1.01795377692391e-05,
+ "loss": 0.4400939345359802,
+ "step": 1486
+ },
+ {
+ "epoch": 0.7801678908709339,
+ "grad_norm": 0.15610899550395085,
+ "learning_rate": 1.0155889816144495e-05,
+ "loss": 0.4546602666378021,
+ "step": 1487
+ },
+ {
+ "epoch": 0.7806925498426023,
+ "grad_norm": 0.1559328242601668,
+ "learning_rate": 1.0132288983114611e-05,
+ "loss": 0.45510709285736084,
+ "step": 1488
+ },
+ {
+ "epoch": 0.7812172088142707,
+ "grad_norm": 0.1537255194600787,
+ "learning_rate": 1.0108735334267639e-05,
+ "loss": 0.4475342631340027,
+ "step": 1489
+ },
+ {
+ "epoch": 0.7817418677859391,
+ "grad_norm": 0.14616575061664402,
+ "learning_rate": 1.0085228933593558e-05,
+ "loss": 0.4378190040588379,
+ "step": 1490
+ },
+ {
+ "epoch": 0.7822665267576075,
+ "grad_norm": 0.1624360967746598,
+ "learning_rate": 1.0061769844953999e-05,
+ "loss": 0.4439539909362793,
+ "step": 1491
+ },
+ {
+ "epoch": 0.782791185729276,
+ "grad_norm": 0.15407336125298413,
+ "learning_rate": 1.003835813208206e-05,
+ "loss": 0.44074907898902893,
+ "step": 1492
+ },
+ {
+ "epoch": 0.7833158447009444,
+ "grad_norm": 0.15786980710136125,
+ "learning_rate": 1.001499385858211e-05,
+ "loss": 0.4491667151451111,
+ "step": 1493
+ },
+ {
+ "epoch": 0.7838405036726128,
+ "grad_norm": 0.15338836738460088,
+ "learning_rate": 9.991677087929665e-06,
+ "loss": 0.446003794670105,
+ "step": 1494
+ },
+ {
+ "epoch": 0.7843651626442812,
+ "grad_norm": 0.156710894892,
+ "learning_rate": 9.968407883471165e-06,
+ "loss": 0.43604689836502075,
+ "step": 1495
+ },
+ {
+ "epoch": 0.7848898216159497,
+ "grad_norm": 0.16100539063191666,
+ "learning_rate": 9.945186308423832e-06,
+ "loss": 0.4431147873401642,
+ "step": 1496
+ },
+ {
+ "epoch": 0.7854144805876181,
+ "grad_norm": 0.15776018851122997,
+ "learning_rate": 9.92201242587548e-06,
+ "loss": 0.4532504975795746,
+ "step": 1497
+ },
+ {
+ "epoch": 0.7859391395592865,
+ "grad_norm": 0.1540333423540934,
+ "learning_rate": 9.898886298784368e-06,
+ "loss": 0.43605154752731323,
+ "step": 1498
+ },
+ {
+ "epoch": 0.7864637985309548,
+ "grad_norm": 0.16009602938972295,
+ "learning_rate": 9.875807989979005e-06,
+ "loss": 0.4605238437652588,
+ "step": 1499
+ },
+ {
+ "epoch": 0.7869884575026233,
+ "grad_norm": 0.15639368804171708,
+ "learning_rate": 9.852777562157983e-06,
+ "loss": 0.4509708881378174,
+ "step": 1500
+ },
+ {
+ "epoch": 0.7875131164742917,
+ "grad_norm": 0.15074011188741518,
+ "learning_rate": 9.829795077889828e-06,
+ "loss": 0.4527304172515869,
+ "step": 1501
+ },
+ {
+ "epoch": 0.7880377754459601,
+ "grad_norm": 0.15155083996490956,
+ "learning_rate": 9.806860599612799e-06,
+ "loss": 0.4306413531303406,
+ "step": 1502
+ },
+ {
+ "epoch": 0.7885624344176285,
+ "grad_norm": 0.1478956630795445,
+ "learning_rate": 9.783974189634734e-06,
+ "loss": 0.4434714913368225,
+ "step": 1503
+ },
+ {
+ "epoch": 0.789087093389297,
+ "grad_norm": 0.15327668434806432,
+ "learning_rate": 9.761135910132902e-06,
+ "loss": 0.44431227445602417,
+ "step": 1504
+ },
+ {
+ "epoch": 0.7896117523609654,
+ "grad_norm": 0.2243177585497212,
+ "learning_rate": 9.738345823153783e-06,
+ "loss": 0.4541007876396179,
+ "step": 1505
+ },
+ {
+ "epoch": 0.7901364113326338,
+ "grad_norm": 0.1600239865812282,
+ "learning_rate": 9.715603990612945e-06,
+ "loss": 0.4456910490989685,
+ "step": 1506
+ },
+ {
+ "epoch": 0.7906610703043022,
+ "grad_norm": 0.15304634075332066,
+ "learning_rate": 9.692910474294858e-06,
+ "loss": 0.43965545296669006,
+ "step": 1507
+ },
+ {
+ "epoch": 0.7911857292759706,
+ "grad_norm": 0.15657752062053032,
+ "learning_rate": 9.670265335852725e-06,
+ "loss": 0.43358081579208374,
+ "step": 1508
+ },
+ {
+ "epoch": 0.7917103882476391,
+ "grad_norm": 0.16110057783374115,
+ "learning_rate": 9.647668636808304e-06,
+ "loss": 0.45711368322372437,
+ "step": 1509
+ },
+ {
+ "epoch": 0.7922350472193075,
+ "grad_norm": 0.1551133690299623,
+ "learning_rate": 9.625120438551791e-06,
+ "loss": 0.4613519608974457,
+ "step": 1510
+ },
+ {
+ "epoch": 0.7927597061909759,
+ "grad_norm": 0.15690897380725946,
+ "learning_rate": 9.602620802341581e-06,
+ "loss": 0.4472261071205139,
+ "step": 1511
+ },
+ {
+ "epoch": 0.7932843651626443,
+ "grad_norm": 0.147796599322911,
+ "learning_rate": 9.580169789304152e-06,
+ "loss": 0.4447203576564789,
+ "step": 1512
+ },
+ {
+ "epoch": 0.7938090241343126,
+ "grad_norm": 0.152019559199633,
+ "learning_rate": 9.557767460433878e-06,
+ "loss": 0.445905864238739,
+ "step": 1513
+ },
+ {
+ "epoch": 0.7943336831059811,
+ "grad_norm": 0.1574763924872037,
+ "learning_rate": 9.535413876592872e-06,
+ "loss": 0.4502032995223999,
+ "step": 1514
+ },
+ {
+ "epoch": 0.7948583420776495,
+ "grad_norm": 0.1524218315499056,
+ "learning_rate": 9.513109098510816e-06,
+ "loss": 0.43753111362457275,
+ "step": 1515
+ },
+ {
+ "epoch": 0.7953830010493179,
+ "grad_norm": 0.15589147437937434,
+ "learning_rate": 9.490853186784801e-06,
+ "loss": 0.44524312019348145,
+ "step": 1516
+ },
+ {
+ "epoch": 0.7959076600209863,
+ "grad_norm": 0.15604009714081407,
+ "learning_rate": 9.468646201879156e-06,
+ "loss": 0.44141292572021484,
+ "step": 1517
+ },
+ {
+ "epoch": 0.7964323189926548,
+ "grad_norm": 0.1562432554004435,
+ "learning_rate": 9.44648820412529e-06,
+ "loss": 0.461681067943573,
+ "step": 1518
+ },
+ {
+ "epoch": 0.7969569779643232,
+ "grad_norm": 0.15233588531495665,
+ "learning_rate": 9.424379253721513e-06,
+ "loss": 0.4529497027397156,
+ "step": 1519
+ },
+ {
+ "epoch": 0.7974816369359916,
+ "grad_norm": 0.16856991685290923,
+ "learning_rate": 9.40231941073291e-06,
+ "loss": 0.4483596086502075,
+ "step": 1520
+ },
+ {
+ "epoch": 0.79800629590766,
+ "grad_norm": 0.1546876712981077,
+ "learning_rate": 9.380308735091125e-06,
+ "loss": 0.43846625089645386,
+ "step": 1521
+ },
+ {
+ "epoch": 0.7985309548793285,
+ "grad_norm": 0.15198637556890449,
+ "learning_rate": 9.35834728659424e-06,
+ "loss": 0.4625553488731384,
+ "step": 1522
+ },
+ {
+ "epoch": 0.7990556138509969,
+ "grad_norm": 0.1519777007198701,
+ "learning_rate": 9.336435124906592e-06,
+ "loss": 0.44183820486068726,
+ "step": 1523
+ },
+ {
+ "epoch": 0.7995802728226653,
+ "grad_norm": 0.16125823074979284,
+ "learning_rate": 9.314572309558616e-06,
+ "loss": 0.46057307720184326,
+ "step": 1524
+ },
+ {
+ "epoch": 0.8001049317943337,
+ "grad_norm": 0.16283323624848034,
+ "learning_rate": 9.2927588999467e-06,
+ "loss": 0.4530155062675476,
+ "step": 1525
+ },
+ {
+ "epoch": 0.8006295907660022,
+ "grad_norm": 0.14998654489027738,
+ "learning_rate": 9.270994955332984e-06,
+ "loss": 0.43155500292778015,
+ "step": 1526
+ },
+ {
+ "epoch": 0.8011542497376705,
+ "grad_norm": 0.15683095030848093,
+ "learning_rate": 9.249280534845241e-06,
+ "loss": 0.45097050070762634,
+ "step": 1527
+ },
+ {
+ "epoch": 0.8016789087093389,
+ "grad_norm": 0.1561578254492378,
+ "learning_rate": 9.227615697476674e-06,
+ "loss": 0.42797762155532837,
+ "step": 1528
+ },
+ {
+ "epoch": 0.8022035676810073,
+ "grad_norm": 0.15642385261469652,
+ "learning_rate": 9.206000502085813e-06,
+ "loss": 0.4560159742832184,
+ "step": 1529
+ },
+ {
+ "epoch": 0.8027282266526757,
+ "grad_norm": 0.1623126279097499,
+ "learning_rate": 9.1844350073963e-06,
+ "loss": 0.4704165458679199,
+ "step": 1530
+ },
+ {
+ "epoch": 0.8032528856243442,
+ "grad_norm": 0.15090934786253773,
+ "learning_rate": 9.162919271996755e-06,
+ "loss": 0.44642555713653564,
+ "step": 1531
+ },
+ {
+ "epoch": 0.8037775445960126,
+ "grad_norm": 0.15305216978534084,
+ "learning_rate": 9.14145335434062e-06,
+ "loss": 0.4634283781051636,
+ "step": 1532
+ },
+ {
+ "epoch": 0.804302203567681,
+ "grad_norm": 0.14653219861345698,
+ "learning_rate": 9.120037312745979e-06,
+ "loss": 0.44544070959091187,
+ "step": 1533
+ },
+ {
+ "epoch": 0.8048268625393494,
+ "grad_norm": 0.16247397938502464,
+ "learning_rate": 9.098671205395422e-06,
+ "loss": 0.4552367925643921,
+ "step": 1534
+ },
+ {
+ "epoch": 0.8053515215110179,
+ "grad_norm": 0.14963174165902007,
+ "learning_rate": 9.077355090335892e-06,
+ "loss": 0.43211787939071655,
+ "step": 1535
+ },
+ {
+ "epoch": 0.8058761804826863,
+ "grad_norm": 0.1554853846772283,
+ "learning_rate": 9.056089025478495e-06,
+ "loss": 0.4319286644458771,
+ "step": 1536
+ },
+ {
+ "epoch": 0.8064008394543547,
+ "grad_norm": 0.15840895821310472,
+ "learning_rate": 9.03487306859837e-06,
+ "loss": 0.4494994878768921,
+ "step": 1537
+ },
+ {
+ "epoch": 0.8069254984260231,
+ "grad_norm": 0.1543496716800982,
+ "learning_rate": 9.013707277334515e-06,
+ "loss": 0.45500510931015015,
+ "step": 1538
+ },
+ {
+ "epoch": 0.8074501573976915,
+ "grad_norm": 0.14797282318550045,
+ "learning_rate": 8.992591709189651e-06,
+ "loss": 0.4385896325111389,
+ "step": 1539
+ },
+ {
+ "epoch": 0.8079748163693599,
+ "grad_norm": 0.17045570682787758,
+ "learning_rate": 8.971526421530066e-06,
+ "loss": 0.4585988521575928,
+ "step": 1540
+ },
+ {
+ "epoch": 0.8084994753410283,
+ "grad_norm": 0.14758220713044884,
+ "learning_rate": 8.950511471585407e-06,
+ "loss": 0.4295082092285156,
+ "step": 1541
+ },
+ {
+ "epoch": 0.8090241343126967,
+ "grad_norm": 0.16005752068175036,
+ "learning_rate": 8.929546916448597e-06,
+ "loss": 0.4507679343223572,
+ "step": 1542
+ },
+ {
+ "epoch": 0.8095487932843651,
+ "grad_norm": 0.14884684083313074,
+ "learning_rate": 8.908632813075641e-06,
+ "loss": 0.43547600507736206,
+ "step": 1543
+ },
+ {
+ "epoch": 0.8100734522560336,
+ "grad_norm": 0.14860107595788136,
+ "learning_rate": 8.88776921828547e-06,
+ "loss": 0.42700642347335815,
+ "step": 1544
+ },
+ {
+ "epoch": 0.810598111227702,
+ "grad_norm": 0.1472630611843838,
+ "learning_rate": 8.86695618875981e-06,
+ "loss": 0.4473339319229126,
+ "step": 1545
+ },
+ {
+ "epoch": 0.8111227701993704,
+ "grad_norm": 0.15105523788317893,
+ "learning_rate": 8.846193781043e-06,
+ "loss": 0.44923150539398193,
+ "step": 1546
+ },
+ {
+ "epoch": 0.8116474291710388,
+ "grad_norm": 0.16121409015263036,
+ "learning_rate": 8.82548205154185e-06,
+ "loss": 0.4653576612472534,
+ "step": 1547
+ },
+ {
+ "epoch": 0.8121720881427072,
+ "grad_norm": 0.1595195150959421,
+ "learning_rate": 8.804821056525497e-06,
+ "loss": 0.4522424638271332,
+ "step": 1548
+ },
+ {
+ "epoch": 0.8126967471143757,
+ "grad_norm": 0.15163510106316194,
+ "learning_rate": 8.784210852125232e-06,
+ "loss": 0.43657630681991577,
+ "step": 1549
+ },
+ {
+ "epoch": 0.8132214060860441,
+ "grad_norm": 0.15155109347013793,
+ "learning_rate": 8.763651494334377e-06,
+ "loss": 0.43183624744415283,
+ "step": 1550
+ },
+ {
+ "epoch": 0.8137460650577125,
+ "grad_norm": 0.14930024958763324,
+ "learning_rate": 8.7431430390081e-06,
+ "loss": 0.43563947081565857,
+ "step": 1551
+ },
+ {
+ "epoch": 0.8142707240293809,
+ "grad_norm": 0.15361200153430712,
+ "learning_rate": 8.722685541863283e-06,
+ "loss": 0.4387405514717102,
+ "step": 1552
+ },
+ {
+ "epoch": 0.8147953830010494,
+ "grad_norm": 0.15549528796193493,
+ "learning_rate": 8.702279058478366e-06,
+ "loss": 0.44347938895225525,
+ "step": 1553
+ },
+ {
+ "epoch": 0.8153200419727177,
+ "grad_norm": 0.16243042867861712,
+ "learning_rate": 8.681923644293194e-06,
+ "loss": 0.44852834939956665,
+ "step": 1554
+ },
+ {
+ "epoch": 0.8158447009443861,
+ "grad_norm": 0.15706129652944026,
+ "learning_rate": 8.661619354608873e-06,
+ "loss": 0.45419150590896606,
+ "step": 1555
+ },
+ {
+ "epoch": 0.8163693599160545,
+ "grad_norm": 0.14854258573134127,
+ "learning_rate": 8.641366244587611e-06,
+ "loss": 0.44866859912872314,
+ "step": 1556
+ },
+ {
+ "epoch": 0.816894018887723,
+ "grad_norm": 0.14896937678017533,
+ "learning_rate": 8.62116436925258e-06,
+ "loss": 0.44596076011657715,
+ "step": 1557
+ },
+ {
+ "epoch": 0.8174186778593914,
+ "grad_norm": 0.15614508495620508,
+ "learning_rate": 8.601013783487746e-06,
+ "loss": 0.45645207166671753,
+ "step": 1558
+ },
+ {
+ "epoch": 0.8179433368310598,
+ "grad_norm": 0.1486538543050606,
+ "learning_rate": 8.580914542037734e-06,
+ "loss": 0.4592300355434418,
+ "step": 1559
+ },
+ {
+ "epoch": 0.8184679958027282,
+ "grad_norm": 0.15475320756830518,
+ "learning_rate": 8.560866699507702e-06,
+ "loss": 0.4605134129524231,
+ "step": 1560
+ },
+ {
+ "epoch": 0.8189926547743966,
+ "grad_norm": 0.14754406728057914,
+ "learning_rate": 8.540870310363137e-06,
+ "loss": 0.43651673197746277,
+ "step": 1561
+ },
+ {
+ "epoch": 0.8195173137460651,
+ "grad_norm": 0.14871990701077825,
+ "learning_rate": 8.520925428929755e-06,
+ "loss": 0.44830837845802307,
+ "step": 1562
+ },
+ {
+ "epoch": 0.8200419727177335,
+ "grad_norm": 0.15358241317712973,
+ "learning_rate": 8.501032109393334e-06,
+ "loss": 0.4546644687652588,
+ "step": 1563
+ },
+ {
+ "epoch": 0.8205666316894019,
+ "grad_norm": 0.15650856948126138,
+ "learning_rate": 8.481190405799569e-06,
+ "loss": 0.4423891305923462,
+ "step": 1564
+ },
+ {
+ "epoch": 0.8210912906610703,
+ "grad_norm": 0.15474996746242287,
+ "learning_rate": 8.461400372053923e-06,
+ "loss": 0.44242313504219055,
+ "step": 1565
+ },
+ {
+ "epoch": 0.8216159496327388,
+ "grad_norm": 0.16544926591962897,
+ "learning_rate": 8.441662061921497e-06,
+ "loss": 0.4487020969390869,
+ "step": 1566
+ },
+ {
+ "epoch": 0.8221406086044072,
+ "grad_norm": 0.1489733586650647,
+ "learning_rate": 8.421975529026855e-06,
+ "loss": 0.4371356964111328,
+ "step": 1567
+ },
+ {
+ "epoch": 0.8226652675760755,
+ "grad_norm": 0.15995455652906124,
+ "learning_rate": 8.402340826853907e-06,
+ "loss": 0.4566021263599396,
+ "step": 1568
+ },
+ {
+ "epoch": 0.8231899265477439,
+ "grad_norm": 0.1500716152622102,
+ "learning_rate": 8.382758008745726e-06,
+ "loss": 0.43888187408447266,
+ "step": 1569
+ },
+ {
+ "epoch": 0.8237145855194123,
+ "grad_norm": 0.15554904485712012,
+ "learning_rate": 8.363227127904464e-06,
+ "loss": 0.4584001898765564,
+ "step": 1570
+ },
+ {
+ "epoch": 0.8242392444910808,
+ "grad_norm": 0.14820917815884277,
+ "learning_rate": 8.343748237391144e-06,
+ "loss": 0.43289071321487427,
+ "step": 1571
+ },
+ {
+ "epoch": 0.8247639034627492,
+ "grad_norm": 0.15801200747859714,
+ "learning_rate": 8.324321390125547e-06,
+ "loss": 0.44196009635925293,
+ "step": 1572
+ },
+ {
+ "epoch": 0.8252885624344176,
+ "grad_norm": 0.15323277975380797,
+ "learning_rate": 8.304946638886078e-06,
+ "loss": 0.43244463205337524,
+ "step": 1573
+ },
+ {
+ "epoch": 0.825813221406086,
+ "grad_norm": 0.15076940176344772,
+ "learning_rate": 8.285624036309583e-06,
+ "loss": 0.44401073455810547,
+ "step": 1574
+ },
+ {
+ "epoch": 0.8263378803777545,
+ "grad_norm": 0.1544958300475393,
+ "learning_rate": 8.266353634891266e-06,
+ "loss": 0.45198434591293335,
+ "step": 1575
+ },
+ {
+ "epoch": 0.8268625393494229,
+ "grad_norm": 0.15093528292068448,
+ "learning_rate": 8.247135486984486e-06,
+ "loss": 0.4403747618198395,
+ "step": 1576
+ },
+ {
+ "epoch": 0.8273871983210913,
+ "grad_norm": 0.14680662784017148,
+ "learning_rate": 8.227969644800648e-06,
+ "loss": 0.4375043213367462,
+ "step": 1577
+ },
+ {
+ "epoch": 0.8279118572927597,
+ "grad_norm": 0.14791704009291914,
+ "learning_rate": 8.208856160409055e-06,
+ "loss": 0.43811577558517456,
+ "step": 1578
+ },
+ {
+ "epoch": 0.8284365162644282,
+ "grad_norm": 0.14924185710518617,
+ "learning_rate": 8.189795085736765e-06,
+ "loss": 0.43853428959846497,
+ "step": 1579
+ },
+ {
+ "epoch": 0.8289611752360966,
+ "grad_norm": 0.15420576936040198,
+ "learning_rate": 8.170786472568451e-06,
+ "loss": 0.445467084646225,
+ "step": 1580
+ },
+ {
+ "epoch": 0.829485834207765,
+ "grad_norm": 0.15577901772605834,
+ "learning_rate": 8.151830372546258e-06,
+ "loss": 0.4529452621936798,
+ "step": 1581
+ },
+ {
+ "epoch": 0.8300104931794333,
+ "grad_norm": 0.15387840488797552,
+ "learning_rate": 8.132926837169676e-06,
+ "loss": 0.46009817719459534,
+ "step": 1582
+ },
+ {
+ "epoch": 0.8305351521511017,
+ "grad_norm": 0.1527913309957562,
+ "learning_rate": 8.11407591779537e-06,
+ "loss": 0.4513983726501465,
+ "step": 1583
+ },
+ {
+ "epoch": 0.8310598111227702,
+ "grad_norm": 0.1537090286922398,
+ "learning_rate": 8.095277665637061e-06,
+ "loss": 0.4537739157676697,
+ "step": 1584
+ },
+ {
+ "epoch": 0.8315844700944386,
+ "grad_norm": 0.15722962107654076,
+ "learning_rate": 8.076532131765408e-06,
+ "loss": 0.45232969522476196,
+ "step": 1585
+ },
+ {
+ "epoch": 0.832109129066107,
+ "grad_norm": 0.1450142396611787,
+ "learning_rate": 8.057839367107817e-06,
+ "loss": 0.4270634651184082,
+ "step": 1586
+ },
+ {
+ "epoch": 0.8326337880377754,
+ "grad_norm": 0.150264001570592,
+ "learning_rate": 8.039199422448355e-06,
+ "loss": 0.43563878536224365,
+ "step": 1587
+ },
+ {
+ "epoch": 0.8331584470094439,
+ "grad_norm": 0.14904943307590535,
+ "learning_rate": 8.020612348427572e-06,
+ "loss": 0.4401509165763855,
+ "step": 1588
+ },
+ {
+ "epoch": 0.8336831059811123,
+ "grad_norm": 0.15453821662860034,
+ "learning_rate": 8.002078195542393e-06,
+ "loss": 0.44545161724090576,
+ "step": 1589
+ },
+ {
+ "epoch": 0.8342077649527807,
+ "grad_norm": 0.1570809678830383,
+ "learning_rate": 7.983597014145951e-06,
+ "loss": 0.45228466391563416,
+ "step": 1590
+ },
+ {
+ "epoch": 0.8347324239244491,
+ "grad_norm": 0.14741240769629818,
+ "learning_rate": 7.965168854447493e-06,
+ "loss": 0.4485381245613098,
+ "step": 1591
+ },
+ {
+ "epoch": 0.8352570828961176,
+ "grad_norm": 0.15362267095533516,
+ "learning_rate": 7.946793766512195e-06,
+ "loss": 0.4358059763908386,
+ "step": 1592
+ },
+ {
+ "epoch": 0.835781741867786,
+ "grad_norm": 0.1544754586495605,
+ "learning_rate": 7.92847180026106e-06,
+ "loss": 0.43327778577804565,
+ "step": 1593
+ },
+ {
+ "epoch": 0.8363064008394544,
+ "grad_norm": 0.15023225026164963,
+ "learning_rate": 7.91020300547077e-06,
+ "loss": 0.44359731674194336,
+ "step": 1594
+ },
+ {
+ "epoch": 0.8368310598111228,
+ "grad_norm": 0.15665530392095345,
+ "learning_rate": 7.891987431773545e-06,
+ "loss": 0.44503387808799744,
+ "step": 1595
+ },
+ {
+ "epoch": 0.8373557187827911,
+ "grad_norm": 0.14717614926392292,
+ "learning_rate": 7.873825128657033e-06,
+ "loss": 0.45314061641693115,
+ "step": 1596
+ },
+ {
+ "epoch": 0.8378803777544596,
+ "grad_norm": 0.15207464934030612,
+ "learning_rate": 7.855716145464131e-06,
+ "loss": 0.4526442885398865,
+ "step": 1597
+ },
+ {
+ "epoch": 0.838405036726128,
+ "grad_norm": 0.15129881541794873,
+ "learning_rate": 7.837660531392904e-06,
+ "loss": 0.45127853751182556,
+ "step": 1598
+ },
+ {
+ "epoch": 0.8389296956977964,
+ "grad_norm": 0.15623362715417285,
+ "learning_rate": 7.819658335496405e-06,
+ "loss": 0.46432793140411377,
+ "step": 1599
+ },
+ {
+ "epoch": 0.8394543546694648,
+ "grad_norm": 0.1529219879642125,
+ "learning_rate": 7.801709606682573e-06,
+ "loss": 0.4475358724594116,
+ "step": 1600
+ },
+ {
+ "epoch": 0.8399790136411333,
+ "grad_norm": 0.14649139854087698,
+ "learning_rate": 7.78381439371409e-06,
+ "loss": 0.4363391101360321,
+ "step": 1601
+ },
+ {
+ "epoch": 0.8405036726128017,
+ "grad_norm": 0.14571233251948087,
+ "learning_rate": 7.765972745208243e-06,
+ "loss": 0.44419488310813904,
+ "step": 1602
+ },
+ {
+ "epoch": 0.8410283315844701,
+ "grad_norm": 0.15168460066406128,
+ "learning_rate": 7.748184709636791e-06,
+ "loss": 0.4433790445327759,
+ "step": 1603
+ },
+ {
+ "epoch": 0.8415529905561385,
+ "grad_norm": 0.14852860876513757,
+ "learning_rate": 7.730450335325853e-06,
+ "loss": 0.4493539333343506,
+ "step": 1604
+ },
+ {
+ "epoch": 0.842077649527807,
+ "grad_norm": 0.1505615491087377,
+ "learning_rate": 7.712769670455741e-06,
+ "loss": 0.4646318554878235,
+ "step": 1605
+ },
+ {
+ "epoch": 0.8426023084994754,
+ "grad_norm": 0.15407610486660114,
+ "learning_rate": 7.695142763060878e-06,
+ "loss": 0.4484049677848816,
+ "step": 1606
+ },
+ {
+ "epoch": 0.8431269674711438,
+ "grad_norm": 0.1521449709584124,
+ "learning_rate": 7.67756966102962e-06,
+ "loss": 0.43366891145706177,
+ "step": 1607
+ },
+ {
+ "epoch": 0.8436516264428122,
+ "grad_norm": 0.14747762647596166,
+ "learning_rate": 7.660050412104153e-06,
+ "loss": 0.43950092792510986,
+ "step": 1608
+ },
+ {
+ "epoch": 0.8441762854144806,
+ "grad_norm": 0.15512610809240027,
+ "learning_rate": 7.642585063880351e-06,
+ "loss": 0.46813055872917175,
+ "step": 1609
+ },
+ {
+ "epoch": 0.844700944386149,
+ "grad_norm": 0.15825480008734216,
+ "learning_rate": 7.625173663807655e-06,
+ "loss": 0.4348755478858948,
+ "step": 1610
+ },
+ {
+ "epoch": 0.8452256033578174,
+ "grad_norm": 0.1573968981854278,
+ "learning_rate": 7.6078162591889434e-06,
+ "loss": 0.45048123598098755,
+ "step": 1611
+ },
+ {
+ "epoch": 0.8457502623294858,
+ "grad_norm": 0.1517038960096417,
+ "learning_rate": 7.590512897180399e-06,
+ "loss": 0.43563610315322876,
+ "step": 1612
+ },
+ {
+ "epoch": 0.8462749213011542,
+ "grad_norm": 0.15503576044724665,
+ "learning_rate": 7.573263624791382e-06,
+ "loss": 0.4467726945877075,
+ "step": 1613
+ },
+ {
+ "epoch": 0.8467995802728226,
+ "grad_norm": 0.1560549927192375,
+ "learning_rate": 7.556068488884304e-06,
+ "loss": 0.4320477545261383,
+ "step": 1614
+ },
+ {
+ "epoch": 0.8473242392444911,
+ "grad_norm": 0.1679830356844654,
+ "learning_rate": 7.538927536174498e-06,
+ "loss": 0.4476034343242645,
+ "step": 1615
+ },
+ {
+ "epoch": 0.8478488982161595,
+ "grad_norm": 0.15169257383162213,
+ "learning_rate": 7.521840813230098e-06,
+ "loss": 0.4325854778289795,
+ "step": 1616
+ },
+ {
+ "epoch": 0.8483735571878279,
+ "grad_norm": 0.15980226324329827,
+ "learning_rate": 7.504808366471909e-06,
+ "loss": 0.4472721815109253,
+ "step": 1617
+ },
+ {
+ "epoch": 0.8488982161594963,
+ "grad_norm": 0.16932727986736365,
+ "learning_rate": 7.48783024217327e-06,
+ "loss": 0.45428287982940674,
+ "step": 1618
+ },
+ {
+ "epoch": 0.8494228751311648,
+ "grad_norm": 0.15065676306311376,
+ "learning_rate": 7.470906486459946e-06,
+ "loss": 0.44418883323669434,
+ "step": 1619
+ },
+ {
+ "epoch": 0.8499475341028332,
+ "grad_norm": 0.1533992998314183,
+ "learning_rate": 7.454037145309986e-06,
+ "loss": 0.43552401661872864,
+ "step": 1620
+ },
+ {
+ "epoch": 0.8504721930745016,
+ "grad_norm": 0.15479480515120017,
+ "learning_rate": 7.437222264553628e-06,
+ "loss": 0.444465696811676,
+ "step": 1621
+ },
+ {
+ "epoch": 0.85099685204617,
+ "grad_norm": 0.15845678337942112,
+ "learning_rate": 7.420461889873129e-06,
+ "loss": 0.44896578788757324,
+ "step": 1622
+ },
+ {
+ "epoch": 0.8515215110178384,
+ "grad_norm": 0.15038940533473208,
+ "learning_rate": 7.4037560668026804e-06,
+ "loss": 0.43062180280685425,
+ "step": 1623
+ },
+ {
+ "epoch": 0.8520461699895068,
+ "grad_norm": 0.1480017115109342,
+ "learning_rate": 7.387104840728259e-06,
+ "loss": 0.4630432724952698,
+ "step": 1624
+ },
+ {
+ "epoch": 0.8525708289611752,
+ "grad_norm": 0.15321305032824092,
+ "learning_rate": 7.370508256887514e-06,
+ "loss": 0.4419487714767456,
+ "step": 1625
+ },
+ {
+ "epoch": 0.8530954879328436,
+ "grad_norm": 0.1568246735699272,
+ "learning_rate": 7.353966360369659e-06,
+ "loss": 0.4483609199523926,
+ "step": 1626
+ },
+ {
+ "epoch": 0.853620146904512,
+ "grad_norm": 0.15539536583814875,
+ "learning_rate": 7.337479196115317e-06,
+ "loss": 0.4567753076553345,
+ "step": 1627
+ },
+ {
+ "epoch": 0.8541448058761805,
+ "grad_norm": 0.15528695146633797,
+ "learning_rate": 7.321046808916426e-06,
+ "loss": 0.4560334086418152,
+ "step": 1628
+ },
+ {
+ "epoch": 0.8546694648478489,
+ "grad_norm": 0.15005728554864303,
+ "learning_rate": 7.304669243416102e-06,
+ "loss": 0.43833765387535095,
+ "step": 1629
+ },
+ {
+ "epoch": 0.8551941238195173,
+ "grad_norm": 0.15959100645189242,
+ "learning_rate": 7.288346544108518e-06,
+ "loss": 0.4713831841945648,
+ "step": 1630
+ },
+ {
+ "epoch": 0.8557187827911857,
+ "grad_norm": 0.14898121463872632,
+ "learning_rate": 7.272078755338807e-06,
+ "loss": 0.44003745913505554,
+ "step": 1631
+ },
+ {
+ "epoch": 0.8562434417628542,
+ "grad_norm": 0.15083110802343333,
+ "learning_rate": 7.255865921302903e-06,
+ "loss": 0.429004043340683,
+ "step": 1632
+ },
+ {
+ "epoch": 0.8567681007345226,
+ "grad_norm": 0.1522757086796616,
+ "learning_rate": 7.239708086047445e-06,
+ "loss": 0.4403839707374573,
+ "step": 1633
+ },
+ {
+ "epoch": 0.857292759706191,
+ "grad_norm": 0.15719824838921245,
+ "learning_rate": 7.223605293469654e-06,
+ "loss": 0.4377821087837219,
+ "step": 1634
+ },
+ {
+ "epoch": 0.8578174186778594,
+ "grad_norm": 0.1450616202708373,
+ "learning_rate": 7.207557587317216e-06,
+ "loss": 0.4357878267765045,
+ "step": 1635
+ },
+ {
+ "epoch": 0.8583420776495279,
+ "grad_norm": 0.15703162220646427,
+ "learning_rate": 7.191565011188149e-06,
+ "loss": 0.44412243366241455,
+ "step": 1636
+ },
+ {
+ "epoch": 0.8588667366211962,
+ "grad_norm": 0.150175313334118,
+ "learning_rate": 7.175627608530716e-06,
+ "loss": 0.4543853998184204,
+ "step": 1637
+ },
+ {
+ "epoch": 0.8593913955928646,
+ "grad_norm": 0.15091711492716975,
+ "learning_rate": 7.159745422643261e-06,
+ "loss": 0.4268660545349121,
+ "step": 1638
+ },
+ {
+ "epoch": 0.859916054564533,
+ "grad_norm": 0.15449835654964011,
+ "learning_rate": 7.1439184966741295e-06,
+ "loss": 0.4507150650024414,
+ "step": 1639
+ },
+ {
+ "epoch": 0.8604407135362014,
+ "grad_norm": 0.15689612593442026,
+ "learning_rate": 7.128146873621533e-06,
+ "loss": 0.45520636439323425,
+ "step": 1640
+ },
+ {
+ "epoch": 0.8609653725078699,
+ "grad_norm": 0.1565226785234465,
+ "learning_rate": 7.11243059633345e-06,
+ "loss": 0.45102012157440186,
+ "step": 1641
+ },
+ {
+ "epoch": 0.8614900314795383,
+ "grad_norm": 0.15500877922230677,
+ "learning_rate": 7.096769707507481e-06,
+ "loss": 0.447115421295166,
+ "step": 1642
+ },
+ {
+ "epoch": 0.8620146904512067,
+ "grad_norm": 0.1595248428000243,
+ "learning_rate": 7.081164249690761e-06,
+ "loss": 0.4627235233783722,
+ "step": 1643
+ },
+ {
+ "epoch": 0.8625393494228751,
+ "grad_norm": 0.15958177510163488,
+ "learning_rate": 7.065614265279818e-06,
+ "loss": 0.4386388063430786,
+ "step": 1644
+ },
+ {
+ "epoch": 0.8630640083945436,
+ "grad_norm": 0.15152834825396286,
+ "learning_rate": 7.050119796520482e-06,
+ "loss": 0.4531296491622925,
+ "step": 1645
+ },
+ {
+ "epoch": 0.863588667366212,
+ "grad_norm": 0.15635733905345023,
+ "learning_rate": 7.034680885507752e-06,
+ "loss": 0.45573747158050537,
+ "step": 1646
+ },
+ {
+ "epoch": 0.8641133263378804,
+ "grad_norm": 0.15666918852633857,
+ "learning_rate": 7.019297574185703e-06,
+ "loss": 0.4499792158603668,
+ "step": 1647
+ },
+ {
+ "epoch": 0.8646379853095488,
+ "grad_norm": 0.156444627477174,
+ "learning_rate": 7.0039699043473405e-06,
+ "loss": 0.4545009136199951,
+ "step": 1648
+ },
+ {
+ "epoch": 0.8651626442812173,
+ "grad_norm": 0.14638989347446527,
+ "learning_rate": 6.988697917634515e-06,
+ "loss": 0.4397326409816742,
+ "step": 1649
+ },
+ {
+ "epoch": 0.8656873032528857,
+ "grad_norm": 0.15492813701343372,
+ "learning_rate": 6.973481655537793e-06,
+ "loss": 0.4402945041656494,
+ "step": 1650
+ },
+ {
+ "epoch": 0.866211962224554,
+ "grad_norm": 0.14466615385057716,
+ "learning_rate": 6.958321159396352e-06,
+ "loss": 0.4479280114173889,
+ "step": 1651
+ },
+ {
+ "epoch": 0.8667366211962224,
+ "grad_norm": 0.15865895330952856,
+ "learning_rate": 6.943216470397871e-06,
+ "loss": 0.4553218483924866,
+ "step": 1652
+ },
+ {
+ "epoch": 0.8672612801678908,
+ "grad_norm": 0.15293327582467067,
+ "learning_rate": 6.928167629578407e-06,
+ "loss": 0.44509196281433105,
+ "step": 1653
+ },
+ {
+ "epoch": 0.8677859391395593,
+ "grad_norm": 0.1515507946195896,
+ "learning_rate": 6.91317467782229e-06,
+ "loss": 0.4498351216316223,
+ "step": 1654
+ },
+ {
+ "epoch": 0.8683105981112277,
+ "grad_norm": 0.15313372799407085,
+ "learning_rate": 6.898237655862007e-06,
+ "loss": 0.4445887804031372,
+ "step": 1655
+ },
+ {
+ "epoch": 0.8688352570828961,
+ "grad_norm": 0.1491970050887588,
+ "learning_rate": 6.883356604278115e-06,
+ "loss": 0.4323808252811432,
+ "step": 1656
+ },
+ {
+ "epoch": 0.8693599160545645,
+ "grad_norm": 0.14449269685472518,
+ "learning_rate": 6.868531563499096e-06,
+ "loss": 0.42508652806282043,
+ "step": 1657
+ },
+ {
+ "epoch": 0.869884575026233,
+ "grad_norm": 0.14997328847830618,
+ "learning_rate": 6.853762573801264e-06,
+ "loss": 0.43655943870544434,
+ "step": 1658
+ },
+ {
+ "epoch": 0.8704092339979014,
+ "grad_norm": 0.15288208370561934,
+ "learning_rate": 6.839049675308662e-06,
+ "loss": 0.4505232274532318,
+ "step": 1659
+ },
+ {
+ "epoch": 0.8709338929695698,
+ "grad_norm": 0.14416201723141706,
+ "learning_rate": 6.824392907992938e-06,
+ "loss": 0.42444318532943726,
+ "step": 1660
+ },
+ {
+ "epoch": 0.8714585519412382,
+ "grad_norm": 0.14468507187223545,
+ "learning_rate": 6.809792311673244e-06,
+ "loss": 0.4251434803009033,
+ "step": 1661
+ },
+ {
+ "epoch": 0.8719832109129066,
+ "grad_norm": 0.15406295659798216,
+ "learning_rate": 6.795247926016142e-06,
+ "loss": 0.45068371295928955,
+ "step": 1662
+ },
+ {
+ "epoch": 0.8725078698845751,
+ "grad_norm": 0.1516721215663676,
+ "learning_rate": 6.780759790535465e-06,
+ "loss": 0.441638320684433,
+ "step": 1663
+ },
+ {
+ "epoch": 0.8730325288562435,
+ "grad_norm": 0.1537661139082594,
+ "learning_rate": 6.766327944592239e-06,
+ "loss": 0.44044923782348633,
+ "step": 1664
+ },
+ {
+ "epoch": 0.8735571878279118,
+ "grad_norm": 0.15175839438218855,
+ "learning_rate": 6.751952427394553e-06,
+ "loss": 0.44454342126846313,
+ "step": 1665
+ },
+ {
+ "epoch": 0.8740818467995802,
+ "grad_norm": 0.1531195282211392,
+ "learning_rate": 6.737633277997471e-06,
+ "loss": 0.4540765881538391,
+ "step": 1666
+ },
+ {
+ "epoch": 0.8746065057712487,
+ "grad_norm": 0.14961744003585584,
+ "learning_rate": 6.723370535302915e-06,
+ "loss": 0.44223541021347046,
+ "step": 1667
+ },
+ {
+ "epoch": 0.8751311647429171,
+ "grad_norm": 0.14816820299503658,
+ "learning_rate": 6.709164238059568e-06,
+ "loss": 0.43701833486557007,
+ "step": 1668
+ },
+ {
+ "epoch": 0.8756558237145855,
+ "grad_norm": 0.14622402294164918,
+ "learning_rate": 6.695014424862756e-06,
+ "loss": 0.42922335863113403,
+ "step": 1669
+ },
+ {
+ "epoch": 0.8761804826862539,
+ "grad_norm": 0.1547291062309004,
+ "learning_rate": 6.680921134154355e-06,
+ "loss": 0.4233800768852234,
+ "step": 1670
+ },
+ {
+ "epoch": 0.8767051416579223,
+ "grad_norm": 0.15378903523440587,
+ "learning_rate": 6.666884404222678e-06,
+ "loss": 0.4432950019836426,
+ "step": 1671
+ },
+ {
+ "epoch": 0.8772298006295908,
+ "grad_norm": 0.15973037860999773,
+ "learning_rate": 6.652904273202384e-06,
+ "loss": 0.4560699462890625,
+ "step": 1672
+ },
+ {
+ "epoch": 0.8777544596012592,
+ "grad_norm": 0.15430233088028492,
+ "learning_rate": 6.638980779074358e-06,
+ "loss": 0.45776498317718506,
+ "step": 1673
+ },
+ {
+ "epoch": 0.8782791185729276,
+ "grad_norm": 0.1587664730992995,
+ "learning_rate": 6.625113959665614e-06,
+ "loss": 0.43579810857772827,
+ "step": 1674
+ },
+ {
+ "epoch": 0.878803777544596,
+ "grad_norm": 0.14370919039693034,
+ "learning_rate": 6.6113038526491946e-06,
+ "loss": 0.4426189064979553,
+ "step": 1675
+ },
+ {
+ "epoch": 0.8793284365162645,
+ "grad_norm": 0.14933009355373217,
+ "learning_rate": 6.59755049554407e-06,
+ "loss": 0.44747287034988403,
+ "step": 1676
+ },
+ {
+ "epoch": 0.8798530954879329,
+ "grad_norm": 0.15713697891738065,
+ "learning_rate": 6.583853925715039e-06,
+ "loss": 0.46320077776908875,
+ "step": 1677
+ },
+ {
+ "epoch": 0.8803777544596013,
+ "grad_norm": 0.1542096833927892,
+ "learning_rate": 6.570214180372609e-06,
+ "loss": 0.4473888874053955,
+ "step": 1678
+ },
+ {
+ "epoch": 0.8809024134312696,
+ "grad_norm": 0.16187259968193946,
+ "learning_rate": 6.556631296572922e-06,
+ "loss": 0.4421869218349457,
+ "step": 1679
+ },
+ {
+ "epoch": 0.881427072402938,
+ "grad_norm": 0.15978032996583555,
+ "learning_rate": 6.54310531121762e-06,
+ "loss": 0.4305853843688965,
+ "step": 1680
+ },
+ {
+ "epoch": 0.8819517313746065,
+ "grad_norm": 0.15631580579775123,
+ "learning_rate": 6.529636261053784e-06,
+ "loss": 0.4340512454509735,
+ "step": 1681
+ },
+ {
+ "epoch": 0.8824763903462749,
+ "grad_norm": 0.150084161390103,
+ "learning_rate": 6.516224182673812e-06,
+ "loss": 0.4383493661880493,
+ "step": 1682
+ },
+ {
+ "epoch": 0.8830010493179433,
+ "grad_norm": 0.15527917215838966,
+ "learning_rate": 6.502869112515314e-06,
+ "loss": 0.45441877841949463,
+ "step": 1683
+ },
+ {
+ "epoch": 0.8835257082896117,
+ "grad_norm": 0.15164994594035522,
+ "learning_rate": 6.489571086861027e-06,
+ "loss": 0.43363988399505615,
+ "step": 1684
+ },
+ {
+ "epoch": 0.8840503672612802,
+ "grad_norm": 0.1467230535452683,
+ "learning_rate": 6.47633014183871e-06,
+ "loss": 0.4263266623020172,
+ "step": 1685
+ },
+ {
+ "epoch": 0.8845750262329486,
+ "grad_norm": 0.14597001769185472,
+ "learning_rate": 6.463146313421044e-06,
+ "loss": 0.4223391115665436,
+ "step": 1686
+ },
+ {
+ "epoch": 0.885099685204617,
+ "grad_norm": 0.15050446002714588,
+ "learning_rate": 6.4500196374255435e-06,
+ "loss": 0.43097931146621704,
+ "step": 1687
+ },
+ {
+ "epoch": 0.8856243441762854,
+ "grad_norm": 0.14918033614543294,
+ "learning_rate": 6.436950149514448e-06,
+ "loss": 0.4587080478668213,
+ "step": 1688
+ },
+ {
+ "epoch": 0.8861490031479539,
+ "grad_norm": 0.15469492243631364,
+ "learning_rate": 6.423937885194631e-06,
+ "loss": 0.45661866664886475,
+ "step": 1689
+ },
+ {
+ "epoch": 0.8866736621196223,
+ "grad_norm": 0.15552798640177423,
+ "learning_rate": 6.410982879817498e-06,
+ "loss": 0.44370236992836,
+ "step": 1690
+ },
+ {
+ "epoch": 0.8871983210912907,
+ "grad_norm": 0.15595593365088142,
+ "learning_rate": 6.398085168578901e-06,
+ "loss": 0.4291878342628479,
+ "step": 1691
+ },
+ {
+ "epoch": 0.887722980062959,
+ "grad_norm": 0.14708046948755318,
+ "learning_rate": 6.385244786519036e-06,
+ "loss": 0.4422011375427246,
+ "step": 1692
+ },
+ {
+ "epoch": 0.8882476390346274,
+ "grad_norm": 0.1499253382866992,
+ "learning_rate": 6.3724617685223404e-06,
+ "loss": 0.4595015048980713,
+ "step": 1693
+ },
+ {
+ "epoch": 0.8887722980062959,
+ "grad_norm": 0.14702868569252941,
+ "learning_rate": 6.3597361493174205e-06,
+ "loss": 0.4347558319568634,
+ "step": 1694
+ },
+ {
+ "epoch": 0.8892969569779643,
+ "grad_norm": 0.14977289247327386,
+ "learning_rate": 6.347067963476928e-06,
+ "loss": 0.43877455592155457,
+ "step": 1695
+ },
+ {
+ "epoch": 0.8898216159496327,
+ "grad_norm": 0.1453902077891248,
+ "learning_rate": 6.334457245417491e-06,
+ "loss": 0.4282415807247162,
+ "step": 1696
+ },
+ {
+ "epoch": 0.8903462749213011,
+ "grad_norm": 0.1545933676895574,
+ "learning_rate": 6.321904029399607e-06,
+ "loss": 0.4429236054420471,
+ "step": 1697
+ },
+ {
+ "epoch": 0.8908709338929696,
+ "grad_norm": 0.15625844051167467,
+ "learning_rate": 6.3094083495275565e-06,
+ "loss": 0.4284421503543854,
+ "step": 1698
+ },
+ {
+ "epoch": 0.891395592864638,
+ "grad_norm": 0.1484963883381567,
+ "learning_rate": 6.2969702397493e-06,
+ "loss": 0.4545463025569916,
+ "step": 1699
+ },
+ {
+ "epoch": 0.8919202518363064,
+ "grad_norm": 0.15122402806869387,
+ "learning_rate": 6.284589733856403e-06,
+ "loss": 0.43840622901916504,
+ "step": 1700
+ },
+ {
+ "epoch": 0.8924449108079748,
+ "grad_norm": 0.14889313167657334,
+ "learning_rate": 6.27226686548392e-06,
+ "loss": 0.4451153874397278,
+ "step": 1701
+ },
+ {
+ "epoch": 0.8929695697796433,
+ "grad_norm": 0.14724940941036385,
+ "learning_rate": 6.260001668110331e-06,
+ "loss": 0.434165358543396,
+ "step": 1702
+ },
+ {
+ "epoch": 0.8934942287513117,
+ "grad_norm": 0.15520838093295836,
+ "learning_rate": 6.247794175057433e-06,
+ "loss": 0.44641873240470886,
+ "step": 1703
+ },
+ {
+ "epoch": 0.8940188877229801,
+ "grad_norm": 0.159167904797364,
+ "learning_rate": 6.235644419490253e-06,
+ "loss": 0.44591957330703735,
+ "step": 1704
+ },
+ {
+ "epoch": 0.8945435466946485,
+ "grad_norm": 0.14843235881370814,
+ "learning_rate": 6.223552434416951e-06,
+ "loss": 0.43010181188583374,
+ "step": 1705
+ },
+ {
+ "epoch": 0.8950682056663168,
+ "grad_norm": 0.1590998153483183,
+ "learning_rate": 6.211518252688749e-06,
+ "loss": 0.4602118134498596,
+ "step": 1706
+ },
+ {
+ "epoch": 0.8955928646379853,
+ "grad_norm": 0.15257025539137767,
+ "learning_rate": 6.19954190699982e-06,
+ "loss": 0.45429927110671997,
+ "step": 1707
+ },
+ {
+ "epoch": 0.8961175236096537,
+ "grad_norm": 0.15304035006408193,
+ "learning_rate": 6.18762342988722e-06,
+ "loss": 0.4405020773410797,
+ "step": 1708
+ },
+ {
+ "epoch": 0.8966421825813221,
+ "grad_norm": 0.15956303033928781,
+ "learning_rate": 6.175762853730777e-06,
+ "loss": 0.4455980062484741,
+ "step": 1709
+ },
+ {
+ "epoch": 0.8971668415529905,
+ "grad_norm": 0.14973831573751925,
+ "learning_rate": 6.163960210753025e-06,
+ "loss": 0.4331505298614502,
+ "step": 1710
+ },
+ {
+ "epoch": 0.897691500524659,
+ "grad_norm": 0.14573687980573624,
+ "learning_rate": 6.1522155330190985e-06,
+ "loss": 0.43307584524154663,
+ "step": 1711
+ },
+ {
+ "epoch": 0.8982161594963274,
+ "grad_norm": 0.1556523345686865,
+ "learning_rate": 6.140528852436665e-06,
+ "loss": 0.4531600773334503,
+ "step": 1712
+ },
+ {
+ "epoch": 0.8987408184679958,
+ "grad_norm": 0.14262442470666953,
+ "learning_rate": 6.128900200755817e-06,
+ "loss": 0.4350634813308716,
+ "step": 1713
+ },
+ {
+ "epoch": 0.8992654774396642,
+ "grad_norm": 0.14709392307082428,
+ "learning_rate": 6.117329609568998e-06,
+ "loss": 0.43465620279312134,
+ "step": 1714
+ },
+ {
+ "epoch": 0.8997901364113327,
+ "grad_norm": 0.15648040137428026,
+ "learning_rate": 6.105817110310909e-06,
+ "loss": 0.4344560503959656,
+ "step": 1715
+ },
+ {
+ "epoch": 0.9003147953830011,
+ "grad_norm": 0.15758204487742147,
+ "learning_rate": 6.094362734258438e-06,
+ "loss": 0.4409951865673065,
+ "step": 1716
+ },
+ {
+ "epoch": 0.9008394543546695,
+ "grad_norm": 0.15755547382189733,
+ "learning_rate": 6.082966512530558e-06,
+ "loss": 0.44050097465515137,
+ "step": 1717
+ },
+ {
+ "epoch": 0.9013641133263379,
+ "grad_norm": 0.1536165045521179,
+ "learning_rate": 6.071628476088257e-06,
+ "loss": 0.43460220098495483,
+ "step": 1718
+ },
+ {
+ "epoch": 0.9018887722980063,
+ "grad_norm": 0.16253540769141986,
+ "learning_rate": 6.060348655734439e-06,
+ "loss": 0.4490512013435364,
+ "step": 1719
+ },
+ {
+ "epoch": 0.9024134312696747,
+ "grad_norm": 0.15003242262029914,
+ "learning_rate": 6.049127082113853e-06,
+ "loss": 0.44427937269210815,
+ "step": 1720
+ },
+ {
+ "epoch": 0.9029380902413431,
+ "grad_norm": 0.1579908758627471,
+ "learning_rate": 6.037963785712995e-06,
+ "loss": 0.4421137571334839,
+ "step": 1721
+ },
+ {
+ "epoch": 0.9034627492130115,
+ "grad_norm": 0.15111957361250286,
+ "learning_rate": 6.026858796860048e-06,
+ "loss": 0.44011324644088745,
+ "step": 1722
+ },
+ {
+ "epoch": 0.9039874081846799,
+ "grad_norm": 0.1446470746259905,
+ "learning_rate": 6.0158121457247845e-06,
+ "loss": 0.43798986077308655,
+ "step": 1723
+ },
+ {
+ "epoch": 0.9045120671563484,
+ "grad_norm": 0.14882152623490427,
+ "learning_rate": 6.004823862318474e-06,
+ "loss": 0.4265901446342468,
+ "step": 1724
+ },
+ {
+ "epoch": 0.9050367261280168,
+ "grad_norm": 0.1506889859372045,
+ "learning_rate": 5.993893976493831e-06,
+ "loss": 0.4503232538700104,
+ "step": 1725
+ },
+ {
+ "epoch": 0.9055613850996852,
+ "grad_norm": 0.16757194686045146,
+ "learning_rate": 5.983022517944901e-06,
+ "loss": 0.4509217441082001,
+ "step": 1726
+ },
+ {
+ "epoch": 0.9060860440713536,
+ "grad_norm": 0.14447292841509515,
+ "learning_rate": 5.972209516207007e-06,
+ "loss": 0.4476190507411957,
+ "step": 1727
+ },
+ {
+ "epoch": 0.906610703043022,
+ "grad_norm": 0.1536564780391058,
+ "learning_rate": 5.961455000656656e-06,
+ "loss": 0.45448166131973267,
+ "step": 1728
+ },
+ {
+ "epoch": 0.9071353620146905,
+ "grad_norm": 0.15436323862710438,
+ "learning_rate": 5.950759000511455e-06,
+ "loss": 0.45128491520881653,
+ "step": 1729
+ },
+ {
+ "epoch": 0.9076600209863589,
+ "grad_norm": 0.15379817460186834,
+ "learning_rate": 5.940121544830047e-06,
+ "loss": 0.4490806460380554,
+ "step": 1730
+ },
+ {
+ "epoch": 0.9081846799580273,
+ "grad_norm": 0.1470796082413195,
+ "learning_rate": 5.929542662512013e-06,
+ "loss": 0.4371582865715027,
+ "step": 1731
+ },
+ {
+ "epoch": 0.9087093389296957,
+ "grad_norm": 0.15107600076433048,
+ "learning_rate": 5.919022382297809e-06,
+ "loss": 0.45175206661224365,
+ "step": 1732
+ },
+ {
+ "epoch": 0.9092339979013642,
+ "grad_norm": 0.14972329317276084,
+ "learning_rate": 5.908560732768684e-06,
+ "loss": 0.43177497386932373,
+ "step": 1733
+ },
+ {
+ "epoch": 0.9097586568730325,
+ "grad_norm": 0.14437715003340176,
+ "learning_rate": 5.898157742346601e-06,
+ "loss": 0.4243622422218323,
+ "step": 1734
+ },
+ {
+ "epoch": 0.9102833158447009,
+ "grad_norm": 0.15200146823831148,
+ "learning_rate": 5.8878134392941465e-06,
+ "loss": 0.47819408774375916,
+ "step": 1735
+ },
+ {
+ "epoch": 0.9108079748163693,
+ "grad_norm": 0.156864616244505,
+ "learning_rate": 5.877527851714479e-06,
+ "loss": 0.44744759798049927,
+ "step": 1736
+ },
+ {
+ "epoch": 0.9113326337880377,
+ "grad_norm": 0.14489667207174767,
+ "learning_rate": 5.867301007551239e-06,
+ "loss": 0.4390789866447449,
+ "step": 1737
+ },
+ {
+ "epoch": 0.9118572927597062,
+ "grad_norm": 0.1528011587102748,
+ "learning_rate": 5.857132934588479e-06,
+ "loss": 0.4338226616382599,
+ "step": 1738
+ },
+ {
+ "epoch": 0.9123819517313746,
+ "grad_norm": 0.14716603285971375,
+ "learning_rate": 5.847023660450568e-06,
+ "loss": 0.43975532054901123,
+ "step": 1739
+ },
+ {
+ "epoch": 0.912906610703043,
+ "grad_norm": 0.15266609903803277,
+ "learning_rate": 5.836973212602143e-06,
+ "loss": 0.449812650680542,
+ "step": 1740
+ },
+ {
+ "epoch": 0.9134312696747114,
+ "grad_norm": 0.1466770441701569,
+ "learning_rate": 5.826981618348025e-06,
+ "loss": 0.43771812319755554,
+ "step": 1741
+ },
+ {
+ "epoch": 0.9139559286463799,
+ "grad_norm": 0.1428112225146131,
+ "learning_rate": 5.8170489048331305e-06,
+ "loss": 0.42403268814086914,
+ "step": 1742
+ },
+ {
+ "epoch": 0.9144805876180483,
+ "grad_norm": 0.1550829416480759,
+ "learning_rate": 5.8071750990424275e-06,
+ "loss": 0.45707154273986816,
+ "step": 1743
+ },
+ {
+ "epoch": 0.9150052465897167,
+ "grad_norm": 0.14687341315819044,
+ "learning_rate": 5.797360227800832e-06,
+ "loss": 0.4558846354484558,
+ "step": 1744
+ },
+ {
+ "epoch": 0.9155299055613851,
+ "grad_norm": 0.15660848922413237,
+ "learning_rate": 5.787604317773154e-06,
+ "loss": 0.44760435819625854,
+ "step": 1745
+ },
+ {
+ "epoch": 0.9160545645330536,
+ "grad_norm": 0.14891742746821762,
+ "learning_rate": 5.777907395464015e-06,
+ "loss": 0.4266620874404907,
+ "step": 1746
+ },
+ {
+ "epoch": 0.916579223504722,
+ "grad_norm": 0.14501110957005472,
+ "learning_rate": 5.768269487217777e-06,
+ "loss": 0.4388393759727478,
+ "step": 1747
+ },
+ {
+ "epoch": 0.9171038824763903,
+ "grad_norm": 0.14651438673799927,
+ "learning_rate": 5.758690619218493e-06,
+ "loss": 0.4416884183883667,
+ "step": 1748
+ },
+ {
+ "epoch": 0.9176285414480587,
+ "grad_norm": 0.14785104880890787,
+ "learning_rate": 5.749170817489792e-06,
+ "loss": 0.4368356466293335,
+ "step": 1749
+ },
+ {
+ "epoch": 0.9181532004197271,
+ "grad_norm": 0.16586567883267808,
+ "learning_rate": 5.739710107894847e-06,
+ "loss": 0.44294649362564087,
+ "step": 1750
+ },
+ {
+ "epoch": 0.9186778593913956,
+ "grad_norm": 0.14377347738227095,
+ "learning_rate": 5.730308516136288e-06,
+ "loss": 0.4405091106891632,
+ "step": 1751
+ },
+ {
+ "epoch": 0.919202518363064,
+ "grad_norm": 0.14703077660050376,
+ "learning_rate": 5.72096606775613e-06,
+ "loss": 0.44809678196907043,
+ "step": 1752
+ },
+ {
+ "epoch": 0.9197271773347324,
+ "grad_norm": 0.15073590419140934,
+ "learning_rate": 5.711682788135726e-06,
+ "loss": 0.43815505504608154,
+ "step": 1753
+ },
+ {
+ "epoch": 0.9202518363064008,
+ "grad_norm": 0.14774775313473323,
+ "learning_rate": 5.702458702495658e-06,
+ "loss": 0.4417206048965454,
+ "step": 1754
+ },
+ {
+ "epoch": 0.9207764952780693,
+ "grad_norm": 0.14289335805130654,
+ "learning_rate": 5.693293835895706e-06,
+ "loss": 0.42886218428611755,
+ "step": 1755
+ },
+ {
+ "epoch": 0.9213011542497377,
+ "grad_norm": 0.15571224276102627,
+ "learning_rate": 5.684188213234763e-06,
+ "loss": 0.44488853216171265,
+ "step": 1756
+ },
+ {
+ "epoch": 0.9218258132214061,
+ "grad_norm": 0.147115948090654,
+ "learning_rate": 5.675141859250763e-06,
+ "loss": 0.44682830572128296,
+ "step": 1757
+ },
+ {
+ "epoch": 0.9223504721930745,
+ "grad_norm": 0.15349533122985531,
+ "learning_rate": 5.666154798520627e-06,
+ "loss": 0.45550864934921265,
+ "step": 1758
+ },
+ {
+ "epoch": 0.922875131164743,
+ "grad_norm": 0.16960595806748266,
+ "learning_rate": 5.657227055460184e-06,
+ "loss": 0.4605559706687927,
+ "step": 1759
+ },
+ {
+ "epoch": 0.9233997901364114,
+ "grad_norm": 0.14665490922506455,
+ "learning_rate": 5.648358654324119e-06,
+ "loss": 0.45442628860473633,
+ "step": 1760
+ },
+ {
+ "epoch": 0.9239244491080797,
+ "grad_norm": 0.14780665809126886,
+ "learning_rate": 5.6395496192058905e-06,
+ "loss": 0.45085081458091736,
+ "step": 1761
+ },
+ {
+ "epoch": 0.9244491080797481,
+ "grad_norm": 0.15599036700169416,
+ "learning_rate": 5.6307999740376714e-06,
+ "loss": 0.4581005573272705,
+ "step": 1762
+ },
+ {
+ "epoch": 0.9249737670514165,
+ "grad_norm": 0.14383027603716222,
+ "learning_rate": 5.622109742590294e-06,
+ "loss": 0.43986010551452637,
+ "step": 1763
+ },
+ {
+ "epoch": 0.925498426023085,
+ "grad_norm": 0.1510369649046927,
+ "learning_rate": 5.613478948473166e-06,
+ "loss": 0.4317719340324402,
+ "step": 1764
+ },
+ {
+ "epoch": 0.9260230849947534,
+ "grad_norm": 0.16608560055258376,
+ "learning_rate": 5.60490761513423e-06,
+ "loss": 0.4500439167022705,
+ "step": 1765
+ },
+ {
+ "epoch": 0.9265477439664218,
+ "grad_norm": 0.15364415343465948,
+ "learning_rate": 5.596395765859875e-06,
+ "loss": 0.43703922629356384,
+ "step": 1766
+ },
+ {
+ "epoch": 0.9270724029380902,
+ "grad_norm": 0.145164767782082,
+ "learning_rate": 5.587943423774886e-06,
+ "loss": 0.4389477074146271,
+ "step": 1767
+ },
+ {
+ "epoch": 0.9275970619097587,
+ "grad_norm": 0.15044099308887526,
+ "learning_rate": 5.579550611842395e-06,
+ "loss": 0.43600398302078247,
+ "step": 1768
+ },
+ {
+ "epoch": 0.9281217208814271,
+ "grad_norm": 0.14694381874423756,
+ "learning_rate": 5.571217352863786e-06,
+ "loss": 0.4525601863861084,
+ "step": 1769
+ },
+ {
+ "epoch": 0.9286463798530955,
+ "grad_norm": 0.15979616589190854,
+ "learning_rate": 5.562943669478661e-06,
+ "loss": 0.44730353355407715,
+ "step": 1770
+ },
+ {
+ "epoch": 0.9291710388247639,
+ "grad_norm": 0.1545596231710442,
+ "learning_rate": 5.554729584164762e-06,
+ "loss": 0.43867969512939453,
+ "step": 1771
+ },
+ {
+ "epoch": 0.9296956977964324,
+ "grad_norm": 0.15923930406000852,
+ "learning_rate": 5.5465751192379235e-06,
+ "loss": 0.4648996889591217,
+ "step": 1772
+ },
+ {
+ "epoch": 0.9302203567681008,
+ "grad_norm": 0.14001241391446362,
+ "learning_rate": 5.538480296852001e-06,
+ "loss": 0.41449663043022156,
+ "step": 1773
+ },
+ {
+ "epoch": 0.9307450157397692,
+ "grad_norm": 0.16538765001279418,
+ "learning_rate": 5.530445138998815e-06,
+ "loss": 0.4503074288368225,
+ "step": 1774
+ },
+ {
+ "epoch": 0.9312696747114375,
+ "grad_norm": 0.15007886784403335,
+ "learning_rate": 5.5224696675080904e-06,
+ "loss": 0.44717323780059814,
+ "step": 1775
+ },
+ {
+ "epoch": 0.9317943336831059,
+ "grad_norm": 0.17834856046133146,
+ "learning_rate": 5.514553904047395e-06,
+ "loss": 0.4374881386756897,
+ "step": 1776
+ },
+ {
+ "epoch": 0.9323189926547744,
+ "grad_norm": 0.1494478926649096,
+ "learning_rate": 5.5066978701220865e-06,
+ "loss": 0.4464038610458374,
+ "step": 1777
+ },
+ {
+ "epoch": 0.9328436516264428,
+ "grad_norm": 0.14887448793632826,
+ "learning_rate": 5.498901587075253e-06,
+ "loss": 0.443928062915802,
+ "step": 1778
+ },
+ {
+ "epoch": 0.9333683105981112,
+ "grad_norm": 0.14192884484464632,
+ "learning_rate": 5.491165076087652e-06,
+ "loss": 0.42002278566360474,
+ "step": 1779
+ },
+ {
+ "epoch": 0.9338929695697796,
+ "grad_norm": 0.14783356724791524,
+ "learning_rate": 5.483488358177647e-06,
+ "loss": 0.43647563457489014,
+ "step": 1780
+ },
+ {
+ "epoch": 0.934417628541448,
+ "grad_norm": 0.15004597148119703,
+ "learning_rate": 5.475871454201163e-06,
+ "loss": 0.4396110773086548,
+ "step": 1781
+ },
+ {
+ "epoch": 0.9349422875131165,
+ "grad_norm": 0.15668918073856763,
+ "learning_rate": 5.468314384851626e-06,
+ "loss": 0.42848116159439087,
+ "step": 1782
+ },
+ {
+ "epoch": 0.9354669464847849,
+ "grad_norm": 0.1449937199040528,
+ "learning_rate": 5.460817170659901e-06,
+ "loss": 0.44262629747390747,
+ "step": 1783
+ },
+ {
+ "epoch": 0.9359916054564533,
+ "grad_norm": 0.15868260486048544,
+ "learning_rate": 5.4533798319942375e-06,
+ "loss": 0.44874119758605957,
+ "step": 1784
+ },
+ {
+ "epoch": 0.9365162644281217,
+ "grad_norm": 0.15004870783258292,
+ "learning_rate": 5.446002389060224e-06,
+ "loss": 0.4575575590133667,
+ "step": 1785
+ },
+ {
+ "epoch": 0.9370409233997902,
+ "grad_norm": 0.14830248639010987,
+ "learning_rate": 5.438684861900721e-06,
+ "loss": 0.44039857387542725,
+ "step": 1786
+ },
+ {
+ "epoch": 0.9375655823714586,
+ "grad_norm": 0.15429570419507257,
+ "learning_rate": 5.431427270395808e-06,
+ "loss": 0.43363744020462036,
+ "step": 1787
+ },
+ {
+ "epoch": 0.938090241343127,
+ "grad_norm": 0.15982575395123055,
+ "learning_rate": 5.4242296342627354e-06,
+ "loss": 0.4484976530075073,
+ "step": 1788
+ },
+ {
+ "epoch": 0.9386149003147953,
+ "grad_norm": 0.1466121447285136,
+ "learning_rate": 5.417091973055878e-06,
+ "loss": 0.44059592485427856,
+ "step": 1789
+ },
+ {
+ "epoch": 0.9391395592864638,
+ "grad_norm": 0.15353848724097463,
+ "learning_rate": 5.410014306166657e-06,
+ "loss": 0.4600589871406555,
+ "step": 1790
+ },
+ {
+ "epoch": 0.9396642182581322,
+ "grad_norm": 0.14968071650065867,
+ "learning_rate": 5.4029966528235065e-06,
+ "loss": 0.4415352940559387,
+ "step": 1791
+ },
+ {
+ "epoch": 0.9401888772298006,
+ "grad_norm": 0.14819069534768417,
+ "learning_rate": 5.396039032091826e-06,
+ "loss": 0.4436836242675781,
+ "step": 1792
+ },
+ {
+ "epoch": 0.940713536201469,
+ "grad_norm": 0.14683654708689367,
+ "learning_rate": 5.389141462873908e-06,
+ "loss": 0.4243587851524353,
+ "step": 1793
+ },
+ {
+ "epoch": 0.9412381951731374,
+ "grad_norm": 0.15431293103561666,
+ "learning_rate": 5.382303963908908e-06,
+ "loss": 0.450438916683197,
+ "step": 1794
+ },
+ {
+ "epoch": 0.9417628541448059,
+ "grad_norm": 0.15125346717879914,
+ "learning_rate": 5.3755265537727775e-06,
+ "loss": 0.4469110369682312,
+ "step": 1795
+ },
+ {
+ "epoch": 0.9422875131164743,
+ "grad_norm": 0.15430864502191044,
+ "learning_rate": 5.368809250878227e-06,
+ "loss": 0.45182669162750244,
+ "step": 1796
+ },
+ {
+ "epoch": 0.9428121720881427,
+ "grad_norm": 0.1515431753176237,
+ "learning_rate": 5.362152073474657e-06,
+ "loss": 0.43683797121047974,
+ "step": 1797
+ },
+ {
+ "epoch": 0.9433368310598111,
+ "grad_norm": 0.1524915522484605,
+ "learning_rate": 5.355555039648133e-06,
+ "loss": 0.44641250371932983,
+ "step": 1798
+ },
+ {
+ "epoch": 0.9438614900314796,
+ "grad_norm": 0.15722684273308743,
+ "learning_rate": 5.3490181673213206e-06,
+ "loss": 0.43558835983276367,
+ "step": 1799
+ },
+ {
+ "epoch": 0.944386149003148,
+ "grad_norm": 0.15527272713567492,
+ "learning_rate": 5.342541474253438e-06,
+ "loss": 0.4423788785934448,
+ "step": 1800
+ },
+ {
+ "epoch": 0.9449108079748164,
+ "grad_norm": 0.1578831929456481,
+ "learning_rate": 5.336124978040208e-06,
+ "loss": 0.4506484866142273,
+ "step": 1801
+ },
+ {
+ "epoch": 0.9454354669464848,
+ "grad_norm": 0.14812076004561775,
+ "learning_rate": 5.329768696113815e-06,
+ "loss": 0.4432883858680725,
+ "step": 1802
+ },
+ {
+ "epoch": 0.9459601259181531,
+ "grad_norm": 0.1511327211030653,
+ "learning_rate": 5.323472645742854e-06,
+ "loss": 0.4321002960205078,
+ "step": 1803
+ },
+ {
+ "epoch": 0.9464847848898216,
+ "grad_norm": 0.1531668263170348,
+ "learning_rate": 5.3172368440322854e-06,
+ "loss": 0.44666337966918945,
+ "step": 1804
+ },
+ {
+ "epoch": 0.94700944386149,
+ "grad_norm": 0.14571994558609092,
+ "learning_rate": 5.311061307923387e-06,
+ "loss": 0.4436221420764923,
+ "step": 1805
+ },
+ {
+ "epoch": 0.9475341028331584,
+ "grad_norm": 0.15158206242691336,
+ "learning_rate": 5.304946054193706e-06,
+ "loss": 0.45979544520378113,
+ "step": 1806
+ },
+ {
+ "epoch": 0.9480587618048268,
+ "grad_norm": 0.15374308675707662,
+ "learning_rate": 5.298891099457018e-06,
+ "loss": 0.44095706939697266,
+ "step": 1807
+ },
+ {
+ "epoch": 0.9485834207764953,
+ "grad_norm": 0.14777162806829072,
+ "learning_rate": 5.292896460163283e-06,
+ "loss": 0.44000717997550964,
+ "step": 1808
+ },
+ {
+ "epoch": 0.9491080797481637,
+ "grad_norm": 0.15416014526696098,
+ "learning_rate": 5.286962152598589e-06,
+ "loss": 0.4367929697036743,
+ "step": 1809
+ },
+ {
+ "epoch": 0.9496327387198321,
+ "grad_norm": 0.1586289813390648,
+ "learning_rate": 5.281088192885125e-06,
+ "loss": 0.44338831305503845,
+ "step": 1810
+ },
+ {
+ "epoch": 0.9501573976915005,
+ "grad_norm": 0.1480189398685719,
+ "learning_rate": 5.275274596981124e-06,
+ "loss": 0.44127213954925537,
+ "step": 1811
+ },
+ {
+ "epoch": 0.950682056663169,
+ "grad_norm": 0.15893037852856393,
+ "learning_rate": 5.269521380680823e-06,
+ "loss": 0.43447786569595337,
+ "step": 1812
+ },
+ {
+ "epoch": 0.9512067156348374,
+ "grad_norm": 0.15257675815194532,
+ "learning_rate": 5.2638285596144236e-06,
+ "loss": 0.4478892385959625,
+ "step": 1813
+ },
+ {
+ "epoch": 0.9517313746065058,
+ "grad_norm": 0.14985387736695718,
+ "learning_rate": 5.258196149248046e-06,
+ "loss": 0.4517829418182373,
+ "step": 1814
+ },
+ {
+ "epoch": 0.9522560335781742,
+ "grad_norm": 0.14441329680781126,
+ "learning_rate": 5.252624164883689e-06,
+ "loss": 0.4337430000305176,
+ "step": 1815
+ },
+ {
+ "epoch": 0.9527806925498427,
+ "grad_norm": 0.15113217761739892,
+ "learning_rate": 5.247112621659186e-06,
+ "loss": 0.44544464349746704,
+ "step": 1816
+ },
+ {
+ "epoch": 0.953305351521511,
+ "grad_norm": 0.15511548841769227,
+ "learning_rate": 5.241661534548169e-06,
+ "loss": 0.4420471787452698,
+ "step": 1817
+ },
+ {
+ "epoch": 0.9538300104931794,
+ "grad_norm": 0.141474135090229,
+ "learning_rate": 5.236270918360016e-06,
+ "loss": 0.4323197603225708,
+ "step": 1818
+ },
+ {
+ "epoch": 0.9543546694648478,
+ "grad_norm": 0.14741159689952743,
+ "learning_rate": 5.23094078773983e-06,
+ "loss": 0.4486733675003052,
+ "step": 1819
+ },
+ {
+ "epoch": 0.9548793284365162,
+ "grad_norm": 0.14938242936939558,
+ "learning_rate": 5.225671157168377e-06,
+ "loss": 0.4469614624977112,
+ "step": 1820
+ },
+ {
+ "epoch": 0.9554039874081847,
+ "grad_norm": 0.14960446458046617,
+ "learning_rate": 5.220462040962068e-06,
+ "loss": 0.4371514320373535,
+ "step": 1821
+ },
+ {
+ "epoch": 0.9559286463798531,
+ "grad_norm": 0.1453826830908917,
+ "learning_rate": 5.215313453272906e-06,
+ "loss": 0.43045082688331604,
+ "step": 1822
+ },
+ {
+ "epoch": 0.9564533053515215,
+ "grad_norm": 0.14561549586732817,
+ "learning_rate": 5.210225408088447e-06,
+ "loss": 0.44164514541625977,
+ "step": 1823
+ },
+ {
+ "epoch": 0.9569779643231899,
+ "grad_norm": 0.15381074610209994,
+ "learning_rate": 5.205197919231773e-06,
+ "loss": 0.45496129989624023,
+ "step": 1824
+ },
+ {
+ "epoch": 0.9575026232948584,
+ "grad_norm": 0.14729572792188594,
+ "learning_rate": 5.200231000361448e-06,
+ "loss": 0.4330272078514099,
+ "step": 1825
+ },
+ {
+ "epoch": 0.9580272822665268,
+ "grad_norm": 0.1549156051160634,
+ "learning_rate": 5.195324664971476e-06,
+ "loss": 0.44893333315849304,
+ "step": 1826
+ },
+ {
+ "epoch": 0.9585519412381952,
+ "grad_norm": 0.1544667437543616,
+ "learning_rate": 5.190478926391271e-06,
+ "loss": 0.4463026821613312,
+ "step": 1827
+ },
+ {
+ "epoch": 0.9590766002098636,
+ "grad_norm": 0.1482681888796153,
+ "learning_rate": 5.1856937977856235e-06,
+ "loss": 0.42915546894073486,
+ "step": 1828
+ },
+ {
+ "epoch": 0.959601259181532,
+ "grad_norm": 0.15393225997035367,
+ "learning_rate": 5.180969292154653e-06,
+ "loss": 0.45655858516693115,
+ "step": 1829
+ },
+ {
+ "epoch": 0.9601259181532005,
+ "grad_norm": 0.15402682858559275,
+ "learning_rate": 5.176305422333788e-06,
+ "loss": 0.45586466789245605,
+ "step": 1830
+ },
+ {
+ "epoch": 0.9606505771248688,
+ "grad_norm": 0.15136206757704734,
+ "learning_rate": 5.171702200993716e-06,
+ "loss": 0.4525001049041748,
+ "step": 1831
+ },
+ {
+ "epoch": 0.9611752360965372,
+ "grad_norm": 0.15443301434325082,
+ "learning_rate": 5.167159640640357e-06,
+ "loss": 0.4500652551651001,
+ "step": 1832
+ },
+ {
+ "epoch": 0.9616998950682056,
+ "grad_norm": 0.1553996888683437,
+ "learning_rate": 5.162677753614834e-06,
+ "loss": 0.4426611363887787,
+ "step": 1833
+ },
+ {
+ "epoch": 0.9622245540398741,
+ "grad_norm": 0.14719352293992807,
+ "learning_rate": 5.158256552093428e-06,
+ "loss": 0.43979009985923767,
+ "step": 1834
+ },
+ {
+ "epoch": 0.9627492130115425,
+ "grad_norm": 0.14547547543617104,
+ "learning_rate": 5.153896048087557e-06,
+ "loss": 0.443497896194458,
+ "step": 1835
+ },
+ {
+ "epoch": 0.9632738719832109,
+ "grad_norm": 0.15665555976593595,
+ "learning_rate": 5.149596253443734e-06,
+ "loss": 0.441650390625,
+ "step": 1836
+ },
+ {
+ "epoch": 0.9637985309548793,
+ "grad_norm": 0.1552107125031637,
+ "learning_rate": 5.145357179843536e-06,
+ "loss": 0.44342389702796936,
+ "step": 1837
+ },
+ {
+ "epoch": 0.9643231899265478,
+ "grad_norm": 0.1450717817395646,
+ "learning_rate": 5.141178838803574e-06,
+ "loss": 0.4459199011325836,
+ "step": 1838
+ },
+ {
+ "epoch": 0.9648478488982162,
+ "grad_norm": 0.15194030785499474,
+ "learning_rate": 5.137061241675472e-06,
+ "loss": 0.43879520893096924,
+ "step": 1839
+ },
+ {
+ "epoch": 0.9653725078698846,
+ "grad_norm": 0.15876627179026034,
+ "learning_rate": 5.133004399645817e-06,
+ "loss": 0.4629112184047699,
+ "step": 1840
+ },
+ {
+ "epoch": 0.965897166841553,
+ "grad_norm": 0.15025078854799365,
+ "learning_rate": 5.129008323736142e-06,
+ "loss": 0.4596017599105835,
+ "step": 1841
+ },
+ {
+ "epoch": 0.9664218258132214,
+ "grad_norm": 0.15108123279475658,
+ "learning_rate": 5.125073024802887e-06,
+ "loss": 0.4394477605819702,
+ "step": 1842
+ },
+ {
+ "epoch": 0.9669464847848899,
+ "grad_norm": 0.15040459550862542,
+ "learning_rate": 5.121198513537382e-06,
+ "loss": 0.44406938552856445,
+ "step": 1843
+ },
+ {
+ "epoch": 0.9674711437565582,
+ "grad_norm": 0.14829125607751426,
+ "learning_rate": 5.1173848004658065e-06,
+ "loss": 0.4245413541793823,
+ "step": 1844
+ },
+ {
+ "epoch": 0.9679958027282266,
+ "grad_norm": 0.15182555299407333,
+ "learning_rate": 5.11363189594916e-06,
+ "loss": 0.44984906911849976,
+ "step": 1845
+ },
+ {
+ "epoch": 0.968520461699895,
+ "grad_norm": 0.1455263692654042,
+ "learning_rate": 5.109939810183251e-06,
+ "loss": 0.41511911153793335,
+ "step": 1846
+ },
+ {
+ "epoch": 0.9690451206715635,
+ "grad_norm": 0.14474034020024001,
+ "learning_rate": 5.106308553198644e-06,
+ "loss": 0.42551544308662415,
+ "step": 1847
+ },
+ {
+ "epoch": 0.9695697796432319,
+ "grad_norm": 0.14650897119054226,
+ "learning_rate": 5.102738134860656e-06,
+ "loss": 0.4347913861274719,
+ "step": 1848
+ },
+ {
+ "epoch": 0.9700944386149003,
+ "grad_norm": 0.15017963706984508,
+ "learning_rate": 5.099228564869313e-06,
+ "loss": 0.442004919052124,
+ "step": 1849
+ },
+ {
+ "epoch": 0.9706190975865687,
+ "grad_norm": 0.14854765100160125,
+ "learning_rate": 5.09577985275933e-06,
+ "loss": 0.4416872262954712,
+ "step": 1850
+ },
+ {
+ "epoch": 0.9711437565582371,
+ "grad_norm": 0.1648726444197388,
+ "learning_rate": 5.092392007900087e-06,
+ "loss": 0.4556121826171875,
+ "step": 1851
+ },
+ {
+ "epoch": 0.9716684155299056,
+ "grad_norm": 0.15795586380007176,
+ "learning_rate": 5.089065039495598e-06,
+ "loss": 0.4360453486442566,
+ "step": 1852
+ },
+ {
+ "epoch": 0.972193074501574,
+ "grad_norm": 0.15841383239716067,
+ "learning_rate": 5.085798956584491e-06,
+ "loss": 0.44655895233154297,
+ "step": 1853
+ },
+ {
+ "epoch": 0.9727177334732424,
+ "grad_norm": 0.14772746394586367,
+ "learning_rate": 5.082593768039989e-06,
+ "loss": 0.44125744700431824,
+ "step": 1854
+ },
+ {
+ "epoch": 0.9732423924449108,
+ "grad_norm": 0.1511843151742375,
+ "learning_rate": 5.079449482569862e-06,
+ "loss": 0.4447266161441803,
+ "step": 1855
+ },
+ {
+ "epoch": 0.9737670514165793,
+ "grad_norm": 0.15676550602119788,
+ "learning_rate": 5.076366108716437e-06,
+ "loss": 0.4479975700378418,
+ "step": 1856
+ },
+ {
+ "epoch": 0.9742917103882477,
+ "grad_norm": 0.15042078359963343,
+ "learning_rate": 5.073343654856548e-06,
+ "loss": 0.442285418510437,
+ "step": 1857
+ },
+ {
+ "epoch": 0.974816369359916,
+ "grad_norm": 0.14908899116909868,
+ "learning_rate": 5.070382129201526e-06,
+ "loss": 0.44102227687835693,
+ "step": 1858
+ },
+ {
+ "epoch": 0.9753410283315844,
+ "grad_norm": 0.14466645777159964,
+ "learning_rate": 5.067481539797172e-06,
+ "loss": 0.42972224950790405,
+ "step": 1859
+ },
+ {
+ "epoch": 0.9758656873032528,
+ "grad_norm": 0.1600541020039413,
+ "learning_rate": 5.064641894523737e-06,
+ "loss": 0.4534302353858948,
+ "step": 1860
+ },
+ {
+ "epoch": 0.9763903462749213,
+ "grad_norm": 0.1599331982060412,
+ "learning_rate": 5.061863201095905e-06,
+ "loss": 0.4575304090976715,
+ "step": 1861
+ },
+ {
+ "epoch": 0.9769150052465897,
+ "grad_norm": 0.14699643507200835,
+ "learning_rate": 5.059145467062763e-06,
+ "loss": 0.4292885959148407,
+ "step": 1862
+ },
+ {
+ "epoch": 0.9774396642182581,
+ "grad_norm": 0.15239951700028798,
+ "learning_rate": 5.056488699807784e-06,
+ "loss": 0.4484747052192688,
+ "step": 1863
+ },
+ {
+ "epoch": 0.9779643231899265,
+ "grad_norm": 0.14925326654404217,
+ "learning_rate": 5.05389290654881e-06,
+ "loss": 0.4315107464790344,
+ "step": 1864
+ },
+ {
+ "epoch": 0.978488982161595,
+ "grad_norm": 0.14557754498869974,
+ "learning_rate": 5.051358094338034e-06,
+ "loss": 0.4448525607585907,
+ "step": 1865
+ },
+ {
+ "epoch": 0.9790136411332634,
+ "grad_norm": 0.14596289150950575,
+ "learning_rate": 5.048884270061969e-06,
+ "loss": 0.4328519403934479,
+ "step": 1866
+ },
+ {
+ "epoch": 0.9795383001049318,
+ "grad_norm": 0.14883370444683877,
+ "learning_rate": 5.046471440441445e-06,
+ "loss": 0.4282095432281494,
+ "step": 1867
+ },
+ {
+ "epoch": 0.9800629590766002,
+ "grad_norm": 0.15233818252290696,
+ "learning_rate": 5.044119612031576e-06,
+ "loss": 0.44765010476112366,
+ "step": 1868
+ },
+ {
+ "epoch": 0.9805876180482687,
+ "grad_norm": 0.15738042523437606,
+ "learning_rate": 5.041828791221759e-06,
+ "loss": 0.44896170496940613,
+ "step": 1869
+ },
+ {
+ "epoch": 0.9811122770199371,
+ "grad_norm": 0.15435191566203477,
+ "learning_rate": 5.039598984235635e-06,
+ "loss": 0.46627920866012573,
+ "step": 1870
+ },
+ {
+ "epoch": 0.9816369359916055,
+ "grad_norm": 0.15330002542858065,
+ "learning_rate": 5.037430197131096e-06,
+ "loss": 0.44774049520492554,
+ "step": 1871
+ },
+ {
+ "epoch": 0.9821615949632738,
+ "grad_norm": 0.15058611877033465,
+ "learning_rate": 5.035322435800247e-06,
+ "loss": 0.4310590624809265,
+ "step": 1872
+ },
+ {
+ "epoch": 0.9826862539349422,
+ "grad_norm": 0.14470485792169563,
+ "learning_rate": 5.033275705969405e-06,
+ "loss": 0.41565707325935364,
+ "step": 1873
+ },
+ {
+ "epoch": 0.9832109129066107,
+ "grad_norm": 0.15361200081510454,
+ "learning_rate": 5.031290013199077e-06,
+ "loss": 0.44648170471191406,
+ "step": 1874
+ },
+ {
+ "epoch": 0.9837355718782791,
+ "grad_norm": 0.14425950736522083,
+ "learning_rate": 5.0293653628839475e-06,
+ "loss": 0.4347214102745056,
+ "step": 1875
+ },
+ {
+ "epoch": 0.9842602308499475,
+ "grad_norm": 0.1565812832501761,
+ "learning_rate": 5.027501760252857e-06,
+ "loss": 0.43630290031433105,
+ "step": 1876
+ },
+ {
+ "epoch": 0.9847848898216159,
+ "grad_norm": 0.14893609351255585,
+ "learning_rate": 5.025699210368801e-06,
+ "loss": 0.45179879665374756,
+ "step": 1877
+ },
+ {
+ "epoch": 0.9853095487932844,
+ "grad_norm": 0.1519863787002136,
+ "learning_rate": 5.023957718128899e-06,
+ "loss": 0.4456064701080322,
+ "step": 1878
+ },
+ {
+ "epoch": 0.9858342077649528,
+ "grad_norm": 0.14899484676330454,
+ "learning_rate": 5.022277288264397e-06,
+ "loss": 0.4311719238758087,
+ "step": 1879
+ },
+ {
+ "epoch": 0.9863588667366212,
+ "grad_norm": 0.15008205295390942,
+ "learning_rate": 5.020657925340649e-06,
+ "loss": 0.43028831481933594,
+ "step": 1880
+ },
+ {
+ "epoch": 0.9868835257082896,
+ "grad_norm": 0.1563139451666976,
+ "learning_rate": 5.019099633757102e-06,
+ "loss": 0.4408448040485382,
+ "step": 1881
+ },
+ {
+ "epoch": 0.9874081846799581,
+ "grad_norm": 0.14428766727134737,
+ "learning_rate": 5.017602417747283e-06,
+ "loss": 0.4480482339859009,
+ "step": 1882
+ },
+ {
+ "epoch": 0.9879328436516265,
+ "grad_norm": 0.14854657656236905,
+ "learning_rate": 5.0161662813787935e-06,
+ "loss": 0.43310457468032837,
+ "step": 1883
+ },
+ {
+ "epoch": 0.9884575026232949,
+ "grad_norm": 0.14979933582255975,
+ "learning_rate": 5.0147912285532935e-06,
+ "loss": 0.46105456352233887,
+ "step": 1884
+ },
+ {
+ "epoch": 0.9889821615949633,
+ "grad_norm": 0.15138043090447337,
+ "learning_rate": 5.0134772630064965e-06,
+ "loss": 0.4424416422843933,
+ "step": 1885
+ },
+ {
+ "epoch": 0.9895068205666316,
+ "grad_norm": 0.15826989524451818,
+ "learning_rate": 5.012224388308148e-06,
+ "loss": 0.45935922861099243,
+ "step": 1886
+ },
+ {
+ "epoch": 0.9900314795383001,
+ "grad_norm": 0.14514140559811425,
+ "learning_rate": 5.011032607862033e-06,
+ "loss": 0.4395596385002136,
+ "step": 1887
+ },
+ {
+ "epoch": 0.9905561385099685,
+ "grad_norm": 0.15650352549659236,
+ "learning_rate": 5.00990192490595e-06,
+ "loss": 0.47225916385650635,
+ "step": 1888
+ },
+ {
+ "epoch": 0.9910807974816369,
+ "grad_norm": 0.15839596050463609,
+ "learning_rate": 5.008832342511712e-06,
+ "loss": 0.43967339396476746,
+ "step": 1889
+ },
+ {
+ "epoch": 0.9916054564533053,
+ "grad_norm": 0.14915540503590238,
+ "learning_rate": 5.007823863585134e-06,
+ "loss": 0.4409275949001312,
+ "step": 1890
+ },
+ {
+ "epoch": 0.9921301154249738,
+ "grad_norm": 0.15382866906925854,
+ "learning_rate": 5.006876490866029e-06,
+ "loss": 0.44721555709838867,
+ "step": 1891
+ },
+ {
+ "epoch": 0.9926547743966422,
+ "grad_norm": 0.14953451753083624,
+ "learning_rate": 5.0059902269281956e-06,
+ "loss": 0.4325152039527893,
+ "step": 1892
+ },
+ {
+ "epoch": 0.9931794333683106,
+ "grad_norm": 0.1540833258535656,
+ "learning_rate": 5.0051650741794135e-06,
+ "loss": 0.4537501931190491,
+ "step": 1893
+ },
+ {
+ "epoch": 0.993704092339979,
+ "grad_norm": 0.14711400847665276,
+ "learning_rate": 5.004401034861441e-06,
+ "loss": 0.4396728277206421,
+ "step": 1894
+ },
+ {
+ "epoch": 0.9942287513116475,
+ "grad_norm": 0.1489662167447431,
+ "learning_rate": 5.003698111049998e-06,
+ "loss": 0.4431758522987366,
+ "step": 1895
+ },
+ {
+ "epoch": 0.9947534102833159,
+ "grad_norm": 0.14660584087882736,
+ "learning_rate": 5.003056304654774e-06,
+ "loss": 0.4379071295261383,
+ "step": 1896
+ },
+ {
+ "epoch": 0.9952780692549843,
+ "grad_norm": 0.14739479545415804,
+ "learning_rate": 5.0024756174194125e-06,
+ "loss": 0.4503260552883148,
+ "step": 1897
+ },
+ {
+ "epoch": 0.9958027282266527,
+ "grad_norm": 0.14953590570482198,
+ "learning_rate": 5.001956050921509e-06,
+ "loss": 0.43298959732055664,
+ "step": 1898
+ },
+ {
+ "epoch": 0.9963273871983211,
+ "grad_norm": 0.14530023273300657,
+ "learning_rate": 5.001497606572614e-06,
+ "loss": 0.4221065938472748,
+ "step": 1899
+ },
+ {
+ "epoch": 0.9968520461699895,
+ "grad_norm": 0.1506918434878359,
+ "learning_rate": 5.001100285618214e-06,
+ "loss": 0.4413088262081146,
+ "step": 1900
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 5.030609902423245e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1900/video_preprocessor_config.json b/checkpoints/checkpoint-1900/video_preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..a933f0ec6941222c33aeba7980bd5acf8dd98e91
--- /dev/null
+++ b/checkpoints/checkpoint-1900/video_preprocessor_config.json
@@ -0,0 +1,30 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "do_sample_frames": true,
+ "fps": 2,
+ "image_mean": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "image_std": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "max_frames": 768,
+ "merge_size": 2,
+ "min_frames": 4,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 786432,
+ "shortest_edge": 131072
+ },
+ "temporal_patch_size": 2,
+ "video_processor_type": "Qwen3VLVideoProcessor"
+}
diff --git a/checkpoints/checkpoint-1906/chat_template.jinja b/checkpoints/checkpoint-1906/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-1906/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-1906/config.json b/checkpoints/checkpoint-1906/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-1906/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-1906/expert_prune_metadata.json b/checkpoints/checkpoint-1906/expert_prune_metadata.json
new file mode 100644
index 0000000000000000000000000000000000000000..16e01c8a10636331e334732fa9629579c68df44e
--- /dev/null
+++ b/checkpoints/checkpoint-1906/expert_prune_metadata.json
@@ -0,0 +1,14 @@
+{
+ "source_model": "/mnt/data/wuwei/model_list/Qwen3.6-35B-A3B/models--Qwen--Qwen3.6-35B-A3B/snapshots/995ad96eacd98c81ed38be0c5b274b04031597b0",
+ "output_model": "/mnt/data/wuwei/model_list/qwen3-19b-a3b",
+ "total_parameters": 18990568816,
+ "parameter_billions_decimal": 18.990568816,
+ "directory_size_label": "19B",
+ "original_num_experts": 256,
+ "kept_experts_per_layer": 128,
+ "num_profiled_layers": 40,
+ "mtp_weights": "removed because Qwen3_5MoeForConditionalGeneration does not instantiate or use the source checkpoint's mtp.* training-only tensors",
+ "expert_order": "ascending original expert index; router rows and expert tensors use the same remapping",
+ "router_statistics": "router_stats.json and router_stats.pt",
+ "selection_score": "top-8 router hit frequency * mean(abs(actual expert output after down_proj and before routing-weight multiplication))"
+}
diff --git a/checkpoints/checkpoint-1906/generation_config.json b/checkpoints/checkpoint-1906/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-1906/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-1906/model.safetensors b/checkpoints/checkpoint-1906/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..b19e9d4b30c5530b6375030ba771707e44d298ce
--- /dev/null
+++ b/checkpoints/checkpoint-1906/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:857b4e46afb68bf9ffb9168248d55c2e47edf70a5efe824287e2a661898a4d44
+size 37983421000
diff --git a/checkpoints/checkpoint-1906/preprocessor_config.json b/checkpoints/checkpoint-1906/preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ec43d1f859ecadf800ebae92fda4a812d34975d3
--- /dev/null
+++ b/checkpoints/checkpoint-1906/preprocessor_config.json
@@ -0,0 +1,26 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "image_mean": [
+ 0.48145466,
+ 0.4578275,
+ 0.40821073
+ ],
+ "image_processor_type": "Qwen2VLImageProcessorFast",
+ "image_std": [
+ 0.26862954,
+ 0.26130258,
+ 0.27577711
+ ],
+ "merge_size": 2,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 1003520,
+ "shortest_edge": 3136
+ },
+ "temporal_patch_size": 2
+}
diff --git a/checkpoints/checkpoint-1906/tokenizer.json b/checkpoints/checkpoint-1906/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-1906/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-1906/tokenizer_config.json b/checkpoints/checkpoint-1906/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-1906/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-1906/trainer_state.json b/checkpoints/checkpoint-1906/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..08d7ed84fc32df1bc43b8022b30817144c547e1f
--- /dev/null
+++ b/checkpoints/checkpoint-1906/trainer_state.json
@@ -0,0 +1,13376 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 1.0,
+ "eval_steps": 500,
+ "global_step": 1906,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ },
+ {
+ "epoch": 0.49895068205666315,
+ "grad_norm": 0.17803498769495077,
+ "learning_rate": 2.757417178035756e-05,
+ "loss": 0.4589846134185791,
+ "step": 951
+ },
+ {
+ "epoch": 0.4994753410283316,
+ "grad_norm": 0.16144143885585,
+ "learning_rate": 2.7537085940555878e-05,
+ "loss": 0.46040093898773193,
+ "step": 952
+ },
+ {
+ "epoch": 0.5,
+ "grad_norm": 0.17150053721325598,
+ "learning_rate": 2.7500000000000004e-05,
+ "loss": 0.470710813999176,
+ "step": 953
+ },
+ {
+ "epoch": 0.5005246589716684,
+ "grad_norm": 0.1689098178014976,
+ "learning_rate": 2.7462914059444128e-05,
+ "loss": 0.45361438393592834,
+ "step": 954
+ },
+ {
+ "epoch": 0.5010493179433368,
+ "grad_norm": 0.16031315983164343,
+ "learning_rate": 2.742582821964244e-05,
+ "loss": 0.4646606743335724,
+ "step": 955
+ },
+ {
+ "epoch": 0.5015739769150053,
+ "grad_norm": 0.168678741064497,
+ "learning_rate": 2.7388742581348892e-05,
+ "loss": 0.4590177536010742,
+ "step": 956
+ },
+ {
+ "epoch": 0.5020986358866737,
+ "grad_norm": 0.16866828326172448,
+ "learning_rate": 2.7351657245316835e-05,
+ "loss": 0.4781872630119324,
+ "step": 957
+ },
+ {
+ "epoch": 0.5026232948583421,
+ "grad_norm": 0.17843804544581887,
+ "learning_rate": 2.731457231229883e-05,
+ "loss": 0.4623679518699646,
+ "step": 958
+ },
+ {
+ "epoch": 0.5031479538300105,
+ "grad_norm": 0.1674437450780801,
+ "learning_rate": 2.7277487883046354e-05,
+ "loss": 0.45383667945861816,
+ "step": 959
+ },
+ {
+ "epoch": 0.5036726128016789,
+ "grad_norm": 0.16821293805727164,
+ "learning_rate": 2.724040405830949e-05,
+ "loss": 0.4695635139942169,
+ "step": 960
+ },
+ {
+ "epoch": 0.5041972717733473,
+ "grad_norm": 0.1691520691167744,
+ "learning_rate": 2.720332093883668e-05,
+ "loss": 0.4580414891242981,
+ "step": 961
+ },
+ {
+ "epoch": 0.5047219307450157,
+ "grad_norm": 0.16762001011924296,
+ "learning_rate": 2.716623862537446e-05,
+ "loss": 0.4468068778514862,
+ "step": 962
+ },
+ {
+ "epoch": 0.5052465897166841,
+ "grad_norm": 0.1656916999758014,
+ "learning_rate": 2.7129157218667183e-05,
+ "loss": 0.45274806022644043,
+ "step": 963
+ },
+ {
+ "epoch": 0.5057712486883525,
+ "grad_norm": 0.1667884084393949,
+ "learning_rate": 2.709207681945673e-05,
+ "loss": 0.45753178000450134,
+ "step": 964
+ },
+ {
+ "epoch": 0.506295907660021,
+ "grad_norm": 0.1703286677805619,
+ "learning_rate": 2.705499752848223e-05,
+ "loss": 0.4668196439743042,
+ "step": 965
+ },
+ {
+ "epoch": 0.5068205666316894,
+ "grad_norm": 0.17151838523084442,
+ "learning_rate": 2.7017919446479838e-05,
+ "loss": 0.45340126752853394,
+ "step": 966
+ },
+ {
+ "epoch": 0.5073452256033578,
+ "grad_norm": 0.1660798409657293,
+ "learning_rate": 2.698084267418239e-05,
+ "loss": 0.4544919431209564,
+ "step": 967
+ },
+ {
+ "epoch": 0.5078698845750262,
+ "grad_norm": 0.17170083554707766,
+ "learning_rate": 2.6943767312319168e-05,
+ "loss": 0.46509140729904175,
+ "step": 968
+ },
+ {
+ "epoch": 0.5083945435466947,
+ "grad_norm": 0.16996696720172885,
+ "learning_rate": 2.690669346161565e-05,
+ "loss": 0.4586217999458313,
+ "step": 969
+ },
+ {
+ "epoch": 0.5089192025183631,
+ "grad_norm": 0.17006350225431985,
+ "learning_rate": 2.6869621222793177e-05,
+ "loss": 0.47401130199432373,
+ "step": 970
+ },
+ {
+ "epoch": 0.5094438614900315,
+ "grad_norm": 0.1773831952645447,
+ "learning_rate": 2.683255069656873e-05,
+ "loss": 0.4695100784301758,
+ "step": 971
+ },
+ {
+ "epoch": 0.5099685204616999,
+ "grad_norm": 0.1775917932807981,
+ "learning_rate": 2.6795481983654617e-05,
+ "loss": 0.46360254287719727,
+ "step": 972
+ },
+ {
+ "epoch": 0.5104931794333684,
+ "grad_norm": 0.17653261198492437,
+ "learning_rate": 2.6758415184758244e-05,
+ "loss": 0.4602853059768677,
+ "step": 973
+ },
+ {
+ "epoch": 0.5110178384050367,
+ "grad_norm": 0.17257610569119436,
+ "learning_rate": 2.6721350400581818e-05,
+ "loss": 0.47050923109054565,
+ "step": 974
+ },
+ {
+ "epoch": 0.5115424973767051,
+ "grad_norm": 0.1676557653266664,
+ "learning_rate": 2.668428773182203e-05,
+ "loss": 0.4552682638168335,
+ "step": 975
+ },
+ {
+ "epoch": 0.5120671563483735,
+ "grad_norm": 0.1774963225534265,
+ "learning_rate": 2.6647227279169874e-05,
+ "loss": 0.4596288502216339,
+ "step": 976
+ },
+ {
+ "epoch": 0.5125918153200419,
+ "grad_norm": 0.160899136643459,
+ "learning_rate": 2.6610169143310287e-05,
+ "loss": 0.4616691470146179,
+ "step": 977
+ },
+ {
+ "epoch": 0.5131164742917104,
+ "grad_norm": 0.17197082121203497,
+ "learning_rate": 2.6573113424921952e-05,
+ "loss": 0.455634742975235,
+ "step": 978
+ },
+ {
+ "epoch": 0.5136411332633788,
+ "grad_norm": 0.1659655138461993,
+ "learning_rate": 2.6536060224676952e-05,
+ "loss": 0.4632532000541687,
+ "step": 979
+ },
+ {
+ "epoch": 0.5141657922350472,
+ "grad_norm": 0.16611632136942467,
+ "learning_rate": 2.6499009643240523e-05,
+ "loss": 0.4387845993041992,
+ "step": 980
+ },
+ {
+ "epoch": 0.5146904512067156,
+ "grad_norm": 0.1648219016226827,
+ "learning_rate": 2.6461961781270822e-05,
+ "loss": 0.45963478088378906,
+ "step": 981
+ },
+ {
+ "epoch": 0.5152151101783841,
+ "grad_norm": 0.17181760034375626,
+ "learning_rate": 2.6424916739418586e-05,
+ "loss": 0.45226943492889404,
+ "step": 982
+ },
+ {
+ "epoch": 0.5157397691500525,
+ "grad_norm": 0.1693836254107806,
+ "learning_rate": 2.638787461832689e-05,
+ "loss": 0.4472505450248718,
+ "step": 983
+ },
+ {
+ "epoch": 0.5162644281217209,
+ "grad_norm": 0.1769203741135292,
+ "learning_rate": 2.6350835518630905e-05,
+ "loss": 0.46511465311050415,
+ "step": 984
+ },
+ {
+ "epoch": 0.5167890870933893,
+ "grad_norm": 0.1675936563542835,
+ "learning_rate": 2.6313799540957567e-05,
+ "loss": 0.438603013753891,
+ "step": 985
+ },
+ {
+ "epoch": 0.5173137460650578,
+ "grad_norm": 0.1777212861804669,
+ "learning_rate": 2.6276766785925327e-05,
+ "loss": 0.4535655975341797,
+ "step": 986
+ },
+ {
+ "epoch": 0.5178384050367262,
+ "grad_norm": 0.17570430219010275,
+ "learning_rate": 2.62397373541439e-05,
+ "loss": 0.46286165714263916,
+ "step": 987
+ },
+ {
+ "epoch": 0.5183630640083945,
+ "grad_norm": 0.16865061203576853,
+ "learning_rate": 2.620271134621397e-05,
+ "loss": 0.45542284846305847,
+ "step": 988
+ },
+ {
+ "epoch": 0.5188877229800629,
+ "grad_norm": 0.16338789272527232,
+ "learning_rate": 2.6165688862726895e-05,
+ "loss": 0.435546875,
+ "step": 989
+ },
+ {
+ "epoch": 0.5194123819517313,
+ "grad_norm": 0.1673882461596605,
+ "learning_rate": 2.6128670004264473e-05,
+ "loss": 0.463608980178833,
+ "step": 990
+ },
+ {
+ "epoch": 0.5199370409233998,
+ "grad_norm": 0.17100774854301634,
+ "learning_rate": 2.6091654871398667e-05,
+ "loss": 0.4598499536514282,
+ "step": 991
+ },
+ {
+ "epoch": 0.5204616998950682,
+ "grad_norm": 0.15887127689261268,
+ "learning_rate": 2.605464356469131e-05,
+ "loss": 0.4522812068462372,
+ "step": 992
+ },
+ {
+ "epoch": 0.5209863588667366,
+ "grad_norm": 0.1745990359763196,
+ "learning_rate": 2.6017636184693828e-05,
+ "loss": 0.4535726308822632,
+ "step": 993
+ },
+ {
+ "epoch": 0.521511017838405,
+ "grad_norm": 0.1665669045299794,
+ "learning_rate": 2.598063283194699e-05,
+ "loss": 0.4681829810142517,
+ "step": 994
+ },
+ {
+ "epoch": 0.5220356768100735,
+ "grad_norm": 0.16020277991841533,
+ "learning_rate": 2.5943633606980615e-05,
+ "loss": 0.4633028507232666,
+ "step": 995
+ },
+ {
+ "epoch": 0.5225603357817419,
+ "grad_norm": 0.16320320145244704,
+ "learning_rate": 2.590663861031333e-05,
+ "loss": 0.4587368965148926,
+ "step": 996
+ },
+ {
+ "epoch": 0.5230849947534103,
+ "grad_norm": 0.1646852443813363,
+ "learning_rate": 2.5869647942452258e-05,
+ "loss": 0.4444723129272461,
+ "step": 997
+ },
+ {
+ "epoch": 0.5236096537250787,
+ "grad_norm": 0.16649446590222183,
+ "learning_rate": 2.5832661703892745e-05,
+ "loss": 0.4698568284511566,
+ "step": 998
+ },
+ {
+ "epoch": 0.5241343126967472,
+ "grad_norm": 0.16382444507376978,
+ "learning_rate": 2.579567999511814e-05,
+ "loss": 0.46485698223114014,
+ "step": 999
+ },
+ {
+ "epoch": 0.5246589716684156,
+ "grad_norm": 0.16665143626895657,
+ "learning_rate": 2.5758702916599458e-05,
+ "loss": 0.46017903089523315,
+ "step": 1000
+ },
+ {
+ "epoch": 0.525183630640084,
+ "grad_norm": 0.16854450510619526,
+ "learning_rate": 2.5721730568795137e-05,
+ "loss": 0.47064143419265747,
+ "step": 1001
+ },
+ {
+ "epoch": 0.5257082896117523,
+ "grad_norm": 0.15934231102978258,
+ "learning_rate": 2.5684763052150796e-05,
+ "loss": 0.4551209509372711,
+ "step": 1002
+ },
+ {
+ "epoch": 0.5262329485834207,
+ "grad_norm": 0.16918453954382703,
+ "learning_rate": 2.564780046709887e-05,
+ "loss": 0.45295917987823486,
+ "step": 1003
+ },
+ {
+ "epoch": 0.5267576075550892,
+ "grad_norm": 0.16073375378306387,
+ "learning_rate": 2.5610842914058448e-05,
+ "loss": 0.45727837085723877,
+ "step": 1004
+ },
+ {
+ "epoch": 0.5272822665267576,
+ "grad_norm": 0.16038893352694694,
+ "learning_rate": 2.5573890493434922e-05,
+ "loss": 0.44097083806991577,
+ "step": 1005
+ },
+ {
+ "epoch": 0.527806925498426,
+ "grad_norm": 0.16328199965840876,
+ "learning_rate": 2.5536943305619744e-05,
+ "loss": 0.4662286937236786,
+ "step": 1006
+ },
+ {
+ "epoch": 0.5283315844700944,
+ "grad_norm": 0.17325178916276576,
+ "learning_rate": 2.5500001450990156e-05,
+ "loss": 0.4626012444496155,
+ "step": 1007
+ },
+ {
+ "epoch": 0.5288562434417629,
+ "grad_norm": 0.16746411961599553,
+ "learning_rate": 2.5463065029908907e-05,
+ "loss": 0.4701128602027893,
+ "step": 1008
+ },
+ {
+ "epoch": 0.5293809024134313,
+ "grad_norm": 0.1756940380379008,
+ "learning_rate": 2.5426134142723984e-05,
+ "loss": 0.48012205958366394,
+ "step": 1009
+ },
+ {
+ "epoch": 0.5299055613850997,
+ "grad_norm": 0.16932842599866052,
+ "learning_rate": 2.538920888976834e-05,
+ "loss": 0.45821353793144226,
+ "step": 1010
+ },
+ {
+ "epoch": 0.5304302203567681,
+ "grad_norm": 0.1681007385601024,
+ "learning_rate": 2.5352289371359618e-05,
+ "loss": 0.4578600823879242,
+ "step": 1011
+ },
+ {
+ "epoch": 0.5309548793284365,
+ "grad_norm": 0.17043434771022006,
+ "learning_rate": 2.5315375687799898e-05,
+ "loss": 0.4651893973350525,
+ "step": 1012
+ },
+ {
+ "epoch": 0.531479538300105,
+ "grad_norm": 0.16270251928699306,
+ "learning_rate": 2.5278467939375373e-05,
+ "loss": 0.456525981426239,
+ "step": 1013
+ },
+ {
+ "epoch": 0.5320041972717734,
+ "grad_norm": 0.17078632830073862,
+ "learning_rate": 2.5241566226356154e-05,
+ "loss": 0.466685026884079,
+ "step": 1014
+ },
+ {
+ "epoch": 0.5325288562434418,
+ "grad_norm": 0.1625702045735913,
+ "learning_rate": 2.5204670648995927e-05,
+ "loss": 0.4558373689651489,
+ "step": 1015
+ },
+ {
+ "epoch": 0.5330535152151101,
+ "grad_norm": 0.1704391221157713,
+ "learning_rate": 2.516778130753171e-05,
+ "loss": 0.4758946895599365,
+ "step": 1016
+ },
+ {
+ "epoch": 0.5335781741867786,
+ "grad_norm": 0.16115108665517358,
+ "learning_rate": 2.5130898302183603e-05,
+ "loss": 0.456872820854187,
+ "step": 1017
+ },
+ {
+ "epoch": 0.534102833158447,
+ "grad_norm": 0.16540130694959762,
+ "learning_rate": 2.5094021733154444e-05,
+ "loss": 0.45968422293663025,
+ "step": 1018
+ },
+ {
+ "epoch": 0.5346274921301154,
+ "grad_norm": 0.17080016840172665,
+ "learning_rate": 2.505715170062965e-05,
+ "loss": 0.458723783493042,
+ "step": 1019
+ },
+ {
+ "epoch": 0.5351521511017838,
+ "grad_norm": 0.1774328646288112,
+ "learning_rate": 2.5020288304776822e-05,
+ "loss": 0.4874815046787262,
+ "step": 1020
+ },
+ {
+ "epoch": 0.5356768100734522,
+ "grad_norm": 0.16815678068833295,
+ "learning_rate": 2.498343164574556e-05,
+ "loss": 0.4765281081199646,
+ "step": 1021
+ },
+ {
+ "epoch": 0.5362014690451207,
+ "grad_norm": 0.17119419603235367,
+ "learning_rate": 2.4946581823667165e-05,
+ "loss": 0.4710816740989685,
+ "step": 1022
+ },
+ {
+ "epoch": 0.5367261280167891,
+ "grad_norm": 0.17187022200465696,
+ "learning_rate": 2.4909738938654343e-05,
+ "loss": 0.46100884675979614,
+ "step": 1023
+ },
+ {
+ "epoch": 0.5372507869884575,
+ "grad_norm": 0.17200118624320845,
+ "learning_rate": 2.4872903090800968e-05,
+ "loss": 0.4666250944137573,
+ "step": 1024
+ },
+ {
+ "epoch": 0.5377754459601259,
+ "grad_norm": 0.17173829895853374,
+ "learning_rate": 2.48360743801818e-05,
+ "loss": 0.46126851439476013,
+ "step": 1025
+ },
+ {
+ "epoch": 0.5383001049317944,
+ "grad_norm": 0.16592322512628485,
+ "learning_rate": 2.479925290685219e-05,
+ "loss": 0.46757131814956665,
+ "step": 1026
+ },
+ {
+ "epoch": 0.5388247639034628,
+ "grad_norm": 0.17156594730450153,
+ "learning_rate": 2.4762438770847852e-05,
+ "loss": 0.472138911485672,
+ "step": 1027
+ },
+ {
+ "epoch": 0.5393494228751312,
+ "grad_norm": 0.18757123043190707,
+ "learning_rate": 2.472563207218454e-05,
+ "loss": 0.4632652699947357,
+ "step": 1028
+ },
+ {
+ "epoch": 0.5398740818467995,
+ "grad_norm": 0.16977357824305928,
+ "learning_rate": 2.4688832910857822e-05,
+ "loss": 0.4638451337814331,
+ "step": 1029
+ },
+ {
+ "epoch": 0.540398740818468,
+ "grad_norm": 0.16682513073505026,
+ "learning_rate": 2.4652041386842772e-05,
+ "loss": 0.4547071158885956,
+ "step": 1030
+ },
+ {
+ "epoch": 0.5409233997901364,
+ "grad_norm": 0.1701947730920509,
+ "learning_rate": 2.4615257600093733e-05,
+ "loss": 0.4719548225402832,
+ "step": 1031
+ },
+ {
+ "epoch": 0.5414480587618048,
+ "grad_norm": 0.1666852040891393,
+ "learning_rate": 2.457848165054402e-05,
+ "loss": 0.47497278451919556,
+ "step": 1032
+ },
+ {
+ "epoch": 0.5419727177334732,
+ "grad_norm": 0.17491620179696482,
+ "learning_rate": 2.4541713638105645e-05,
+ "loss": 0.4707529544830322,
+ "step": 1033
+ },
+ {
+ "epoch": 0.5424973767051416,
+ "grad_norm": 0.16416692627941032,
+ "learning_rate": 2.4504953662669082e-05,
+ "loss": 0.4550850987434387,
+ "step": 1034
+ },
+ {
+ "epoch": 0.5430220356768101,
+ "grad_norm": 0.17472986117515302,
+ "learning_rate": 2.446820182410295e-05,
+ "loss": 0.47211822867393494,
+ "step": 1035
+ },
+ {
+ "epoch": 0.5435466946484785,
+ "grad_norm": 0.15886538942718104,
+ "learning_rate": 2.4431458222253754e-05,
+ "loss": 0.4502398371696472,
+ "step": 1036
+ },
+ {
+ "epoch": 0.5440713536201469,
+ "grad_norm": 0.176519810790353,
+ "learning_rate": 2.4394722956945652e-05,
+ "loss": 0.4647930860519409,
+ "step": 1037
+ },
+ {
+ "epoch": 0.5445960125918153,
+ "grad_norm": 0.1639300373024075,
+ "learning_rate": 2.435799612798013e-05,
+ "loss": 0.46724116802215576,
+ "step": 1038
+ },
+ {
+ "epoch": 0.5451206715634838,
+ "grad_norm": 0.17144573217657566,
+ "learning_rate": 2.4321277835135757e-05,
+ "loss": 0.45073503255844116,
+ "step": 1039
+ },
+ {
+ "epoch": 0.5456453305351522,
+ "grad_norm": 0.1687915468277188,
+ "learning_rate": 2.4284568178167923e-05,
+ "loss": 0.4591040015220642,
+ "step": 1040
+ },
+ {
+ "epoch": 0.5461699895068206,
+ "grad_norm": 0.17053171275539503,
+ "learning_rate": 2.4247867256808536e-05,
+ "loss": 0.4502571225166321,
+ "step": 1041
+ },
+ {
+ "epoch": 0.546694648478489,
+ "grad_norm": 0.17125742415678732,
+ "learning_rate": 2.4211175170765797e-05,
+ "loss": 0.443480908870697,
+ "step": 1042
+ },
+ {
+ "epoch": 0.5472193074501573,
+ "grad_norm": 0.16797028359369434,
+ "learning_rate": 2.4174492019723892e-05,
+ "loss": 0.45717787742614746,
+ "step": 1043
+ },
+ {
+ "epoch": 0.5477439664218258,
+ "grad_norm": 0.173042670970041,
+ "learning_rate": 2.413781790334272e-05,
+ "loss": 0.4510262608528137,
+ "step": 1044
+ },
+ {
+ "epoch": 0.5482686253934942,
+ "grad_norm": 0.163069552280238,
+ "learning_rate": 2.4101152921257636e-05,
+ "loss": 0.4488248825073242,
+ "step": 1045
+ },
+ {
+ "epoch": 0.5487932843651626,
+ "grad_norm": 0.1603346875007337,
+ "learning_rate": 2.4064497173079204e-05,
+ "loss": 0.4401039481163025,
+ "step": 1046
+ },
+ {
+ "epoch": 0.549317943336831,
+ "grad_norm": 0.1693292044826663,
+ "learning_rate": 2.4027850758392885e-05,
+ "loss": 0.45918384194374084,
+ "step": 1047
+ },
+ {
+ "epoch": 0.5498426023084995,
+ "grad_norm": 0.16586524007132455,
+ "learning_rate": 2.3991213776758775e-05,
+ "loss": 0.48708122968673706,
+ "step": 1048
+ },
+ {
+ "epoch": 0.5503672612801679,
+ "grad_norm": 0.16761507557611402,
+ "learning_rate": 2.3954586327711363e-05,
+ "loss": 0.45938029885292053,
+ "step": 1049
+ },
+ {
+ "epoch": 0.5508919202518363,
+ "grad_norm": 0.15507168956542738,
+ "learning_rate": 2.391796851075923e-05,
+ "loss": 0.4463997781276703,
+ "step": 1050
+ },
+ {
+ "epoch": 0.5514165792235047,
+ "grad_norm": 0.16401550913586452,
+ "learning_rate": 2.388136042538477e-05,
+ "loss": 0.47262513637542725,
+ "step": 1051
+ },
+ {
+ "epoch": 0.5519412381951732,
+ "grad_norm": 0.16309515575711464,
+ "learning_rate": 2.384476217104397e-05,
+ "loss": 0.4534379839897156,
+ "step": 1052
+ },
+ {
+ "epoch": 0.5524658971668416,
+ "grad_norm": 0.16168451905031236,
+ "learning_rate": 2.3808173847166097e-05,
+ "loss": 0.4452853202819824,
+ "step": 1053
+ },
+ {
+ "epoch": 0.55299055613851,
+ "grad_norm": 0.1669659787110618,
+ "learning_rate": 2.377159555315342e-05,
+ "loss": 0.4565466642379761,
+ "step": 1054
+ },
+ {
+ "epoch": 0.5535152151101784,
+ "grad_norm": 0.16655292320798284,
+ "learning_rate": 2.3735027388380994e-05,
+ "loss": 0.46129778027534485,
+ "step": 1055
+ },
+ {
+ "epoch": 0.5540398740818469,
+ "grad_norm": 0.16647174335134302,
+ "learning_rate": 2.3698469452196327e-05,
+ "loss": 0.4564422369003296,
+ "step": 1056
+ },
+ {
+ "epoch": 0.5545645330535152,
+ "grad_norm": 0.16078152741988602,
+ "learning_rate": 2.3661921843919155e-05,
+ "loss": 0.4478951692581177,
+ "step": 1057
+ },
+ {
+ "epoch": 0.5550891920251836,
+ "grad_norm": 0.16599106977468595,
+ "learning_rate": 2.3625384662841132e-05,
+ "loss": 0.46234452724456787,
+ "step": 1058
+ },
+ {
+ "epoch": 0.555613850996852,
+ "grad_norm": 0.16547625723108067,
+ "learning_rate": 2.3588858008225617e-05,
+ "loss": 0.46375685930252075,
+ "step": 1059
+ },
+ {
+ "epoch": 0.5561385099685204,
+ "grad_norm": 0.16417349009686877,
+ "learning_rate": 2.3552341979307347e-05,
+ "loss": 0.45512670278549194,
+ "step": 1060
+ },
+ {
+ "epoch": 0.5566631689401889,
+ "grad_norm": 0.16460457851307972,
+ "learning_rate": 2.3515836675292187e-05,
+ "loss": 0.4700523614883423,
+ "step": 1061
+ },
+ {
+ "epoch": 0.5571878279118573,
+ "grad_norm": 0.2070392684158882,
+ "learning_rate": 2.3479342195356896e-05,
+ "loss": 0.45811665058135986,
+ "step": 1062
+ },
+ {
+ "epoch": 0.5577124868835257,
+ "grad_norm": 0.16997059677904428,
+ "learning_rate": 2.3442858638648796e-05,
+ "loss": 0.4636073708534241,
+ "step": 1063
+ },
+ {
+ "epoch": 0.5582371458551941,
+ "grad_norm": 0.16551460271499602,
+ "learning_rate": 2.340638610428554e-05,
+ "loss": 0.45901966094970703,
+ "step": 1064
+ },
+ {
+ "epoch": 0.5587618048268626,
+ "grad_norm": 0.15753502642408268,
+ "learning_rate": 2.336992469135485e-05,
+ "loss": 0.4570973813533783,
+ "step": 1065
+ },
+ {
+ "epoch": 0.559286463798531,
+ "grad_norm": 0.16972166998460111,
+ "learning_rate": 2.333347449891421e-05,
+ "loss": 0.44186559319496155,
+ "step": 1066
+ },
+ {
+ "epoch": 0.5598111227701994,
+ "grad_norm": 0.16012632536422983,
+ "learning_rate": 2.3297035625990643e-05,
+ "loss": 0.45937681198120117,
+ "step": 1067
+ },
+ {
+ "epoch": 0.5603357817418678,
+ "grad_norm": 0.16132277641147474,
+ "learning_rate": 2.3260608171580412e-05,
+ "loss": 0.47288697957992554,
+ "step": 1068
+ },
+ {
+ "epoch": 0.5608604407135362,
+ "grad_norm": 0.1645087207282203,
+ "learning_rate": 2.3224192234648738e-05,
+ "loss": 0.45268645882606506,
+ "step": 1069
+ },
+ {
+ "epoch": 0.5613850996852047,
+ "grad_norm": 0.1600285098330049,
+ "learning_rate": 2.3187787914129588e-05,
+ "loss": 0.45290541648864746,
+ "step": 1070
+ },
+ {
+ "epoch": 0.561909758656873,
+ "grad_norm": 0.1713735388083292,
+ "learning_rate": 2.3151395308925357e-05,
+ "loss": 0.45641642808914185,
+ "step": 1071
+ },
+ {
+ "epoch": 0.5624344176285414,
+ "grad_norm": 0.16076163994202516,
+ "learning_rate": 2.311501451790658e-05,
+ "loss": 0.43980205059051514,
+ "step": 1072
+ },
+ {
+ "epoch": 0.5629590766002098,
+ "grad_norm": 0.1683846829455339,
+ "learning_rate": 2.3078645639911744e-05,
+ "loss": 0.4654780328273773,
+ "step": 1073
+ },
+ {
+ "epoch": 0.5634837355718783,
+ "grad_norm": 0.17147138009150142,
+ "learning_rate": 2.3042288773746933e-05,
+ "loss": 0.45051276683807373,
+ "step": 1074
+ },
+ {
+ "epoch": 0.5640083945435467,
+ "grad_norm": 0.1705090352117141,
+ "learning_rate": 2.3005944018185628e-05,
+ "loss": 0.4673599302768707,
+ "step": 1075
+ },
+ {
+ "epoch": 0.5645330535152151,
+ "grad_norm": 0.1674460275860359,
+ "learning_rate": 2.2969611471968382e-05,
+ "loss": 0.4713797867298126,
+ "step": 1076
+ },
+ {
+ "epoch": 0.5650577124868835,
+ "grad_norm": 0.1687831459870983,
+ "learning_rate": 2.29332912338026e-05,
+ "loss": 0.46389394998550415,
+ "step": 1077
+ },
+ {
+ "epoch": 0.565582371458552,
+ "grad_norm": 0.15976572678594333,
+ "learning_rate": 2.2896983402362234e-05,
+ "loss": 0.4625714421272278,
+ "step": 1078
+ },
+ {
+ "epoch": 0.5661070304302204,
+ "grad_norm": 0.16276239425946248,
+ "learning_rate": 2.2860688076287528e-05,
+ "loss": 0.4624657928943634,
+ "step": 1079
+ },
+ {
+ "epoch": 0.5666316894018888,
+ "grad_norm": 0.1743525592442312,
+ "learning_rate": 2.2824405354184775e-05,
+ "loss": 0.45578494668006897,
+ "step": 1080
+ },
+ {
+ "epoch": 0.5671563483735572,
+ "grad_norm": 0.1660454992376916,
+ "learning_rate": 2.2788135334625992e-05,
+ "loss": 0.4678182005882263,
+ "step": 1081
+ },
+ {
+ "epoch": 0.5676810073452256,
+ "grad_norm": 0.17229894593055467,
+ "learning_rate": 2.2751878116148706e-05,
+ "loss": 0.4494876563549042,
+ "step": 1082
+ },
+ {
+ "epoch": 0.5682056663168941,
+ "grad_norm": 0.1737439892096869,
+ "learning_rate": 2.271563379725567e-05,
+ "loss": 0.4581034779548645,
+ "step": 1083
+ },
+ {
+ "epoch": 0.5687303252885625,
+ "grad_norm": 0.17117021542919106,
+ "learning_rate": 2.267940247641458e-05,
+ "loss": 0.45302605628967285,
+ "step": 1084
+ },
+ {
+ "epoch": 0.5692549842602308,
+ "grad_norm": 0.18016699255331403,
+ "learning_rate": 2.264318425205783e-05,
+ "loss": 0.4793376326560974,
+ "step": 1085
+ },
+ {
+ "epoch": 0.5697796432318992,
+ "grad_norm": 0.17090795453695037,
+ "learning_rate": 2.2606979222582208e-05,
+ "loss": 0.44775575399398804,
+ "step": 1086
+ },
+ {
+ "epoch": 0.5703043022035676,
+ "grad_norm": 0.17023784814953102,
+ "learning_rate": 2.25707874863487e-05,
+ "loss": 0.4730002284049988,
+ "step": 1087
+ },
+ {
+ "epoch": 0.5708289611752361,
+ "grad_norm": 0.1771558721338525,
+ "learning_rate": 2.2534609141682134e-05,
+ "loss": 0.4453514814376831,
+ "step": 1088
+ },
+ {
+ "epoch": 0.5713536201469045,
+ "grad_norm": 0.16396052310380732,
+ "learning_rate": 2.249844428687098e-05,
+ "loss": 0.46599841117858887,
+ "step": 1089
+ },
+ {
+ "epoch": 0.5718782791185729,
+ "grad_norm": 0.17458205812689367,
+ "learning_rate": 2.2462293020167052e-05,
+ "loss": 0.4596468210220337,
+ "step": 1090
+ },
+ {
+ "epoch": 0.5724029380902413,
+ "grad_norm": 0.17421286364398095,
+ "learning_rate": 2.242615543978525e-05,
+ "loss": 0.4522855877876282,
+ "step": 1091
+ },
+ {
+ "epoch": 0.5729275970619098,
+ "grad_norm": 0.16957892614104514,
+ "learning_rate": 2.2390031643903285e-05,
+ "loss": 0.4533312916755676,
+ "step": 1092
+ },
+ {
+ "epoch": 0.5734522560335782,
+ "grad_norm": 0.180432192546227,
+ "learning_rate": 2.2353921730661436e-05,
+ "loss": 0.465215265750885,
+ "step": 1093
+ },
+ {
+ "epoch": 0.5739769150052466,
+ "grad_norm": 0.17340613196675436,
+ "learning_rate": 2.2317825798162233e-05,
+ "loss": 0.4453255236148834,
+ "step": 1094
+ },
+ {
+ "epoch": 0.574501573976915,
+ "grad_norm": 0.18003266614794608,
+ "learning_rate": 2.2281743944470272e-05,
+ "loss": 0.4505641460418701,
+ "step": 1095
+ },
+ {
+ "epoch": 0.5750262329485835,
+ "grad_norm": 0.172035768572177,
+ "learning_rate": 2.2245676267611847e-05,
+ "loss": 0.46140533685684204,
+ "step": 1096
+ },
+ {
+ "epoch": 0.5755508919202519,
+ "grad_norm": 0.16713628890831114,
+ "learning_rate": 2.2209622865574775e-05,
+ "loss": 0.43574708700180054,
+ "step": 1097
+ },
+ {
+ "epoch": 0.5760755508919203,
+ "grad_norm": 0.17017611071190292,
+ "learning_rate": 2.2173583836308078e-05,
+ "loss": 0.46617692708969116,
+ "step": 1098
+ },
+ {
+ "epoch": 0.5766002098635886,
+ "grad_norm": 0.17140413725613413,
+ "learning_rate": 2.213755927772173e-05,
+ "loss": 0.45045268535614014,
+ "step": 1099
+ },
+ {
+ "epoch": 0.577124868835257,
+ "grad_norm": 0.16760828903438224,
+ "learning_rate": 2.2101549287686384e-05,
+ "loss": 0.45884764194488525,
+ "step": 1100
+ },
+ {
+ "epoch": 0.5776495278069255,
+ "grad_norm": 0.1862053032854098,
+ "learning_rate": 2.2065553964033126e-05,
+ "loss": 0.4816875457763672,
+ "step": 1101
+ },
+ {
+ "epoch": 0.5781741867785939,
+ "grad_norm": 0.16996015894363117,
+ "learning_rate": 2.20295734045532e-05,
+ "loss": 0.4471197724342346,
+ "step": 1102
+ },
+ {
+ "epoch": 0.5786988457502623,
+ "grad_norm": 0.16889227008390692,
+ "learning_rate": 2.1993607706997725e-05,
+ "loss": 0.468391478061676,
+ "step": 1103
+ },
+ {
+ "epoch": 0.5792235047219307,
+ "grad_norm": 0.16670931681562373,
+ "learning_rate": 2.195765696907745e-05,
+ "loss": 0.44934162497520447,
+ "step": 1104
+ },
+ {
+ "epoch": 0.5797481636935992,
+ "grad_norm": 0.16328418741223716,
+ "learning_rate": 2.1921721288462482e-05,
+ "loss": 0.4558270275592804,
+ "step": 1105
+ },
+ {
+ "epoch": 0.5802728226652676,
+ "grad_norm": 0.17364660408703936,
+ "learning_rate": 2.188580076278203e-05,
+ "loss": 0.4597751796245575,
+ "step": 1106
+ },
+ {
+ "epoch": 0.580797481636936,
+ "grad_norm": 0.1628618453800601,
+ "learning_rate": 2.18498954896241e-05,
+ "loss": 0.4521859288215637,
+ "step": 1107
+ },
+ {
+ "epoch": 0.5813221406086044,
+ "grad_norm": 0.16813817951871482,
+ "learning_rate": 2.1814005566535302e-05,
+ "loss": 0.4581005573272705,
+ "step": 1108
+ },
+ {
+ "epoch": 0.5818467995802729,
+ "grad_norm": 0.16825923629726025,
+ "learning_rate": 2.1778131091020516e-05,
+ "loss": 0.4433916211128235,
+ "step": 1109
+ },
+ {
+ "epoch": 0.5823714585519413,
+ "grad_norm": 0.17194743467883944,
+ "learning_rate": 2.1742272160542654e-05,
+ "loss": 0.47296425700187683,
+ "step": 1110
+ },
+ {
+ "epoch": 0.5828961175236097,
+ "grad_norm": 0.16947730546039969,
+ "learning_rate": 2.170642887252241e-05,
+ "loss": 0.4815937876701355,
+ "step": 1111
+ },
+ {
+ "epoch": 0.583420776495278,
+ "grad_norm": 0.1657449928668473,
+ "learning_rate": 2.167060132433798e-05,
+ "loss": 0.4479164779186249,
+ "step": 1112
+ },
+ {
+ "epoch": 0.5839454354669464,
+ "grad_norm": 0.1658952516607364,
+ "learning_rate": 2.1634789613324778e-05,
+ "loss": 0.4637463688850403,
+ "step": 1113
+ },
+ {
+ "epoch": 0.5844700944386149,
+ "grad_norm": 0.16777002188998236,
+ "learning_rate": 2.1598993836775218e-05,
+ "loss": 0.457524836063385,
+ "step": 1114
+ },
+ {
+ "epoch": 0.5849947534102833,
+ "grad_norm": 0.15837566447341322,
+ "learning_rate": 2.1563214091938405e-05,
+ "loss": 0.44938087463378906,
+ "step": 1115
+ },
+ {
+ "epoch": 0.5855194123819517,
+ "grad_norm": 0.17162206617035314,
+ "learning_rate": 2.1527450476019896e-05,
+ "loss": 0.4393450617790222,
+ "step": 1116
+ },
+ {
+ "epoch": 0.5860440713536201,
+ "grad_norm": 0.16224101921433814,
+ "learning_rate": 2.1491703086181426e-05,
+ "loss": 0.451913982629776,
+ "step": 1117
+ },
+ {
+ "epoch": 0.5865687303252886,
+ "grad_norm": 0.1736253504575144,
+ "learning_rate": 2.145597201954066e-05,
+ "loss": 0.4611290395259857,
+ "step": 1118
+ },
+ {
+ "epoch": 0.587093389296957,
+ "grad_norm": 0.16308408333737234,
+ "learning_rate": 2.1420257373170903e-05,
+ "loss": 0.45898568630218506,
+ "step": 1119
+ },
+ {
+ "epoch": 0.5876180482686254,
+ "grad_norm": 0.16948634794175524,
+ "learning_rate": 2.1384559244100848e-05,
+ "loss": 0.45982882380485535,
+ "step": 1120
+ },
+ {
+ "epoch": 0.5881427072402938,
+ "grad_norm": 0.16676366538565252,
+ "learning_rate": 2.1348877729314336e-05,
+ "loss": 0.44826576113700867,
+ "step": 1121
+ },
+ {
+ "epoch": 0.5886673662119623,
+ "grad_norm": 0.16431125630712295,
+ "learning_rate": 2.1313212925750037e-05,
+ "loss": 0.4616445302963257,
+ "step": 1122
+ },
+ {
+ "epoch": 0.5891920251836307,
+ "grad_norm": 0.1631659061616955,
+ "learning_rate": 2.1277564930301252e-05,
+ "loss": 0.45651188492774963,
+ "step": 1123
+ },
+ {
+ "epoch": 0.5897166841552991,
+ "grad_norm": 0.15805950502689237,
+ "learning_rate": 2.1241933839815604e-05,
+ "loss": 0.4486045837402344,
+ "step": 1124
+ },
+ {
+ "epoch": 0.5902413431269675,
+ "grad_norm": 0.16530288038603502,
+ "learning_rate": 2.1206319751094773e-05,
+ "loss": 0.4485066533088684,
+ "step": 1125
+ },
+ {
+ "epoch": 0.5907660020986358,
+ "grad_norm": 0.16061903194239047,
+ "learning_rate": 2.1170722760894295e-05,
+ "loss": 0.4565330743789673,
+ "step": 1126
+ },
+ {
+ "epoch": 0.5912906610703043,
+ "grad_norm": 0.1643346999353627,
+ "learning_rate": 2.1135142965923195e-05,
+ "loss": 0.45560717582702637,
+ "step": 1127
+ },
+ {
+ "epoch": 0.5918153200419727,
+ "grad_norm": 0.17484644315215753,
+ "learning_rate": 2.109958046284382e-05,
+ "loss": 0.4795891046524048,
+ "step": 1128
+ },
+ {
+ "epoch": 0.5923399790136411,
+ "grad_norm": 0.1623762678622651,
+ "learning_rate": 2.1064035348271528e-05,
+ "loss": 0.47395604848861694,
+ "step": 1129
+ },
+ {
+ "epoch": 0.5928646379853095,
+ "grad_norm": 0.16888824241669015,
+ "learning_rate": 2.1028507718774456e-05,
+ "loss": 0.4601455330848694,
+ "step": 1130
+ },
+ {
+ "epoch": 0.593389296956978,
+ "grad_norm": 0.16291744568806302,
+ "learning_rate": 2.0992997670873193e-05,
+ "loss": 0.46248236298561096,
+ "step": 1131
+ },
+ {
+ "epoch": 0.5939139559286464,
+ "grad_norm": 0.16185598691236988,
+ "learning_rate": 2.0957505301040603e-05,
+ "loss": 0.45085301995277405,
+ "step": 1132
+ },
+ {
+ "epoch": 0.5944386149003148,
+ "grad_norm": 0.1678000806117814,
+ "learning_rate": 2.0922030705701518e-05,
+ "loss": 0.44939014315605164,
+ "step": 1133
+ },
+ {
+ "epoch": 0.5949632738719832,
+ "grad_norm": 0.16104474282288994,
+ "learning_rate": 2.088657398123246e-05,
+ "loss": 0.4313976764678955,
+ "step": 1134
+ },
+ {
+ "epoch": 0.5954879328436516,
+ "grad_norm": 0.1631128968199669,
+ "learning_rate": 2.085113522396141e-05,
+ "loss": 0.4508278965950012,
+ "step": 1135
+ },
+ {
+ "epoch": 0.5960125918153201,
+ "grad_norm": 0.16307688195844533,
+ "learning_rate": 2.0815714530167557e-05,
+ "loss": 0.4467576742172241,
+ "step": 1136
+ },
+ {
+ "epoch": 0.5965372507869885,
+ "grad_norm": 0.16456832304532207,
+ "learning_rate": 2.0780311996080976e-05,
+ "loss": 0.4621703624725342,
+ "step": 1137
+ },
+ {
+ "epoch": 0.5970619097586569,
+ "grad_norm": 0.16346217748325367,
+ "learning_rate": 2.0744927717882435e-05,
+ "loss": 0.4589449465274811,
+ "step": 1138
+ },
+ {
+ "epoch": 0.5975865687303253,
+ "grad_norm": 0.16180691501076994,
+ "learning_rate": 2.0709561791703097e-05,
+ "loss": 0.43698400259017944,
+ "step": 1139
+ },
+ {
+ "epoch": 0.5981112277019937,
+ "grad_norm": 0.15695459976898835,
+ "learning_rate": 2.067421431362427e-05,
+ "loss": 0.452271044254303,
+ "step": 1140
+ },
+ {
+ "epoch": 0.5986358866736621,
+ "grad_norm": 0.15930893553721054,
+ "learning_rate": 2.063888537967713e-05,
+ "loss": 0.44868963956832886,
+ "step": 1141
+ },
+ {
+ "epoch": 0.5991605456453305,
+ "grad_norm": 0.15963802284017928,
+ "learning_rate": 2.0603575085842485e-05,
+ "loss": 0.4376864731311798,
+ "step": 1142
+ },
+ {
+ "epoch": 0.5996852046169989,
+ "grad_norm": 0.16610266900507178,
+ "learning_rate": 2.05682835280505e-05,
+ "loss": 0.44676727056503296,
+ "step": 1143
+ },
+ {
+ "epoch": 0.6002098635886673,
+ "grad_norm": 0.15978253236669612,
+ "learning_rate": 2.053301080218044e-05,
+ "loss": 0.434773713350296,
+ "step": 1144
+ },
+ {
+ "epoch": 0.6007345225603358,
+ "grad_norm": 0.16305141803526324,
+ "learning_rate": 2.0497757004060403e-05,
+ "loss": 0.4529780149459839,
+ "step": 1145
+ },
+ {
+ "epoch": 0.6012591815320042,
+ "grad_norm": 0.1649737864101215,
+ "learning_rate": 2.046252222946707e-05,
+ "loss": 0.4571434259414673,
+ "step": 1146
+ },
+ {
+ "epoch": 0.6017838405036726,
+ "grad_norm": 0.15738824968326598,
+ "learning_rate": 2.0427306574125427e-05,
+ "loss": 0.4446289539337158,
+ "step": 1147
+ },
+ {
+ "epoch": 0.602308499475341,
+ "grad_norm": 0.1731139535185187,
+ "learning_rate": 2.039211013370853e-05,
+ "loss": 0.4551090598106384,
+ "step": 1148
+ },
+ {
+ "epoch": 0.6028331584470095,
+ "grad_norm": 0.15857251900043037,
+ "learning_rate": 2.0356933003837237e-05,
+ "loss": 0.4644010663032532,
+ "step": 1149
+ },
+ {
+ "epoch": 0.6033578174186779,
+ "grad_norm": 0.1589582861134211,
+ "learning_rate": 2.032177528007992e-05,
+ "loss": 0.45741650462150574,
+ "step": 1150
+ },
+ {
+ "epoch": 0.6038824763903463,
+ "grad_norm": 0.15639092706575122,
+ "learning_rate": 2.0286637057952257e-05,
+ "loss": 0.4375547766685486,
+ "step": 1151
+ },
+ {
+ "epoch": 0.6044071353620147,
+ "grad_norm": 0.16335605581337187,
+ "learning_rate": 2.0251518432916933e-05,
+ "loss": 0.4636140465736389,
+ "step": 1152
+ },
+ {
+ "epoch": 0.6049317943336832,
+ "grad_norm": 0.15437437113359356,
+ "learning_rate": 2.0216419500383372e-05,
+ "loss": 0.44632142782211304,
+ "step": 1153
+ },
+ {
+ "epoch": 0.6054564533053515,
+ "grad_norm": 0.15522039198698664,
+ "learning_rate": 2.0181340355707545e-05,
+ "loss": 0.449049174785614,
+ "step": 1154
+ },
+ {
+ "epoch": 0.6059811122770199,
+ "grad_norm": 0.1599951917909323,
+ "learning_rate": 2.0146281094191604e-05,
+ "loss": 0.44395962357521057,
+ "step": 1155
+ },
+ {
+ "epoch": 0.6065057712486883,
+ "grad_norm": 0.16866621620527716,
+ "learning_rate": 2.011124181108373e-05,
+ "loss": 0.45693716406822205,
+ "step": 1156
+ },
+ {
+ "epoch": 0.6070304302203567,
+ "grad_norm": 0.15508935650135075,
+ "learning_rate": 2.0076222601577815e-05,
+ "loss": 0.4525151252746582,
+ "step": 1157
+ },
+ {
+ "epoch": 0.6075550891920252,
+ "grad_norm": 0.16508998118005971,
+ "learning_rate": 2.004122356081321e-05,
+ "loss": 0.466006875038147,
+ "step": 1158
+ },
+ {
+ "epoch": 0.6080797481636936,
+ "grad_norm": 0.15817418864022237,
+ "learning_rate": 2.0006244783874466e-05,
+ "loss": 0.4496110677719116,
+ "step": 1159
+ },
+ {
+ "epoch": 0.608604407135362,
+ "grad_norm": 0.16127270259749824,
+ "learning_rate": 1.997128636579109e-05,
+ "loss": 0.45918628573417664,
+ "step": 1160
+ },
+ {
+ "epoch": 0.6091290661070304,
+ "grad_norm": 0.1702599924159677,
+ "learning_rate": 1.9936348401537292e-05,
+ "loss": 0.47218555212020874,
+ "step": 1161
+ },
+ {
+ "epoch": 0.6096537250786989,
+ "grad_norm": 0.15690176248846127,
+ "learning_rate": 1.9901430986031695e-05,
+ "loss": 0.4516097903251648,
+ "step": 1162
+ },
+ {
+ "epoch": 0.6101783840503673,
+ "grad_norm": 0.1713614405614144,
+ "learning_rate": 1.9866534214137085e-05,
+ "loss": 0.45030856132507324,
+ "step": 1163
+ },
+ {
+ "epoch": 0.6107030430220357,
+ "grad_norm": 0.1670865563026229,
+ "learning_rate": 1.9831658180660195e-05,
+ "loss": 0.45603489875793457,
+ "step": 1164
+ },
+ {
+ "epoch": 0.6112277019937041,
+ "grad_norm": 0.16897539113401952,
+ "learning_rate": 1.9796802980351392e-05,
+ "loss": 0.46314537525177,
+ "step": 1165
+ },
+ {
+ "epoch": 0.6117523609653726,
+ "grad_norm": 0.15832868870697644,
+ "learning_rate": 1.976196870790446e-05,
+ "loss": 0.44784611463546753,
+ "step": 1166
+ },
+ {
+ "epoch": 0.612277019937041,
+ "grad_norm": 0.16595365551891114,
+ "learning_rate": 1.9727155457956315e-05,
+ "loss": 0.4409624934196472,
+ "step": 1167
+ },
+ {
+ "epoch": 0.6128016789087093,
+ "grad_norm": 0.1642630411212017,
+ "learning_rate": 1.969236332508677e-05,
+ "loss": 0.46120351552963257,
+ "step": 1168
+ },
+ {
+ "epoch": 0.6133263378803777,
+ "grad_norm": 0.16481796135868812,
+ "learning_rate": 1.9657592403818253e-05,
+ "loss": 0.45544904470443726,
+ "step": 1169
+ },
+ {
+ "epoch": 0.6138509968520461,
+ "grad_norm": 0.1629832779852364,
+ "learning_rate": 1.962284278861557e-05,
+ "loss": 0.4578462839126587,
+ "step": 1170
+ },
+ {
+ "epoch": 0.6143756558237146,
+ "grad_norm": 0.15916219613603688,
+ "learning_rate": 1.9588114573885664e-05,
+ "loss": 0.4546983540058136,
+ "step": 1171
+ },
+ {
+ "epoch": 0.614900314795383,
+ "grad_norm": 0.1660520165166075,
+ "learning_rate": 1.95534078539773e-05,
+ "loss": 0.4601942300796509,
+ "step": 1172
+ },
+ {
+ "epoch": 0.6154249737670514,
+ "grad_norm": 0.16924794578577768,
+ "learning_rate": 1.9518722723180877e-05,
+ "loss": 0.45997846126556396,
+ "step": 1173
+ },
+ {
+ "epoch": 0.6159496327387198,
+ "grad_norm": 0.16597646133045382,
+ "learning_rate": 1.9484059275728127e-05,
+ "loss": 0.4594605565071106,
+ "step": 1174
+ },
+ {
+ "epoch": 0.6164742917103883,
+ "grad_norm": 0.16964243792519748,
+ "learning_rate": 1.9449417605791876e-05,
+ "loss": 0.4638690650463104,
+ "step": 1175
+ },
+ {
+ "epoch": 0.6169989506820567,
+ "grad_norm": 0.1758591426791696,
+ "learning_rate": 1.9414797807485796e-05,
+ "loss": 0.4666447639465332,
+ "step": 1176
+ },
+ {
+ "epoch": 0.6175236096537251,
+ "grad_norm": 0.16217661227891753,
+ "learning_rate": 1.9380199974864126e-05,
+ "loss": 0.44252535700798035,
+ "step": 1177
+ },
+ {
+ "epoch": 0.6180482686253935,
+ "grad_norm": 0.1576446184499019,
+ "learning_rate": 1.934562420192142e-05,
+ "loss": 0.4499641954898834,
+ "step": 1178
+ },
+ {
+ "epoch": 0.618572927597062,
+ "grad_norm": 0.16231008092636912,
+ "learning_rate": 1.9311070582592327e-05,
+ "loss": 0.4561542570590973,
+ "step": 1179
+ },
+ {
+ "epoch": 0.6190975865687304,
+ "grad_norm": 0.16790382249577043,
+ "learning_rate": 1.9276539210751292e-05,
+ "loss": 0.4484325647354126,
+ "step": 1180
+ },
+ {
+ "epoch": 0.6196222455403987,
+ "grad_norm": 0.18794729932605855,
+ "learning_rate": 1.9242030180212323e-05,
+ "loss": 0.44169026613235474,
+ "step": 1181
+ },
+ {
+ "epoch": 0.6201469045120671,
+ "grad_norm": 0.16948051243889814,
+ "learning_rate": 1.920754358472872e-05,
+ "loss": 0.4592876434326172,
+ "step": 1182
+ },
+ {
+ "epoch": 0.6206715634837355,
+ "grad_norm": 0.16762558710029374,
+ "learning_rate": 1.917307951799286e-05,
+ "loss": 0.4402095079421997,
+ "step": 1183
+ },
+ {
+ "epoch": 0.621196222455404,
+ "grad_norm": 0.18170280116387003,
+ "learning_rate": 1.9138638073635885e-05,
+ "loss": 0.4633026123046875,
+ "step": 1184
+ },
+ {
+ "epoch": 0.6217208814270724,
+ "grad_norm": 0.1718051709136108,
+ "learning_rate": 1.9104219345227486e-05,
+ "loss": 0.4510066509246826,
+ "step": 1185
+ },
+ {
+ "epoch": 0.6222455403987408,
+ "grad_norm": 0.16829066326345576,
+ "learning_rate": 1.906982342627565e-05,
+ "loss": 0.4632919430732727,
+ "step": 1186
+ },
+ {
+ "epoch": 0.6227701993704092,
+ "grad_norm": 0.16669141808548274,
+ "learning_rate": 1.9035450410226395e-05,
+ "loss": 0.455544650554657,
+ "step": 1187
+ },
+ {
+ "epoch": 0.6232948583420777,
+ "grad_norm": 0.16679996559722232,
+ "learning_rate": 1.90011003904635e-05,
+ "loss": 0.459994375705719,
+ "step": 1188
+ },
+ {
+ "epoch": 0.6238195173137461,
+ "grad_norm": 0.17379754713278164,
+ "learning_rate": 1.8966773460308284e-05,
+ "loss": 0.4576749801635742,
+ "step": 1189
+ },
+ {
+ "epoch": 0.6243441762854145,
+ "grad_norm": 0.17088539733389246,
+ "learning_rate": 1.8932469713019335e-05,
+ "loss": 0.44853052496910095,
+ "step": 1190
+ },
+ {
+ "epoch": 0.6248688352570829,
+ "grad_norm": 0.15556883889201015,
+ "learning_rate": 1.8898189241792248e-05,
+ "loss": 0.44060447812080383,
+ "step": 1191
+ },
+ {
+ "epoch": 0.6253934942287513,
+ "grad_norm": 0.1640093917439215,
+ "learning_rate": 1.8863932139759388e-05,
+ "loss": 0.4531911611557007,
+ "step": 1192
+ },
+ {
+ "epoch": 0.6259181532004198,
+ "grad_norm": 0.16444284826591576,
+ "learning_rate": 1.882969849998964e-05,
+ "loss": 0.44176363945007324,
+ "step": 1193
+ },
+ {
+ "epoch": 0.6264428121720882,
+ "grad_norm": 0.1696462520995642,
+ "learning_rate": 1.8795488415488143e-05,
+ "loss": 0.44516652822494507,
+ "step": 1194
+ },
+ {
+ "epoch": 0.6269674711437565,
+ "grad_norm": 0.16810724705191796,
+ "learning_rate": 1.8761301979196032e-05,
+ "loss": 0.4542563855648041,
+ "step": 1195
+ },
+ {
+ "epoch": 0.6274921301154249,
+ "grad_norm": 0.17334813313855824,
+ "learning_rate": 1.8727139283990204e-05,
+ "loss": 0.47769248485565186,
+ "step": 1196
+ },
+ {
+ "epoch": 0.6280167890870934,
+ "grad_norm": 0.16308212035303235,
+ "learning_rate": 1.8693000422683043e-05,
+ "loss": 0.4407799243927002,
+ "step": 1197
+ },
+ {
+ "epoch": 0.6285414480587618,
+ "grad_norm": 0.16853319139928977,
+ "learning_rate": 1.8658885488022204e-05,
+ "loss": 0.4565497040748596,
+ "step": 1198
+ },
+ {
+ "epoch": 0.6290661070304302,
+ "grad_norm": 0.16694968781284703,
+ "learning_rate": 1.8624794572690328e-05,
+ "loss": 0.44522374868392944,
+ "step": 1199
+ },
+ {
+ "epoch": 0.6295907660020986,
+ "grad_norm": 0.16418579468339103,
+ "learning_rate": 1.8590727769304788e-05,
+ "loss": 0.4441048502922058,
+ "step": 1200
+ },
+ {
+ "epoch": 0.630115424973767,
+ "grad_norm": 0.1661930777124156,
+ "learning_rate": 1.8556685170417486e-05,
+ "loss": 0.457327663898468,
+ "step": 1201
+ },
+ {
+ "epoch": 0.6306400839454355,
+ "grad_norm": 0.16536630027039814,
+ "learning_rate": 1.852266686851452e-05,
+ "loss": 0.462494432926178,
+ "step": 1202
+ },
+ {
+ "epoch": 0.6311647429171039,
+ "grad_norm": 0.16611474088307454,
+ "learning_rate": 1.8488672956016013e-05,
+ "loss": 0.4557701647281647,
+ "step": 1203
+ },
+ {
+ "epoch": 0.6316894018887723,
+ "grad_norm": 0.16759343079749697,
+ "learning_rate": 1.8454703525275813e-05,
+ "loss": 0.45069995522499084,
+ "step": 1204
+ },
+ {
+ "epoch": 0.6322140608604407,
+ "grad_norm": 0.16186676294485336,
+ "learning_rate": 1.8420758668581255e-05,
+ "loss": 0.44954800605773926,
+ "step": 1205
+ },
+ {
+ "epoch": 0.6327387198321092,
+ "grad_norm": 0.15687159835368553,
+ "learning_rate": 1.8386838478152916e-05,
+ "loss": 0.4490429759025574,
+ "step": 1206
+ },
+ {
+ "epoch": 0.6332633788037776,
+ "grad_norm": 0.16242951976115577,
+ "learning_rate": 1.8352943046144372e-05,
+ "loss": 0.45819586515426636,
+ "step": 1207
+ },
+ {
+ "epoch": 0.633788037775446,
+ "grad_norm": 0.16516087266517232,
+ "learning_rate": 1.831907246464191e-05,
+ "loss": 0.44769883155822754,
+ "step": 1208
+ },
+ {
+ "epoch": 0.6343126967471143,
+ "grad_norm": 0.16709475871512644,
+ "learning_rate": 1.8285226825664332e-05,
+ "loss": 0.4699128568172455,
+ "step": 1209
+ },
+ {
+ "epoch": 0.6348373557187827,
+ "grad_norm": 0.16288623150595774,
+ "learning_rate": 1.8251406221162636e-05,
+ "loss": 0.45322558283805847,
+ "step": 1210
+ },
+ {
+ "epoch": 0.6353620146904512,
+ "grad_norm": 0.1675722848054237,
+ "learning_rate": 1.821761074301986e-05,
+ "loss": 0.4752347469329834,
+ "step": 1211
+ },
+ {
+ "epoch": 0.6358866736621196,
+ "grad_norm": 0.163285963514504,
+ "learning_rate": 1.818384048305074e-05,
+ "loss": 0.441958487033844,
+ "step": 1212
+ },
+ {
+ "epoch": 0.636411332633788,
+ "grad_norm": 0.17078102287557764,
+ "learning_rate": 1.8150095533001517e-05,
+ "loss": 0.4551333487033844,
+ "step": 1213
+ },
+ {
+ "epoch": 0.6369359916054564,
+ "grad_norm": 0.16188540746097138,
+ "learning_rate": 1.811637598454966e-05,
+ "loss": 0.4470779597759247,
+ "step": 1214
+ },
+ {
+ "epoch": 0.6374606505771249,
+ "grad_norm": 0.15235628218958286,
+ "learning_rate": 1.808268192930364e-05,
+ "loss": 0.43604153394699097,
+ "step": 1215
+ },
+ {
+ "epoch": 0.6379853095487933,
+ "grad_norm": 0.1604601949959104,
+ "learning_rate": 1.8049013458802665e-05,
+ "loss": 0.4445713758468628,
+ "step": 1216
+ },
+ {
+ "epoch": 0.6385099685204617,
+ "grad_norm": 0.15511797131215901,
+ "learning_rate": 1.8015370664516427e-05,
+ "loss": 0.4455301761627197,
+ "step": 1217
+ },
+ {
+ "epoch": 0.6390346274921301,
+ "grad_norm": 0.16815669546944095,
+ "learning_rate": 1.798175363784487e-05,
+ "loss": 0.46135014295578003,
+ "step": 1218
+ },
+ {
+ "epoch": 0.6395592864637986,
+ "grad_norm": 0.161558177688922,
+ "learning_rate": 1.7948162470117917e-05,
+ "loss": 0.4519304037094116,
+ "step": 1219
+ },
+ {
+ "epoch": 0.640083945435467,
+ "grad_norm": 0.16677842130048232,
+ "learning_rate": 1.7914597252595265e-05,
+ "loss": 0.4550550878047943,
+ "step": 1220
+ },
+ {
+ "epoch": 0.6406086044071354,
+ "grad_norm": 0.1597331480423138,
+ "learning_rate": 1.7881058076466095e-05,
+ "loss": 0.441120445728302,
+ "step": 1221
+ },
+ {
+ "epoch": 0.6411332633788038,
+ "grad_norm": 0.15951106545246824,
+ "learning_rate": 1.7847545032848844e-05,
+ "loss": 0.4471382200717926,
+ "step": 1222
+ },
+ {
+ "epoch": 0.6416579223504721,
+ "grad_norm": 0.16157787674667565,
+ "learning_rate": 1.7814058212790947e-05,
+ "loss": 0.4550648331642151,
+ "step": 1223
+ },
+ {
+ "epoch": 0.6421825813221406,
+ "grad_norm": 0.16989186763352415,
+ "learning_rate": 1.7780597707268583e-05,
+ "loss": 0.4515089988708496,
+ "step": 1224
+ },
+ {
+ "epoch": 0.642707240293809,
+ "grad_norm": 0.15952153223099508,
+ "learning_rate": 1.7747163607186468e-05,
+ "loss": 0.4658638834953308,
+ "step": 1225
+ },
+ {
+ "epoch": 0.6432318992654774,
+ "grad_norm": 0.15205926699647218,
+ "learning_rate": 1.7713756003377567e-05,
+ "loss": 0.4330868422985077,
+ "step": 1226
+ },
+ {
+ "epoch": 0.6437565582371458,
+ "grad_norm": 0.15784806480936794,
+ "learning_rate": 1.7680374986602865e-05,
+ "loss": 0.4506068527698517,
+ "step": 1227
+ },
+ {
+ "epoch": 0.6442812172088143,
+ "grad_norm": 0.16006709280686687,
+ "learning_rate": 1.7647020647551105e-05,
+ "loss": 0.446941077709198,
+ "step": 1228
+ },
+ {
+ "epoch": 0.6448058761804827,
+ "grad_norm": 0.16064631158222062,
+ "learning_rate": 1.761369307683856e-05,
+ "loss": 0.4418525695800781,
+ "step": 1229
+ },
+ {
+ "epoch": 0.6453305351521511,
+ "grad_norm": 0.16454136480444392,
+ "learning_rate": 1.7580392365008767e-05,
+ "loss": 0.46268391609191895,
+ "step": 1230
+ },
+ {
+ "epoch": 0.6458551941238195,
+ "grad_norm": 0.15401246535771462,
+ "learning_rate": 1.754711860253233e-05,
+ "loss": 0.4372989535331726,
+ "step": 1231
+ },
+ {
+ "epoch": 0.646379853095488,
+ "grad_norm": 0.1544567975028383,
+ "learning_rate": 1.7513871879806603e-05,
+ "loss": 0.44806575775146484,
+ "step": 1232
+ },
+ {
+ "epoch": 0.6469045120671564,
+ "grad_norm": 0.16411225112730965,
+ "learning_rate": 1.748065228715549e-05,
+ "loss": 0.43782058358192444,
+ "step": 1233
+ },
+ {
+ "epoch": 0.6474291710388248,
+ "grad_norm": 0.1552767832972093,
+ "learning_rate": 1.7447459914829183e-05,
+ "loss": 0.44531378149986267,
+ "step": 1234
+ },
+ {
+ "epoch": 0.6479538300104932,
+ "grad_norm": 0.16214695257123513,
+ "learning_rate": 1.741429485300392e-05,
+ "loss": 0.47399842739105225,
+ "step": 1235
+ },
+ {
+ "epoch": 0.6484784889821616,
+ "grad_norm": 0.16684861111484112,
+ "learning_rate": 1.7381157191781766e-05,
+ "loss": 0.4654282331466675,
+ "step": 1236
+ },
+ {
+ "epoch": 0.64900314795383,
+ "grad_norm": 0.15901864178630273,
+ "learning_rate": 1.7348047021190334e-05,
+ "loss": 0.45128414034843445,
+ "step": 1237
+ },
+ {
+ "epoch": 0.6495278069254984,
+ "grad_norm": 0.15360991961742707,
+ "learning_rate": 1.7314964431182523e-05,
+ "loss": 0.4303375482559204,
+ "step": 1238
+ },
+ {
+ "epoch": 0.6500524658971668,
+ "grad_norm": 0.1633151574385741,
+ "learning_rate": 1.728190951163633e-05,
+ "loss": 0.45568403601646423,
+ "step": 1239
+ },
+ {
+ "epoch": 0.6505771248688352,
+ "grad_norm": 0.16377853571058645,
+ "learning_rate": 1.724888235235457e-05,
+ "loss": 0.46503543853759766,
+ "step": 1240
+ },
+ {
+ "epoch": 0.6511017838405037,
+ "grad_norm": 0.1595023801275368,
+ "learning_rate": 1.7215883043064655e-05,
+ "loss": 0.46337294578552246,
+ "step": 1241
+ },
+ {
+ "epoch": 0.6516264428121721,
+ "grad_norm": 0.16324796182212292,
+ "learning_rate": 1.7182911673418313e-05,
+ "loss": 0.46197354793548584,
+ "step": 1242
+ },
+ {
+ "epoch": 0.6521511017838405,
+ "grad_norm": 0.16353734830879765,
+ "learning_rate": 1.714996833299138e-05,
+ "loss": 0.45741671323776245,
+ "step": 1243
+ },
+ {
+ "epoch": 0.6526757607555089,
+ "grad_norm": 0.15941587678263386,
+ "learning_rate": 1.7117053111283528e-05,
+ "loss": 0.4514503479003906,
+ "step": 1244
+ },
+ {
+ "epoch": 0.6532004197271774,
+ "grad_norm": 0.1642653230145008,
+ "learning_rate": 1.708416609771804e-05,
+ "loss": 0.4437655210494995,
+ "step": 1245
+ },
+ {
+ "epoch": 0.6537250786988458,
+ "grad_norm": 0.15814155519130152,
+ "learning_rate": 1.7051307381641587e-05,
+ "loss": 0.4314163327217102,
+ "step": 1246
+ },
+ {
+ "epoch": 0.6542497376705142,
+ "grad_norm": 0.16086078370768056,
+ "learning_rate": 1.701847705232394e-05,
+ "loss": 0.4526430666446686,
+ "step": 1247
+ },
+ {
+ "epoch": 0.6547743966421826,
+ "grad_norm": 0.16389360190715183,
+ "learning_rate": 1.698567519895775e-05,
+ "loss": 0.4278966188430786,
+ "step": 1248
+ },
+ {
+ "epoch": 0.655299055613851,
+ "grad_norm": 0.16303142287804262,
+ "learning_rate": 1.6952901910658316e-05,
+ "loss": 0.45778220891952515,
+ "step": 1249
+ },
+ {
+ "epoch": 0.6558237145855194,
+ "grad_norm": 0.1675209956904048,
+ "learning_rate": 1.6920157276463318e-05,
+ "loss": 0.4489403963088989,
+ "step": 1250
+ },
+ {
+ "epoch": 0.6563483735571878,
+ "grad_norm": 0.16228311551074323,
+ "learning_rate": 1.6887441385332604e-05,
+ "loss": 0.4534342885017395,
+ "step": 1251
+ },
+ {
+ "epoch": 0.6568730325288562,
+ "grad_norm": 0.16408957851046913,
+ "learning_rate": 1.6854754326147916e-05,
+ "loss": 0.4549208879470825,
+ "step": 1252
+ },
+ {
+ "epoch": 0.6573976915005246,
+ "grad_norm": 0.15967922111432845,
+ "learning_rate": 1.682209618771268e-05,
+ "loss": 0.4451141357421875,
+ "step": 1253
+ },
+ {
+ "epoch": 0.657922350472193,
+ "grad_norm": 0.16440698722204208,
+ "learning_rate": 1.678946705875176e-05,
+ "loss": 0.454200804233551,
+ "step": 1254
+ },
+ {
+ "epoch": 0.6584470094438615,
+ "grad_norm": 0.16251120184938497,
+ "learning_rate": 1.675686702791117e-05,
+ "loss": 0.4580027461051941,
+ "step": 1255
+ },
+ {
+ "epoch": 0.6589716684155299,
+ "grad_norm": 0.1643717386150457,
+ "learning_rate": 1.6724296183757924e-05,
+ "loss": 0.4422311782836914,
+ "step": 1256
+ },
+ {
+ "epoch": 0.6594963273871983,
+ "grad_norm": 0.1644586414378722,
+ "learning_rate": 1.66917546147797e-05,
+ "loss": 0.4593179225921631,
+ "step": 1257
+ },
+ {
+ "epoch": 0.6600209863588667,
+ "grad_norm": 0.16751074186967685,
+ "learning_rate": 1.6659242409384657e-05,
+ "loss": 0.457930326461792,
+ "step": 1258
+ },
+ {
+ "epoch": 0.6605456453305352,
+ "grad_norm": 0.15398418609155334,
+ "learning_rate": 1.6626759655901182e-05,
+ "loss": 0.44002899527549744,
+ "step": 1259
+ },
+ {
+ "epoch": 0.6610703043022036,
+ "grad_norm": 0.16328113175790226,
+ "learning_rate": 1.659430644257763e-05,
+ "loss": 0.4500438868999481,
+ "step": 1260
+ },
+ {
+ "epoch": 0.661594963273872,
+ "grad_norm": 0.1694244340615106,
+ "learning_rate": 1.656188285758214e-05,
+ "loss": 0.4818427264690399,
+ "step": 1261
+ },
+ {
+ "epoch": 0.6621196222455404,
+ "grad_norm": 0.1578320701754368,
+ "learning_rate": 1.652948898900232e-05,
+ "loss": 0.4492436647415161,
+ "step": 1262
+ },
+ {
+ "epoch": 0.6626442812172089,
+ "grad_norm": 0.16538865917828469,
+ "learning_rate": 1.6497124924845064e-05,
+ "loss": 0.45070338249206543,
+ "step": 1263
+ },
+ {
+ "epoch": 0.6631689401888772,
+ "grad_norm": 0.15670817848256852,
+ "learning_rate": 1.6464790753036296e-05,
+ "loss": 0.45439285039901733,
+ "step": 1264
+ },
+ {
+ "epoch": 0.6636935991605456,
+ "grad_norm": 0.16184786585965885,
+ "learning_rate": 1.64324865614207e-05,
+ "loss": 0.446522980928421,
+ "step": 1265
+ },
+ {
+ "epoch": 0.664218258132214,
+ "grad_norm": 0.15719726462251984,
+ "learning_rate": 1.6400212437761554e-05,
+ "loss": 0.4495280981063843,
+ "step": 1266
+ },
+ {
+ "epoch": 0.6647429171038824,
+ "grad_norm": 0.1680804414134405,
+ "learning_rate": 1.6367968469740416e-05,
+ "loss": 0.4652596712112427,
+ "step": 1267
+ },
+ {
+ "epoch": 0.6652675760755509,
+ "grad_norm": 0.15919541846683566,
+ "learning_rate": 1.6335754744956937e-05,
+ "loss": 0.43842798471450806,
+ "step": 1268
+ },
+ {
+ "epoch": 0.6657922350472193,
+ "grad_norm": 0.16322634187139418,
+ "learning_rate": 1.630357135092859e-05,
+ "loss": 0.45242664217948914,
+ "step": 1269
+ },
+ {
+ "epoch": 0.6663168940188877,
+ "grad_norm": 0.15644867734734783,
+ "learning_rate": 1.6271418375090445e-05,
+ "loss": 0.44801655411720276,
+ "step": 1270
+ },
+ {
+ "epoch": 0.6668415529905561,
+ "grad_norm": 0.1514717416124423,
+ "learning_rate": 1.623929590479496e-05,
+ "loss": 0.4506194293498993,
+ "step": 1271
+ },
+ {
+ "epoch": 0.6673662119622246,
+ "grad_norm": 0.1740345085495889,
+ "learning_rate": 1.6207204027311683e-05,
+ "loss": 0.4818177819252014,
+ "step": 1272
+ },
+ {
+ "epoch": 0.667890870933893,
+ "grad_norm": 0.15797701900765457,
+ "learning_rate": 1.6175142829827067e-05,
+ "loss": 0.4639382064342499,
+ "step": 1273
+ },
+ {
+ "epoch": 0.6684155299055614,
+ "grad_norm": 0.15663812316653875,
+ "learning_rate": 1.6143112399444198e-05,
+ "loss": 0.46217501163482666,
+ "step": 1274
+ },
+ {
+ "epoch": 0.6689401888772298,
+ "grad_norm": 0.15474018977043758,
+ "learning_rate": 1.6111112823182592e-05,
+ "loss": 0.43691688776016235,
+ "step": 1275
+ },
+ {
+ "epoch": 0.6694648478488983,
+ "grad_norm": 0.1615634942529431,
+ "learning_rate": 1.6079144187977923e-05,
+ "loss": 0.4412601590156555,
+ "step": 1276
+ },
+ {
+ "epoch": 0.6699895068205667,
+ "grad_norm": 0.15440022842754444,
+ "learning_rate": 1.604720658068184e-05,
+ "loss": 0.4523273706436157,
+ "step": 1277
+ },
+ {
+ "epoch": 0.670514165792235,
+ "grad_norm": 0.1634985665776242,
+ "learning_rate": 1.6015300088061664e-05,
+ "loss": 0.46264398097991943,
+ "step": 1278
+ },
+ {
+ "epoch": 0.6710388247639034,
+ "grad_norm": 0.15037886370689527,
+ "learning_rate": 1.598342479680018e-05,
+ "loss": 0.451465904712677,
+ "step": 1279
+ },
+ {
+ "epoch": 0.6715634837355718,
+ "grad_norm": 0.16895242437746613,
+ "learning_rate": 1.595158079349542e-05,
+ "loss": 0.4466145634651184,
+ "step": 1280
+ },
+ {
+ "epoch": 0.6720881427072403,
+ "grad_norm": 0.1629548326999642,
+ "learning_rate": 1.5919768164660425e-05,
+ "loss": 0.45415782928466797,
+ "step": 1281
+ },
+ {
+ "epoch": 0.6726128016789087,
+ "grad_norm": 0.1563872102268991,
+ "learning_rate": 1.588798699672298e-05,
+ "loss": 0.4386737048625946,
+ "step": 1282
+ },
+ {
+ "epoch": 0.6731374606505771,
+ "grad_norm": 0.16038172773344717,
+ "learning_rate": 1.5856237376025403e-05,
+ "loss": 0.446260929107666,
+ "step": 1283
+ },
+ {
+ "epoch": 0.6736621196222455,
+ "grad_norm": 0.16189780714407623,
+ "learning_rate": 1.582451938882431e-05,
+ "loss": 0.45458072423934937,
+ "step": 1284
+ },
+ {
+ "epoch": 0.674186778593914,
+ "grad_norm": 0.15204521785049285,
+ "learning_rate": 1.5792833121290367e-05,
+ "loss": 0.4425617754459381,
+ "step": 1285
+ },
+ {
+ "epoch": 0.6747114375655824,
+ "grad_norm": 0.16237376704792778,
+ "learning_rate": 1.5761178659508063e-05,
+ "loss": 0.44631800055503845,
+ "step": 1286
+ },
+ {
+ "epoch": 0.6752360965372508,
+ "grad_norm": 0.1582063618466069,
+ "learning_rate": 1.57295560894755e-05,
+ "loss": 0.44280534982681274,
+ "step": 1287
+ },
+ {
+ "epoch": 0.6757607555089192,
+ "grad_norm": 0.16587032349847394,
+ "learning_rate": 1.5697965497104116e-05,
+ "loss": 0.4495997130870819,
+ "step": 1288
+ },
+ {
+ "epoch": 0.6762854144805877,
+ "grad_norm": 0.1557269262755563,
+ "learning_rate": 1.566640696821848e-05,
+ "loss": 0.4493359625339508,
+ "step": 1289
+ },
+ {
+ "epoch": 0.6768100734522561,
+ "grad_norm": 0.16038490357012256,
+ "learning_rate": 1.5634880588556043e-05,
+ "loss": 0.4447544813156128,
+ "step": 1290
+ },
+ {
+ "epoch": 0.6773347324239245,
+ "grad_norm": 0.15355201908546406,
+ "learning_rate": 1.560338644376692e-05,
+ "loss": 0.4468432664871216,
+ "step": 1291
+ },
+ {
+ "epoch": 0.6778593913955928,
+ "grad_norm": 0.15077009515589776,
+ "learning_rate": 1.5571924619413668e-05,
+ "loss": 0.44959044456481934,
+ "step": 1292
+ },
+ {
+ "epoch": 0.6783840503672612,
+ "grad_norm": 0.1583038146613965,
+ "learning_rate": 1.5540495200971012e-05,
+ "loss": 0.42450425028800964,
+ "step": 1293
+ },
+ {
+ "epoch": 0.6789087093389297,
+ "grad_norm": 0.15950687190598087,
+ "learning_rate": 1.5509098273825635e-05,
+ "loss": 0.4527238607406616,
+ "step": 1294
+ },
+ {
+ "epoch": 0.6794333683105981,
+ "grad_norm": 0.1526262208847808,
+ "learning_rate": 1.5477733923275968e-05,
+ "loss": 0.4379724860191345,
+ "step": 1295
+ },
+ {
+ "epoch": 0.6799580272822665,
+ "grad_norm": 0.15541579650021778,
+ "learning_rate": 1.5446402234531934e-05,
+ "loss": 0.4462166130542755,
+ "step": 1296
+ },
+ {
+ "epoch": 0.6804826862539349,
+ "grad_norm": 0.15483930734313786,
+ "learning_rate": 1.5415103292714718e-05,
+ "loss": 0.4390607178211212,
+ "step": 1297
+ },
+ {
+ "epoch": 0.6810073452256034,
+ "grad_norm": 0.1578005731149154,
+ "learning_rate": 1.5383837182856532e-05,
+ "loss": 0.44462254643440247,
+ "step": 1298
+ },
+ {
+ "epoch": 0.6815320041972718,
+ "grad_norm": 0.15830183884958823,
+ "learning_rate": 1.5352603989900406e-05,
+ "loss": 0.4563537538051605,
+ "step": 1299
+ },
+ {
+ "epoch": 0.6820566631689402,
+ "grad_norm": 0.15605472163445522,
+ "learning_rate": 1.5321403798699923e-05,
+ "loss": 0.43852412700653076,
+ "step": 1300
+ },
+ {
+ "epoch": 0.6825813221406086,
+ "grad_norm": 0.15770470227141278,
+ "learning_rate": 1.5290236694019017e-05,
+ "loss": 0.4438081979751587,
+ "step": 1301
+ },
+ {
+ "epoch": 0.683105981112277,
+ "grad_norm": 0.16175339845012077,
+ "learning_rate": 1.5259102760531747e-05,
+ "loss": 0.46670734882354736,
+ "step": 1302
+ },
+ {
+ "epoch": 0.6836306400839455,
+ "grad_norm": 0.16359504716154477,
+ "learning_rate": 1.5228002082822036e-05,
+ "loss": 0.4590187668800354,
+ "step": 1303
+ },
+ {
+ "epoch": 0.6841552990556139,
+ "grad_norm": 0.15422879500254713,
+ "learning_rate": 1.5196934745383457e-05,
+ "loss": 0.45287445187568665,
+ "step": 1304
+ },
+ {
+ "epoch": 0.6846799580272823,
+ "grad_norm": 0.15929902143574254,
+ "learning_rate": 1.5165900832619015e-05,
+ "loss": 0.44286859035491943,
+ "step": 1305
+ },
+ {
+ "epoch": 0.6852046169989506,
+ "grad_norm": 0.14838868714016124,
+ "learning_rate": 1.5134900428840904e-05,
+ "loss": 0.45155346393585205,
+ "step": 1306
+ },
+ {
+ "epoch": 0.6857292759706191,
+ "grad_norm": 0.15494288640397672,
+ "learning_rate": 1.5103933618270283e-05,
+ "loss": 0.44010141491889954,
+ "step": 1307
+ },
+ {
+ "epoch": 0.6862539349422875,
+ "grad_norm": 0.1686262360598338,
+ "learning_rate": 1.5073000485037042e-05,
+ "loss": 0.446555495262146,
+ "step": 1308
+ },
+ {
+ "epoch": 0.6867785939139559,
+ "grad_norm": 0.16122110823982602,
+ "learning_rate": 1.5042101113179579e-05,
+ "loss": 0.44640374183654785,
+ "step": 1309
+ },
+ {
+ "epoch": 0.6873032528856243,
+ "grad_norm": 0.16849937797919867,
+ "learning_rate": 1.5011235586644572e-05,
+ "loss": 0.4465785026550293,
+ "step": 1310
+ },
+ {
+ "epoch": 0.6878279118572928,
+ "grad_norm": 0.16062971262800754,
+ "learning_rate": 1.498040398928674e-05,
+ "loss": 0.445321649312973,
+ "step": 1311
+ },
+ {
+ "epoch": 0.6883525708289612,
+ "grad_norm": 0.1608953185993034,
+ "learning_rate": 1.4949606404868638e-05,
+ "loss": 0.46518227458000183,
+ "step": 1312
+ },
+ {
+ "epoch": 0.6888772298006296,
+ "grad_norm": 0.15880255672830915,
+ "learning_rate": 1.491884291706041e-05,
+ "loss": 0.44582217931747437,
+ "step": 1313
+ },
+ {
+ "epoch": 0.689401888772298,
+ "grad_norm": 0.16715145200789822,
+ "learning_rate": 1.4888113609439563e-05,
+ "loss": 0.4435983896255493,
+ "step": 1314
+ },
+ {
+ "epoch": 0.6899265477439664,
+ "grad_norm": 0.15263740340694454,
+ "learning_rate": 1.485741856549075e-05,
+ "loss": 0.4405066967010498,
+ "step": 1315
+ },
+ {
+ "epoch": 0.6904512067156349,
+ "grad_norm": 0.16247034518813966,
+ "learning_rate": 1.4826757868605517e-05,
+ "loss": 0.4544215500354767,
+ "step": 1316
+ },
+ {
+ "epoch": 0.6909758656873033,
+ "grad_norm": 0.15649146705978162,
+ "learning_rate": 1.4796131602082142e-05,
+ "loss": 0.4454413056373596,
+ "step": 1317
+ },
+ {
+ "epoch": 0.6915005246589717,
+ "grad_norm": 0.16286718636058178,
+ "learning_rate": 1.4765539849125318e-05,
+ "loss": 0.4497183561325073,
+ "step": 1318
+ },
+ {
+ "epoch": 0.6920251836306401,
+ "grad_norm": 0.15812342085005582,
+ "learning_rate": 1.473498269284599e-05,
+ "loss": 0.445032000541687,
+ "step": 1319
+ },
+ {
+ "epoch": 0.6925498426023085,
+ "grad_norm": 0.1558101290511678,
+ "learning_rate": 1.4704460216261123e-05,
+ "loss": 0.45769619941711426,
+ "step": 1320
+ },
+ {
+ "epoch": 0.6930745015739769,
+ "grad_norm": 0.15642889059495207,
+ "learning_rate": 1.4673972502293428e-05,
+ "loss": 0.44788360595703125,
+ "step": 1321
+ },
+ {
+ "epoch": 0.6935991605456453,
+ "grad_norm": 0.15722513518113074,
+ "learning_rate": 1.4643519633771222e-05,
+ "loss": 0.449631929397583,
+ "step": 1322
+ },
+ {
+ "epoch": 0.6941238195173137,
+ "grad_norm": 0.15587633748081542,
+ "learning_rate": 1.4613101693428125e-05,
+ "loss": 0.4450187087059021,
+ "step": 1323
+ },
+ {
+ "epoch": 0.6946484784889821,
+ "grad_norm": 0.1577107074882479,
+ "learning_rate": 1.458271876390287e-05,
+ "loss": 0.45108890533447266,
+ "step": 1324
+ },
+ {
+ "epoch": 0.6951731374606506,
+ "grad_norm": 0.15649839486990696,
+ "learning_rate": 1.4552370927739073e-05,
+ "loss": 0.46430301666259766,
+ "step": 1325
+ },
+ {
+ "epoch": 0.695697796432319,
+ "grad_norm": 0.18573600420690636,
+ "learning_rate": 1.4522058267385008e-05,
+ "loss": 0.4701972007751465,
+ "step": 1326
+ },
+ {
+ "epoch": 0.6962224554039874,
+ "grad_norm": 0.16296497084037542,
+ "learning_rate": 1.4491780865193402e-05,
+ "loss": 0.4607710838317871,
+ "step": 1327
+ },
+ {
+ "epoch": 0.6967471143756558,
+ "grad_norm": 0.161124007217343,
+ "learning_rate": 1.446153880342117e-05,
+ "loss": 0.45095115900039673,
+ "step": 1328
+ },
+ {
+ "epoch": 0.6972717733473243,
+ "grad_norm": 0.1537884039166766,
+ "learning_rate": 1.4431332164229222e-05,
+ "loss": 0.424068808555603,
+ "step": 1329
+ },
+ {
+ "epoch": 0.6977964323189927,
+ "grad_norm": 0.16588834503456087,
+ "learning_rate": 1.4401161029682237e-05,
+ "loss": 0.4622839093208313,
+ "step": 1330
+ },
+ {
+ "epoch": 0.6983210912906611,
+ "grad_norm": 0.16100510604015544,
+ "learning_rate": 1.4371025481748435e-05,
+ "loss": 0.4660705626010895,
+ "step": 1331
+ },
+ {
+ "epoch": 0.6988457502623295,
+ "grad_norm": 0.15559888940590075,
+ "learning_rate": 1.4340925602299337e-05,
+ "loss": 0.4484976530075073,
+ "step": 1332
+ },
+ {
+ "epoch": 0.6993704092339978,
+ "grad_norm": 0.15503821966213488,
+ "learning_rate": 1.4310861473109616e-05,
+ "loss": 0.45636600255966187,
+ "step": 1333
+ },
+ {
+ "epoch": 0.6998950682056663,
+ "grad_norm": 0.15539074896109334,
+ "learning_rate": 1.4280833175856743e-05,
+ "loss": 0.45863068103790283,
+ "step": 1334
+ },
+ {
+ "epoch": 0.7004197271773347,
+ "grad_norm": 0.15664821009154872,
+ "learning_rate": 1.4250840792120896e-05,
+ "loss": 0.44043463468551636,
+ "step": 1335
+ },
+ {
+ "epoch": 0.7009443861490031,
+ "grad_norm": 0.15825478808520285,
+ "learning_rate": 1.4220884403384655e-05,
+ "loss": 0.46522051095962524,
+ "step": 1336
+ },
+ {
+ "epoch": 0.7014690451206715,
+ "grad_norm": 0.16004268568141544,
+ "learning_rate": 1.4190964091032843e-05,
+ "loss": 0.46114686131477356,
+ "step": 1337
+ },
+ {
+ "epoch": 0.70199370409234,
+ "grad_norm": 0.16413139949945635,
+ "learning_rate": 1.4161079936352234e-05,
+ "loss": 0.4457515478134155,
+ "step": 1338
+ },
+ {
+ "epoch": 0.7025183630640084,
+ "grad_norm": 0.1538890577972702,
+ "learning_rate": 1.4131232020531398e-05,
+ "loss": 0.4616093337535858,
+ "step": 1339
+ },
+ {
+ "epoch": 0.7030430220356768,
+ "grad_norm": 0.16240737854323284,
+ "learning_rate": 1.4101420424660434e-05,
+ "loss": 0.44631537795066833,
+ "step": 1340
+ },
+ {
+ "epoch": 0.7035676810073452,
+ "grad_norm": 0.15984708603406073,
+ "learning_rate": 1.4071645229730762e-05,
+ "loss": 0.4366717040538788,
+ "step": 1341
+ },
+ {
+ "epoch": 0.7040923399790137,
+ "grad_norm": 0.15149695695635904,
+ "learning_rate": 1.4041906516634946e-05,
+ "loss": 0.44129428267478943,
+ "step": 1342
+ },
+ {
+ "epoch": 0.7046169989506821,
+ "grad_norm": 0.15041167501768415,
+ "learning_rate": 1.4012204366166399e-05,
+ "loss": 0.43654775619506836,
+ "step": 1343
+ },
+ {
+ "epoch": 0.7051416579223505,
+ "grad_norm": 0.1596097627963756,
+ "learning_rate": 1.3982538859019217e-05,
+ "loss": 0.4652445614337921,
+ "step": 1344
+ },
+ {
+ "epoch": 0.7056663168940189,
+ "grad_norm": 0.15621704205027884,
+ "learning_rate": 1.3952910075787941e-05,
+ "loss": 0.44577041268348694,
+ "step": 1345
+ },
+ {
+ "epoch": 0.7061909758656874,
+ "grad_norm": 0.16073041098008783,
+ "learning_rate": 1.3923318096967341e-05,
+ "loss": 0.4391976594924927,
+ "step": 1346
+ },
+ {
+ "epoch": 0.7067156348373557,
+ "grad_norm": 0.14793853946372448,
+ "learning_rate": 1.3893763002952204e-05,
+ "loss": 0.4463799297809601,
+ "step": 1347
+ },
+ {
+ "epoch": 0.7072402938090241,
+ "grad_norm": 0.16176080358867354,
+ "learning_rate": 1.3864244874037097e-05,
+ "loss": 0.45183271169662476,
+ "step": 1348
+ },
+ {
+ "epoch": 0.7077649527806925,
+ "grad_norm": 0.16372771179026924,
+ "learning_rate": 1.3834763790416171e-05,
+ "loss": 0.44205576181411743,
+ "step": 1349
+ },
+ {
+ "epoch": 0.7082896117523609,
+ "grad_norm": 0.15135222516127125,
+ "learning_rate": 1.3805319832182937e-05,
+ "loss": 0.4390261173248291,
+ "step": 1350
+ },
+ {
+ "epoch": 0.7088142707240294,
+ "grad_norm": 0.15493183225792365,
+ "learning_rate": 1.377591307933002e-05,
+ "loss": 0.45837974548339844,
+ "step": 1351
+ },
+ {
+ "epoch": 0.7093389296956978,
+ "grad_norm": 0.16405906906516504,
+ "learning_rate": 1.3746543611749008e-05,
+ "loss": 0.45032668113708496,
+ "step": 1352
+ },
+ {
+ "epoch": 0.7098635886673662,
+ "grad_norm": 0.15725677976524505,
+ "learning_rate": 1.3717211509230155e-05,
+ "loss": 0.4486224949359894,
+ "step": 1353
+ },
+ {
+ "epoch": 0.7103882476390346,
+ "grad_norm": 0.15781469066983164,
+ "learning_rate": 1.3687916851462227e-05,
+ "loss": 0.4502638578414917,
+ "step": 1354
+ },
+ {
+ "epoch": 0.710912906610703,
+ "grad_norm": 0.16115007729054298,
+ "learning_rate": 1.3658659718032241e-05,
+ "loss": 0.4508914351463318,
+ "step": 1355
+ },
+ {
+ "epoch": 0.7114375655823715,
+ "grad_norm": 0.16755060475968744,
+ "learning_rate": 1.3629440188425286e-05,
+ "loss": 0.45653945207595825,
+ "step": 1356
+ },
+ {
+ "epoch": 0.7119622245540399,
+ "grad_norm": 0.1574994294756123,
+ "learning_rate": 1.3600258342024268e-05,
+ "loss": 0.4349849820137024,
+ "step": 1357
+ },
+ {
+ "epoch": 0.7124868835257083,
+ "grad_norm": 0.16104440922657715,
+ "learning_rate": 1.3571114258109758e-05,
+ "loss": 0.4501233696937561,
+ "step": 1358
+ },
+ {
+ "epoch": 0.7130115424973767,
+ "grad_norm": 0.15163184290939022,
+ "learning_rate": 1.3542008015859691e-05,
+ "loss": 0.44171035289764404,
+ "step": 1359
+ },
+ {
+ "epoch": 0.7135362014690452,
+ "grad_norm": 0.15955896674456582,
+ "learning_rate": 1.3512939694349211e-05,
+ "loss": 0.4558671712875366,
+ "step": 1360
+ },
+ {
+ "epoch": 0.7140608604407135,
+ "grad_norm": 0.1589376751536802,
+ "learning_rate": 1.3483909372550446e-05,
+ "loss": 0.4430835247039795,
+ "step": 1361
+ },
+ {
+ "epoch": 0.7145855194123819,
+ "grad_norm": 0.15979088397822155,
+ "learning_rate": 1.3454917129332279e-05,
+ "loss": 0.4426245391368866,
+ "step": 1362
+ },
+ {
+ "epoch": 0.7151101783840503,
+ "grad_norm": 0.15493179714072167,
+ "learning_rate": 1.3425963043460143e-05,
+ "loss": 0.4550474286079407,
+ "step": 1363
+ },
+ {
+ "epoch": 0.7156348373557188,
+ "grad_norm": 0.15396498968995032,
+ "learning_rate": 1.3397047193595808e-05,
+ "loss": 0.44220536947250366,
+ "step": 1364
+ },
+ {
+ "epoch": 0.7161594963273872,
+ "grad_norm": 0.1489346822667273,
+ "learning_rate": 1.336816965829717e-05,
+ "loss": 0.4437323212623596,
+ "step": 1365
+ },
+ {
+ "epoch": 0.7166841552990556,
+ "grad_norm": 0.16311964931734477,
+ "learning_rate": 1.333933051601802e-05,
+ "loss": 0.45308777689933777,
+ "step": 1366
+ },
+ {
+ "epoch": 0.717208814270724,
+ "grad_norm": 0.1534354619021179,
+ "learning_rate": 1.331052984510785e-05,
+ "loss": 0.4334886372089386,
+ "step": 1367
+ },
+ {
+ "epoch": 0.7177334732423925,
+ "grad_norm": 0.15541089758065404,
+ "learning_rate": 1.328176772381165e-05,
+ "loss": 0.45248323678970337,
+ "step": 1368
+ },
+ {
+ "epoch": 0.7182581322140609,
+ "grad_norm": 0.1617921305700596,
+ "learning_rate": 1.3253044230269651e-05,
+ "loss": 0.44092947244644165,
+ "step": 1369
+ },
+ {
+ "epoch": 0.7187827911857293,
+ "grad_norm": 0.15047353793760151,
+ "learning_rate": 1.3224359442517165e-05,
+ "loss": 0.4424414336681366,
+ "step": 1370
+ },
+ {
+ "epoch": 0.7193074501573977,
+ "grad_norm": 0.15958089102258743,
+ "learning_rate": 1.319571343848433e-05,
+ "loss": 0.4458580017089844,
+ "step": 1371
+ },
+ {
+ "epoch": 0.7198321091290661,
+ "grad_norm": 0.15349258306968733,
+ "learning_rate": 1.3167106295995924e-05,
+ "loss": 0.4285386800765991,
+ "step": 1372
+ },
+ {
+ "epoch": 0.7203567681007346,
+ "grad_norm": 0.16042529563759886,
+ "learning_rate": 1.3138538092771164e-05,
+ "loss": 0.4497690200805664,
+ "step": 1373
+ },
+ {
+ "epoch": 0.720881427072403,
+ "grad_norm": 0.1565113736588754,
+ "learning_rate": 1.3110008906423454e-05,
+ "loss": 0.43562617897987366,
+ "step": 1374
+ },
+ {
+ "epoch": 0.7214060860440713,
+ "grad_norm": 0.15177291086871506,
+ "learning_rate": 1.3081518814460215e-05,
+ "loss": 0.44456908106803894,
+ "step": 1375
+ },
+ {
+ "epoch": 0.7219307450157397,
+ "grad_norm": 0.15928490480439225,
+ "learning_rate": 1.3053067894282622e-05,
+ "loss": 0.46170780062675476,
+ "step": 1376
+ },
+ {
+ "epoch": 0.7224554039874082,
+ "grad_norm": 0.1554814469217892,
+ "learning_rate": 1.302465622318549e-05,
+ "loss": 0.4422169327735901,
+ "step": 1377
+ },
+ {
+ "epoch": 0.7229800629590766,
+ "grad_norm": 0.15375370214966386,
+ "learning_rate": 1.2996283878356947e-05,
+ "loss": 0.4564763903617859,
+ "step": 1378
+ },
+ {
+ "epoch": 0.723504721930745,
+ "grad_norm": 0.15942841969395802,
+ "learning_rate": 1.2967950936878311e-05,
+ "loss": 0.45121875405311584,
+ "step": 1379
+ },
+ {
+ "epoch": 0.7240293809024134,
+ "grad_norm": 0.15731308716774903,
+ "learning_rate": 1.2939657475723844e-05,
+ "loss": 0.4285027086734772,
+ "step": 1380
+ },
+ {
+ "epoch": 0.7245540398740818,
+ "grad_norm": 0.15997753613932147,
+ "learning_rate": 1.291140357176054e-05,
+ "loss": 0.4566386342048645,
+ "step": 1381
+ },
+ {
+ "epoch": 0.7250786988457503,
+ "grad_norm": 0.15333715267473727,
+ "learning_rate": 1.2883189301747927e-05,
+ "loss": 0.45093846321105957,
+ "step": 1382
+ },
+ {
+ "epoch": 0.7256033578174187,
+ "grad_norm": 0.1514510304709061,
+ "learning_rate": 1.2855014742337874e-05,
+ "loss": 0.44256967306137085,
+ "step": 1383
+ },
+ {
+ "epoch": 0.7261280167890871,
+ "grad_norm": 0.15624301298995746,
+ "learning_rate": 1.2826879970074346e-05,
+ "loss": 0.4361875653266907,
+ "step": 1384
+ },
+ {
+ "epoch": 0.7266526757607555,
+ "grad_norm": 0.15767850122006694,
+ "learning_rate": 1.279878506139322e-05,
+ "loss": 0.4477671980857849,
+ "step": 1385
+ },
+ {
+ "epoch": 0.727177334732424,
+ "grad_norm": 0.15537676969694064,
+ "learning_rate": 1.2770730092622072e-05,
+ "loss": 0.4339832663536072,
+ "step": 1386
+ },
+ {
+ "epoch": 0.7277019937040924,
+ "grad_norm": 0.15944113633281293,
+ "learning_rate": 1.2742715139979965e-05,
+ "loss": 0.45407935976982117,
+ "step": 1387
+ },
+ {
+ "epoch": 0.7282266526757608,
+ "grad_norm": 0.16763519518606218,
+ "learning_rate": 1.271474027957727e-05,
+ "loss": 0.46442392468452454,
+ "step": 1388
+ },
+ {
+ "epoch": 0.7287513116474291,
+ "grad_norm": 0.1564572774601902,
+ "learning_rate": 1.2686805587415423e-05,
+ "loss": 0.4463968276977539,
+ "step": 1389
+ },
+ {
+ "epoch": 0.7292759706190975,
+ "grad_norm": 0.16384626314986128,
+ "learning_rate": 1.2658911139386709e-05,
+ "loss": 0.45519885420799255,
+ "step": 1390
+ },
+ {
+ "epoch": 0.729800629590766,
+ "grad_norm": 0.1562023802820964,
+ "learning_rate": 1.2631057011274106e-05,
+ "loss": 0.4365222454071045,
+ "step": 1391
+ },
+ {
+ "epoch": 0.7303252885624344,
+ "grad_norm": 0.15597287232940926,
+ "learning_rate": 1.2603243278751042e-05,
+ "loss": 0.4328266680240631,
+ "step": 1392
+ },
+ {
+ "epoch": 0.7308499475341028,
+ "grad_norm": 0.14762833706132167,
+ "learning_rate": 1.2575470017381214e-05,
+ "loss": 0.43902426958084106,
+ "step": 1393
+ },
+ {
+ "epoch": 0.7313746065057712,
+ "grad_norm": 0.1688371515856315,
+ "learning_rate": 1.2547737302618351e-05,
+ "loss": 0.4649963676929474,
+ "step": 1394
+ },
+ {
+ "epoch": 0.7318992654774397,
+ "grad_norm": 0.15501514509586015,
+ "learning_rate": 1.2520045209806024e-05,
+ "loss": 0.4501558542251587,
+ "step": 1395
+ },
+ {
+ "epoch": 0.7324239244491081,
+ "grad_norm": 0.15516748917693537,
+ "learning_rate": 1.2492393814177457e-05,
+ "loss": 0.4405498206615448,
+ "step": 1396
+ },
+ {
+ "epoch": 0.7329485834207765,
+ "grad_norm": 0.17115400468199243,
+ "learning_rate": 1.2464783190855295e-05,
+ "loss": 0.4445379674434662,
+ "step": 1397
+ },
+ {
+ "epoch": 0.7334732423924449,
+ "grad_norm": 0.15966386341604186,
+ "learning_rate": 1.243721341485143e-05,
+ "loss": 0.4769889712333679,
+ "step": 1398
+ },
+ {
+ "epoch": 0.7339979013641134,
+ "grad_norm": 0.15614706359781472,
+ "learning_rate": 1.2409684561066766e-05,
+ "loss": 0.43643918633461,
+ "step": 1399
+ },
+ {
+ "epoch": 0.7345225603357818,
+ "grad_norm": 0.17467968960064226,
+ "learning_rate": 1.238219670429104e-05,
+ "loss": 0.45413005352020264,
+ "step": 1400
+ },
+ {
+ "epoch": 0.7350472193074502,
+ "grad_norm": 0.1544517430810819,
+ "learning_rate": 1.2354749919202598e-05,
+ "loss": 0.45218425989151,
+ "step": 1401
+ },
+ {
+ "epoch": 0.7355718782791185,
+ "grad_norm": 0.16551083673374484,
+ "learning_rate": 1.2327344280368217e-05,
+ "loss": 0.4459628462791443,
+ "step": 1402
+ },
+ {
+ "epoch": 0.7360965372507869,
+ "grad_norm": 0.17614490459613263,
+ "learning_rate": 1.2299979862242882e-05,
+ "loss": 0.4406280815601349,
+ "step": 1403
+ },
+ {
+ "epoch": 0.7366211962224554,
+ "grad_norm": 0.15554925394435032,
+ "learning_rate": 1.2272656739169588e-05,
+ "loss": 0.4380159378051758,
+ "step": 1404
+ },
+ {
+ "epoch": 0.7371458551941238,
+ "grad_norm": 0.16475466799870916,
+ "learning_rate": 1.2245374985379147e-05,
+ "loss": 0.4442812204360962,
+ "step": 1405
+ },
+ {
+ "epoch": 0.7376705141657922,
+ "grad_norm": 0.16877301168875541,
+ "learning_rate": 1.2218134674989978e-05,
+ "loss": 0.4589429199695587,
+ "step": 1406
+ },
+ {
+ "epoch": 0.7381951731374606,
+ "grad_norm": 0.1583278589457438,
+ "learning_rate": 1.2190935882007894e-05,
+ "loss": 0.44896745681762695,
+ "step": 1407
+ },
+ {
+ "epoch": 0.7387198321091291,
+ "grad_norm": 0.1582022541656192,
+ "learning_rate": 1.2163778680325947e-05,
+ "loss": 0.4367997646331787,
+ "step": 1408
+ },
+ {
+ "epoch": 0.7392444910807975,
+ "grad_norm": 0.1669573982300068,
+ "learning_rate": 1.2136663143724163e-05,
+ "loss": 0.4503386616706848,
+ "step": 1409
+ },
+ {
+ "epoch": 0.7397691500524659,
+ "grad_norm": 0.1637877523924539,
+ "learning_rate": 1.2109589345869387e-05,
+ "loss": 0.4660584330558777,
+ "step": 1410
+ },
+ {
+ "epoch": 0.7402938090241343,
+ "grad_norm": 0.1640136159943701,
+ "learning_rate": 1.2082557360315065e-05,
+ "loss": 0.45206528902053833,
+ "step": 1411
+ },
+ {
+ "epoch": 0.7408184679958028,
+ "grad_norm": 0.1596580168682554,
+ "learning_rate": 1.2055567260501054e-05,
+ "loss": 0.4405035376548767,
+ "step": 1412
+ },
+ {
+ "epoch": 0.7413431269674712,
+ "grad_norm": 0.15224511741945293,
+ "learning_rate": 1.2028619119753402e-05,
+ "loss": 0.42810291051864624,
+ "step": 1413
+ },
+ {
+ "epoch": 0.7418677859391396,
+ "grad_norm": 0.15631381931817098,
+ "learning_rate": 1.2001713011284193e-05,
+ "loss": 0.4421626925468445,
+ "step": 1414
+ },
+ {
+ "epoch": 0.742392444910808,
+ "grad_norm": 0.15721819981265503,
+ "learning_rate": 1.1974849008191288e-05,
+ "loss": 0.4540696144104004,
+ "step": 1415
+ },
+ {
+ "epoch": 0.7429171038824763,
+ "grad_norm": 0.15904478578744463,
+ "learning_rate": 1.1948027183458182e-05,
+ "loss": 0.4601830244064331,
+ "step": 1416
+ },
+ {
+ "epoch": 0.7434417628541448,
+ "grad_norm": 0.1530755605235123,
+ "learning_rate": 1.1921247609953747e-05,
+ "loss": 0.4552589952945709,
+ "step": 1417
+ },
+ {
+ "epoch": 0.7439664218258132,
+ "grad_norm": 0.15375751072258462,
+ "learning_rate": 1.1894510360432106e-05,
+ "loss": 0.42965373396873474,
+ "step": 1418
+ },
+ {
+ "epoch": 0.7444910807974816,
+ "grad_norm": 0.15678686023937333,
+ "learning_rate": 1.1867815507532377e-05,
+ "loss": 0.44684016704559326,
+ "step": 1419
+ },
+ {
+ "epoch": 0.74501573976915,
+ "grad_norm": 0.15486952983783606,
+ "learning_rate": 1.1841163123778502e-05,
+ "loss": 0.44415023922920227,
+ "step": 1420
+ },
+ {
+ "epoch": 0.7455403987408185,
+ "grad_norm": 0.15342611490545593,
+ "learning_rate": 1.1814553281579034e-05,
+ "loss": 0.4501304626464844,
+ "step": 1421
+ },
+ {
+ "epoch": 0.7460650577124869,
+ "grad_norm": 0.15374260012788166,
+ "learning_rate": 1.1787986053226954e-05,
+ "loss": 0.4521749019622803,
+ "step": 1422
+ },
+ {
+ "epoch": 0.7465897166841553,
+ "grad_norm": 0.1543553258370361,
+ "learning_rate": 1.176146151089949e-05,
+ "loss": 0.44753748178482056,
+ "step": 1423
+ },
+ {
+ "epoch": 0.7471143756558237,
+ "grad_norm": 0.15921670193848306,
+ "learning_rate": 1.1734979726657878e-05,
+ "loss": 0.4473114311695099,
+ "step": 1424
+ },
+ {
+ "epoch": 0.7476390346274921,
+ "grad_norm": 0.15023498686186612,
+ "learning_rate": 1.17085407724472e-05,
+ "loss": 0.4568778872489929,
+ "step": 1425
+ },
+ {
+ "epoch": 0.7481636935991606,
+ "grad_norm": 0.14601156920274583,
+ "learning_rate": 1.1682144720096168e-05,
+ "loss": 0.4389464855194092,
+ "step": 1426
+ },
+ {
+ "epoch": 0.748688352570829,
+ "grad_norm": 0.1520548095149398,
+ "learning_rate": 1.1655791641316954e-05,
+ "loss": 0.440733402967453,
+ "step": 1427
+ },
+ {
+ "epoch": 0.7492130115424974,
+ "grad_norm": 0.1591915163097993,
+ "learning_rate": 1.1629481607704963e-05,
+ "loss": 0.46245887875556946,
+ "step": 1428
+ },
+ {
+ "epoch": 0.7497376705141658,
+ "grad_norm": 0.14867295658447263,
+ "learning_rate": 1.1603214690738683e-05,
+ "loss": 0.4332764744758606,
+ "step": 1429
+ },
+ {
+ "epoch": 0.7502623294858342,
+ "grad_norm": 0.1498957762903134,
+ "learning_rate": 1.1576990961779447e-05,
+ "loss": 0.45215466618537903,
+ "step": 1430
+ },
+ {
+ "epoch": 0.7507869884575026,
+ "grad_norm": 0.15430555803491824,
+ "learning_rate": 1.1550810492071243e-05,
+ "loss": 0.4454396963119507,
+ "step": 1431
+ },
+ {
+ "epoch": 0.751311647429171,
+ "grad_norm": 0.1517461021936533,
+ "learning_rate": 1.1524673352740543e-05,
+ "loss": 0.43869835138320923,
+ "step": 1432
+ },
+ {
+ "epoch": 0.7518363064008394,
+ "grad_norm": 0.14541657994692936,
+ "learning_rate": 1.1498579614796121e-05,
+ "loss": 0.4373054504394531,
+ "step": 1433
+ },
+ {
+ "epoch": 0.7523609653725079,
+ "grad_norm": 0.15539513763786977,
+ "learning_rate": 1.1472529349128812e-05,
+ "loss": 0.45451128482818604,
+ "step": 1434
+ },
+ {
+ "epoch": 0.7528856243441763,
+ "grad_norm": 0.1482043113690946,
+ "learning_rate": 1.144652262651136e-05,
+ "loss": 0.43644773960113525,
+ "step": 1435
+ },
+ {
+ "epoch": 0.7534102833158447,
+ "grad_norm": 0.16293628094190815,
+ "learning_rate": 1.1420559517598205e-05,
+ "loss": 0.4687269330024719,
+ "step": 1436
+ },
+ {
+ "epoch": 0.7539349422875131,
+ "grad_norm": 0.1553393974390287,
+ "learning_rate": 1.1394640092925302e-05,
+ "loss": 0.4501398205757141,
+ "step": 1437
+ },
+ {
+ "epoch": 0.7544596012591815,
+ "grad_norm": 0.16214837694535802,
+ "learning_rate": 1.1368764422909919e-05,
+ "loss": 0.4542577862739563,
+ "step": 1438
+ },
+ {
+ "epoch": 0.75498426023085,
+ "grad_norm": 0.15838878372428297,
+ "learning_rate": 1.1342932577850473e-05,
+ "loss": 0.46228161454200745,
+ "step": 1439
+ },
+ {
+ "epoch": 0.7555089192025184,
+ "grad_norm": 0.15090419497122687,
+ "learning_rate": 1.13171446279263e-05,
+ "loss": 0.4495932161808014,
+ "step": 1440
+ },
+ {
+ "epoch": 0.7560335781741868,
+ "grad_norm": 0.1543376786458394,
+ "learning_rate": 1.1291400643197484e-05,
+ "loss": 0.451630175113678,
+ "step": 1441
+ },
+ {
+ "epoch": 0.7565582371458552,
+ "grad_norm": 0.1539470751610696,
+ "learning_rate": 1.1265700693604668e-05,
+ "loss": 0.447547972202301,
+ "step": 1442
+ },
+ {
+ "epoch": 0.7570828961175237,
+ "grad_norm": 0.15761047703305445,
+ "learning_rate": 1.1240044848968856e-05,
+ "loss": 0.4516204595565796,
+ "step": 1443
+ },
+ {
+ "epoch": 0.757607555089192,
+ "grad_norm": 0.15675279683704735,
+ "learning_rate": 1.1214433178991258e-05,
+ "loss": 0.4597971737384796,
+ "step": 1444
+ },
+ {
+ "epoch": 0.7581322140608604,
+ "grad_norm": 0.15592110994904113,
+ "learning_rate": 1.1188865753253022e-05,
+ "loss": 0.443898469209671,
+ "step": 1445
+ },
+ {
+ "epoch": 0.7586568730325288,
+ "grad_norm": 0.15597501691030619,
+ "learning_rate": 1.116334264121513e-05,
+ "loss": 0.4581306278705597,
+ "step": 1446
+ },
+ {
+ "epoch": 0.7591815320041972,
+ "grad_norm": 0.15552304728589958,
+ "learning_rate": 1.1137863912218154e-05,
+ "loss": 0.4475254416465759,
+ "step": 1447
+ },
+ {
+ "epoch": 0.7597061909758657,
+ "grad_norm": 0.15759990123898035,
+ "learning_rate": 1.1112429635482118e-05,
+ "loss": 0.43585264682769775,
+ "step": 1448
+ },
+ {
+ "epoch": 0.7602308499475341,
+ "grad_norm": 0.15964657264676635,
+ "learning_rate": 1.1087039880106247e-05,
+ "loss": 0.46786290407180786,
+ "step": 1449
+ },
+ {
+ "epoch": 0.7607555089192025,
+ "grad_norm": 0.15155804051218852,
+ "learning_rate": 1.106169471506882e-05,
+ "loss": 0.42292726039886475,
+ "step": 1450
+ },
+ {
+ "epoch": 0.7612801678908709,
+ "grad_norm": 0.15519270966930884,
+ "learning_rate": 1.103639420922698e-05,
+ "loss": 0.4248972237110138,
+ "step": 1451
+ },
+ {
+ "epoch": 0.7618048268625394,
+ "grad_norm": 0.1526330511128821,
+ "learning_rate": 1.1011138431316537e-05,
+ "loss": 0.4317784309387207,
+ "step": 1452
+ },
+ {
+ "epoch": 0.7623294858342078,
+ "grad_norm": 0.15797682979272226,
+ "learning_rate": 1.0985927449951779e-05,
+ "loss": 0.44506698846817017,
+ "step": 1453
+ },
+ {
+ "epoch": 0.7628541448058762,
+ "grad_norm": 0.15479979838505462,
+ "learning_rate": 1.096076133362531e-05,
+ "loss": 0.4471631348133087,
+ "step": 1454
+ },
+ {
+ "epoch": 0.7633788037775446,
+ "grad_norm": 0.15712267274447342,
+ "learning_rate": 1.093564015070783e-05,
+ "loss": 0.4321915805339813,
+ "step": 1455
+ },
+ {
+ "epoch": 0.7639034627492131,
+ "grad_norm": 0.15857431494180016,
+ "learning_rate": 1.0910563969447969e-05,
+ "loss": 0.44455766677856445,
+ "step": 1456
+ },
+ {
+ "epoch": 0.7644281217208815,
+ "grad_norm": 0.15377057080997789,
+ "learning_rate": 1.0885532857972097e-05,
+ "loss": 0.4567924439907074,
+ "step": 1457
+ },
+ {
+ "epoch": 0.7649527806925498,
+ "grad_norm": 0.15544844005009104,
+ "learning_rate": 1.0860546884284143e-05,
+ "loss": 0.4491572976112366,
+ "step": 1458
+ },
+ {
+ "epoch": 0.7654774396642182,
+ "grad_norm": 0.16210735566479478,
+ "learning_rate": 1.08356061162654e-05,
+ "loss": 0.43823331594467163,
+ "step": 1459
+ },
+ {
+ "epoch": 0.7660020986358866,
+ "grad_norm": 0.15213559486379452,
+ "learning_rate": 1.0810710621674355e-05,
+ "loss": 0.45375770330429077,
+ "step": 1460
+ },
+ {
+ "epoch": 0.7665267576075551,
+ "grad_norm": 0.1727832306402992,
+ "learning_rate": 1.0785860468146494e-05,
+ "loss": 0.4639572501182556,
+ "step": 1461
+ },
+ {
+ "epoch": 0.7670514165792235,
+ "grad_norm": 0.1637555902013663,
+ "learning_rate": 1.076105572319412e-05,
+ "loss": 0.4355923533439636,
+ "step": 1462
+ },
+ {
+ "epoch": 0.7675760755508919,
+ "grad_norm": 0.16646936030470866,
+ "learning_rate": 1.0736296454206166e-05,
+ "loss": 0.4650861322879791,
+ "step": 1463
+ },
+ {
+ "epoch": 0.7681007345225603,
+ "grad_norm": 0.14987620615431385,
+ "learning_rate": 1.0711582728448033e-05,
+ "loss": 0.45445430278778076,
+ "step": 1464
+ },
+ {
+ "epoch": 0.7686253934942288,
+ "grad_norm": 0.16428618403547932,
+ "learning_rate": 1.0686914613061377e-05,
+ "loss": 0.46050935983657837,
+ "step": 1465
+ },
+ {
+ "epoch": 0.7691500524658972,
+ "grad_norm": 0.15052999053128632,
+ "learning_rate": 1.0662292175063949e-05,
+ "loss": 0.43385741114616394,
+ "step": 1466
+ },
+ {
+ "epoch": 0.7696747114375656,
+ "grad_norm": 0.1654949402740152,
+ "learning_rate": 1.0637715481349397e-05,
+ "loss": 0.4434410333633423,
+ "step": 1467
+ },
+ {
+ "epoch": 0.770199370409234,
+ "grad_norm": 0.15923836508490036,
+ "learning_rate": 1.0613184598687088e-05,
+ "loss": 0.4410185217857361,
+ "step": 1468
+ },
+ {
+ "epoch": 0.7707240293809025,
+ "grad_norm": 0.15813007111170696,
+ "learning_rate": 1.0588699593721957e-05,
+ "loss": 0.43588417768478394,
+ "step": 1469
+ },
+ {
+ "epoch": 0.7712486883525709,
+ "grad_norm": 0.16978817898431764,
+ "learning_rate": 1.0564260532974274e-05,
+ "loss": 0.4514843225479126,
+ "step": 1470
+ },
+ {
+ "epoch": 0.7717733473242392,
+ "grad_norm": 0.1613844852697828,
+ "learning_rate": 1.0539867482839499e-05,
+ "loss": 0.4554520845413208,
+ "step": 1471
+ },
+ {
+ "epoch": 0.7722980062959076,
+ "grad_norm": 0.1587788181984633,
+ "learning_rate": 1.0515520509588075e-05,
+ "loss": 0.43200671672821045,
+ "step": 1472
+ },
+ {
+ "epoch": 0.772822665267576,
+ "grad_norm": 0.1539930824827353,
+ "learning_rate": 1.0491219679365283e-05,
+ "loss": 0.4343116283416748,
+ "step": 1473
+ },
+ {
+ "epoch": 0.7733473242392445,
+ "grad_norm": 0.1561612537773709,
+ "learning_rate": 1.0466965058191049e-05,
+ "loss": 0.4477815330028534,
+ "step": 1474
+ },
+ {
+ "epoch": 0.7738719832109129,
+ "grad_norm": 0.169990777691657,
+ "learning_rate": 1.0442756711959745e-05,
+ "loss": 0.4476611614227295,
+ "step": 1475
+ },
+ {
+ "epoch": 0.7743966421825813,
+ "grad_norm": 0.1624313049193793,
+ "learning_rate": 1.041859470644002e-05,
+ "loss": 0.47136399149894714,
+ "step": 1476
+ },
+ {
+ "epoch": 0.7749213011542497,
+ "grad_norm": 0.15993315458994328,
+ "learning_rate": 1.0394479107274645e-05,
+ "loss": 0.4672670364379883,
+ "step": 1477
+ },
+ {
+ "epoch": 0.7754459601259182,
+ "grad_norm": 0.16755349272347217,
+ "learning_rate": 1.0370409979980294e-05,
+ "loss": 0.45135706663131714,
+ "step": 1478
+ },
+ {
+ "epoch": 0.7759706190975866,
+ "grad_norm": 0.15665059815830362,
+ "learning_rate": 1.0346387389947407e-05,
+ "loss": 0.45178359746932983,
+ "step": 1479
+ },
+ {
+ "epoch": 0.776495278069255,
+ "grad_norm": 0.1554612731452571,
+ "learning_rate": 1.0322411402439983e-05,
+ "loss": 0.4412590265274048,
+ "step": 1480
+ },
+ {
+ "epoch": 0.7770199370409234,
+ "grad_norm": 0.1532496869312065,
+ "learning_rate": 1.0298482082595408e-05,
+ "loss": 0.44104987382888794,
+ "step": 1481
+ },
+ {
+ "epoch": 0.7775445960125918,
+ "grad_norm": 0.15065868316960426,
+ "learning_rate": 1.0274599495424286e-05,
+ "loss": 0.4336603283882141,
+ "step": 1482
+ },
+ {
+ "epoch": 0.7780692549842603,
+ "grad_norm": 0.1548373981760097,
+ "learning_rate": 1.0250763705810267e-05,
+ "loss": 0.4658375680446625,
+ "step": 1483
+ },
+ {
+ "epoch": 0.7785939139559287,
+ "grad_norm": 0.15470181175866582,
+ "learning_rate": 1.0226974778509846e-05,
+ "loss": 0.45232075452804565,
+ "step": 1484
+ },
+ {
+ "epoch": 0.779118572927597,
+ "grad_norm": 0.15507371187727106,
+ "learning_rate": 1.0203232778152228e-05,
+ "loss": 0.43640458583831787,
+ "step": 1485
+ },
+ {
+ "epoch": 0.7796432318992654,
+ "grad_norm": 0.15269440736544085,
+ "learning_rate": 1.01795377692391e-05,
+ "loss": 0.4400939345359802,
+ "step": 1486
+ },
+ {
+ "epoch": 0.7801678908709339,
+ "grad_norm": 0.15610899550395085,
+ "learning_rate": 1.0155889816144495e-05,
+ "loss": 0.4546602666378021,
+ "step": 1487
+ },
+ {
+ "epoch": 0.7806925498426023,
+ "grad_norm": 0.1559328242601668,
+ "learning_rate": 1.0132288983114611e-05,
+ "loss": 0.45510709285736084,
+ "step": 1488
+ },
+ {
+ "epoch": 0.7812172088142707,
+ "grad_norm": 0.1537255194600787,
+ "learning_rate": 1.0108735334267639e-05,
+ "loss": 0.4475342631340027,
+ "step": 1489
+ },
+ {
+ "epoch": 0.7817418677859391,
+ "grad_norm": 0.14616575061664402,
+ "learning_rate": 1.0085228933593558e-05,
+ "loss": 0.4378190040588379,
+ "step": 1490
+ },
+ {
+ "epoch": 0.7822665267576075,
+ "grad_norm": 0.1624360967746598,
+ "learning_rate": 1.0061769844953999e-05,
+ "loss": 0.4439539909362793,
+ "step": 1491
+ },
+ {
+ "epoch": 0.782791185729276,
+ "grad_norm": 0.15407336125298413,
+ "learning_rate": 1.003835813208206e-05,
+ "loss": 0.44074907898902893,
+ "step": 1492
+ },
+ {
+ "epoch": 0.7833158447009444,
+ "grad_norm": 0.15786980710136125,
+ "learning_rate": 1.001499385858211e-05,
+ "loss": 0.4491667151451111,
+ "step": 1493
+ },
+ {
+ "epoch": 0.7838405036726128,
+ "grad_norm": 0.15338836738460088,
+ "learning_rate": 9.991677087929665e-06,
+ "loss": 0.446003794670105,
+ "step": 1494
+ },
+ {
+ "epoch": 0.7843651626442812,
+ "grad_norm": 0.156710894892,
+ "learning_rate": 9.968407883471165e-06,
+ "loss": 0.43604689836502075,
+ "step": 1495
+ },
+ {
+ "epoch": 0.7848898216159497,
+ "grad_norm": 0.16100539063191666,
+ "learning_rate": 9.945186308423832e-06,
+ "loss": 0.4431147873401642,
+ "step": 1496
+ },
+ {
+ "epoch": 0.7854144805876181,
+ "grad_norm": 0.15776018851122997,
+ "learning_rate": 9.92201242587548e-06,
+ "loss": 0.4532504975795746,
+ "step": 1497
+ },
+ {
+ "epoch": 0.7859391395592865,
+ "grad_norm": 0.1540333423540934,
+ "learning_rate": 9.898886298784368e-06,
+ "loss": 0.43605154752731323,
+ "step": 1498
+ },
+ {
+ "epoch": 0.7864637985309548,
+ "grad_norm": 0.16009602938972295,
+ "learning_rate": 9.875807989979005e-06,
+ "loss": 0.4605238437652588,
+ "step": 1499
+ },
+ {
+ "epoch": 0.7869884575026233,
+ "grad_norm": 0.15639368804171708,
+ "learning_rate": 9.852777562157983e-06,
+ "loss": 0.4509708881378174,
+ "step": 1500
+ },
+ {
+ "epoch": 0.7875131164742917,
+ "grad_norm": 0.15074011188741518,
+ "learning_rate": 9.829795077889828e-06,
+ "loss": 0.4527304172515869,
+ "step": 1501
+ },
+ {
+ "epoch": 0.7880377754459601,
+ "grad_norm": 0.15155083996490956,
+ "learning_rate": 9.806860599612799e-06,
+ "loss": 0.4306413531303406,
+ "step": 1502
+ },
+ {
+ "epoch": 0.7885624344176285,
+ "grad_norm": 0.1478956630795445,
+ "learning_rate": 9.783974189634734e-06,
+ "loss": 0.4434714913368225,
+ "step": 1503
+ },
+ {
+ "epoch": 0.789087093389297,
+ "grad_norm": 0.15327668434806432,
+ "learning_rate": 9.761135910132902e-06,
+ "loss": 0.44431227445602417,
+ "step": 1504
+ },
+ {
+ "epoch": 0.7896117523609654,
+ "grad_norm": 0.2243177585497212,
+ "learning_rate": 9.738345823153783e-06,
+ "loss": 0.4541007876396179,
+ "step": 1505
+ },
+ {
+ "epoch": 0.7901364113326338,
+ "grad_norm": 0.1600239865812282,
+ "learning_rate": 9.715603990612945e-06,
+ "loss": 0.4456910490989685,
+ "step": 1506
+ },
+ {
+ "epoch": 0.7906610703043022,
+ "grad_norm": 0.15304634075332066,
+ "learning_rate": 9.692910474294858e-06,
+ "loss": 0.43965545296669006,
+ "step": 1507
+ },
+ {
+ "epoch": 0.7911857292759706,
+ "grad_norm": 0.15657752062053032,
+ "learning_rate": 9.670265335852725e-06,
+ "loss": 0.43358081579208374,
+ "step": 1508
+ },
+ {
+ "epoch": 0.7917103882476391,
+ "grad_norm": 0.16110057783374115,
+ "learning_rate": 9.647668636808304e-06,
+ "loss": 0.45711368322372437,
+ "step": 1509
+ },
+ {
+ "epoch": 0.7922350472193075,
+ "grad_norm": 0.1551133690299623,
+ "learning_rate": 9.625120438551791e-06,
+ "loss": 0.4613519608974457,
+ "step": 1510
+ },
+ {
+ "epoch": 0.7927597061909759,
+ "grad_norm": 0.15690897380725946,
+ "learning_rate": 9.602620802341581e-06,
+ "loss": 0.4472261071205139,
+ "step": 1511
+ },
+ {
+ "epoch": 0.7932843651626443,
+ "grad_norm": 0.147796599322911,
+ "learning_rate": 9.580169789304152e-06,
+ "loss": 0.4447203576564789,
+ "step": 1512
+ },
+ {
+ "epoch": 0.7938090241343126,
+ "grad_norm": 0.152019559199633,
+ "learning_rate": 9.557767460433878e-06,
+ "loss": 0.445905864238739,
+ "step": 1513
+ },
+ {
+ "epoch": 0.7943336831059811,
+ "grad_norm": 0.1574763924872037,
+ "learning_rate": 9.535413876592872e-06,
+ "loss": 0.4502032995223999,
+ "step": 1514
+ },
+ {
+ "epoch": 0.7948583420776495,
+ "grad_norm": 0.1524218315499056,
+ "learning_rate": 9.513109098510816e-06,
+ "loss": 0.43753111362457275,
+ "step": 1515
+ },
+ {
+ "epoch": 0.7953830010493179,
+ "grad_norm": 0.15589147437937434,
+ "learning_rate": 9.490853186784801e-06,
+ "loss": 0.44524312019348145,
+ "step": 1516
+ },
+ {
+ "epoch": 0.7959076600209863,
+ "grad_norm": 0.15604009714081407,
+ "learning_rate": 9.468646201879156e-06,
+ "loss": 0.44141292572021484,
+ "step": 1517
+ },
+ {
+ "epoch": 0.7964323189926548,
+ "grad_norm": 0.1562432554004435,
+ "learning_rate": 9.44648820412529e-06,
+ "loss": 0.461681067943573,
+ "step": 1518
+ },
+ {
+ "epoch": 0.7969569779643232,
+ "grad_norm": 0.15233588531495665,
+ "learning_rate": 9.424379253721513e-06,
+ "loss": 0.4529497027397156,
+ "step": 1519
+ },
+ {
+ "epoch": 0.7974816369359916,
+ "grad_norm": 0.16856991685290923,
+ "learning_rate": 9.40231941073291e-06,
+ "loss": 0.4483596086502075,
+ "step": 1520
+ },
+ {
+ "epoch": 0.79800629590766,
+ "grad_norm": 0.1546876712981077,
+ "learning_rate": 9.380308735091125e-06,
+ "loss": 0.43846625089645386,
+ "step": 1521
+ },
+ {
+ "epoch": 0.7985309548793285,
+ "grad_norm": 0.15198637556890449,
+ "learning_rate": 9.35834728659424e-06,
+ "loss": 0.4625553488731384,
+ "step": 1522
+ },
+ {
+ "epoch": 0.7990556138509969,
+ "grad_norm": 0.1519777007198701,
+ "learning_rate": 9.336435124906592e-06,
+ "loss": 0.44183820486068726,
+ "step": 1523
+ },
+ {
+ "epoch": 0.7995802728226653,
+ "grad_norm": 0.16125823074979284,
+ "learning_rate": 9.314572309558616e-06,
+ "loss": 0.46057307720184326,
+ "step": 1524
+ },
+ {
+ "epoch": 0.8001049317943337,
+ "grad_norm": 0.16283323624848034,
+ "learning_rate": 9.2927588999467e-06,
+ "loss": 0.4530155062675476,
+ "step": 1525
+ },
+ {
+ "epoch": 0.8006295907660022,
+ "grad_norm": 0.14998654489027738,
+ "learning_rate": 9.270994955332984e-06,
+ "loss": 0.43155500292778015,
+ "step": 1526
+ },
+ {
+ "epoch": 0.8011542497376705,
+ "grad_norm": 0.15683095030848093,
+ "learning_rate": 9.249280534845241e-06,
+ "loss": 0.45097050070762634,
+ "step": 1527
+ },
+ {
+ "epoch": 0.8016789087093389,
+ "grad_norm": 0.1561578254492378,
+ "learning_rate": 9.227615697476674e-06,
+ "loss": 0.42797762155532837,
+ "step": 1528
+ },
+ {
+ "epoch": 0.8022035676810073,
+ "grad_norm": 0.15642385261469652,
+ "learning_rate": 9.206000502085813e-06,
+ "loss": 0.4560159742832184,
+ "step": 1529
+ },
+ {
+ "epoch": 0.8027282266526757,
+ "grad_norm": 0.1623126279097499,
+ "learning_rate": 9.1844350073963e-06,
+ "loss": 0.4704165458679199,
+ "step": 1530
+ },
+ {
+ "epoch": 0.8032528856243442,
+ "grad_norm": 0.15090934786253773,
+ "learning_rate": 9.162919271996755e-06,
+ "loss": 0.44642555713653564,
+ "step": 1531
+ },
+ {
+ "epoch": 0.8037775445960126,
+ "grad_norm": 0.15305216978534084,
+ "learning_rate": 9.14145335434062e-06,
+ "loss": 0.4634283781051636,
+ "step": 1532
+ },
+ {
+ "epoch": 0.804302203567681,
+ "grad_norm": 0.14653219861345698,
+ "learning_rate": 9.120037312745979e-06,
+ "loss": 0.44544070959091187,
+ "step": 1533
+ },
+ {
+ "epoch": 0.8048268625393494,
+ "grad_norm": 0.16247397938502464,
+ "learning_rate": 9.098671205395422e-06,
+ "loss": 0.4552367925643921,
+ "step": 1534
+ },
+ {
+ "epoch": 0.8053515215110179,
+ "grad_norm": 0.14963174165902007,
+ "learning_rate": 9.077355090335892e-06,
+ "loss": 0.43211787939071655,
+ "step": 1535
+ },
+ {
+ "epoch": 0.8058761804826863,
+ "grad_norm": 0.1554853846772283,
+ "learning_rate": 9.056089025478495e-06,
+ "loss": 0.4319286644458771,
+ "step": 1536
+ },
+ {
+ "epoch": 0.8064008394543547,
+ "grad_norm": 0.15840895821310472,
+ "learning_rate": 9.03487306859837e-06,
+ "loss": 0.4494994878768921,
+ "step": 1537
+ },
+ {
+ "epoch": 0.8069254984260231,
+ "grad_norm": 0.1543496716800982,
+ "learning_rate": 9.013707277334515e-06,
+ "loss": 0.45500510931015015,
+ "step": 1538
+ },
+ {
+ "epoch": 0.8074501573976915,
+ "grad_norm": 0.14797282318550045,
+ "learning_rate": 8.992591709189651e-06,
+ "loss": 0.4385896325111389,
+ "step": 1539
+ },
+ {
+ "epoch": 0.8079748163693599,
+ "grad_norm": 0.17045570682787758,
+ "learning_rate": 8.971526421530066e-06,
+ "loss": 0.4585988521575928,
+ "step": 1540
+ },
+ {
+ "epoch": 0.8084994753410283,
+ "grad_norm": 0.14758220713044884,
+ "learning_rate": 8.950511471585407e-06,
+ "loss": 0.4295082092285156,
+ "step": 1541
+ },
+ {
+ "epoch": 0.8090241343126967,
+ "grad_norm": 0.16005752068175036,
+ "learning_rate": 8.929546916448597e-06,
+ "loss": 0.4507679343223572,
+ "step": 1542
+ },
+ {
+ "epoch": 0.8095487932843651,
+ "grad_norm": 0.14884684083313074,
+ "learning_rate": 8.908632813075641e-06,
+ "loss": 0.43547600507736206,
+ "step": 1543
+ },
+ {
+ "epoch": 0.8100734522560336,
+ "grad_norm": 0.14860107595788136,
+ "learning_rate": 8.88776921828547e-06,
+ "loss": 0.42700642347335815,
+ "step": 1544
+ },
+ {
+ "epoch": 0.810598111227702,
+ "grad_norm": 0.1472630611843838,
+ "learning_rate": 8.86695618875981e-06,
+ "loss": 0.4473339319229126,
+ "step": 1545
+ },
+ {
+ "epoch": 0.8111227701993704,
+ "grad_norm": 0.15105523788317893,
+ "learning_rate": 8.846193781043e-06,
+ "loss": 0.44923150539398193,
+ "step": 1546
+ },
+ {
+ "epoch": 0.8116474291710388,
+ "grad_norm": 0.16121409015263036,
+ "learning_rate": 8.82548205154185e-06,
+ "loss": 0.4653576612472534,
+ "step": 1547
+ },
+ {
+ "epoch": 0.8121720881427072,
+ "grad_norm": 0.1595195150959421,
+ "learning_rate": 8.804821056525497e-06,
+ "loss": 0.4522424638271332,
+ "step": 1548
+ },
+ {
+ "epoch": 0.8126967471143757,
+ "grad_norm": 0.15163510106316194,
+ "learning_rate": 8.784210852125232e-06,
+ "loss": 0.43657630681991577,
+ "step": 1549
+ },
+ {
+ "epoch": 0.8132214060860441,
+ "grad_norm": 0.15155109347013793,
+ "learning_rate": 8.763651494334377e-06,
+ "loss": 0.43183624744415283,
+ "step": 1550
+ },
+ {
+ "epoch": 0.8137460650577125,
+ "grad_norm": 0.14930024958763324,
+ "learning_rate": 8.7431430390081e-06,
+ "loss": 0.43563947081565857,
+ "step": 1551
+ },
+ {
+ "epoch": 0.8142707240293809,
+ "grad_norm": 0.15361200153430712,
+ "learning_rate": 8.722685541863283e-06,
+ "loss": 0.4387405514717102,
+ "step": 1552
+ },
+ {
+ "epoch": 0.8147953830010494,
+ "grad_norm": 0.15549528796193493,
+ "learning_rate": 8.702279058478366e-06,
+ "loss": 0.44347938895225525,
+ "step": 1553
+ },
+ {
+ "epoch": 0.8153200419727177,
+ "grad_norm": 0.16243042867861712,
+ "learning_rate": 8.681923644293194e-06,
+ "loss": 0.44852834939956665,
+ "step": 1554
+ },
+ {
+ "epoch": 0.8158447009443861,
+ "grad_norm": 0.15706129652944026,
+ "learning_rate": 8.661619354608873e-06,
+ "loss": 0.45419150590896606,
+ "step": 1555
+ },
+ {
+ "epoch": 0.8163693599160545,
+ "grad_norm": 0.14854258573134127,
+ "learning_rate": 8.641366244587611e-06,
+ "loss": 0.44866859912872314,
+ "step": 1556
+ },
+ {
+ "epoch": 0.816894018887723,
+ "grad_norm": 0.14896937678017533,
+ "learning_rate": 8.62116436925258e-06,
+ "loss": 0.44596076011657715,
+ "step": 1557
+ },
+ {
+ "epoch": 0.8174186778593914,
+ "grad_norm": 0.15614508495620508,
+ "learning_rate": 8.601013783487746e-06,
+ "loss": 0.45645207166671753,
+ "step": 1558
+ },
+ {
+ "epoch": 0.8179433368310598,
+ "grad_norm": 0.1486538543050606,
+ "learning_rate": 8.580914542037734e-06,
+ "loss": 0.4592300355434418,
+ "step": 1559
+ },
+ {
+ "epoch": 0.8184679958027282,
+ "grad_norm": 0.15475320756830518,
+ "learning_rate": 8.560866699507702e-06,
+ "loss": 0.4605134129524231,
+ "step": 1560
+ },
+ {
+ "epoch": 0.8189926547743966,
+ "grad_norm": 0.14754406728057914,
+ "learning_rate": 8.540870310363137e-06,
+ "loss": 0.43651673197746277,
+ "step": 1561
+ },
+ {
+ "epoch": 0.8195173137460651,
+ "grad_norm": 0.14871990701077825,
+ "learning_rate": 8.520925428929755e-06,
+ "loss": 0.44830837845802307,
+ "step": 1562
+ },
+ {
+ "epoch": 0.8200419727177335,
+ "grad_norm": 0.15358241317712973,
+ "learning_rate": 8.501032109393334e-06,
+ "loss": 0.4546644687652588,
+ "step": 1563
+ },
+ {
+ "epoch": 0.8205666316894019,
+ "grad_norm": 0.15650856948126138,
+ "learning_rate": 8.481190405799569e-06,
+ "loss": 0.4423891305923462,
+ "step": 1564
+ },
+ {
+ "epoch": 0.8210912906610703,
+ "grad_norm": 0.15474996746242287,
+ "learning_rate": 8.461400372053923e-06,
+ "loss": 0.44242313504219055,
+ "step": 1565
+ },
+ {
+ "epoch": 0.8216159496327388,
+ "grad_norm": 0.16544926591962897,
+ "learning_rate": 8.441662061921497e-06,
+ "loss": 0.4487020969390869,
+ "step": 1566
+ },
+ {
+ "epoch": 0.8221406086044072,
+ "grad_norm": 0.1489733586650647,
+ "learning_rate": 8.421975529026855e-06,
+ "loss": 0.4371356964111328,
+ "step": 1567
+ },
+ {
+ "epoch": 0.8226652675760755,
+ "grad_norm": 0.15995455652906124,
+ "learning_rate": 8.402340826853907e-06,
+ "loss": 0.4566021263599396,
+ "step": 1568
+ },
+ {
+ "epoch": 0.8231899265477439,
+ "grad_norm": 0.1500716152622102,
+ "learning_rate": 8.382758008745726e-06,
+ "loss": 0.43888187408447266,
+ "step": 1569
+ },
+ {
+ "epoch": 0.8237145855194123,
+ "grad_norm": 0.15554904485712012,
+ "learning_rate": 8.363227127904464e-06,
+ "loss": 0.4584001898765564,
+ "step": 1570
+ },
+ {
+ "epoch": 0.8242392444910808,
+ "grad_norm": 0.14820917815884277,
+ "learning_rate": 8.343748237391144e-06,
+ "loss": 0.43289071321487427,
+ "step": 1571
+ },
+ {
+ "epoch": 0.8247639034627492,
+ "grad_norm": 0.15801200747859714,
+ "learning_rate": 8.324321390125547e-06,
+ "loss": 0.44196009635925293,
+ "step": 1572
+ },
+ {
+ "epoch": 0.8252885624344176,
+ "grad_norm": 0.15323277975380797,
+ "learning_rate": 8.304946638886078e-06,
+ "loss": 0.43244463205337524,
+ "step": 1573
+ },
+ {
+ "epoch": 0.825813221406086,
+ "grad_norm": 0.15076940176344772,
+ "learning_rate": 8.285624036309583e-06,
+ "loss": 0.44401073455810547,
+ "step": 1574
+ },
+ {
+ "epoch": 0.8263378803777545,
+ "grad_norm": 0.1544958300475393,
+ "learning_rate": 8.266353634891266e-06,
+ "loss": 0.45198434591293335,
+ "step": 1575
+ },
+ {
+ "epoch": 0.8268625393494229,
+ "grad_norm": 0.15093528292068448,
+ "learning_rate": 8.247135486984486e-06,
+ "loss": 0.4403747618198395,
+ "step": 1576
+ },
+ {
+ "epoch": 0.8273871983210913,
+ "grad_norm": 0.14680662784017148,
+ "learning_rate": 8.227969644800648e-06,
+ "loss": 0.4375043213367462,
+ "step": 1577
+ },
+ {
+ "epoch": 0.8279118572927597,
+ "grad_norm": 0.14791704009291914,
+ "learning_rate": 8.208856160409055e-06,
+ "loss": 0.43811577558517456,
+ "step": 1578
+ },
+ {
+ "epoch": 0.8284365162644282,
+ "grad_norm": 0.14924185710518617,
+ "learning_rate": 8.189795085736765e-06,
+ "loss": 0.43853428959846497,
+ "step": 1579
+ },
+ {
+ "epoch": 0.8289611752360966,
+ "grad_norm": 0.15420576936040198,
+ "learning_rate": 8.170786472568451e-06,
+ "loss": 0.445467084646225,
+ "step": 1580
+ },
+ {
+ "epoch": 0.829485834207765,
+ "grad_norm": 0.15577901772605834,
+ "learning_rate": 8.151830372546258e-06,
+ "loss": 0.4529452621936798,
+ "step": 1581
+ },
+ {
+ "epoch": 0.8300104931794333,
+ "grad_norm": 0.15387840488797552,
+ "learning_rate": 8.132926837169676e-06,
+ "loss": 0.46009817719459534,
+ "step": 1582
+ },
+ {
+ "epoch": 0.8305351521511017,
+ "grad_norm": 0.1527913309957562,
+ "learning_rate": 8.11407591779537e-06,
+ "loss": 0.4513983726501465,
+ "step": 1583
+ },
+ {
+ "epoch": 0.8310598111227702,
+ "grad_norm": 0.1537090286922398,
+ "learning_rate": 8.095277665637061e-06,
+ "loss": 0.4537739157676697,
+ "step": 1584
+ },
+ {
+ "epoch": 0.8315844700944386,
+ "grad_norm": 0.15722962107654076,
+ "learning_rate": 8.076532131765408e-06,
+ "loss": 0.45232969522476196,
+ "step": 1585
+ },
+ {
+ "epoch": 0.832109129066107,
+ "grad_norm": 0.1450142396611787,
+ "learning_rate": 8.057839367107817e-06,
+ "loss": 0.4270634651184082,
+ "step": 1586
+ },
+ {
+ "epoch": 0.8326337880377754,
+ "grad_norm": 0.150264001570592,
+ "learning_rate": 8.039199422448355e-06,
+ "loss": 0.43563878536224365,
+ "step": 1587
+ },
+ {
+ "epoch": 0.8331584470094439,
+ "grad_norm": 0.14904943307590535,
+ "learning_rate": 8.020612348427572e-06,
+ "loss": 0.4401509165763855,
+ "step": 1588
+ },
+ {
+ "epoch": 0.8336831059811123,
+ "grad_norm": 0.15453821662860034,
+ "learning_rate": 8.002078195542393e-06,
+ "loss": 0.44545161724090576,
+ "step": 1589
+ },
+ {
+ "epoch": 0.8342077649527807,
+ "grad_norm": 0.1570809678830383,
+ "learning_rate": 7.983597014145951e-06,
+ "loss": 0.45228466391563416,
+ "step": 1590
+ },
+ {
+ "epoch": 0.8347324239244491,
+ "grad_norm": 0.14741240769629818,
+ "learning_rate": 7.965168854447493e-06,
+ "loss": 0.4485381245613098,
+ "step": 1591
+ },
+ {
+ "epoch": 0.8352570828961176,
+ "grad_norm": 0.15362267095533516,
+ "learning_rate": 7.946793766512195e-06,
+ "loss": 0.4358059763908386,
+ "step": 1592
+ },
+ {
+ "epoch": 0.835781741867786,
+ "grad_norm": 0.1544754586495605,
+ "learning_rate": 7.92847180026106e-06,
+ "loss": 0.43327778577804565,
+ "step": 1593
+ },
+ {
+ "epoch": 0.8363064008394544,
+ "grad_norm": 0.15023225026164963,
+ "learning_rate": 7.91020300547077e-06,
+ "loss": 0.44359731674194336,
+ "step": 1594
+ },
+ {
+ "epoch": 0.8368310598111228,
+ "grad_norm": 0.15665530392095345,
+ "learning_rate": 7.891987431773545e-06,
+ "loss": 0.44503387808799744,
+ "step": 1595
+ },
+ {
+ "epoch": 0.8373557187827911,
+ "grad_norm": 0.14717614926392292,
+ "learning_rate": 7.873825128657033e-06,
+ "loss": 0.45314061641693115,
+ "step": 1596
+ },
+ {
+ "epoch": 0.8378803777544596,
+ "grad_norm": 0.15207464934030612,
+ "learning_rate": 7.855716145464131e-06,
+ "loss": 0.4526442885398865,
+ "step": 1597
+ },
+ {
+ "epoch": 0.838405036726128,
+ "grad_norm": 0.15129881541794873,
+ "learning_rate": 7.837660531392904e-06,
+ "loss": 0.45127853751182556,
+ "step": 1598
+ },
+ {
+ "epoch": 0.8389296956977964,
+ "grad_norm": 0.15623362715417285,
+ "learning_rate": 7.819658335496405e-06,
+ "loss": 0.46432793140411377,
+ "step": 1599
+ },
+ {
+ "epoch": 0.8394543546694648,
+ "grad_norm": 0.1529219879642125,
+ "learning_rate": 7.801709606682573e-06,
+ "loss": 0.4475358724594116,
+ "step": 1600
+ },
+ {
+ "epoch": 0.8399790136411333,
+ "grad_norm": 0.14649139854087698,
+ "learning_rate": 7.78381439371409e-06,
+ "loss": 0.4363391101360321,
+ "step": 1601
+ },
+ {
+ "epoch": 0.8405036726128017,
+ "grad_norm": 0.14571233251948087,
+ "learning_rate": 7.765972745208243e-06,
+ "loss": 0.44419488310813904,
+ "step": 1602
+ },
+ {
+ "epoch": 0.8410283315844701,
+ "grad_norm": 0.15168460066406128,
+ "learning_rate": 7.748184709636791e-06,
+ "loss": 0.4433790445327759,
+ "step": 1603
+ },
+ {
+ "epoch": 0.8415529905561385,
+ "grad_norm": 0.14852860876513757,
+ "learning_rate": 7.730450335325853e-06,
+ "loss": 0.4493539333343506,
+ "step": 1604
+ },
+ {
+ "epoch": 0.842077649527807,
+ "grad_norm": 0.1505615491087377,
+ "learning_rate": 7.712769670455741e-06,
+ "loss": 0.4646318554878235,
+ "step": 1605
+ },
+ {
+ "epoch": 0.8426023084994754,
+ "grad_norm": 0.15407610486660114,
+ "learning_rate": 7.695142763060878e-06,
+ "loss": 0.4484049677848816,
+ "step": 1606
+ },
+ {
+ "epoch": 0.8431269674711438,
+ "grad_norm": 0.1521449709584124,
+ "learning_rate": 7.67756966102962e-06,
+ "loss": 0.43366891145706177,
+ "step": 1607
+ },
+ {
+ "epoch": 0.8436516264428122,
+ "grad_norm": 0.14747762647596166,
+ "learning_rate": 7.660050412104153e-06,
+ "loss": 0.43950092792510986,
+ "step": 1608
+ },
+ {
+ "epoch": 0.8441762854144806,
+ "grad_norm": 0.15512610809240027,
+ "learning_rate": 7.642585063880351e-06,
+ "loss": 0.46813055872917175,
+ "step": 1609
+ },
+ {
+ "epoch": 0.844700944386149,
+ "grad_norm": 0.15825480008734216,
+ "learning_rate": 7.625173663807655e-06,
+ "loss": 0.4348755478858948,
+ "step": 1610
+ },
+ {
+ "epoch": 0.8452256033578174,
+ "grad_norm": 0.1573968981854278,
+ "learning_rate": 7.6078162591889434e-06,
+ "loss": 0.45048123598098755,
+ "step": 1611
+ },
+ {
+ "epoch": 0.8457502623294858,
+ "grad_norm": 0.1517038960096417,
+ "learning_rate": 7.590512897180399e-06,
+ "loss": 0.43563610315322876,
+ "step": 1612
+ },
+ {
+ "epoch": 0.8462749213011542,
+ "grad_norm": 0.15503576044724665,
+ "learning_rate": 7.573263624791382e-06,
+ "loss": 0.4467726945877075,
+ "step": 1613
+ },
+ {
+ "epoch": 0.8467995802728226,
+ "grad_norm": 0.1560549927192375,
+ "learning_rate": 7.556068488884304e-06,
+ "loss": 0.4320477545261383,
+ "step": 1614
+ },
+ {
+ "epoch": 0.8473242392444911,
+ "grad_norm": 0.1679830356844654,
+ "learning_rate": 7.538927536174498e-06,
+ "loss": 0.4476034343242645,
+ "step": 1615
+ },
+ {
+ "epoch": 0.8478488982161595,
+ "grad_norm": 0.15169257383162213,
+ "learning_rate": 7.521840813230098e-06,
+ "loss": 0.4325854778289795,
+ "step": 1616
+ },
+ {
+ "epoch": 0.8483735571878279,
+ "grad_norm": 0.15980226324329827,
+ "learning_rate": 7.504808366471909e-06,
+ "loss": 0.4472721815109253,
+ "step": 1617
+ },
+ {
+ "epoch": 0.8488982161594963,
+ "grad_norm": 0.16932727986736365,
+ "learning_rate": 7.48783024217327e-06,
+ "loss": 0.45428287982940674,
+ "step": 1618
+ },
+ {
+ "epoch": 0.8494228751311648,
+ "grad_norm": 0.15065676306311376,
+ "learning_rate": 7.470906486459946e-06,
+ "loss": 0.44418883323669434,
+ "step": 1619
+ },
+ {
+ "epoch": 0.8499475341028332,
+ "grad_norm": 0.1533992998314183,
+ "learning_rate": 7.454037145309986e-06,
+ "loss": 0.43552401661872864,
+ "step": 1620
+ },
+ {
+ "epoch": 0.8504721930745016,
+ "grad_norm": 0.15479480515120017,
+ "learning_rate": 7.437222264553628e-06,
+ "loss": 0.444465696811676,
+ "step": 1621
+ },
+ {
+ "epoch": 0.85099685204617,
+ "grad_norm": 0.15845678337942112,
+ "learning_rate": 7.420461889873129e-06,
+ "loss": 0.44896578788757324,
+ "step": 1622
+ },
+ {
+ "epoch": 0.8515215110178384,
+ "grad_norm": 0.15038940533473208,
+ "learning_rate": 7.4037560668026804e-06,
+ "loss": 0.43062180280685425,
+ "step": 1623
+ },
+ {
+ "epoch": 0.8520461699895068,
+ "grad_norm": 0.1480017115109342,
+ "learning_rate": 7.387104840728259e-06,
+ "loss": 0.4630432724952698,
+ "step": 1624
+ },
+ {
+ "epoch": 0.8525708289611752,
+ "grad_norm": 0.15321305032824092,
+ "learning_rate": 7.370508256887514e-06,
+ "loss": 0.4419487714767456,
+ "step": 1625
+ },
+ {
+ "epoch": 0.8530954879328436,
+ "grad_norm": 0.1568246735699272,
+ "learning_rate": 7.353966360369659e-06,
+ "loss": 0.4483609199523926,
+ "step": 1626
+ },
+ {
+ "epoch": 0.853620146904512,
+ "grad_norm": 0.15539536583814875,
+ "learning_rate": 7.337479196115317e-06,
+ "loss": 0.4567753076553345,
+ "step": 1627
+ },
+ {
+ "epoch": 0.8541448058761805,
+ "grad_norm": 0.15528695146633797,
+ "learning_rate": 7.321046808916426e-06,
+ "loss": 0.4560334086418152,
+ "step": 1628
+ },
+ {
+ "epoch": 0.8546694648478489,
+ "grad_norm": 0.15005728554864303,
+ "learning_rate": 7.304669243416102e-06,
+ "loss": 0.43833765387535095,
+ "step": 1629
+ },
+ {
+ "epoch": 0.8551941238195173,
+ "grad_norm": 0.15959100645189242,
+ "learning_rate": 7.288346544108518e-06,
+ "loss": 0.4713831841945648,
+ "step": 1630
+ },
+ {
+ "epoch": 0.8557187827911857,
+ "grad_norm": 0.14898121463872632,
+ "learning_rate": 7.272078755338807e-06,
+ "loss": 0.44003745913505554,
+ "step": 1631
+ },
+ {
+ "epoch": 0.8562434417628542,
+ "grad_norm": 0.15083110802343333,
+ "learning_rate": 7.255865921302903e-06,
+ "loss": 0.429004043340683,
+ "step": 1632
+ },
+ {
+ "epoch": 0.8567681007345226,
+ "grad_norm": 0.1522757086796616,
+ "learning_rate": 7.239708086047445e-06,
+ "loss": 0.4403839707374573,
+ "step": 1633
+ },
+ {
+ "epoch": 0.857292759706191,
+ "grad_norm": 0.15719824838921245,
+ "learning_rate": 7.223605293469654e-06,
+ "loss": 0.4377821087837219,
+ "step": 1634
+ },
+ {
+ "epoch": 0.8578174186778594,
+ "grad_norm": 0.1450616202708373,
+ "learning_rate": 7.207557587317216e-06,
+ "loss": 0.4357878267765045,
+ "step": 1635
+ },
+ {
+ "epoch": 0.8583420776495279,
+ "grad_norm": 0.15703162220646427,
+ "learning_rate": 7.191565011188149e-06,
+ "loss": 0.44412243366241455,
+ "step": 1636
+ },
+ {
+ "epoch": 0.8588667366211962,
+ "grad_norm": 0.150175313334118,
+ "learning_rate": 7.175627608530716e-06,
+ "loss": 0.4543853998184204,
+ "step": 1637
+ },
+ {
+ "epoch": 0.8593913955928646,
+ "grad_norm": 0.15091711492716975,
+ "learning_rate": 7.159745422643261e-06,
+ "loss": 0.4268660545349121,
+ "step": 1638
+ },
+ {
+ "epoch": 0.859916054564533,
+ "grad_norm": 0.15449835654964011,
+ "learning_rate": 7.1439184966741295e-06,
+ "loss": 0.4507150650024414,
+ "step": 1639
+ },
+ {
+ "epoch": 0.8604407135362014,
+ "grad_norm": 0.15689612593442026,
+ "learning_rate": 7.128146873621533e-06,
+ "loss": 0.45520636439323425,
+ "step": 1640
+ },
+ {
+ "epoch": 0.8609653725078699,
+ "grad_norm": 0.1565226785234465,
+ "learning_rate": 7.11243059633345e-06,
+ "loss": 0.45102012157440186,
+ "step": 1641
+ },
+ {
+ "epoch": 0.8614900314795383,
+ "grad_norm": 0.15500877922230677,
+ "learning_rate": 7.096769707507481e-06,
+ "loss": 0.447115421295166,
+ "step": 1642
+ },
+ {
+ "epoch": 0.8620146904512067,
+ "grad_norm": 0.1595248428000243,
+ "learning_rate": 7.081164249690761e-06,
+ "loss": 0.4627235233783722,
+ "step": 1643
+ },
+ {
+ "epoch": 0.8625393494228751,
+ "grad_norm": 0.15958177510163488,
+ "learning_rate": 7.065614265279818e-06,
+ "loss": 0.4386388063430786,
+ "step": 1644
+ },
+ {
+ "epoch": 0.8630640083945436,
+ "grad_norm": 0.15152834825396286,
+ "learning_rate": 7.050119796520482e-06,
+ "loss": 0.4531296491622925,
+ "step": 1645
+ },
+ {
+ "epoch": 0.863588667366212,
+ "grad_norm": 0.15635733905345023,
+ "learning_rate": 7.034680885507752e-06,
+ "loss": 0.45573747158050537,
+ "step": 1646
+ },
+ {
+ "epoch": 0.8641133263378804,
+ "grad_norm": 0.15666918852633857,
+ "learning_rate": 7.019297574185703e-06,
+ "loss": 0.4499792158603668,
+ "step": 1647
+ },
+ {
+ "epoch": 0.8646379853095488,
+ "grad_norm": 0.156444627477174,
+ "learning_rate": 7.0039699043473405e-06,
+ "loss": 0.4545009136199951,
+ "step": 1648
+ },
+ {
+ "epoch": 0.8651626442812173,
+ "grad_norm": 0.14638989347446527,
+ "learning_rate": 6.988697917634515e-06,
+ "loss": 0.4397326409816742,
+ "step": 1649
+ },
+ {
+ "epoch": 0.8656873032528857,
+ "grad_norm": 0.15492813701343372,
+ "learning_rate": 6.973481655537793e-06,
+ "loss": 0.4402945041656494,
+ "step": 1650
+ },
+ {
+ "epoch": 0.866211962224554,
+ "grad_norm": 0.14466615385057716,
+ "learning_rate": 6.958321159396352e-06,
+ "loss": 0.4479280114173889,
+ "step": 1651
+ },
+ {
+ "epoch": 0.8667366211962224,
+ "grad_norm": 0.15865895330952856,
+ "learning_rate": 6.943216470397871e-06,
+ "loss": 0.4553218483924866,
+ "step": 1652
+ },
+ {
+ "epoch": 0.8672612801678908,
+ "grad_norm": 0.15293327582467067,
+ "learning_rate": 6.928167629578407e-06,
+ "loss": 0.44509196281433105,
+ "step": 1653
+ },
+ {
+ "epoch": 0.8677859391395593,
+ "grad_norm": 0.1515507946195896,
+ "learning_rate": 6.91317467782229e-06,
+ "loss": 0.4498351216316223,
+ "step": 1654
+ },
+ {
+ "epoch": 0.8683105981112277,
+ "grad_norm": 0.15313372799407085,
+ "learning_rate": 6.898237655862007e-06,
+ "loss": 0.4445887804031372,
+ "step": 1655
+ },
+ {
+ "epoch": 0.8688352570828961,
+ "grad_norm": 0.1491970050887588,
+ "learning_rate": 6.883356604278115e-06,
+ "loss": 0.4323808252811432,
+ "step": 1656
+ },
+ {
+ "epoch": 0.8693599160545645,
+ "grad_norm": 0.14449269685472518,
+ "learning_rate": 6.868531563499096e-06,
+ "loss": 0.42508652806282043,
+ "step": 1657
+ },
+ {
+ "epoch": 0.869884575026233,
+ "grad_norm": 0.14997328847830618,
+ "learning_rate": 6.853762573801264e-06,
+ "loss": 0.43655943870544434,
+ "step": 1658
+ },
+ {
+ "epoch": 0.8704092339979014,
+ "grad_norm": 0.15288208370561934,
+ "learning_rate": 6.839049675308662e-06,
+ "loss": 0.4505232274532318,
+ "step": 1659
+ },
+ {
+ "epoch": 0.8709338929695698,
+ "grad_norm": 0.14416201723141706,
+ "learning_rate": 6.824392907992938e-06,
+ "loss": 0.42444318532943726,
+ "step": 1660
+ },
+ {
+ "epoch": 0.8714585519412382,
+ "grad_norm": 0.14468507187223545,
+ "learning_rate": 6.809792311673244e-06,
+ "loss": 0.4251434803009033,
+ "step": 1661
+ },
+ {
+ "epoch": 0.8719832109129066,
+ "grad_norm": 0.15406295659798216,
+ "learning_rate": 6.795247926016142e-06,
+ "loss": 0.45068371295928955,
+ "step": 1662
+ },
+ {
+ "epoch": 0.8725078698845751,
+ "grad_norm": 0.1516721215663676,
+ "learning_rate": 6.780759790535465e-06,
+ "loss": 0.441638320684433,
+ "step": 1663
+ },
+ {
+ "epoch": 0.8730325288562435,
+ "grad_norm": 0.1537661139082594,
+ "learning_rate": 6.766327944592239e-06,
+ "loss": 0.44044923782348633,
+ "step": 1664
+ },
+ {
+ "epoch": 0.8735571878279118,
+ "grad_norm": 0.15175839438218855,
+ "learning_rate": 6.751952427394553e-06,
+ "loss": 0.44454342126846313,
+ "step": 1665
+ },
+ {
+ "epoch": 0.8740818467995802,
+ "grad_norm": 0.1531195282211392,
+ "learning_rate": 6.737633277997471e-06,
+ "loss": 0.4540765881538391,
+ "step": 1666
+ },
+ {
+ "epoch": 0.8746065057712487,
+ "grad_norm": 0.14961744003585584,
+ "learning_rate": 6.723370535302915e-06,
+ "loss": 0.44223541021347046,
+ "step": 1667
+ },
+ {
+ "epoch": 0.8751311647429171,
+ "grad_norm": 0.14816820299503658,
+ "learning_rate": 6.709164238059568e-06,
+ "loss": 0.43701833486557007,
+ "step": 1668
+ },
+ {
+ "epoch": 0.8756558237145855,
+ "grad_norm": 0.14622402294164918,
+ "learning_rate": 6.695014424862756e-06,
+ "loss": 0.42922335863113403,
+ "step": 1669
+ },
+ {
+ "epoch": 0.8761804826862539,
+ "grad_norm": 0.1547291062309004,
+ "learning_rate": 6.680921134154355e-06,
+ "loss": 0.4233800768852234,
+ "step": 1670
+ },
+ {
+ "epoch": 0.8767051416579223,
+ "grad_norm": 0.15378903523440587,
+ "learning_rate": 6.666884404222678e-06,
+ "loss": 0.4432950019836426,
+ "step": 1671
+ },
+ {
+ "epoch": 0.8772298006295908,
+ "grad_norm": 0.15973037860999773,
+ "learning_rate": 6.652904273202384e-06,
+ "loss": 0.4560699462890625,
+ "step": 1672
+ },
+ {
+ "epoch": 0.8777544596012592,
+ "grad_norm": 0.15430233088028492,
+ "learning_rate": 6.638980779074358e-06,
+ "loss": 0.45776498317718506,
+ "step": 1673
+ },
+ {
+ "epoch": 0.8782791185729276,
+ "grad_norm": 0.1587664730992995,
+ "learning_rate": 6.625113959665614e-06,
+ "loss": 0.43579810857772827,
+ "step": 1674
+ },
+ {
+ "epoch": 0.878803777544596,
+ "grad_norm": 0.14370919039693034,
+ "learning_rate": 6.6113038526491946e-06,
+ "loss": 0.4426189064979553,
+ "step": 1675
+ },
+ {
+ "epoch": 0.8793284365162645,
+ "grad_norm": 0.14933009355373217,
+ "learning_rate": 6.59755049554407e-06,
+ "loss": 0.44747287034988403,
+ "step": 1676
+ },
+ {
+ "epoch": 0.8798530954879329,
+ "grad_norm": 0.15713697891738065,
+ "learning_rate": 6.583853925715039e-06,
+ "loss": 0.46320077776908875,
+ "step": 1677
+ },
+ {
+ "epoch": 0.8803777544596013,
+ "grad_norm": 0.1542096833927892,
+ "learning_rate": 6.570214180372609e-06,
+ "loss": 0.4473888874053955,
+ "step": 1678
+ },
+ {
+ "epoch": 0.8809024134312696,
+ "grad_norm": 0.16187259968193946,
+ "learning_rate": 6.556631296572922e-06,
+ "loss": 0.4421869218349457,
+ "step": 1679
+ },
+ {
+ "epoch": 0.881427072402938,
+ "grad_norm": 0.15978032996583555,
+ "learning_rate": 6.54310531121762e-06,
+ "loss": 0.4305853843688965,
+ "step": 1680
+ },
+ {
+ "epoch": 0.8819517313746065,
+ "grad_norm": 0.15631580579775123,
+ "learning_rate": 6.529636261053784e-06,
+ "loss": 0.4340512454509735,
+ "step": 1681
+ },
+ {
+ "epoch": 0.8824763903462749,
+ "grad_norm": 0.150084161390103,
+ "learning_rate": 6.516224182673812e-06,
+ "loss": 0.4383493661880493,
+ "step": 1682
+ },
+ {
+ "epoch": 0.8830010493179433,
+ "grad_norm": 0.15527917215838966,
+ "learning_rate": 6.502869112515314e-06,
+ "loss": 0.45441877841949463,
+ "step": 1683
+ },
+ {
+ "epoch": 0.8835257082896117,
+ "grad_norm": 0.15164994594035522,
+ "learning_rate": 6.489571086861027e-06,
+ "loss": 0.43363988399505615,
+ "step": 1684
+ },
+ {
+ "epoch": 0.8840503672612802,
+ "grad_norm": 0.1467230535452683,
+ "learning_rate": 6.47633014183871e-06,
+ "loss": 0.4263266623020172,
+ "step": 1685
+ },
+ {
+ "epoch": 0.8845750262329486,
+ "grad_norm": 0.14597001769185472,
+ "learning_rate": 6.463146313421044e-06,
+ "loss": 0.4223391115665436,
+ "step": 1686
+ },
+ {
+ "epoch": 0.885099685204617,
+ "grad_norm": 0.15050446002714588,
+ "learning_rate": 6.4500196374255435e-06,
+ "loss": 0.43097931146621704,
+ "step": 1687
+ },
+ {
+ "epoch": 0.8856243441762854,
+ "grad_norm": 0.14918033614543294,
+ "learning_rate": 6.436950149514448e-06,
+ "loss": 0.4587080478668213,
+ "step": 1688
+ },
+ {
+ "epoch": 0.8861490031479539,
+ "grad_norm": 0.15469492243631364,
+ "learning_rate": 6.423937885194631e-06,
+ "loss": 0.45661866664886475,
+ "step": 1689
+ },
+ {
+ "epoch": 0.8866736621196223,
+ "grad_norm": 0.15552798640177423,
+ "learning_rate": 6.410982879817498e-06,
+ "loss": 0.44370236992836,
+ "step": 1690
+ },
+ {
+ "epoch": 0.8871983210912907,
+ "grad_norm": 0.15595593365088142,
+ "learning_rate": 6.398085168578901e-06,
+ "loss": 0.4291878342628479,
+ "step": 1691
+ },
+ {
+ "epoch": 0.887722980062959,
+ "grad_norm": 0.14708046948755318,
+ "learning_rate": 6.385244786519036e-06,
+ "loss": 0.4422011375427246,
+ "step": 1692
+ },
+ {
+ "epoch": 0.8882476390346274,
+ "grad_norm": 0.1499253382866992,
+ "learning_rate": 6.3724617685223404e-06,
+ "loss": 0.4595015048980713,
+ "step": 1693
+ },
+ {
+ "epoch": 0.8887722980062959,
+ "grad_norm": 0.14702868569252941,
+ "learning_rate": 6.3597361493174205e-06,
+ "loss": 0.4347558319568634,
+ "step": 1694
+ },
+ {
+ "epoch": 0.8892969569779643,
+ "grad_norm": 0.14977289247327386,
+ "learning_rate": 6.347067963476928e-06,
+ "loss": 0.43877455592155457,
+ "step": 1695
+ },
+ {
+ "epoch": 0.8898216159496327,
+ "grad_norm": 0.1453902077891248,
+ "learning_rate": 6.334457245417491e-06,
+ "loss": 0.4282415807247162,
+ "step": 1696
+ },
+ {
+ "epoch": 0.8903462749213011,
+ "grad_norm": 0.1545933676895574,
+ "learning_rate": 6.321904029399607e-06,
+ "loss": 0.4429236054420471,
+ "step": 1697
+ },
+ {
+ "epoch": 0.8908709338929696,
+ "grad_norm": 0.15625844051167467,
+ "learning_rate": 6.3094083495275565e-06,
+ "loss": 0.4284421503543854,
+ "step": 1698
+ },
+ {
+ "epoch": 0.891395592864638,
+ "grad_norm": 0.1484963883381567,
+ "learning_rate": 6.2969702397493e-06,
+ "loss": 0.4545463025569916,
+ "step": 1699
+ },
+ {
+ "epoch": 0.8919202518363064,
+ "grad_norm": 0.15122402806869387,
+ "learning_rate": 6.284589733856403e-06,
+ "loss": 0.43840622901916504,
+ "step": 1700
+ },
+ {
+ "epoch": 0.8924449108079748,
+ "grad_norm": 0.14889313167657334,
+ "learning_rate": 6.27226686548392e-06,
+ "loss": 0.4451153874397278,
+ "step": 1701
+ },
+ {
+ "epoch": 0.8929695697796433,
+ "grad_norm": 0.14724940941036385,
+ "learning_rate": 6.260001668110331e-06,
+ "loss": 0.434165358543396,
+ "step": 1702
+ },
+ {
+ "epoch": 0.8934942287513117,
+ "grad_norm": 0.15520838093295836,
+ "learning_rate": 6.247794175057433e-06,
+ "loss": 0.44641873240470886,
+ "step": 1703
+ },
+ {
+ "epoch": 0.8940188877229801,
+ "grad_norm": 0.159167904797364,
+ "learning_rate": 6.235644419490253e-06,
+ "loss": 0.44591957330703735,
+ "step": 1704
+ },
+ {
+ "epoch": 0.8945435466946485,
+ "grad_norm": 0.14843235881370814,
+ "learning_rate": 6.223552434416951e-06,
+ "loss": 0.43010181188583374,
+ "step": 1705
+ },
+ {
+ "epoch": 0.8950682056663168,
+ "grad_norm": 0.1590998153483183,
+ "learning_rate": 6.211518252688749e-06,
+ "loss": 0.4602118134498596,
+ "step": 1706
+ },
+ {
+ "epoch": 0.8955928646379853,
+ "grad_norm": 0.15257025539137767,
+ "learning_rate": 6.19954190699982e-06,
+ "loss": 0.45429927110671997,
+ "step": 1707
+ },
+ {
+ "epoch": 0.8961175236096537,
+ "grad_norm": 0.15304035006408193,
+ "learning_rate": 6.18762342988722e-06,
+ "loss": 0.4405020773410797,
+ "step": 1708
+ },
+ {
+ "epoch": 0.8966421825813221,
+ "grad_norm": 0.15956303033928781,
+ "learning_rate": 6.175762853730777e-06,
+ "loss": 0.4455980062484741,
+ "step": 1709
+ },
+ {
+ "epoch": 0.8971668415529905,
+ "grad_norm": 0.14973831573751925,
+ "learning_rate": 6.163960210753025e-06,
+ "loss": 0.4331505298614502,
+ "step": 1710
+ },
+ {
+ "epoch": 0.897691500524659,
+ "grad_norm": 0.14573687980573624,
+ "learning_rate": 6.1522155330190985e-06,
+ "loss": 0.43307584524154663,
+ "step": 1711
+ },
+ {
+ "epoch": 0.8982161594963274,
+ "grad_norm": 0.1556523345686865,
+ "learning_rate": 6.140528852436665e-06,
+ "loss": 0.4531600773334503,
+ "step": 1712
+ },
+ {
+ "epoch": 0.8987408184679958,
+ "grad_norm": 0.14262442470666953,
+ "learning_rate": 6.128900200755817e-06,
+ "loss": 0.4350634813308716,
+ "step": 1713
+ },
+ {
+ "epoch": 0.8992654774396642,
+ "grad_norm": 0.14709392307082428,
+ "learning_rate": 6.117329609568998e-06,
+ "loss": 0.43465620279312134,
+ "step": 1714
+ },
+ {
+ "epoch": 0.8997901364113327,
+ "grad_norm": 0.15648040137428026,
+ "learning_rate": 6.105817110310909e-06,
+ "loss": 0.4344560503959656,
+ "step": 1715
+ },
+ {
+ "epoch": 0.9003147953830011,
+ "grad_norm": 0.15758204487742147,
+ "learning_rate": 6.094362734258438e-06,
+ "loss": 0.4409951865673065,
+ "step": 1716
+ },
+ {
+ "epoch": 0.9008394543546695,
+ "grad_norm": 0.15755547382189733,
+ "learning_rate": 6.082966512530558e-06,
+ "loss": 0.44050097465515137,
+ "step": 1717
+ },
+ {
+ "epoch": 0.9013641133263379,
+ "grad_norm": 0.1536165045521179,
+ "learning_rate": 6.071628476088257e-06,
+ "loss": 0.43460220098495483,
+ "step": 1718
+ },
+ {
+ "epoch": 0.9018887722980063,
+ "grad_norm": 0.16253540769141986,
+ "learning_rate": 6.060348655734439e-06,
+ "loss": 0.4490512013435364,
+ "step": 1719
+ },
+ {
+ "epoch": 0.9024134312696747,
+ "grad_norm": 0.15003242262029914,
+ "learning_rate": 6.049127082113853e-06,
+ "loss": 0.44427937269210815,
+ "step": 1720
+ },
+ {
+ "epoch": 0.9029380902413431,
+ "grad_norm": 0.1579908758627471,
+ "learning_rate": 6.037963785712995e-06,
+ "loss": 0.4421137571334839,
+ "step": 1721
+ },
+ {
+ "epoch": 0.9034627492130115,
+ "grad_norm": 0.15111957361250286,
+ "learning_rate": 6.026858796860048e-06,
+ "loss": 0.44011324644088745,
+ "step": 1722
+ },
+ {
+ "epoch": 0.9039874081846799,
+ "grad_norm": 0.1446470746259905,
+ "learning_rate": 6.0158121457247845e-06,
+ "loss": 0.43798986077308655,
+ "step": 1723
+ },
+ {
+ "epoch": 0.9045120671563484,
+ "grad_norm": 0.14882152623490427,
+ "learning_rate": 6.004823862318474e-06,
+ "loss": 0.4265901446342468,
+ "step": 1724
+ },
+ {
+ "epoch": 0.9050367261280168,
+ "grad_norm": 0.1506889859372045,
+ "learning_rate": 5.993893976493831e-06,
+ "loss": 0.4503232538700104,
+ "step": 1725
+ },
+ {
+ "epoch": 0.9055613850996852,
+ "grad_norm": 0.16757194686045146,
+ "learning_rate": 5.983022517944901e-06,
+ "loss": 0.4509217441082001,
+ "step": 1726
+ },
+ {
+ "epoch": 0.9060860440713536,
+ "grad_norm": 0.14447292841509515,
+ "learning_rate": 5.972209516207007e-06,
+ "loss": 0.4476190507411957,
+ "step": 1727
+ },
+ {
+ "epoch": 0.906610703043022,
+ "grad_norm": 0.1536564780391058,
+ "learning_rate": 5.961455000656656e-06,
+ "loss": 0.45448166131973267,
+ "step": 1728
+ },
+ {
+ "epoch": 0.9071353620146905,
+ "grad_norm": 0.15436323862710438,
+ "learning_rate": 5.950759000511455e-06,
+ "loss": 0.45128491520881653,
+ "step": 1729
+ },
+ {
+ "epoch": 0.9076600209863589,
+ "grad_norm": 0.15379817460186834,
+ "learning_rate": 5.940121544830047e-06,
+ "loss": 0.4490806460380554,
+ "step": 1730
+ },
+ {
+ "epoch": 0.9081846799580273,
+ "grad_norm": 0.1470796082413195,
+ "learning_rate": 5.929542662512013e-06,
+ "loss": 0.4371582865715027,
+ "step": 1731
+ },
+ {
+ "epoch": 0.9087093389296957,
+ "grad_norm": 0.15107600076433048,
+ "learning_rate": 5.919022382297809e-06,
+ "loss": 0.45175206661224365,
+ "step": 1732
+ },
+ {
+ "epoch": 0.9092339979013642,
+ "grad_norm": 0.14972329317276084,
+ "learning_rate": 5.908560732768684e-06,
+ "loss": 0.43177497386932373,
+ "step": 1733
+ },
+ {
+ "epoch": 0.9097586568730325,
+ "grad_norm": 0.14437715003340176,
+ "learning_rate": 5.898157742346601e-06,
+ "loss": 0.4243622422218323,
+ "step": 1734
+ },
+ {
+ "epoch": 0.9102833158447009,
+ "grad_norm": 0.15200146823831148,
+ "learning_rate": 5.8878134392941465e-06,
+ "loss": 0.47819408774375916,
+ "step": 1735
+ },
+ {
+ "epoch": 0.9108079748163693,
+ "grad_norm": 0.156864616244505,
+ "learning_rate": 5.877527851714479e-06,
+ "loss": 0.44744759798049927,
+ "step": 1736
+ },
+ {
+ "epoch": 0.9113326337880377,
+ "grad_norm": 0.14489667207174767,
+ "learning_rate": 5.867301007551239e-06,
+ "loss": 0.4390789866447449,
+ "step": 1737
+ },
+ {
+ "epoch": 0.9118572927597062,
+ "grad_norm": 0.1528011587102748,
+ "learning_rate": 5.857132934588479e-06,
+ "loss": 0.4338226616382599,
+ "step": 1738
+ },
+ {
+ "epoch": 0.9123819517313746,
+ "grad_norm": 0.14716603285971375,
+ "learning_rate": 5.847023660450568e-06,
+ "loss": 0.43975532054901123,
+ "step": 1739
+ },
+ {
+ "epoch": 0.912906610703043,
+ "grad_norm": 0.15266609903803277,
+ "learning_rate": 5.836973212602143e-06,
+ "loss": 0.449812650680542,
+ "step": 1740
+ },
+ {
+ "epoch": 0.9134312696747114,
+ "grad_norm": 0.1466770441701569,
+ "learning_rate": 5.826981618348025e-06,
+ "loss": 0.43771812319755554,
+ "step": 1741
+ },
+ {
+ "epoch": 0.9139559286463799,
+ "grad_norm": 0.1428112225146131,
+ "learning_rate": 5.8170489048331305e-06,
+ "loss": 0.42403268814086914,
+ "step": 1742
+ },
+ {
+ "epoch": 0.9144805876180483,
+ "grad_norm": 0.1550829416480759,
+ "learning_rate": 5.8071750990424275e-06,
+ "loss": 0.45707154273986816,
+ "step": 1743
+ },
+ {
+ "epoch": 0.9150052465897167,
+ "grad_norm": 0.14687341315819044,
+ "learning_rate": 5.797360227800832e-06,
+ "loss": 0.4558846354484558,
+ "step": 1744
+ },
+ {
+ "epoch": 0.9155299055613851,
+ "grad_norm": 0.15660848922413237,
+ "learning_rate": 5.787604317773154e-06,
+ "loss": 0.44760435819625854,
+ "step": 1745
+ },
+ {
+ "epoch": 0.9160545645330536,
+ "grad_norm": 0.14891742746821762,
+ "learning_rate": 5.777907395464015e-06,
+ "loss": 0.4266620874404907,
+ "step": 1746
+ },
+ {
+ "epoch": 0.916579223504722,
+ "grad_norm": 0.14501110957005472,
+ "learning_rate": 5.768269487217777e-06,
+ "loss": 0.4388393759727478,
+ "step": 1747
+ },
+ {
+ "epoch": 0.9171038824763903,
+ "grad_norm": 0.14651438673799927,
+ "learning_rate": 5.758690619218493e-06,
+ "loss": 0.4416884183883667,
+ "step": 1748
+ },
+ {
+ "epoch": 0.9176285414480587,
+ "grad_norm": 0.14785104880890787,
+ "learning_rate": 5.749170817489792e-06,
+ "loss": 0.4368356466293335,
+ "step": 1749
+ },
+ {
+ "epoch": 0.9181532004197271,
+ "grad_norm": 0.16586567883267808,
+ "learning_rate": 5.739710107894847e-06,
+ "loss": 0.44294649362564087,
+ "step": 1750
+ },
+ {
+ "epoch": 0.9186778593913956,
+ "grad_norm": 0.14377347738227095,
+ "learning_rate": 5.730308516136288e-06,
+ "loss": 0.4405091106891632,
+ "step": 1751
+ },
+ {
+ "epoch": 0.919202518363064,
+ "grad_norm": 0.14703077660050376,
+ "learning_rate": 5.72096606775613e-06,
+ "loss": 0.44809678196907043,
+ "step": 1752
+ },
+ {
+ "epoch": 0.9197271773347324,
+ "grad_norm": 0.15073590419140934,
+ "learning_rate": 5.711682788135726e-06,
+ "loss": 0.43815505504608154,
+ "step": 1753
+ },
+ {
+ "epoch": 0.9202518363064008,
+ "grad_norm": 0.14774775313473323,
+ "learning_rate": 5.702458702495658e-06,
+ "loss": 0.4417206048965454,
+ "step": 1754
+ },
+ {
+ "epoch": 0.9207764952780693,
+ "grad_norm": 0.14289335805130654,
+ "learning_rate": 5.693293835895706e-06,
+ "loss": 0.42886218428611755,
+ "step": 1755
+ },
+ {
+ "epoch": 0.9213011542497377,
+ "grad_norm": 0.15571224276102627,
+ "learning_rate": 5.684188213234763e-06,
+ "loss": 0.44488853216171265,
+ "step": 1756
+ },
+ {
+ "epoch": 0.9218258132214061,
+ "grad_norm": 0.147115948090654,
+ "learning_rate": 5.675141859250763e-06,
+ "loss": 0.44682830572128296,
+ "step": 1757
+ },
+ {
+ "epoch": 0.9223504721930745,
+ "grad_norm": 0.15349533122985531,
+ "learning_rate": 5.666154798520627e-06,
+ "loss": 0.45550864934921265,
+ "step": 1758
+ },
+ {
+ "epoch": 0.922875131164743,
+ "grad_norm": 0.16960595806748266,
+ "learning_rate": 5.657227055460184e-06,
+ "loss": 0.4605559706687927,
+ "step": 1759
+ },
+ {
+ "epoch": 0.9233997901364114,
+ "grad_norm": 0.14665490922506455,
+ "learning_rate": 5.648358654324119e-06,
+ "loss": 0.45442628860473633,
+ "step": 1760
+ },
+ {
+ "epoch": 0.9239244491080797,
+ "grad_norm": 0.14780665809126886,
+ "learning_rate": 5.6395496192058905e-06,
+ "loss": 0.45085081458091736,
+ "step": 1761
+ },
+ {
+ "epoch": 0.9244491080797481,
+ "grad_norm": 0.15599036700169416,
+ "learning_rate": 5.6307999740376714e-06,
+ "loss": 0.4581005573272705,
+ "step": 1762
+ },
+ {
+ "epoch": 0.9249737670514165,
+ "grad_norm": 0.14383027603716222,
+ "learning_rate": 5.622109742590294e-06,
+ "loss": 0.43986010551452637,
+ "step": 1763
+ },
+ {
+ "epoch": 0.925498426023085,
+ "grad_norm": 0.1510369649046927,
+ "learning_rate": 5.613478948473166e-06,
+ "loss": 0.4317719340324402,
+ "step": 1764
+ },
+ {
+ "epoch": 0.9260230849947534,
+ "grad_norm": 0.16608560055258376,
+ "learning_rate": 5.60490761513423e-06,
+ "loss": 0.4500439167022705,
+ "step": 1765
+ },
+ {
+ "epoch": 0.9265477439664218,
+ "grad_norm": 0.15364415343465948,
+ "learning_rate": 5.596395765859875e-06,
+ "loss": 0.43703922629356384,
+ "step": 1766
+ },
+ {
+ "epoch": 0.9270724029380902,
+ "grad_norm": 0.145164767782082,
+ "learning_rate": 5.587943423774886e-06,
+ "loss": 0.4389477074146271,
+ "step": 1767
+ },
+ {
+ "epoch": 0.9275970619097587,
+ "grad_norm": 0.15044099308887526,
+ "learning_rate": 5.579550611842395e-06,
+ "loss": 0.43600398302078247,
+ "step": 1768
+ },
+ {
+ "epoch": 0.9281217208814271,
+ "grad_norm": 0.14694381874423756,
+ "learning_rate": 5.571217352863786e-06,
+ "loss": 0.4525601863861084,
+ "step": 1769
+ },
+ {
+ "epoch": 0.9286463798530955,
+ "grad_norm": 0.15979616589190854,
+ "learning_rate": 5.562943669478661e-06,
+ "loss": 0.44730353355407715,
+ "step": 1770
+ },
+ {
+ "epoch": 0.9291710388247639,
+ "grad_norm": 0.1545596231710442,
+ "learning_rate": 5.554729584164762e-06,
+ "loss": 0.43867969512939453,
+ "step": 1771
+ },
+ {
+ "epoch": 0.9296956977964324,
+ "grad_norm": 0.15923930406000852,
+ "learning_rate": 5.5465751192379235e-06,
+ "loss": 0.4648996889591217,
+ "step": 1772
+ },
+ {
+ "epoch": 0.9302203567681008,
+ "grad_norm": 0.14001241391446362,
+ "learning_rate": 5.538480296852001e-06,
+ "loss": 0.41449663043022156,
+ "step": 1773
+ },
+ {
+ "epoch": 0.9307450157397692,
+ "grad_norm": 0.16538765001279418,
+ "learning_rate": 5.530445138998815e-06,
+ "loss": 0.4503074288368225,
+ "step": 1774
+ },
+ {
+ "epoch": 0.9312696747114375,
+ "grad_norm": 0.15007886784403335,
+ "learning_rate": 5.5224696675080904e-06,
+ "loss": 0.44717323780059814,
+ "step": 1775
+ },
+ {
+ "epoch": 0.9317943336831059,
+ "grad_norm": 0.17834856046133146,
+ "learning_rate": 5.514553904047395e-06,
+ "loss": 0.4374881386756897,
+ "step": 1776
+ },
+ {
+ "epoch": 0.9323189926547744,
+ "grad_norm": 0.1494478926649096,
+ "learning_rate": 5.5066978701220865e-06,
+ "loss": 0.4464038610458374,
+ "step": 1777
+ },
+ {
+ "epoch": 0.9328436516264428,
+ "grad_norm": 0.14887448793632826,
+ "learning_rate": 5.498901587075253e-06,
+ "loss": 0.443928062915802,
+ "step": 1778
+ },
+ {
+ "epoch": 0.9333683105981112,
+ "grad_norm": 0.14192884484464632,
+ "learning_rate": 5.491165076087652e-06,
+ "loss": 0.42002278566360474,
+ "step": 1779
+ },
+ {
+ "epoch": 0.9338929695697796,
+ "grad_norm": 0.14783356724791524,
+ "learning_rate": 5.483488358177647e-06,
+ "loss": 0.43647563457489014,
+ "step": 1780
+ },
+ {
+ "epoch": 0.934417628541448,
+ "grad_norm": 0.15004597148119703,
+ "learning_rate": 5.475871454201163e-06,
+ "loss": 0.4396110773086548,
+ "step": 1781
+ },
+ {
+ "epoch": 0.9349422875131165,
+ "grad_norm": 0.15668918073856763,
+ "learning_rate": 5.468314384851626e-06,
+ "loss": 0.42848116159439087,
+ "step": 1782
+ },
+ {
+ "epoch": 0.9354669464847849,
+ "grad_norm": 0.1449937199040528,
+ "learning_rate": 5.460817170659901e-06,
+ "loss": 0.44262629747390747,
+ "step": 1783
+ },
+ {
+ "epoch": 0.9359916054564533,
+ "grad_norm": 0.15868260486048544,
+ "learning_rate": 5.4533798319942375e-06,
+ "loss": 0.44874119758605957,
+ "step": 1784
+ },
+ {
+ "epoch": 0.9365162644281217,
+ "grad_norm": 0.15004870783258292,
+ "learning_rate": 5.446002389060224e-06,
+ "loss": 0.4575575590133667,
+ "step": 1785
+ },
+ {
+ "epoch": 0.9370409233997902,
+ "grad_norm": 0.14830248639010987,
+ "learning_rate": 5.438684861900721e-06,
+ "loss": 0.44039857387542725,
+ "step": 1786
+ },
+ {
+ "epoch": 0.9375655823714586,
+ "grad_norm": 0.15429570419507257,
+ "learning_rate": 5.431427270395808e-06,
+ "loss": 0.43363744020462036,
+ "step": 1787
+ },
+ {
+ "epoch": 0.938090241343127,
+ "grad_norm": 0.15982575395123055,
+ "learning_rate": 5.4242296342627354e-06,
+ "loss": 0.4484976530075073,
+ "step": 1788
+ },
+ {
+ "epoch": 0.9386149003147953,
+ "grad_norm": 0.1466121447285136,
+ "learning_rate": 5.417091973055878e-06,
+ "loss": 0.44059592485427856,
+ "step": 1789
+ },
+ {
+ "epoch": 0.9391395592864638,
+ "grad_norm": 0.15353848724097463,
+ "learning_rate": 5.410014306166657e-06,
+ "loss": 0.4600589871406555,
+ "step": 1790
+ },
+ {
+ "epoch": 0.9396642182581322,
+ "grad_norm": 0.14968071650065867,
+ "learning_rate": 5.4029966528235065e-06,
+ "loss": 0.4415352940559387,
+ "step": 1791
+ },
+ {
+ "epoch": 0.9401888772298006,
+ "grad_norm": 0.14819069534768417,
+ "learning_rate": 5.396039032091826e-06,
+ "loss": 0.4436836242675781,
+ "step": 1792
+ },
+ {
+ "epoch": 0.940713536201469,
+ "grad_norm": 0.14683654708689367,
+ "learning_rate": 5.389141462873908e-06,
+ "loss": 0.4243587851524353,
+ "step": 1793
+ },
+ {
+ "epoch": 0.9412381951731374,
+ "grad_norm": 0.15431293103561666,
+ "learning_rate": 5.382303963908908e-06,
+ "loss": 0.450438916683197,
+ "step": 1794
+ },
+ {
+ "epoch": 0.9417628541448059,
+ "grad_norm": 0.15125346717879914,
+ "learning_rate": 5.3755265537727775e-06,
+ "loss": 0.4469110369682312,
+ "step": 1795
+ },
+ {
+ "epoch": 0.9422875131164743,
+ "grad_norm": 0.15430864502191044,
+ "learning_rate": 5.368809250878227e-06,
+ "loss": 0.45182669162750244,
+ "step": 1796
+ },
+ {
+ "epoch": 0.9428121720881427,
+ "grad_norm": 0.1515431753176237,
+ "learning_rate": 5.362152073474657e-06,
+ "loss": 0.43683797121047974,
+ "step": 1797
+ },
+ {
+ "epoch": 0.9433368310598111,
+ "grad_norm": 0.1524915522484605,
+ "learning_rate": 5.355555039648133e-06,
+ "loss": 0.44641250371932983,
+ "step": 1798
+ },
+ {
+ "epoch": 0.9438614900314796,
+ "grad_norm": 0.15722684273308743,
+ "learning_rate": 5.3490181673213206e-06,
+ "loss": 0.43558835983276367,
+ "step": 1799
+ },
+ {
+ "epoch": 0.944386149003148,
+ "grad_norm": 0.15527272713567492,
+ "learning_rate": 5.342541474253438e-06,
+ "loss": 0.4423788785934448,
+ "step": 1800
+ },
+ {
+ "epoch": 0.9449108079748164,
+ "grad_norm": 0.1578831929456481,
+ "learning_rate": 5.336124978040208e-06,
+ "loss": 0.4506484866142273,
+ "step": 1801
+ },
+ {
+ "epoch": 0.9454354669464848,
+ "grad_norm": 0.14812076004561775,
+ "learning_rate": 5.329768696113815e-06,
+ "loss": 0.4432883858680725,
+ "step": 1802
+ },
+ {
+ "epoch": 0.9459601259181531,
+ "grad_norm": 0.1511327211030653,
+ "learning_rate": 5.323472645742854e-06,
+ "loss": 0.4321002960205078,
+ "step": 1803
+ },
+ {
+ "epoch": 0.9464847848898216,
+ "grad_norm": 0.1531668263170348,
+ "learning_rate": 5.3172368440322854e-06,
+ "loss": 0.44666337966918945,
+ "step": 1804
+ },
+ {
+ "epoch": 0.94700944386149,
+ "grad_norm": 0.14571994558609092,
+ "learning_rate": 5.311061307923387e-06,
+ "loss": 0.4436221420764923,
+ "step": 1805
+ },
+ {
+ "epoch": 0.9475341028331584,
+ "grad_norm": 0.15158206242691336,
+ "learning_rate": 5.304946054193706e-06,
+ "loss": 0.45979544520378113,
+ "step": 1806
+ },
+ {
+ "epoch": 0.9480587618048268,
+ "grad_norm": 0.15374308675707662,
+ "learning_rate": 5.298891099457018e-06,
+ "loss": 0.44095706939697266,
+ "step": 1807
+ },
+ {
+ "epoch": 0.9485834207764953,
+ "grad_norm": 0.14777162806829072,
+ "learning_rate": 5.292896460163283e-06,
+ "loss": 0.44000717997550964,
+ "step": 1808
+ },
+ {
+ "epoch": 0.9491080797481637,
+ "grad_norm": 0.15416014526696098,
+ "learning_rate": 5.286962152598589e-06,
+ "loss": 0.4367929697036743,
+ "step": 1809
+ },
+ {
+ "epoch": 0.9496327387198321,
+ "grad_norm": 0.1586289813390648,
+ "learning_rate": 5.281088192885125e-06,
+ "loss": 0.44338831305503845,
+ "step": 1810
+ },
+ {
+ "epoch": 0.9501573976915005,
+ "grad_norm": 0.1480189398685719,
+ "learning_rate": 5.275274596981124e-06,
+ "loss": 0.44127213954925537,
+ "step": 1811
+ },
+ {
+ "epoch": 0.950682056663169,
+ "grad_norm": 0.15893037852856393,
+ "learning_rate": 5.269521380680823e-06,
+ "loss": 0.43447786569595337,
+ "step": 1812
+ },
+ {
+ "epoch": 0.9512067156348374,
+ "grad_norm": 0.15257675815194532,
+ "learning_rate": 5.2638285596144236e-06,
+ "loss": 0.4478892385959625,
+ "step": 1813
+ },
+ {
+ "epoch": 0.9517313746065058,
+ "grad_norm": 0.14985387736695718,
+ "learning_rate": 5.258196149248046e-06,
+ "loss": 0.4517829418182373,
+ "step": 1814
+ },
+ {
+ "epoch": 0.9522560335781742,
+ "grad_norm": 0.14441329680781126,
+ "learning_rate": 5.252624164883689e-06,
+ "loss": 0.4337430000305176,
+ "step": 1815
+ },
+ {
+ "epoch": 0.9527806925498427,
+ "grad_norm": 0.15113217761739892,
+ "learning_rate": 5.247112621659186e-06,
+ "loss": 0.44544464349746704,
+ "step": 1816
+ },
+ {
+ "epoch": 0.953305351521511,
+ "grad_norm": 0.15511548841769227,
+ "learning_rate": 5.241661534548169e-06,
+ "loss": 0.4420471787452698,
+ "step": 1817
+ },
+ {
+ "epoch": 0.9538300104931794,
+ "grad_norm": 0.141474135090229,
+ "learning_rate": 5.236270918360016e-06,
+ "loss": 0.4323197603225708,
+ "step": 1818
+ },
+ {
+ "epoch": 0.9543546694648478,
+ "grad_norm": 0.14741159689952743,
+ "learning_rate": 5.23094078773983e-06,
+ "loss": 0.4486733675003052,
+ "step": 1819
+ },
+ {
+ "epoch": 0.9548793284365162,
+ "grad_norm": 0.14938242936939558,
+ "learning_rate": 5.225671157168377e-06,
+ "loss": 0.4469614624977112,
+ "step": 1820
+ },
+ {
+ "epoch": 0.9554039874081847,
+ "grad_norm": 0.14960446458046617,
+ "learning_rate": 5.220462040962068e-06,
+ "loss": 0.4371514320373535,
+ "step": 1821
+ },
+ {
+ "epoch": 0.9559286463798531,
+ "grad_norm": 0.1453826830908917,
+ "learning_rate": 5.215313453272906e-06,
+ "loss": 0.43045082688331604,
+ "step": 1822
+ },
+ {
+ "epoch": 0.9564533053515215,
+ "grad_norm": 0.14561549586732817,
+ "learning_rate": 5.210225408088447e-06,
+ "loss": 0.44164514541625977,
+ "step": 1823
+ },
+ {
+ "epoch": 0.9569779643231899,
+ "grad_norm": 0.15381074610209994,
+ "learning_rate": 5.205197919231773e-06,
+ "loss": 0.45496129989624023,
+ "step": 1824
+ },
+ {
+ "epoch": 0.9575026232948584,
+ "grad_norm": 0.14729572792188594,
+ "learning_rate": 5.200231000361448e-06,
+ "loss": 0.4330272078514099,
+ "step": 1825
+ },
+ {
+ "epoch": 0.9580272822665268,
+ "grad_norm": 0.1549156051160634,
+ "learning_rate": 5.195324664971476e-06,
+ "loss": 0.44893333315849304,
+ "step": 1826
+ },
+ {
+ "epoch": 0.9585519412381952,
+ "grad_norm": 0.1544667437543616,
+ "learning_rate": 5.190478926391271e-06,
+ "loss": 0.4463026821613312,
+ "step": 1827
+ },
+ {
+ "epoch": 0.9590766002098636,
+ "grad_norm": 0.1482681888796153,
+ "learning_rate": 5.1856937977856235e-06,
+ "loss": 0.42915546894073486,
+ "step": 1828
+ },
+ {
+ "epoch": 0.959601259181532,
+ "grad_norm": 0.15393225997035367,
+ "learning_rate": 5.180969292154653e-06,
+ "loss": 0.45655858516693115,
+ "step": 1829
+ },
+ {
+ "epoch": 0.9601259181532005,
+ "grad_norm": 0.15402682858559275,
+ "learning_rate": 5.176305422333788e-06,
+ "loss": 0.45586466789245605,
+ "step": 1830
+ },
+ {
+ "epoch": 0.9606505771248688,
+ "grad_norm": 0.15136206757704734,
+ "learning_rate": 5.171702200993716e-06,
+ "loss": 0.4525001049041748,
+ "step": 1831
+ },
+ {
+ "epoch": 0.9611752360965372,
+ "grad_norm": 0.15443301434325082,
+ "learning_rate": 5.167159640640357e-06,
+ "loss": 0.4500652551651001,
+ "step": 1832
+ },
+ {
+ "epoch": 0.9616998950682056,
+ "grad_norm": 0.1553996888683437,
+ "learning_rate": 5.162677753614834e-06,
+ "loss": 0.4426611363887787,
+ "step": 1833
+ },
+ {
+ "epoch": 0.9622245540398741,
+ "grad_norm": 0.14719352293992807,
+ "learning_rate": 5.158256552093428e-06,
+ "loss": 0.43979009985923767,
+ "step": 1834
+ },
+ {
+ "epoch": 0.9627492130115425,
+ "grad_norm": 0.14547547543617104,
+ "learning_rate": 5.153896048087557e-06,
+ "loss": 0.443497896194458,
+ "step": 1835
+ },
+ {
+ "epoch": 0.9632738719832109,
+ "grad_norm": 0.15665555976593595,
+ "learning_rate": 5.149596253443734e-06,
+ "loss": 0.441650390625,
+ "step": 1836
+ },
+ {
+ "epoch": 0.9637985309548793,
+ "grad_norm": 0.1552107125031637,
+ "learning_rate": 5.145357179843536e-06,
+ "loss": 0.44342389702796936,
+ "step": 1837
+ },
+ {
+ "epoch": 0.9643231899265478,
+ "grad_norm": 0.1450717817395646,
+ "learning_rate": 5.141178838803574e-06,
+ "loss": 0.4459199011325836,
+ "step": 1838
+ },
+ {
+ "epoch": 0.9648478488982162,
+ "grad_norm": 0.15194030785499474,
+ "learning_rate": 5.137061241675472e-06,
+ "loss": 0.43879520893096924,
+ "step": 1839
+ },
+ {
+ "epoch": 0.9653725078698846,
+ "grad_norm": 0.15876627179026034,
+ "learning_rate": 5.133004399645817e-06,
+ "loss": 0.4629112184047699,
+ "step": 1840
+ },
+ {
+ "epoch": 0.965897166841553,
+ "grad_norm": 0.15025078854799365,
+ "learning_rate": 5.129008323736142e-06,
+ "loss": 0.4596017599105835,
+ "step": 1841
+ },
+ {
+ "epoch": 0.9664218258132214,
+ "grad_norm": 0.15108123279475658,
+ "learning_rate": 5.125073024802887e-06,
+ "loss": 0.4394477605819702,
+ "step": 1842
+ },
+ {
+ "epoch": 0.9669464847848899,
+ "grad_norm": 0.15040459550862542,
+ "learning_rate": 5.121198513537382e-06,
+ "loss": 0.44406938552856445,
+ "step": 1843
+ },
+ {
+ "epoch": 0.9674711437565582,
+ "grad_norm": 0.14829125607751426,
+ "learning_rate": 5.1173848004658065e-06,
+ "loss": 0.4245413541793823,
+ "step": 1844
+ },
+ {
+ "epoch": 0.9679958027282266,
+ "grad_norm": 0.15182555299407333,
+ "learning_rate": 5.11363189594916e-06,
+ "loss": 0.44984906911849976,
+ "step": 1845
+ },
+ {
+ "epoch": 0.968520461699895,
+ "grad_norm": 0.1455263692654042,
+ "learning_rate": 5.109939810183251e-06,
+ "loss": 0.41511911153793335,
+ "step": 1846
+ },
+ {
+ "epoch": 0.9690451206715635,
+ "grad_norm": 0.14474034020024001,
+ "learning_rate": 5.106308553198644e-06,
+ "loss": 0.42551544308662415,
+ "step": 1847
+ },
+ {
+ "epoch": 0.9695697796432319,
+ "grad_norm": 0.14650897119054226,
+ "learning_rate": 5.102738134860656e-06,
+ "loss": 0.4347913861274719,
+ "step": 1848
+ },
+ {
+ "epoch": 0.9700944386149003,
+ "grad_norm": 0.15017963706984508,
+ "learning_rate": 5.099228564869313e-06,
+ "loss": 0.442004919052124,
+ "step": 1849
+ },
+ {
+ "epoch": 0.9706190975865687,
+ "grad_norm": 0.14854765100160125,
+ "learning_rate": 5.09577985275933e-06,
+ "loss": 0.4416872262954712,
+ "step": 1850
+ },
+ {
+ "epoch": 0.9711437565582371,
+ "grad_norm": 0.1648726444197388,
+ "learning_rate": 5.092392007900087e-06,
+ "loss": 0.4556121826171875,
+ "step": 1851
+ },
+ {
+ "epoch": 0.9716684155299056,
+ "grad_norm": 0.15795586380007176,
+ "learning_rate": 5.089065039495598e-06,
+ "loss": 0.4360453486442566,
+ "step": 1852
+ },
+ {
+ "epoch": 0.972193074501574,
+ "grad_norm": 0.15841383239716067,
+ "learning_rate": 5.085798956584491e-06,
+ "loss": 0.44655895233154297,
+ "step": 1853
+ },
+ {
+ "epoch": 0.9727177334732424,
+ "grad_norm": 0.14772746394586367,
+ "learning_rate": 5.082593768039989e-06,
+ "loss": 0.44125744700431824,
+ "step": 1854
+ },
+ {
+ "epoch": 0.9732423924449108,
+ "grad_norm": 0.1511843151742375,
+ "learning_rate": 5.079449482569862e-06,
+ "loss": 0.4447266161441803,
+ "step": 1855
+ },
+ {
+ "epoch": 0.9737670514165793,
+ "grad_norm": 0.15676550602119788,
+ "learning_rate": 5.076366108716437e-06,
+ "loss": 0.4479975700378418,
+ "step": 1856
+ },
+ {
+ "epoch": 0.9742917103882477,
+ "grad_norm": 0.15042078359963343,
+ "learning_rate": 5.073343654856548e-06,
+ "loss": 0.442285418510437,
+ "step": 1857
+ },
+ {
+ "epoch": 0.974816369359916,
+ "grad_norm": 0.14908899116909868,
+ "learning_rate": 5.070382129201526e-06,
+ "loss": 0.44102227687835693,
+ "step": 1858
+ },
+ {
+ "epoch": 0.9753410283315844,
+ "grad_norm": 0.14466645777159964,
+ "learning_rate": 5.067481539797172e-06,
+ "loss": 0.42972224950790405,
+ "step": 1859
+ },
+ {
+ "epoch": 0.9758656873032528,
+ "grad_norm": 0.1600541020039413,
+ "learning_rate": 5.064641894523737e-06,
+ "loss": 0.4534302353858948,
+ "step": 1860
+ },
+ {
+ "epoch": 0.9763903462749213,
+ "grad_norm": 0.1599331982060412,
+ "learning_rate": 5.061863201095905e-06,
+ "loss": 0.4575304090976715,
+ "step": 1861
+ },
+ {
+ "epoch": 0.9769150052465897,
+ "grad_norm": 0.14699643507200835,
+ "learning_rate": 5.059145467062763e-06,
+ "loss": 0.4292885959148407,
+ "step": 1862
+ },
+ {
+ "epoch": 0.9774396642182581,
+ "grad_norm": 0.15239951700028798,
+ "learning_rate": 5.056488699807784e-06,
+ "loss": 0.4484747052192688,
+ "step": 1863
+ },
+ {
+ "epoch": 0.9779643231899265,
+ "grad_norm": 0.14925326654404217,
+ "learning_rate": 5.05389290654881e-06,
+ "loss": 0.4315107464790344,
+ "step": 1864
+ },
+ {
+ "epoch": 0.978488982161595,
+ "grad_norm": 0.14557754498869974,
+ "learning_rate": 5.051358094338034e-06,
+ "loss": 0.4448525607585907,
+ "step": 1865
+ },
+ {
+ "epoch": 0.9790136411332634,
+ "grad_norm": 0.14596289150950575,
+ "learning_rate": 5.048884270061969e-06,
+ "loss": 0.4328519403934479,
+ "step": 1866
+ },
+ {
+ "epoch": 0.9795383001049318,
+ "grad_norm": 0.14883370444683877,
+ "learning_rate": 5.046471440441445e-06,
+ "loss": 0.4282095432281494,
+ "step": 1867
+ },
+ {
+ "epoch": 0.9800629590766002,
+ "grad_norm": 0.15233818252290696,
+ "learning_rate": 5.044119612031576e-06,
+ "loss": 0.44765010476112366,
+ "step": 1868
+ },
+ {
+ "epoch": 0.9805876180482687,
+ "grad_norm": 0.15738042523437606,
+ "learning_rate": 5.041828791221759e-06,
+ "loss": 0.44896170496940613,
+ "step": 1869
+ },
+ {
+ "epoch": 0.9811122770199371,
+ "grad_norm": 0.15435191566203477,
+ "learning_rate": 5.039598984235635e-06,
+ "loss": 0.46627920866012573,
+ "step": 1870
+ },
+ {
+ "epoch": 0.9816369359916055,
+ "grad_norm": 0.15330002542858065,
+ "learning_rate": 5.037430197131096e-06,
+ "loss": 0.44774049520492554,
+ "step": 1871
+ },
+ {
+ "epoch": 0.9821615949632738,
+ "grad_norm": 0.15058611877033465,
+ "learning_rate": 5.035322435800247e-06,
+ "loss": 0.4310590624809265,
+ "step": 1872
+ },
+ {
+ "epoch": 0.9826862539349422,
+ "grad_norm": 0.14470485792169563,
+ "learning_rate": 5.033275705969405e-06,
+ "loss": 0.41565707325935364,
+ "step": 1873
+ },
+ {
+ "epoch": 0.9832109129066107,
+ "grad_norm": 0.15361200081510454,
+ "learning_rate": 5.031290013199077e-06,
+ "loss": 0.44648170471191406,
+ "step": 1874
+ },
+ {
+ "epoch": 0.9837355718782791,
+ "grad_norm": 0.14425950736522083,
+ "learning_rate": 5.0293653628839475e-06,
+ "loss": 0.4347214102745056,
+ "step": 1875
+ },
+ {
+ "epoch": 0.9842602308499475,
+ "grad_norm": 0.1565812832501761,
+ "learning_rate": 5.027501760252857e-06,
+ "loss": 0.43630290031433105,
+ "step": 1876
+ },
+ {
+ "epoch": 0.9847848898216159,
+ "grad_norm": 0.14893609351255585,
+ "learning_rate": 5.025699210368801e-06,
+ "loss": 0.45179879665374756,
+ "step": 1877
+ },
+ {
+ "epoch": 0.9853095487932844,
+ "grad_norm": 0.1519863787002136,
+ "learning_rate": 5.023957718128899e-06,
+ "loss": 0.4456064701080322,
+ "step": 1878
+ },
+ {
+ "epoch": 0.9858342077649528,
+ "grad_norm": 0.14899484676330454,
+ "learning_rate": 5.022277288264397e-06,
+ "loss": 0.4311719238758087,
+ "step": 1879
+ },
+ {
+ "epoch": 0.9863588667366212,
+ "grad_norm": 0.15008205295390942,
+ "learning_rate": 5.020657925340649e-06,
+ "loss": 0.43028831481933594,
+ "step": 1880
+ },
+ {
+ "epoch": 0.9868835257082896,
+ "grad_norm": 0.1563139451666976,
+ "learning_rate": 5.019099633757102e-06,
+ "loss": 0.4408448040485382,
+ "step": 1881
+ },
+ {
+ "epoch": 0.9874081846799581,
+ "grad_norm": 0.14428766727134737,
+ "learning_rate": 5.017602417747283e-06,
+ "loss": 0.4480482339859009,
+ "step": 1882
+ },
+ {
+ "epoch": 0.9879328436516265,
+ "grad_norm": 0.14854657656236905,
+ "learning_rate": 5.0161662813787935e-06,
+ "loss": 0.43310457468032837,
+ "step": 1883
+ },
+ {
+ "epoch": 0.9884575026232949,
+ "grad_norm": 0.14979933582255975,
+ "learning_rate": 5.0147912285532935e-06,
+ "loss": 0.46105456352233887,
+ "step": 1884
+ },
+ {
+ "epoch": 0.9889821615949633,
+ "grad_norm": 0.15138043090447337,
+ "learning_rate": 5.0134772630064965e-06,
+ "loss": 0.4424416422843933,
+ "step": 1885
+ },
+ {
+ "epoch": 0.9895068205666316,
+ "grad_norm": 0.15826989524451818,
+ "learning_rate": 5.012224388308148e-06,
+ "loss": 0.45935922861099243,
+ "step": 1886
+ },
+ {
+ "epoch": 0.9900314795383001,
+ "grad_norm": 0.14514140559811425,
+ "learning_rate": 5.011032607862033e-06,
+ "loss": 0.4395596385002136,
+ "step": 1887
+ },
+ {
+ "epoch": 0.9905561385099685,
+ "grad_norm": 0.15650352549659236,
+ "learning_rate": 5.00990192490595e-06,
+ "loss": 0.47225916385650635,
+ "step": 1888
+ },
+ {
+ "epoch": 0.9910807974816369,
+ "grad_norm": 0.15839596050463609,
+ "learning_rate": 5.008832342511712e-06,
+ "loss": 0.43967339396476746,
+ "step": 1889
+ },
+ {
+ "epoch": 0.9916054564533053,
+ "grad_norm": 0.14915540503590238,
+ "learning_rate": 5.007823863585134e-06,
+ "loss": 0.4409275949001312,
+ "step": 1890
+ },
+ {
+ "epoch": 0.9921301154249738,
+ "grad_norm": 0.15382866906925854,
+ "learning_rate": 5.006876490866029e-06,
+ "loss": 0.44721555709838867,
+ "step": 1891
+ },
+ {
+ "epoch": 0.9926547743966422,
+ "grad_norm": 0.14953451753083624,
+ "learning_rate": 5.0059902269281956e-06,
+ "loss": 0.4325152039527893,
+ "step": 1892
+ },
+ {
+ "epoch": 0.9931794333683106,
+ "grad_norm": 0.1540833258535656,
+ "learning_rate": 5.0051650741794135e-06,
+ "loss": 0.4537501931190491,
+ "step": 1893
+ },
+ {
+ "epoch": 0.993704092339979,
+ "grad_norm": 0.14711400847665276,
+ "learning_rate": 5.004401034861441e-06,
+ "loss": 0.4396728277206421,
+ "step": 1894
+ },
+ {
+ "epoch": 0.9942287513116475,
+ "grad_norm": 0.1489662167447431,
+ "learning_rate": 5.003698111049998e-06,
+ "loss": 0.4431758522987366,
+ "step": 1895
+ },
+ {
+ "epoch": 0.9947534102833159,
+ "grad_norm": 0.14660584087882736,
+ "learning_rate": 5.003056304654774e-06,
+ "loss": 0.4379071295261383,
+ "step": 1896
+ },
+ {
+ "epoch": 0.9952780692549843,
+ "grad_norm": 0.14739479545415804,
+ "learning_rate": 5.0024756174194125e-06,
+ "loss": 0.4503260552883148,
+ "step": 1897
+ },
+ {
+ "epoch": 0.9958027282266527,
+ "grad_norm": 0.14953590570482198,
+ "learning_rate": 5.001956050921509e-06,
+ "loss": 0.43298959732055664,
+ "step": 1898
+ },
+ {
+ "epoch": 0.9963273871983211,
+ "grad_norm": 0.14530023273300657,
+ "learning_rate": 5.001497606572614e-06,
+ "loss": 0.4221065938472748,
+ "step": 1899
+ },
+ {
+ "epoch": 0.9968520461699895,
+ "grad_norm": 0.1506918434878359,
+ "learning_rate": 5.001100285618214e-06,
+ "loss": 0.4413088262081146,
+ "step": 1900
+ },
+ {
+ "epoch": 0.9973767051416579,
+ "grad_norm": 0.1523639630038553,
+ "learning_rate": 5.000764089137742e-06,
+ "loss": 0.44044029712677,
+ "step": 1901
+ },
+ {
+ "epoch": 0.9979013641133263,
+ "grad_norm": 0.1541210997698642,
+ "learning_rate": 5.000489018044571e-06,
+ "loss": 0.4456636905670166,
+ "step": 1902
+ },
+ {
+ "epoch": 0.9984260230849947,
+ "grad_norm": 0.15136191547860647,
+ "learning_rate": 5.000275073086002e-06,
+ "loss": 0.4322631061077118,
+ "step": 1903
+ },
+ {
+ "epoch": 0.9989506820566632,
+ "grad_norm": 0.1502650378306557,
+ "learning_rate": 5.000122254843282e-06,
+ "loss": 0.4421358108520508,
+ "step": 1904
+ },
+ {
+ "epoch": 0.9994753410283316,
+ "grad_norm": 0.1449475211228009,
+ "learning_rate": 5.0000305637315785e-06,
+ "loss": 0.43860897421836853,
+ "step": 1905
+ },
+ {
+ "epoch": 1.0,
+ "grad_norm": 0.16118507844975188,
+ "learning_rate": 5e-06,
+ "loss": 0.4352246820926666,
+ "step": 1906
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": true
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 5.04636066758656e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-1906/video_preprocessor_config.json b/checkpoints/checkpoint-1906/video_preprocessor_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..a933f0ec6941222c33aeba7980bd5acf8dd98e91
--- /dev/null
+++ b/checkpoints/checkpoint-1906/video_preprocessor_config.json
@@ -0,0 +1,30 @@
+{
+ "do_convert_rgb": true,
+ "do_normalize": true,
+ "do_rescale": true,
+ "do_resize": true,
+ "do_sample_frames": true,
+ "fps": 2,
+ "image_mean": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "image_std": [
+ 0.5,
+ 0.5,
+ 0.5
+ ],
+ "max_frames": 768,
+ "merge_size": 2,
+ "min_frames": 4,
+ "patch_size": 16,
+ "resample": 3,
+ "rescale_factor": 0.00392156862745098,
+ "size": {
+ "longest_edge": 786432,
+ "shortest_edge": 131072
+ },
+ "temporal_patch_size": 2,
+ "video_processor_type": "Qwen3VLVideoProcessor"
+}
diff --git a/checkpoints/checkpoint-50/chat_template.jinja b/checkpoints/checkpoint-50/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-50/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-50/config.json b/checkpoints/checkpoint-50/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-50/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-50/generation_config.json b/checkpoints/checkpoint-50/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-50/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-50/model.safetensors b/checkpoints/checkpoint-50/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..9620336f35a095cc2929e882bc49fbfa8993473d
--- /dev/null
+++ b/checkpoints/checkpoint-50/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e74eacf8d0f20da3ba2dc65671437a68012da40941c3477ee60d27fad55fdb97
+size 37983421000
diff --git a/checkpoints/checkpoint-50/tokenizer.json b/checkpoints/checkpoint-50/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-50/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-50/tokenizer_config.json b/checkpoints/checkpoint-50/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-50/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-50/trainer_state.json b/checkpoints/checkpoint-50/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..397d7f9b57c54cacde92f06ddec6e21e33480e40
--- /dev/null
+++ b/checkpoints/checkpoint-50/trainer_state.json
@@ -0,0 +1,384 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.013117334557617892,
+ "eval_steps": 500,
+ "global_step": 50,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.00026234669115235784,
+ "grad_norm": 2.2714586704272426,
+ "learning_rate": 4.9999992359065805e-05,
+ "loss": 0.7856928110122681,
+ "step": 1
+ },
+ {
+ "epoch": 0.0005246933823047157,
+ "grad_norm": 6.399851861699839,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.9817550778388977,
+ "step": 2
+ },
+ {
+ "epoch": 0.0007870400734570736,
+ "grad_norm": 3.2243286363133743,
+ "learning_rate": 4.999993123162341e-05,
+ "loss": 0.8669183254241943,
+ "step": 3
+ },
+ {
+ "epoch": 0.0010493867646094314,
+ "grad_norm": 1.651084570952225,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.7457306981086731,
+ "step": 4
+ },
+ {
+ "epoch": 0.0013117334557617892,
+ "grad_norm": 1.8846646841836883,
+ "learning_rate": 4.9999808976904684e-05,
+ "loss": 0.7109265327453613,
+ "step": 5
+ },
+ {
+ "epoch": 0.0015740801469141471,
+ "grad_norm": 0.8685009639291987,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.6621891260147095,
+ "step": 6
+ },
+ {
+ "epoch": 0.0018364268380665048,
+ "grad_norm": 0.8936972800186528,
+ "learning_rate": 4.999962559524176e-05,
+ "loss": 0.6629632711410522,
+ "step": 7
+ },
+ {
+ "epoch": 0.0020987735292188627,
+ "grad_norm": 0.6209976030487597,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.6567398309707642,
+ "step": 8
+ },
+ {
+ "epoch": 0.0023611202203712204,
+ "grad_norm": 0.6252473545934151,
+ "learning_rate": 4.999938108713286e-05,
+ "loss": 0.6359516978263855,
+ "step": 9
+ },
+ {
+ "epoch": 0.0026234669115235785,
+ "grad_norm": 0.511999855315358,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.6289734840393066,
+ "step": 10
+ },
+ {
+ "epoch": 0.002885813602675936,
+ "grad_norm": 0.5313423923886501,
+ "learning_rate": 4.999907545324224e-05,
+ "loss": 0.6182207465171814,
+ "step": 11
+ },
+ {
+ "epoch": 0.0031481602938282943,
+ "grad_norm": 0.42062600638930797,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6219775676727295,
+ "step": 12
+ },
+ {
+ "epoch": 0.003410506984980652,
+ "grad_norm": 0.44391063858033153,
+ "learning_rate": 4.9998708694400254e-05,
+ "loss": 0.5877193212509155,
+ "step": 13
+ },
+ {
+ "epoch": 0.0036728536761330096,
+ "grad_norm": 0.37761669960853983,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.614560604095459,
+ "step": 14
+ },
+ {
+ "epoch": 0.003935200367285368,
+ "grad_norm": 0.3756418760272285,
+ "learning_rate": 4.99982808116033e-05,
+ "loss": 0.6007119417190552,
+ "step": 15
+ },
+ {
+ "epoch": 0.004197547058437725,
+ "grad_norm": 0.39558369142666805,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6008801460266113,
+ "step": 16
+ },
+ {
+ "epoch": 0.004459893749590083,
+ "grad_norm": 0.38513447223129527,
+ "learning_rate": 4.999779180601384e-05,
+ "loss": 0.6442829370498657,
+ "step": 17
+ },
+ {
+ "epoch": 0.004722240440742441,
+ "grad_norm": 0.34453280343588777,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6031531095504761,
+ "step": 18
+ },
+ {
+ "epoch": 0.004984587131894799,
+ "grad_norm": 0.3643772843918706,
+ "learning_rate": 4.999724167896039e-05,
+ "loss": 0.585789144039154,
+ "step": 19
+ },
+ {
+ "epoch": 0.005246933823047157,
+ "grad_norm": 0.35287348923163614,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.5892359614372253,
+ "step": 20
+ },
+ {
+ "epoch": 0.005509280514199515,
+ "grad_norm": 0.33436796314895195,
+ "learning_rate": 4.999663043193751e-05,
+ "loss": 0.6053206920623779,
+ "step": 21
+ },
+ {
+ "epoch": 0.005771627205351872,
+ "grad_norm": 0.3347022500072737,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.5888890027999878,
+ "step": 22
+ },
+ {
+ "epoch": 0.00603397389650423,
+ "grad_norm": 0.31959982532987413,
+ "learning_rate": 4.9995958066605845e-05,
+ "loss": 0.613573431968689,
+ "step": 23
+ },
+ {
+ "epoch": 0.0062963205876565885,
+ "grad_norm": 0.3153719515894866,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.5705129504203796,
+ "step": 24
+ },
+ {
+ "epoch": 0.006558667278808946,
+ "grad_norm": 0.33921987175859747,
+ "learning_rate": 4.999522458479205e-05,
+ "loss": 0.5836034417152405,
+ "step": 25
+ },
+ {
+ "epoch": 0.006821013969961304,
+ "grad_norm": 0.33492843381904264,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.5898736715316772,
+ "step": 26
+ },
+ {
+ "epoch": 0.0070833606611136616,
+ "grad_norm": 0.3322820760244329,
+ "learning_rate": 4.999442998848883e-05,
+ "loss": 0.6031719446182251,
+ "step": 27
+ },
+ {
+ "epoch": 0.007345707352266019,
+ "grad_norm": 0.31805926413306335,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.5801360607147217,
+ "step": 28
+ },
+ {
+ "epoch": 0.007608054043418378,
+ "grad_norm": 0.2924327104211072,
+ "learning_rate": 4.9993574279854926e-05,
+ "loss": 0.5841026306152344,
+ "step": 29
+ },
+ {
+ "epoch": 0.007870400734570735,
+ "grad_norm": 0.32419872693133733,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5635761618614197,
+ "step": 30
+ },
+ {
+ "epoch": 0.008132747425723093,
+ "grad_norm": 0.3226233138332213,
+ "learning_rate": 4.999265746121511e-05,
+ "loss": 0.5694052577018738,
+ "step": 31
+ },
+ {
+ "epoch": 0.00839509411687545,
+ "grad_norm": 0.27757723505773657,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5479917526245117,
+ "step": 32
+ },
+ {
+ "epoch": 0.008657440808027808,
+ "grad_norm": 0.2856810760831006,
+ "learning_rate": 4.999167953506016e-05,
+ "loss": 0.5649982690811157,
+ "step": 33
+ },
+ {
+ "epoch": 0.008919787499180166,
+ "grad_norm": 0.3079655566370459,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.5785400867462158,
+ "step": 34
+ },
+ {
+ "epoch": 0.009182134190332524,
+ "grad_norm": 0.38052049760913864,
+ "learning_rate": 4.999064050404691e-05,
+ "loss": 0.5726369619369507,
+ "step": 35
+ },
+ {
+ "epoch": 0.009444480881484882,
+ "grad_norm": 0.27886201170397207,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5590238571166992,
+ "step": 36
+ },
+ {
+ "epoch": 0.009706827572637241,
+ "grad_norm": 0.295534166971022,
+ "learning_rate": 4.998954037099814e-05,
+ "loss": 0.5798577070236206,
+ "step": 37
+ },
+ {
+ "epoch": 0.009969174263789599,
+ "grad_norm": 0.27038408490120824,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5620362162590027,
+ "step": 38
+ },
+ {
+ "epoch": 0.010231520954941956,
+ "grad_norm": 0.2871369097168852,
+ "learning_rate": 4.99883791389027e-05,
+ "loss": 0.5668522715568542,
+ "step": 39
+ },
+ {
+ "epoch": 0.010493867646094314,
+ "grad_norm": 0.28270332707392204,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5634897947311401,
+ "step": 40
+ },
+ {
+ "epoch": 0.010756214337246672,
+ "grad_norm": 0.2834128081666473,
+ "learning_rate": 4.998715681091536e-05,
+ "loss": 0.5633743405342102,
+ "step": 41
+ },
+ {
+ "epoch": 0.01101856102839903,
+ "grad_norm": 0.2713974891334188,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.5614528656005859,
+ "step": 42
+ },
+ {
+ "epoch": 0.011280907719551387,
+ "grad_norm": 0.27728181950502956,
+ "learning_rate": 4.998587339035695e-05,
+ "loss": 0.5788841247558594,
+ "step": 43
+ },
+ {
+ "epoch": 0.011543254410703745,
+ "grad_norm": 0.28984987697140707,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5365938544273376,
+ "step": 44
+ },
+ {
+ "epoch": 0.011805601101856102,
+ "grad_norm": 0.26457276236895516,
+ "learning_rate": 4.99845288807142e-05,
+ "loss": 0.5527335405349731,
+ "step": 45
+ },
+ {
+ "epoch": 0.01206794779300846,
+ "grad_norm": 0.2852856853378878,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5504194498062134,
+ "step": 46
+ },
+ {
+ "epoch": 0.012330294484160818,
+ "grad_norm": 0.29211982306814627,
+ "learning_rate": 4.9983123285639875e-05,
+ "loss": 0.5598798990249634,
+ "step": 47
+ },
+ {
+ "epoch": 0.012592641175313177,
+ "grad_norm": 1.380352477274431,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5437134504318237,
+ "step": 48
+ },
+ {
+ "epoch": 0.012854987866465535,
+ "grad_norm": 0.3067242638125492,
+ "learning_rate": 4.9981656608952636e-05,
+ "loss": 0.5705212354660034,
+ "step": 49
+ },
+ {
+ "epoch": 0.013117334557617892,
+ "grad_norm": 0.3032625186204062,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5649327039718628,
+ "step": 50
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 3812,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 307360162119680.0,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoints/checkpoint-950/chat_template.jinja b/checkpoints/checkpoint-950/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..a8755d827c0a7b614c246c4060dfd58ab352a8ff
--- /dev/null
+++ b/checkpoints/checkpoint-950/chat_template.jinja
@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+ {%- if content is string %}
+ {{- content }}
+ {%- elif content is iterable and content is not mapping %}
+ {%- for item in content %}
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain images.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set image_count.value = image_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+ {%- elif 'video' in item or item.type == 'video' %}
+ {%- if is_system_content %}
+ {{- raise_exception('System message cannot contain videos.') }}
+ {%- endif %}
+ {%- if do_vision_count %}
+ {%- set video_count.value = video_count.value + 1 %}
+ {%- endif %}
+ {%- if add_vision_id %}
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
+ {%- endif %}
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+ {%- elif 'text' in item %}
+ {{- item.text }}
+ {%- else %}
+ {{- raise_exception('Unexpected item type in content.') }}
+ {%- endif %}
+ {%- endfor %}
+ {%- elif content is none or content is undefined %}
+ {{- '' }}
+ {%- else %}
+ {{- raise_exception('Unexpected content type.') }}
+ {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+ {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+ {{- '<|im_start|>system\n' }}
+ {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n" }}
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n\n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n\n\n\n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n' }}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {%- if content %}
+ {{- '\n\n' + content }}
+ {%- endif %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" %}
+ {%- set content = render_content(message.content, false)|trim %}
+ {%- if not(content.startswith('') and content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+ {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+ {%- set content = render_content(message.content, true)|trim %}
+ {%- if message.role == "system" %}
+ {%- if not loop.first %}
+ {{- raise_exception('System message must be at the beginning.') }}
+ {%- endif %}
+ {%- elif message.role == "user" %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- set reasoning_content = reasoning_content|trim %}
+ {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n\n\n' + content }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if tool_call.function is defined %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {%- if loop.first %}
+ {%- if content|trim %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- '\n\n\n' }}
+ {%- endif %}
+ {%- if tool_call.arguments is defined %}
+ {%- for args_name, args_value in tool_call.arguments|items %}
+ {{- '\n' }}
+ {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+ {{- args_value }}
+ {{- '\n\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
+ {{- '<|im_end|>\n' }}
+ {%- elif loop.last %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- else %}
+ {{- raise_exception('Unexpected message role.') }}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- else %}
+ {{- '\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoints/checkpoint-950/config.json b/checkpoints/checkpoint-950/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff28f8be559aa17cd8539b37673ff62e215c6b24
--- /dev/null
+++ b/checkpoints/checkpoint-950/config.json
@@ -0,0 +1,125 @@
+{
+ "architectures": [
+ "Qwen3_5MoeForConditionalGeneration"
+ ],
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 248046,
+ "image_token_id": 248056,
+ "model_type": "qwen3_5_moe",
+ "pad_token_id": 248044,
+ "text_config": {
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "attn_output_gate": true,
+ "bos_token_id": 248044,
+ "dtype": "bfloat16",
+ "eos_token_id": 248044,
+ "full_attention_interval": 4,
+ "head_dim": 256,
+ "hidden_act": "silu",
+ "hidden_size": 2048,
+ "initializer_range": 0.02,
+ "layer_types": [
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention",
+ "linear_attention",
+ "linear_attention",
+ "linear_attention",
+ "full_attention"
+ ],
+ "linear_conv_kernel_dim": 4,
+ "linear_key_head_dim": 128,
+ "linear_num_key_heads": 16,
+ "linear_num_value_heads": 32,
+ "linear_value_head_dim": 128,
+ "mamba_ssm_dtype": "float32",
+ "max_position_embeddings": 262144,
+ "model_type": "qwen3_5_moe_text",
+ "moe_intermediate_size": 512,
+ "mtp_num_hidden_layers": 1,
+ "mtp_use_dedicated_embeddings": false,
+ "num_attention_heads": 16,
+ "num_experts": 128,
+ "num_experts_per_tok": 8,
+ "num_hidden_layers": 40,
+ "num_key_value_heads": 2,
+ "output_router_logits": false,
+ "pad_token_id": 248044,
+ "partial_rotary_factor": 0.25,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "mrope_interleaved": true,
+ "mrope_section": [
+ 11,
+ 11,
+ 10
+ ],
+ "partial_rotary_factor": 0.25,
+ "rope_theta": 10000000,
+ "rope_type": "default"
+ },
+ "router_aux_loss_coef": 0.001,
+ "shared_expert_intermediate_size": 512,
+ "tie_word_embeddings": false,
+ "use_cache": false,
+ "vocab_size": 248320
+ },
+ "tie_word_embeddings": false,
+ "transformers_version": "5.14.1",
+ "use_cache": false,
+ "video_token_id": 248057,
+ "vision_config": {
+ "deepstack_visual_indexes": [],
+ "depth": 27,
+ "dtype": "bfloat16",
+ "hidden_act": "gelu_pytorch_tanh",
+ "hidden_size": 1152,
+ "in_channels": 3,
+ "initializer_range": 0.02,
+ "intermediate_size": 4304,
+ "model_type": "qwen3_5_moe_vision",
+ "num_heads": 16,
+ "num_position_embeddings": 2304,
+ "out_hidden_size": 2048,
+ "patch_size": 16,
+ "spatial_merge_size": 2,
+ "temporal_patch_size": 2
+ },
+ "vision_end_token_id": 248054,
+ "vision_start_token_id": 248053
+}
diff --git a/checkpoints/checkpoint-950/generation_config.json b/checkpoints/checkpoint-950/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..24986107a0eb467a0a50aa7a8cd494d8e1f3d72b
--- /dev/null
+++ b/checkpoints/checkpoint-950/generation_config.json
@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 248046,
+ 248046,
+ 248044
+ ],
+ "pad_token_id": 248044,
+ "temperature": 1.0,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.14.1"
+}
diff --git a/checkpoints/checkpoint-950/model.safetensors b/checkpoints/checkpoint-950/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..c689ac65479314dfd30d0fdd6b8b43765eced639
--- /dev/null
+++ b/checkpoints/checkpoint-950/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f6b2c7ea27525ce29f3e600ad2a6c916ceb8c9a4ace37d3d85f6af57fc415e2
+size 37983421000
diff --git a/checkpoints/checkpoint-950/tokenizer.json b/checkpoints/checkpoint-950/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..5520bfd2dd834ce386c1312c410fa71af56db5ad
--- /dev/null
+++ b/checkpoints/checkpoint-950/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
+size 19989325
diff --git a/checkpoints/checkpoint-950/tokenizer_config.json b/checkpoints/checkpoint-950/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1a20cc3a8838d08eca9de8a7b84d510d60a0524
--- /dev/null
+++ b/checkpoints/checkpoint-950/tokenizer_config.json
@@ -0,0 +1,32 @@
+{
+ "add_prefix_space": false,
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "image_token": "<|image_pad|>",
+ "is_local": true,
+ "local_files_only": false,
+ "model_max_length": 262144,
+ "model_specific_special_tokens": {
+ "audio_bos_token": "<|audio_start|>",
+ "audio_eos_token": "<|audio_end|>",
+ "audio_token": "<|audio_pad|>",
+ "image_token": "<|image_pad|>",
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+ },
+ "pad_token": "<|endoftext|>",
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null,
+ "video_token": "<|video_pad|>",
+ "vision_bos_token": "<|vision_start|>",
+ "vision_eos_token": "<|vision_end|>"
+}
diff --git a/checkpoints/checkpoint-950/trainer_state.json b/checkpoints/checkpoint-950/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..49683f1c883d0bc8176c2bf470719dbfe7952ece
--- /dev/null
+++ b/checkpoints/checkpoint-950/trainer_state.json
@@ -0,0 +1,6684 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.49842602308499473,
+ "eval_steps": 500,
+ "global_step": 950,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.0005246589716684155,
+ "grad_norm": 2.2354415641373153,
+ "learning_rate": 4.999996943626843e-05,
+ "loss": 0.7954971790313721,
+ "step": 1
+ },
+ {
+ "epoch": 0.001049317943336831,
+ "grad_norm": 6.545498835151217,
+ "learning_rate": 4.999987774515672e-05,
+ "loss": 0.9905959963798523,
+ "step": 2
+ },
+ {
+ "epoch": 0.0015739769150052466,
+ "grad_norm": 2.9806283977452166,
+ "learning_rate": 4.9999724926914e-05,
+ "loss": 0.836991548538208,
+ "step": 3
+ },
+ {
+ "epoch": 0.002098635886673662,
+ "grad_norm": 1.9023976323647784,
+ "learning_rate": 4.999951098195543e-05,
+ "loss": 0.759330153465271,
+ "step": 4
+ },
+ {
+ "epoch": 0.0026232948583420775,
+ "grad_norm": 1.8831472974943915,
+ "learning_rate": 4.999923591086226e-05,
+ "loss": 0.7026219367980957,
+ "step": 5
+ },
+ {
+ "epoch": 0.0031479538300104933,
+ "grad_norm": 0.8897649784754303,
+ "learning_rate": 4.9998899714381784e-05,
+ "loss": 0.6762715578079224,
+ "step": 6
+ },
+ {
+ "epoch": 0.0036726128016789086,
+ "grad_norm": 0.8282743100536848,
+ "learning_rate": 4.999850239342739e-05,
+ "loss": 0.6472997069358826,
+ "step": 7
+ },
+ {
+ "epoch": 0.004197271773347324,
+ "grad_norm": 0.5530534008696922,
+ "learning_rate": 4.999804394907849e-05,
+ "loss": 0.6370899677276611,
+ "step": 8
+ },
+ {
+ "epoch": 0.00472193074501574,
+ "grad_norm": 0.5552721871462486,
+ "learning_rate": 4.9997524382580594e-05,
+ "loss": 0.6554895639419556,
+ "step": 9
+ },
+ {
+ "epoch": 0.005246589716684155,
+ "grad_norm": 0.5058280115950572,
+ "learning_rate": 4.999694369534523e-05,
+ "loss": 0.6161210536956787,
+ "step": 10
+ },
+ {
+ "epoch": 0.005771248688352571,
+ "grad_norm": 0.41439693586879883,
+ "learning_rate": 4.999630188895e-05,
+ "loss": 0.6199569702148438,
+ "step": 11
+ },
+ {
+ "epoch": 0.0062959076600209865,
+ "grad_norm": 0.4095246818227211,
+ "learning_rate": 4.999559896513856e-05,
+ "loss": 0.6158584356307983,
+ "step": 12
+ },
+ {
+ "epoch": 0.006820566631689402,
+ "grad_norm": 0.3919369878888828,
+ "learning_rate": 4.999483492582059e-05,
+ "loss": 0.6056086421012878,
+ "step": 13
+ },
+ {
+ "epoch": 0.007345225603357817,
+ "grad_norm": 0.3464084213351009,
+ "learning_rate": 4.9994009773071806e-05,
+ "loss": 0.6100624203681946,
+ "step": 14
+ },
+ {
+ "epoch": 0.007869884575026232,
+ "grad_norm": 0.3686210287238475,
+ "learning_rate": 4.999312350913398e-05,
+ "loss": 0.5915063619613647,
+ "step": 15
+ },
+ {
+ "epoch": 0.008394543546694649,
+ "grad_norm": 0.3293320952324157,
+ "learning_rate": 4.999217613641487e-05,
+ "loss": 0.5765297412872314,
+ "step": 16
+ },
+ {
+ "epoch": 0.008919202518363064,
+ "grad_norm": 0.30944871417248837,
+ "learning_rate": 4.999116765748829e-05,
+ "loss": 0.585294246673584,
+ "step": 17
+ },
+ {
+ "epoch": 0.00944386149003148,
+ "grad_norm": 0.29758476071343837,
+ "learning_rate": 4.9990098075094054e-05,
+ "loss": 0.5796446800231934,
+ "step": 18
+ },
+ {
+ "epoch": 0.009968520461699895,
+ "grad_norm": 0.2995083294642699,
+ "learning_rate": 4.998896739213797e-05,
+ "loss": 0.5860245227813721,
+ "step": 19
+ },
+ {
+ "epoch": 0.01049317943336831,
+ "grad_norm": 0.2835359863674351,
+ "learning_rate": 4.998777561169186e-05,
+ "loss": 0.5776117444038391,
+ "step": 20
+ },
+ {
+ "epoch": 0.011017838405036727,
+ "grad_norm": 0.2798477132594057,
+ "learning_rate": 4.998652273699351e-05,
+ "loss": 0.574779748916626,
+ "step": 21
+ },
+ {
+ "epoch": 0.011542497376705142,
+ "grad_norm": 0.2594899659764607,
+ "learning_rate": 4.998520877144671e-05,
+ "loss": 0.5683784484863281,
+ "step": 22
+ },
+ {
+ "epoch": 0.012067156348373556,
+ "grad_norm": 0.260429456778487,
+ "learning_rate": 4.998383371862121e-05,
+ "loss": 0.5610731840133667,
+ "step": 23
+ },
+ {
+ "epoch": 0.012591815320041973,
+ "grad_norm": 0.24973987619359944,
+ "learning_rate": 4.998239758225272e-05,
+ "loss": 0.5583471059799194,
+ "step": 24
+ },
+ {
+ "epoch": 0.013116474291710388,
+ "grad_norm": 0.25029886960791803,
+ "learning_rate": 4.9980900366242906e-05,
+ "loss": 0.5736842751502991,
+ "step": 25
+ },
+ {
+ "epoch": 0.013641133263378805,
+ "grad_norm": 0.26262467578242815,
+ "learning_rate": 4.9979342074659353e-05,
+ "loss": 0.5847593545913696,
+ "step": 26
+ },
+ {
+ "epoch": 0.01416579223504722,
+ "grad_norm": 0.23722487597044217,
+ "learning_rate": 4.997772271173561e-05,
+ "loss": 0.583673894405365,
+ "step": 27
+ },
+ {
+ "epoch": 0.014690451206715634,
+ "grad_norm": 0.23217509906514594,
+ "learning_rate": 4.99760422818711e-05,
+ "loss": 0.5572662353515625,
+ "step": 28
+ },
+ {
+ "epoch": 0.015215110178384051,
+ "grad_norm": 0.2313130681837237,
+ "learning_rate": 4.9974300789631205e-05,
+ "loss": 0.5564031600952148,
+ "step": 29
+ },
+ {
+ "epoch": 0.015739769150052464,
+ "grad_norm": 0.24558830853519514,
+ "learning_rate": 4.997249823974715e-05,
+ "loss": 0.5657411813735962,
+ "step": 30
+ },
+ {
+ "epoch": 0.01626442812172088,
+ "grad_norm": 0.21937264249300642,
+ "learning_rate": 4.997063463711606e-05,
+ "loss": 0.5597659349441528,
+ "step": 31
+ },
+ {
+ "epoch": 0.016789087093389297,
+ "grad_norm": 0.2156580234996722,
+ "learning_rate": 4.996870998680092e-05,
+ "loss": 0.5514897108078003,
+ "step": 32
+ },
+ {
+ "epoch": 0.017313746065057714,
+ "grad_norm": 0.23475686548139646,
+ "learning_rate": 4.9966724294030604e-05,
+ "loss": 0.548216700553894,
+ "step": 33
+ },
+ {
+ "epoch": 0.017838405036726127,
+ "grad_norm": 0.21959118317176146,
+ "learning_rate": 4.996467756419975e-05,
+ "loss": 0.560903787612915,
+ "step": 34
+ },
+ {
+ "epoch": 0.018363064008394544,
+ "grad_norm": 0.2215412651004435,
+ "learning_rate": 4.996256980286891e-05,
+ "loss": 0.5571032762527466,
+ "step": 35
+ },
+ {
+ "epoch": 0.01888772298006296,
+ "grad_norm": 0.21856495782321173,
+ "learning_rate": 4.996040101576436e-05,
+ "loss": 0.5423357486724854,
+ "step": 36
+ },
+ {
+ "epoch": 0.019412381951731374,
+ "grad_norm": 0.23290224945871943,
+ "learning_rate": 4.9958171208778246e-05,
+ "loss": 0.552675187587738,
+ "step": 37
+ },
+ {
+ "epoch": 0.01993704092339979,
+ "grad_norm": 0.22887163505707167,
+ "learning_rate": 4.9955880387968424e-05,
+ "loss": 0.5600690245628357,
+ "step": 38
+ },
+ {
+ "epoch": 0.020461699895068207,
+ "grad_norm": 0.214971746309691,
+ "learning_rate": 4.9953528559558556e-05,
+ "loss": 0.5562878847122192,
+ "step": 39
+ },
+ {
+ "epoch": 0.02098635886673662,
+ "grad_norm": 0.22603658876737284,
+ "learning_rate": 4.995111572993803e-05,
+ "loss": 0.5502439737319946,
+ "step": 40
+ },
+ {
+ "epoch": 0.021511017838405037,
+ "grad_norm": 0.21901395377043378,
+ "learning_rate": 4.994864190566196e-05,
+ "loss": 0.5443499088287354,
+ "step": 41
+ },
+ {
+ "epoch": 0.022035676810073453,
+ "grad_norm": 0.22045674759068756,
+ "learning_rate": 4.994610709345119e-05,
+ "loss": 0.5651016235351562,
+ "step": 42
+ },
+ {
+ "epoch": 0.022560335781741866,
+ "grad_norm": 0.21997338360702298,
+ "learning_rate": 4.994351130019222e-05,
+ "loss": 0.5430660247802734,
+ "step": 43
+ },
+ {
+ "epoch": 0.023084994753410283,
+ "grad_norm": 0.21518647430254678,
+ "learning_rate": 4.994085453293724e-05,
+ "loss": 0.5679960250854492,
+ "step": 44
+ },
+ {
+ "epoch": 0.0236096537250787,
+ "grad_norm": 0.21368529766163033,
+ "learning_rate": 4.99381367989041e-05,
+ "loss": 0.5193968415260315,
+ "step": 45
+ },
+ {
+ "epoch": 0.024134312696747113,
+ "grad_norm": 0.2099415505214506,
+ "learning_rate": 4.993535810547627e-05,
+ "loss": 0.5413987040519714,
+ "step": 46
+ },
+ {
+ "epoch": 0.02465897166841553,
+ "grad_norm": 0.2244194894598741,
+ "learning_rate": 4.993251846020284e-05,
+ "loss": 0.5505638122558594,
+ "step": 47
+ },
+ {
+ "epoch": 0.025183630640083946,
+ "grad_norm": 0.2158217142288702,
+ "learning_rate": 4.992961787079848e-05,
+ "loss": 0.5466951131820679,
+ "step": 48
+ },
+ {
+ "epoch": 0.02570828961175236,
+ "grad_norm": 0.22301635158609115,
+ "learning_rate": 4.992665634514345e-05,
+ "loss": 0.5419921875,
+ "step": 49
+ },
+ {
+ "epoch": 0.026232948583420776,
+ "grad_norm": 0.24506503088700576,
+ "learning_rate": 4.992363389128357e-05,
+ "loss": 0.5718627572059631,
+ "step": 50
+ },
+ {
+ "epoch": 0.026757607555089193,
+ "grad_norm": 0.21244165911127053,
+ "learning_rate": 4.992055051743014e-05,
+ "loss": 0.5287284255027771,
+ "step": 51
+ },
+ {
+ "epoch": 0.02728226652675761,
+ "grad_norm": 0.2260756033963987,
+ "learning_rate": 4.9917406231960014e-05,
+ "loss": 0.5291695594787598,
+ "step": 52
+ },
+ {
+ "epoch": 0.027806925498426022,
+ "grad_norm": 0.2061998730850307,
+ "learning_rate": 4.9914201043415514e-05,
+ "loss": 0.5288651585578918,
+ "step": 53
+ },
+ {
+ "epoch": 0.02833158447009444,
+ "grad_norm": 0.2296754342503533,
+ "learning_rate": 4.9910934960504405e-05,
+ "loss": 0.5032019019126892,
+ "step": 54
+ },
+ {
+ "epoch": 0.028856243441762856,
+ "grad_norm": 0.22850545911342737,
+ "learning_rate": 4.990760799209992e-05,
+ "loss": 0.5328084230422974,
+ "step": 55
+ },
+ {
+ "epoch": 0.02938090241343127,
+ "grad_norm": 0.2133015276784329,
+ "learning_rate": 4.990422014724067e-05,
+ "loss": 0.5356483459472656,
+ "step": 56
+ },
+ {
+ "epoch": 0.029905561385099685,
+ "grad_norm": 0.25432432016514744,
+ "learning_rate": 4.9900771435130694e-05,
+ "loss": 0.5211013555526733,
+ "step": 57
+ },
+ {
+ "epoch": 0.030430220356768102,
+ "grad_norm": 0.2455826426741049,
+ "learning_rate": 4.989726186513935e-05,
+ "loss": 0.556255578994751,
+ "step": 58
+ },
+ {
+ "epoch": 0.030954879328436515,
+ "grad_norm": 0.20829914558922374,
+ "learning_rate": 4.989369144680136e-05,
+ "loss": 0.5227149724960327,
+ "step": 59
+ },
+ {
+ "epoch": 0.03147953830010493,
+ "grad_norm": 0.2412696538409334,
+ "learning_rate": 4.989006018981675e-05,
+ "loss": 0.5249186158180237,
+ "step": 60
+ },
+ {
+ "epoch": 0.032004197271773345,
+ "grad_norm": 0.23010200074979875,
+ "learning_rate": 4.9886368104050845e-05,
+ "loss": 0.5433767437934875,
+ "step": 61
+ },
+ {
+ "epoch": 0.03252885624344176,
+ "grad_norm": 0.22687921985456583,
+ "learning_rate": 4.98826151995342e-05,
+ "loss": 0.5285462141036987,
+ "step": 62
+ },
+ {
+ "epoch": 0.03305351521511018,
+ "grad_norm": 0.24500458944944026,
+ "learning_rate": 4.9878801486462626e-05,
+ "loss": 0.5401079058647156,
+ "step": 63
+ },
+ {
+ "epoch": 0.033578174186778595,
+ "grad_norm": 0.21913800757036783,
+ "learning_rate": 4.987492697519712e-05,
+ "loss": 0.5190115571022034,
+ "step": 64
+ },
+ {
+ "epoch": 0.03410283315844701,
+ "grad_norm": 0.2418283669079999,
+ "learning_rate": 4.987099167626386e-05,
+ "loss": 0.5514561533927917,
+ "step": 65
+ },
+ {
+ "epoch": 0.03462749213011543,
+ "grad_norm": 0.203488726313704,
+ "learning_rate": 4.986699560035418e-05,
+ "loss": 0.5108942985534668,
+ "step": 66
+ },
+ {
+ "epoch": 0.03515215110178384,
+ "grad_norm": 0.22612089636497223,
+ "learning_rate": 4.986293875832453e-05,
+ "loss": 0.5352374911308289,
+ "step": 67
+ },
+ {
+ "epoch": 0.035676810073452254,
+ "grad_norm": 0.22445061386427623,
+ "learning_rate": 4.9858821161196437e-05,
+ "loss": 0.5404313802719116,
+ "step": 68
+ },
+ {
+ "epoch": 0.03620146904512067,
+ "grad_norm": 0.2322237251795982,
+ "learning_rate": 4.9854642820156474e-05,
+ "loss": 0.5575121641159058,
+ "step": 69
+ },
+ {
+ "epoch": 0.03672612801678909,
+ "grad_norm": 0.235412830124128,
+ "learning_rate": 4.985040374655627e-05,
+ "loss": 0.5186446905136108,
+ "step": 70
+ },
+ {
+ "epoch": 0.037250786988457504,
+ "grad_norm": 0.21389419385401767,
+ "learning_rate": 4.984610395191244e-05,
+ "loss": 0.5176467895507812,
+ "step": 71
+ },
+ {
+ "epoch": 0.03777544596012592,
+ "grad_norm": 0.22060463043917883,
+ "learning_rate": 4.9841743447906576e-05,
+ "loss": 0.538845419883728,
+ "step": 72
+ },
+ {
+ "epoch": 0.03830010493179433,
+ "grad_norm": 0.23093557378737242,
+ "learning_rate": 4.983732224638517e-05,
+ "loss": 0.5405384302139282,
+ "step": 73
+ },
+ {
+ "epoch": 0.03882476390346275,
+ "grad_norm": 0.22490106972589463,
+ "learning_rate": 4.983284035935965e-05,
+ "loss": 0.5374243855476379,
+ "step": 74
+ },
+ {
+ "epoch": 0.039349422875131164,
+ "grad_norm": 0.22335259891372472,
+ "learning_rate": 4.982829779900629e-05,
+ "loss": 0.5149869322776794,
+ "step": 75
+ },
+ {
+ "epoch": 0.03987408184679958,
+ "grad_norm": 0.2281062576578094,
+ "learning_rate": 4.9823694577666216e-05,
+ "loss": 0.525192379951477,
+ "step": 76
+ },
+ {
+ "epoch": 0.040398740818468,
+ "grad_norm": 0.26041474702954615,
+ "learning_rate": 4.9819030707845357e-05,
+ "loss": 0.5436732769012451,
+ "step": 77
+ },
+ {
+ "epoch": 0.040923399790136414,
+ "grad_norm": 0.221533455955,
+ "learning_rate": 4.9814306202214376e-05,
+ "loss": 0.5022153854370117,
+ "step": 78
+ },
+ {
+ "epoch": 0.041448058761804824,
+ "grad_norm": 0.22577026100012787,
+ "learning_rate": 4.980952107360873e-05,
+ "loss": 0.5170741677284241,
+ "step": 79
+ },
+ {
+ "epoch": 0.04197271773347324,
+ "grad_norm": 0.2521301983769309,
+ "learning_rate": 4.9804675335028527e-05,
+ "loss": 0.5379733443260193,
+ "step": 80
+ },
+ {
+ "epoch": 0.04249737670514166,
+ "grad_norm": 0.2348617707157391,
+ "learning_rate": 4.979976899963856e-05,
+ "loss": 0.5332635641098022,
+ "step": 81
+ },
+ {
+ "epoch": 0.04302203567681007,
+ "grad_norm": 0.23742835644998692,
+ "learning_rate": 4.979480208076823e-05,
+ "loss": 0.5211970806121826,
+ "step": 82
+ },
+ {
+ "epoch": 0.04354669464847849,
+ "grad_norm": 0.2585088182130228,
+ "learning_rate": 4.9789774591911554e-05,
+ "loss": 0.528963565826416,
+ "step": 83
+ },
+ {
+ "epoch": 0.04407135362014691,
+ "grad_norm": 0.21864407338975772,
+ "learning_rate": 4.978468654672709e-05,
+ "loss": 0.5321191549301147,
+ "step": 84
+ },
+ {
+ "epoch": 0.04459601259181532,
+ "grad_norm": 0.228766800309278,
+ "learning_rate": 4.977953795903794e-05,
+ "loss": 0.5213573575019836,
+ "step": 85
+ },
+ {
+ "epoch": 0.04512067156348373,
+ "grad_norm": 0.2486348376134939,
+ "learning_rate": 4.977432884283162e-05,
+ "loss": 0.5288217067718506,
+ "step": 86
+ },
+ {
+ "epoch": 0.04564533053515215,
+ "grad_norm": 0.2040146812241684,
+ "learning_rate": 4.9769059212260174e-05,
+ "loss": 0.534426212310791,
+ "step": 87
+ },
+ {
+ "epoch": 0.046169989506820566,
+ "grad_norm": 0.2430172776877213,
+ "learning_rate": 4.976372908163999e-05,
+ "loss": 0.5217506885528564,
+ "step": 88
+ },
+ {
+ "epoch": 0.04669464847848898,
+ "grad_norm": 0.2240532727571746,
+ "learning_rate": 4.975833846545183e-05,
+ "loss": 0.5273749828338623,
+ "step": 89
+ },
+ {
+ "epoch": 0.0472193074501574,
+ "grad_norm": 0.2329298721775431,
+ "learning_rate": 4.975288737834081e-05,
+ "loss": 0.5111975073814392,
+ "step": 90
+ },
+ {
+ "epoch": 0.047743966421825816,
+ "grad_norm": 0.23936582978557824,
+ "learning_rate": 4.9747375835116314e-05,
+ "loss": 0.522817075252533,
+ "step": 91
+ },
+ {
+ "epoch": 0.048268625393494226,
+ "grad_norm": 0.2203271348841118,
+ "learning_rate": 4.974180385075196e-05,
+ "loss": 0.5198397636413574,
+ "step": 92
+ },
+ {
+ "epoch": 0.04879328436516264,
+ "grad_norm": 0.26898506337489986,
+ "learning_rate": 4.973617144038558e-05,
+ "loss": 0.5347901582717896,
+ "step": 93
+ },
+ {
+ "epoch": 0.04931794333683106,
+ "grad_norm": 0.2244077209335566,
+ "learning_rate": 4.973047861931918e-05,
+ "loss": 0.5209472179412842,
+ "step": 94
+ },
+ {
+ "epoch": 0.049842602308499476,
+ "grad_norm": 0.23564320422388416,
+ "learning_rate": 4.9724725403018876e-05,
+ "loss": 0.5071642994880676,
+ "step": 95
+ },
+ {
+ "epoch": 0.05036726128016789,
+ "grad_norm": 0.24960130961180893,
+ "learning_rate": 4.971891180711488e-05,
+ "loss": 0.5236656069755554,
+ "step": 96
+ },
+ {
+ "epoch": 0.05089192025183631,
+ "grad_norm": 0.25724576936503896,
+ "learning_rate": 4.971303784740141e-05,
+ "loss": 0.5381065607070923,
+ "step": 97
+ },
+ {
+ "epoch": 0.05141657922350472,
+ "grad_norm": 0.23261451235153804,
+ "learning_rate": 4.970710353983672e-05,
+ "loss": 0.5220236778259277,
+ "step": 98
+ },
+ {
+ "epoch": 0.051941238195173135,
+ "grad_norm": 0.22622545828425783,
+ "learning_rate": 4.9701108900542986e-05,
+ "loss": 0.5250488519668579,
+ "step": 99
+ },
+ {
+ "epoch": 0.05246589716684155,
+ "grad_norm": 0.2879752941314558,
+ "learning_rate": 4.9695053945806295e-05,
+ "loss": 0.5213186740875244,
+ "step": 100
+ },
+ {
+ "epoch": 0.05299055613850997,
+ "grad_norm": 0.21266767479960802,
+ "learning_rate": 4.968893869207661e-05,
+ "loss": 0.5395318865776062,
+ "step": 101
+ },
+ {
+ "epoch": 0.053515215110178385,
+ "grad_norm": 0.2289013387019485,
+ "learning_rate": 4.968276315596772e-05,
+ "loss": 0.5267707109451294,
+ "step": 102
+ },
+ {
+ "epoch": 0.0540398740818468,
+ "grad_norm": 0.27034678869545914,
+ "learning_rate": 4.967652735425715e-05,
+ "loss": 0.5251526832580566,
+ "step": 103
+ },
+ {
+ "epoch": 0.05456453305351522,
+ "grad_norm": 0.23197270289572672,
+ "learning_rate": 4.967023130388619e-05,
+ "loss": 0.5144383907318115,
+ "step": 104
+ },
+ {
+ "epoch": 0.05508919202518363,
+ "grad_norm": 0.23320222192603857,
+ "learning_rate": 4.966387502195979e-05,
+ "loss": 0.5134050250053406,
+ "step": 105
+ },
+ {
+ "epoch": 0.055613850996852045,
+ "grad_norm": 0.2552812453771761,
+ "learning_rate": 4.965745852574656e-05,
+ "loss": 0.5225021243095398,
+ "step": 106
+ },
+ {
+ "epoch": 0.05613850996852046,
+ "grad_norm": 0.2322884371893299,
+ "learning_rate": 4.9650981832678684e-05,
+ "loss": 0.5003984570503235,
+ "step": 107
+ },
+ {
+ "epoch": 0.05666316894018888,
+ "grad_norm": 0.22714174562634631,
+ "learning_rate": 4.964444496035186e-05,
+ "loss": 0.5158190131187439,
+ "step": 108
+ },
+ {
+ "epoch": 0.057187827911857295,
+ "grad_norm": 0.22879578407883067,
+ "learning_rate": 4.9637847926525346e-05,
+ "loss": 0.5036455392837524,
+ "step": 109
+ },
+ {
+ "epoch": 0.05771248688352571,
+ "grad_norm": 0.2166892406632336,
+ "learning_rate": 4.963119074912178e-05,
+ "loss": 0.5308535099029541,
+ "step": 110
+ },
+ {
+ "epoch": 0.05823714585519412,
+ "grad_norm": 0.21036625667891182,
+ "learning_rate": 4.962447344622723e-05,
+ "loss": 0.5025012493133545,
+ "step": 111
+ },
+ {
+ "epoch": 0.05876180482686254,
+ "grad_norm": 0.22481506458552633,
+ "learning_rate": 4.961769603609109e-05,
+ "loss": 0.5202745199203491,
+ "step": 112
+ },
+ {
+ "epoch": 0.059286463798530954,
+ "grad_norm": 0.21462119221993772,
+ "learning_rate": 4.96108585371261e-05,
+ "loss": 0.48475122451782227,
+ "step": 113
+ },
+ {
+ "epoch": 0.05981112277019937,
+ "grad_norm": 0.23321086180729633,
+ "learning_rate": 4.960396096790817e-05,
+ "loss": 0.5212013721466064,
+ "step": 114
+ },
+ {
+ "epoch": 0.06033578174186779,
+ "grad_norm": 0.2141551679969153,
+ "learning_rate": 4.95970033471765e-05,
+ "loss": 0.4899653196334839,
+ "step": 115
+ },
+ {
+ "epoch": 0.060860440713536204,
+ "grad_norm": 0.24432098841788752,
+ "learning_rate": 4.958998569383335e-05,
+ "loss": 0.530897855758667,
+ "step": 116
+ },
+ {
+ "epoch": 0.061385099685204614,
+ "grad_norm": 0.2244062539959418,
+ "learning_rate": 4.958290802694412e-05,
+ "loss": 0.5234667658805847,
+ "step": 117
+ },
+ {
+ "epoch": 0.06190975865687303,
+ "grad_norm": 0.22346281532015827,
+ "learning_rate": 4.9575770365737265e-05,
+ "loss": 0.5311257243156433,
+ "step": 118
+ },
+ {
+ "epoch": 0.06243441762854145,
+ "grad_norm": 0.2888198836799783,
+ "learning_rate": 4.956857272960419e-05,
+ "loss": 0.5221022963523865,
+ "step": 119
+ },
+ {
+ "epoch": 0.06295907660020986,
+ "grad_norm": 0.2367678287225713,
+ "learning_rate": 4.956131513809928e-05,
+ "loss": 0.5127741098403931,
+ "step": 120
+ },
+ {
+ "epoch": 0.06348373557187828,
+ "grad_norm": 0.20956211467189642,
+ "learning_rate": 4.955399761093978e-05,
+ "loss": 0.5230909585952759,
+ "step": 121
+ },
+ {
+ "epoch": 0.06400839454354669,
+ "grad_norm": 0.2193688027719425,
+ "learning_rate": 4.9546620168005765e-05,
+ "loss": 0.5042833089828491,
+ "step": 122
+ },
+ {
+ "epoch": 0.06453305351521511,
+ "grad_norm": 0.23260210878827017,
+ "learning_rate": 4.95391828293401e-05,
+ "loss": 0.51006019115448,
+ "step": 123
+ },
+ {
+ "epoch": 0.06505771248688352,
+ "grad_norm": 0.24015857583456987,
+ "learning_rate": 4.953168561514838e-05,
+ "loss": 0.514384388923645,
+ "step": 124
+ },
+ {
+ "epoch": 0.06558237145855195,
+ "grad_norm": 0.20528055378876914,
+ "learning_rate": 4.952412854579884e-05,
+ "loss": 0.49396130442619324,
+ "step": 125
+ },
+ {
+ "epoch": 0.06610703043022036,
+ "grad_norm": 0.2166772494469198,
+ "learning_rate": 4.951651164182236e-05,
+ "loss": 0.4965657591819763,
+ "step": 126
+ },
+ {
+ "epoch": 0.06663168940188877,
+ "grad_norm": 0.25397760382807155,
+ "learning_rate": 4.950883492391236e-05,
+ "loss": 0.524378776550293,
+ "step": 127
+ },
+ {
+ "epoch": 0.06715634837355719,
+ "grad_norm": 0.20346923724942226,
+ "learning_rate": 4.9501098412924754e-05,
+ "loss": 0.5099847316741943,
+ "step": 128
+ },
+ {
+ "epoch": 0.0676810073452256,
+ "grad_norm": 0.2183870932732748,
+ "learning_rate": 4.9493302129877915e-05,
+ "loss": 0.5425564050674438,
+ "step": 129
+ },
+ {
+ "epoch": 0.06820566631689402,
+ "grad_norm": 0.21209367279802577,
+ "learning_rate": 4.9485446095952615e-05,
+ "loss": 0.48642098903656006,
+ "step": 130
+ },
+ {
+ "epoch": 0.06873032528856243,
+ "grad_norm": 0.21179917083725588,
+ "learning_rate": 4.947753033249192e-05,
+ "loss": 0.5301004648208618,
+ "step": 131
+ },
+ {
+ "epoch": 0.06925498426023086,
+ "grad_norm": 0.21033541966956984,
+ "learning_rate": 4.946955486100119e-05,
+ "loss": 0.5264711380004883,
+ "step": 132
+ },
+ {
+ "epoch": 0.06977964323189927,
+ "grad_norm": 0.21916394320437205,
+ "learning_rate": 4.9461519703148e-05,
+ "loss": 0.527003288269043,
+ "step": 133
+ },
+ {
+ "epoch": 0.07030430220356768,
+ "grad_norm": 0.21028309641902634,
+ "learning_rate": 4.945342488076208e-05,
+ "loss": 0.50738525390625,
+ "step": 134
+ },
+ {
+ "epoch": 0.0708289611752361,
+ "grad_norm": 0.20427806833116693,
+ "learning_rate": 4.944527041583524e-05,
+ "loss": 0.5049724578857422,
+ "step": 135
+ },
+ {
+ "epoch": 0.07135362014690451,
+ "grad_norm": 0.22381987551737148,
+ "learning_rate": 4.9437056330521344e-05,
+ "loss": 0.5334229469299316,
+ "step": 136
+ },
+ {
+ "epoch": 0.07187827911857293,
+ "grad_norm": 0.20919152799100757,
+ "learning_rate": 4.9428782647136215e-05,
+ "loss": 0.5008583068847656,
+ "step": 137
+ },
+ {
+ "epoch": 0.07240293809024134,
+ "grad_norm": 0.23260328724401666,
+ "learning_rate": 4.942044938815761e-05,
+ "loss": 0.48956719040870667,
+ "step": 138
+ },
+ {
+ "epoch": 0.07292759706190975,
+ "grad_norm": 0.21749158403760774,
+ "learning_rate": 4.941205657622512e-05,
+ "loss": 0.5170283913612366,
+ "step": 139
+ },
+ {
+ "epoch": 0.07345225603357818,
+ "grad_norm": 0.252859812938656,
+ "learning_rate": 4.9403604234140134e-05,
+ "loss": 0.5194739103317261,
+ "step": 140
+ },
+ {
+ "epoch": 0.07397691500524659,
+ "grad_norm": 0.2415992713618842,
+ "learning_rate": 4.939509238486577e-05,
+ "loss": 0.5247742533683777,
+ "step": 141
+ },
+ {
+ "epoch": 0.07450157397691501,
+ "grad_norm": 0.22977174448684123,
+ "learning_rate": 4.938652105152683e-05,
+ "loss": 0.5246601700782776,
+ "step": 142
+ },
+ {
+ "epoch": 0.07502623294858342,
+ "grad_norm": 0.30547188416500454,
+ "learning_rate": 4.937789025740971e-05,
+ "loss": 0.500076949596405,
+ "step": 143
+ },
+ {
+ "epoch": 0.07555089192025184,
+ "grad_norm": 0.2891987803145593,
+ "learning_rate": 4.9369200025962334e-05,
+ "loss": 0.5170419216156006,
+ "step": 144
+ },
+ {
+ "epoch": 0.07607555089192025,
+ "grad_norm": 0.204395530696793,
+ "learning_rate": 4.9360450380794117e-05,
+ "loss": 0.5127545595169067,
+ "step": 145
+ },
+ {
+ "epoch": 0.07660020986358866,
+ "grad_norm": 0.28029537359032003,
+ "learning_rate": 4.935164134567589e-05,
+ "loss": 0.524910032749176,
+ "step": 146
+ },
+ {
+ "epoch": 0.07712486883525708,
+ "grad_norm": 0.2709250211112369,
+ "learning_rate": 4.9342772944539824e-05,
+ "loss": 0.4903565049171448,
+ "step": 147
+ },
+ {
+ "epoch": 0.0776495278069255,
+ "grad_norm": 0.2073473540606605,
+ "learning_rate": 4.933384520147937e-05,
+ "loss": 0.5083538889884949,
+ "step": 148
+ },
+ {
+ "epoch": 0.07817418677859392,
+ "grad_norm": 0.22431851122273783,
+ "learning_rate": 4.932485814074924e-05,
+ "loss": 0.5073282718658447,
+ "step": 149
+ },
+ {
+ "epoch": 0.07869884575026233,
+ "grad_norm": 0.26180972861077034,
+ "learning_rate": 4.931581178676524e-05,
+ "loss": 0.5109356641769409,
+ "step": 150
+ },
+ {
+ "epoch": 0.07922350472193075,
+ "grad_norm": 0.22251520290809648,
+ "learning_rate": 4.9306706164104295e-05,
+ "loss": 0.5166893005371094,
+ "step": 151
+ },
+ {
+ "epoch": 0.07974816369359916,
+ "grad_norm": 0.21380115992979964,
+ "learning_rate": 4.929754129750434e-05,
+ "loss": 0.5094020366668701,
+ "step": 152
+ },
+ {
+ "epoch": 0.08027282266526757,
+ "grad_norm": 0.25317541372355157,
+ "learning_rate": 4.9288317211864286e-05,
+ "loss": 0.5010517835617065,
+ "step": 153
+ },
+ {
+ "epoch": 0.080797481636936,
+ "grad_norm": 0.22194749433196714,
+ "learning_rate": 4.927903393224387e-05,
+ "loss": 0.4915440082550049,
+ "step": 154
+ },
+ {
+ "epoch": 0.0813221406086044,
+ "grad_norm": 0.21407034391897298,
+ "learning_rate": 4.926969148386372e-05,
+ "loss": 0.5005519986152649,
+ "step": 155
+ },
+ {
+ "epoch": 0.08184679958027283,
+ "grad_norm": 0.2546125343376434,
+ "learning_rate": 4.926028989210516e-05,
+ "loss": 0.5167546272277832,
+ "step": 156
+ },
+ {
+ "epoch": 0.08237145855194124,
+ "grad_norm": 0.2329587981774556,
+ "learning_rate": 4.925082918251021e-05,
+ "loss": 0.5154269933700562,
+ "step": 157
+ },
+ {
+ "epoch": 0.08289611752360965,
+ "grad_norm": 0.22108085309109965,
+ "learning_rate": 4.924130938078151e-05,
+ "loss": 0.5080733895301819,
+ "step": 158
+ },
+ {
+ "epoch": 0.08342077649527807,
+ "grad_norm": 0.2651559584149174,
+ "learning_rate": 4.923173051278223e-05,
+ "loss": 0.5120258331298828,
+ "step": 159
+ },
+ {
+ "epoch": 0.08394543546694648,
+ "grad_norm": 0.23458414436293315,
+ "learning_rate": 4.922209260453599e-05,
+ "loss": 0.5022101402282715,
+ "step": 160
+ },
+ {
+ "epoch": 0.0844700944386149,
+ "grad_norm": 0.2446638948059357,
+ "learning_rate": 4.921239568222685e-05,
+ "loss": 0.5126672983169556,
+ "step": 161
+ },
+ {
+ "epoch": 0.08499475341028331,
+ "grad_norm": 0.2310469793232587,
+ "learning_rate": 4.920263977219917e-05,
+ "loss": 0.5054258704185486,
+ "step": 162
+ },
+ {
+ "epoch": 0.08551941238195174,
+ "grad_norm": 0.2384997159826766,
+ "learning_rate": 4.919282490095757e-05,
+ "loss": 0.5190805196762085,
+ "step": 163
+ },
+ {
+ "epoch": 0.08604407135362015,
+ "grad_norm": 0.2326069379441097,
+ "learning_rate": 4.9182951095166875e-05,
+ "loss": 0.5004345178604126,
+ "step": 164
+ },
+ {
+ "epoch": 0.08656873032528856,
+ "grad_norm": 0.22035795902394487,
+ "learning_rate": 4.917301838165198e-05,
+ "loss": 0.5224157571792603,
+ "step": 165
+ },
+ {
+ "epoch": 0.08709338929695698,
+ "grad_norm": 0.2246927535161224,
+ "learning_rate": 4.916302678739786e-05,
+ "loss": 0.5029760599136353,
+ "step": 166
+ },
+ {
+ "epoch": 0.08761804826862539,
+ "grad_norm": 0.25648353178321964,
+ "learning_rate": 4.9152976339549436e-05,
+ "loss": 0.4898151457309723,
+ "step": 167
+ },
+ {
+ "epoch": 0.08814270724029381,
+ "grad_norm": 0.22250684833170264,
+ "learning_rate": 4.9142867065411526e-05,
+ "loss": 0.5149059295654297,
+ "step": 168
+ },
+ {
+ "epoch": 0.08866736621196222,
+ "grad_norm": 0.2070229361127925,
+ "learning_rate": 4.9132698992448764e-05,
+ "loss": 0.5157654881477356,
+ "step": 169
+ },
+ {
+ "epoch": 0.08919202518363065,
+ "grad_norm": 0.2275155272949515,
+ "learning_rate": 4.9122472148285524e-05,
+ "loss": 0.5038745403289795,
+ "step": 170
+ },
+ {
+ "epoch": 0.08971668415529906,
+ "grad_norm": 0.22703137619933889,
+ "learning_rate": 4.9112186560705855e-05,
+ "loss": 0.5031003952026367,
+ "step": 171
+ },
+ {
+ "epoch": 0.09024134312696747,
+ "grad_norm": 0.21967376293140634,
+ "learning_rate": 4.91018422576534e-05,
+ "loss": 0.4934842586517334,
+ "step": 172
+ },
+ {
+ "epoch": 0.09076600209863589,
+ "grad_norm": 0.24955642046140025,
+ "learning_rate": 4.9091439267231315e-05,
+ "loss": 0.5058956742286682,
+ "step": 173
+ },
+ {
+ "epoch": 0.0912906610703043,
+ "grad_norm": 0.22133414408844335,
+ "learning_rate": 4.908097761770219e-05,
+ "loss": 0.5094645619392395,
+ "step": 174
+ },
+ {
+ "epoch": 0.09181532004197272,
+ "grad_norm": 0.21365544683422769,
+ "learning_rate": 4.907045733748799e-05,
+ "loss": 0.5109387636184692,
+ "step": 175
+ },
+ {
+ "epoch": 0.09233997901364113,
+ "grad_norm": 0.21377053096310691,
+ "learning_rate": 4.905987845516996e-05,
+ "loss": 0.5152729749679565,
+ "step": 176
+ },
+ {
+ "epoch": 0.09286463798530954,
+ "grad_norm": 0.22518302750830563,
+ "learning_rate": 4.904924099948855e-05,
+ "loss": 0.5165923237800598,
+ "step": 177
+ },
+ {
+ "epoch": 0.09338929695697797,
+ "grad_norm": 0.2107705920567083,
+ "learning_rate": 4.903854499934335e-05,
+ "loss": 0.5165438652038574,
+ "step": 178
+ },
+ {
+ "epoch": 0.09391395592864638,
+ "grad_norm": 0.20518792387882806,
+ "learning_rate": 4.9027790483792996e-05,
+ "loss": 0.4956984519958496,
+ "step": 179
+ },
+ {
+ "epoch": 0.0944386149003148,
+ "grad_norm": 0.23150703445003076,
+ "learning_rate": 4.90169774820551e-05,
+ "loss": 0.49541592597961426,
+ "step": 180
+ },
+ {
+ "epoch": 0.09496327387198321,
+ "grad_norm": 0.23096057489180669,
+ "learning_rate": 4.9006106023506175e-05,
+ "loss": 0.506559431552887,
+ "step": 181
+ },
+ {
+ "epoch": 0.09548793284365163,
+ "grad_norm": 0.21391359463240722,
+ "learning_rate": 4.899517613768153e-05,
+ "loss": 0.5075074434280396,
+ "step": 182
+ },
+ {
+ "epoch": 0.09601259181532004,
+ "grad_norm": 0.21167720103110035,
+ "learning_rate": 4.8984187854275223e-05,
+ "loss": 0.5093033909797668,
+ "step": 183
+ },
+ {
+ "epoch": 0.09653725078698845,
+ "grad_norm": 0.22442243951240298,
+ "learning_rate": 4.8973141203139954e-05,
+ "loss": 0.5153380632400513,
+ "step": 184
+ },
+ {
+ "epoch": 0.09706190975865688,
+ "grad_norm": 0.21165160822454496,
+ "learning_rate": 4.8962036214287004e-05,
+ "loss": 0.5167010426521301,
+ "step": 185
+ },
+ {
+ "epoch": 0.09758656873032528,
+ "grad_norm": 0.2154572342704065,
+ "learning_rate": 4.895087291788615e-05,
+ "loss": 0.5220831632614136,
+ "step": 186
+ },
+ {
+ "epoch": 0.09811122770199371,
+ "grad_norm": 0.20407873014033961,
+ "learning_rate": 4.8939651344265565e-05,
+ "loss": 0.5085086822509766,
+ "step": 187
+ },
+ {
+ "epoch": 0.09863588667366212,
+ "grad_norm": 0.21245583404945137,
+ "learning_rate": 4.892837152391174e-05,
+ "loss": 0.5149598121643066,
+ "step": 188
+ },
+ {
+ "epoch": 0.09916054564533054,
+ "grad_norm": 0.21281283014647004,
+ "learning_rate": 4.891703348746944e-05,
+ "loss": 0.5080534219741821,
+ "step": 189
+ },
+ {
+ "epoch": 0.09968520461699895,
+ "grad_norm": 0.2124159176652017,
+ "learning_rate": 4.890563726574157e-05,
+ "loss": 0.5082594156265259,
+ "step": 190
+ },
+ {
+ "epoch": 0.10020986358866736,
+ "grad_norm": 0.2149512540650875,
+ "learning_rate": 4.8894182889689094e-05,
+ "loss": 0.5026682615280151,
+ "step": 191
+ },
+ {
+ "epoch": 0.10073452256033578,
+ "grad_norm": 0.20288484980072988,
+ "learning_rate": 4.888267039043101e-05,
+ "loss": 0.495418518781662,
+ "step": 192
+ },
+ {
+ "epoch": 0.1012591815320042,
+ "grad_norm": 0.22946010947962428,
+ "learning_rate": 4.887109979924419e-05,
+ "loss": 0.5350666046142578,
+ "step": 193
+ },
+ {
+ "epoch": 0.10178384050367262,
+ "grad_norm": 0.21412922978876073,
+ "learning_rate": 4.8859471147563334e-05,
+ "loss": 0.5078902244567871,
+ "step": 194
+ },
+ {
+ "epoch": 0.10230849947534103,
+ "grad_norm": 0.2230983644346339,
+ "learning_rate": 4.88477844669809e-05,
+ "loss": 0.499994158744812,
+ "step": 195
+ },
+ {
+ "epoch": 0.10283315844700944,
+ "grad_norm": 0.22215749848369246,
+ "learning_rate": 4.883603978924698e-05,
+ "loss": 0.4992578625679016,
+ "step": 196
+ },
+ {
+ "epoch": 0.10335781741867786,
+ "grad_norm": 0.21242156260092737,
+ "learning_rate": 4.882423714626923e-05,
+ "loss": 0.5196024775505066,
+ "step": 197
+ },
+ {
+ "epoch": 0.10388247639034627,
+ "grad_norm": 0.21173603443443906,
+ "learning_rate": 4.8812376570112786e-05,
+ "loss": 0.5052409172058105,
+ "step": 198
+ },
+ {
+ "epoch": 0.1044071353620147,
+ "grad_norm": 0.21408855227581017,
+ "learning_rate": 4.8800458093000186e-05,
+ "loss": 0.5189114809036255,
+ "step": 199
+ },
+ {
+ "epoch": 0.1049317943336831,
+ "grad_norm": 0.21246535161690677,
+ "learning_rate": 4.878848174731126e-05,
+ "loss": 0.5028671026229858,
+ "step": 200
+ },
+ {
+ "epoch": 0.10545645330535153,
+ "grad_norm": 0.204372770965071,
+ "learning_rate": 4.877644756558305e-05,
+ "loss": 0.49684005975723267,
+ "step": 201
+ },
+ {
+ "epoch": 0.10598111227701994,
+ "grad_norm": 0.21290259979832005,
+ "learning_rate": 4.8764355580509755e-05,
+ "loss": 0.5039458870887756,
+ "step": 202
+ },
+ {
+ "epoch": 0.10650577124868835,
+ "grad_norm": 0.2117143608609357,
+ "learning_rate": 4.875220582494257e-05,
+ "loss": 0.5215772986412048,
+ "step": 203
+ },
+ {
+ "epoch": 0.10703043022035677,
+ "grad_norm": 0.20960185194467149,
+ "learning_rate": 4.8739998331889666e-05,
+ "loss": 0.5018843412399292,
+ "step": 204
+ },
+ {
+ "epoch": 0.10755508919202518,
+ "grad_norm": 0.19614871887812332,
+ "learning_rate": 4.872773313451608e-05,
+ "loss": 0.4955351948738098,
+ "step": 205
+ },
+ {
+ "epoch": 0.1080797481636936,
+ "grad_norm": 0.20752989026796015,
+ "learning_rate": 4.8715410266143605e-05,
+ "loss": 0.496296226978302,
+ "step": 206
+ },
+ {
+ "epoch": 0.10860440713536201,
+ "grad_norm": 0.21493210321580256,
+ "learning_rate": 4.87030297602507e-05,
+ "loss": 0.5092719793319702,
+ "step": 207
+ },
+ {
+ "epoch": 0.10912906610703044,
+ "grad_norm": 0.21388940324614214,
+ "learning_rate": 4.869059165047245e-05,
+ "loss": 0.5000571012496948,
+ "step": 208
+ },
+ {
+ "epoch": 0.10965372507869885,
+ "grad_norm": 0.22463723620435969,
+ "learning_rate": 4.8678095970600394e-05,
+ "loss": 0.489624559879303,
+ "step": 209
+ },
+ {
+ "epoch": 0.11017838405036726,
+ "grad_norm": 0.2078348469979708,
+ "learning_rate": 4.866554275458252e-05,
+ "loss": 0.5135753154754639,
+ "step": 210
+ },
+ {
+ "epoch": 0.11070304302203568,
+ "grad_norm": 0.22672243186891075,
+ "learning_rate": 4.8652932036523075e-05,
+ "loss": 0.5013821125030518,
+ "step": 211
+ },
+ {
+ "epoch": 0.11122770199370409,
+ "grad_norm": 0.21000848519694862,
+ "learning_rate": 4.864026385068259e-05,
+ "loss": 0.4991903007030487,
+ "step": 212
+ },
+ {
+ "epoch": 0.11175236096537251,
+ "grad_norm": 0.22481756061489036,
+ "learning_rate": 4.862753823147766e-05,
+ "loss": 0.4913789629936218,
+ "step": 213
+ },
+ {
+ "epoch": 0.11227701993704092,
+ "grad_norm": 0.23243435581868596,
+ "learning_rate": 4.8614755213480964e-05,
+ "loss": 0.4873138666152954,
+ "step": 214
+ },
+ {
+ "epoch": 0.11280167890870933,
+ "grad_norm": 0.20773568721432215,
+ "learning_rate": 4.8601914831421105e-05,
+ "loss": 0.48931920528411865,
+ "step": 215
+ },
+ {
+ "epoch": 0.11332633788037776,
+ "grad_norm": 0.22587538106795335,
+ "learning_rate": 4.85890171201825e-05,
+ "loss": 0.5019721984863281,
+ "step": 216
+ },
+ {
+ "epoch": 0.11385099685204617,
+ "grad_norm": 0.22090931269953926,
+ "learning_rate": 4.8576062114805375e-05,
+ "loss": 0.4983716309070587,
+ "step": 217
+ },
+ {
+ "epoch": 0.11437565582371459,
+ "grad_norm": 0.23135790381711147,
+ "learning_rate": 4.856304985048555e-05,
+ "loss": 0.49929213523864746,
+ "step": 218
+ },
+ {
+ "epoch": 0.114900314795383,
+ "grad_norm": 0.20141970500097148,
+ "learning_rate": 4.854998036257446e-05,
+ "loss": 0.5061747431755066,
+ "step": 219
+ },
+ {
+ "epoch": 0.11542497376705142,
+ "grad_norm": 0.2276305065122677,
+ "learning_rate": 4.853685368657896e-05,
+ "loss": 0.5087096691131592,
+ "step": 220
+ },
+ {
+ "epoch": 0.11594963273871983,
+ "grad_norm": 0.21662008709396724,
+ "learning_rate": 4.85236698581613e-05,
+ "loss": 0.4714198708534241,
+ "step": 221
+ },
+ {
+ "epoch": 0.11647429171038824,
+ "grad_norm": 0.20490548513016843,
+ "learning_rate": 4.8510428913138975e-05,
+ "loss": 0.4982035756111145,
+ "step": 222
+ },
+ {
+ "epoch": 0.11699895068205667,
+ "grad_norm": 0.21096446203066194,
+ "learning_rate": 4.849713088748469e-05,
+ "loss": 0.501306414604187,
+ "step": 223
+ },
+ {
+ "epoch": 0.11752360965372508,
+ "grad_norm": 0.2141638900206865,
+ "learning_rate": 4.848377581732619e-05,
+ "loss": 0.4849642217159271,
+ "step": 224
+ },
+ {
+ "epoch": 0.1180482686253935,
+ "grad_norm": 0.2130390168933917,
+ "learning_rate": 4.847036373894622e-05,
+ "loss": 0.4944990873336792,
+ "step": 225
+ },
+ {
+ "epoch": 0.11857292759706191,
+ "grad_norm": 0.2010142480319778,
+ "learning_rate": 4.845689468878238e-05,
+ "loss": 0.4936804175376892,
+ "step": 226
+ },
+ {
+ "epoch": 0.11909758656873033,
+ "grad_norm": 0.21222278524608443,
+ "learning_rate": 4.8443368703427087e-05,
+ "loss": 0.5042952299118042,
+ "step": 227
+ },
+ {
+ "epoch": 0.11962224554039874,
+ "grad_norm": 0.21725247300914316,
+ "learning_rate": 4.8429785819627393e-05,
+ "loss": 0.4913664162158966,
+ "step": 228
+ },
+ {
+ "epoch": 0.12014690451206715,
+ "grad_norm": 0.20413867499875316,
+ "learning_rate": 4.8416146074284964e-05,
+ "loss": 0.49605828523635864,
+ "step": 229
+ },
+ {
+ "epoch": 0.12067156348373557,
+ "grad_norm": 0.2043327492615352,
+ "learning_rate": 4.840244950445593e-05,
+ "loss": 0.5114777088165283,
+ "step": 230
+ },
+ {
+ "epoch": 0.12119622245540398,
+ "grad_norm": 0.19719405748044208,
+ "learning_rate": 4.838869614735081e-05,
+ "loss": 0.500380277633667,
+ "step": 231
+ },
+ {
+ "epoch": 0.12172088142707241,
+ "grad_norm": 0.21428024536826415,
+ "learning_rate": 4.837488604033439e-05,
+ "loss": 0.5037723183631897,
+ "step": 232
+ },
+ {
+ "epoch": 0.12224554039874082,
+ "grad_norm": 0.2018208533608718,
+ "learning_rate": 4.8361019220925646e-05,
+ "loss": 0.48484015464782715,
+ "step": 233
+ },
+ {
+ "epoch": 0.12277019937040923,
+ "grad_norm": 0.20954218265596586,
+ "learning_rate": 4.834709572679762e-05,
+ "loss": 0.48974648118019104,
+ "step": 234
+ },
+ {
+ "epoch": 0.12329485834207765,
+ "grad_norm": 0.21528975577905585,
+ "learning_rate": 4.833311559577732e-05,
+ "loss": 0.49275070428848267,
+ "step": 235
+ },
+ {
+ "epoch": 0.12381951731374606,
+ "grad_norm": 0.213862353527007,
+ "learning_rate": 4.8319078865845653e-05,
+ "loss": 0.5088430643081665,
+ "step": 236
+ },
+ {
+ "epoch": 0.12434417628541448,
+ "grad_norm": 0.21539776648038736,
+ "learning_rate": 4.830498557513725e-05,
+ "loss": 0.50986647605896,
+ "step": 237
+ },
+ {
+ "epoch": 0.1248688352570829,
+ "grad_norm": 0.2249759362737276,
+ "learning_rate": 4.8290835761940435e-05,
+ "loss": 0.5068402290344238,
+ "step": 238
+ },
+ {
+ "epoch": 0.12539349422875132,
+ "grad_norm": 0.22445974527083887,
+ "learning_rate": 4.8276629464697085e-05,
+ "loss": 0.5031242370605469,
+ "step": 239
+ },
+ {
+ "epoch": 0.1259181532004197,
+ "grad_norm": 0.2273192490996244,
+ "learning_rate": 4.826236672200253e-05,
+ "loss": 0.5126334428787231,
+ "step": 240
+ },
+ {
+ "epoch": 0.12644281217208814,
+ "grad_norm": 0.207187361029785,
+ "learning_rate": 4.824804757260545e-05,
+ "loss": 0.48009148240089417,
+ "step": 241
+ },
+ {
+ "epoch": 0.12696747114375656,
+ "grad_norm": 0.2104671545211291,
+ "learning_rate": 4.823367205540776e-05,
+ "loss": 0.5050241947174072,
+ "step": 242
+ },
+ {
+ "epoch": 0.12749213011542498,
+ "grad_norm": 0.2124476964001143,
+ "learning_rate": 4.8219240209464536e-05,
+ "loss": 0.5104424953460693,
+ "step": 243
+ },
+ {
+ "epoch": 0.12801678908709338,
+ "grad_norm": 0.2091591091815355,
+ "learning_rate": 4.8204752073983854e-05,
+ "loss": 0.5073314309120178,
+ "step": 244
+ },
+ {
+ "epoch": 0.1285414480587618,
+ "grad_norm": 0.1944372243616572,
+ "learning_rate": 4.819020768832675e-05,
+ "loss": 0.48875677585601807,
+ "step": 245
+ },
+ {
+ "epoch": 0.12906610703043023,
+ "grad_norm": 0.21654766569004938,
+ "learning_rate": 4.817560709200707e-05,
+ "loss": 0.4805977940559387,
+ "step": 246
+ },
+ {
+ "epoch": 0.12959076600209862,
+ "grad_norm": 0.2108915028659573,
+ "learning_rate": 4.816095032469134e-05,
+ "loss": 0.49152469635009766,
+ "step": 247
+ },
+ {
+ "epoch": 0.13011542497376705,
+ "grad_norm": 0.20689611001281927,
+ "learning_rate": 4.814623742619874e-05,
+ "loss": 0.4831926226615906,
+ "step": 248
+ },
+ {
+ "epoch": 0.13064008394543547,
+ "grad_norm": 0.1985452705406765,
+ "learning_rate": 4.813146843650091e-05,
+ "loss": 0.5019830465316772,
+ "step": 249
+ },
+ {
+ "epoch": 0.1311647429171039,
+ "grad_norm": 0.23754724572080882,
+ "learning_rate": 4.811664339572189e-05,
+ "loss": 0.48042142391204834,
+ "step": 250
+ },
+ {
+ "epoch": 0.1316894018887723,
+ "grad_norm": 0.21389487742120458,
+ "learning_rate": 4.8101762344138e-05,
+ "loss": 0.5036448240280151,
+ "step": 251
+ },
+ {
+ "epoch": 0.1322140608604407,
+ "grad_norm": 0.21311035268619086,
+ "learning_rate": 4.808682532217772e-05,
+ "loss": 0.4970911145210266,
+ "step": 252
+ },
+ {
+ "epoch": 0.13273871983210914,
+ "grad_norm": 0.22918034580875016,
+ "learning_rate": 4.80718323704216e-05,
+ "loss": 0.4878022074699402,
+ "step": 253
+ },
+ {
+ "epoch": 0.13326337880377753,
+ "grad_norm": 0.22389522912197427,
+ "learning_rate": 4.8056783529602137e-05,
+ "loss": 0.48091578483581543,
+ "step": 254
+ },
+ {
+ "epoch": 0.13378803777544596,
+ "grad_norm": 0.2178556232330264,
+ "learning_rate": 4.8041678840603646e-05,
+ "loss": 0.49805814027786255,
+ "step": 255
+ },
+ {
+ "epoch": 0.13431269674711438,
+ "grad_norm": 0.24973060737806424,
+ "learning_rate": 4.802651834446221e-05,
+ "loss": 0.49214544892311096,
+ "step": 256
+ },
+ {
+ "epoch": 0.1348373557187828,
+ "grad_norm": 0.24336805559494964,
+ "learning_rate": 4.8011302082365495e-05,
+ "loss": 0.5148031711578369,
+ "step": 257
+ },
+ {
+ "epoch": 0.1353620146904512,
+ "grad_norm": 0.23032848996708188,
+ "learning_rate": 4.799603009565266e-05,
+ "loss": 0.5113047361373901,
+ "step": 258
+ },
+ {
+ "epoch": 0.13588667366211962,
+ "grad_norm": 0.2151292845892052,
+ "learning_rate": 4.79807024258143e-05,
+ "loss": 0.48844093084335327,
+ "step": 259
+ },
+ {
+ "epoch": 0.13641133263378805,
+ "grad_norm": 0.25062002830467534,
+ "learning_rate": 4.796531911449225e-05,
+ "loss": 0.49366849660873413,
+ "step": 260
+ },
+ {
+ "epoch": 0.13693599160545644,
+ "grad_norm": 0.22872577435643346,
+ "learning_rate": 4.7949880203479525e-05,
+ "loss": 0.50008225440979,
+ "step": 261
+ },
+ {
+ "epoch": 0.13746065057712487,
+ "grad_norm": 0.22414490489780955,
+ "learning_rate": 4.7934385734720185e-05,
+ "loss": 0.4895337224006653,
+ "step": 262
+ },
+ {
+ "epoch": 0.1379853095487933,
+ "grad_norm": 0.2814935999129343,
+ "learning_rate": 4.791883575030924e-05,
+ "loss": 0.4864671230316162,
+ "step": 263
+ },
+ {
+ "epoch": 0.1385099685204617,
+ "grad_norm": 0.22174651207639173,
+ "learning_rate": 4.790323029249252e-05,
+ "loss": 0.49753478169441223,
+ "step": 264
+ },
+ {
+ "epoch": 0.1390346274921301,
+ "grad_norm": 0.24997778960811773,
+ "learning_rate": 4.7887569403666555e-05,
+ "loss": 0.4836593270301819,
+ "step": 265
+ },
+ {
+ "epoch": 0.13955928646379853,
+ "grad_norm": 0.2654347615420452,
+ "learning_rate": 4.787185312637847e-05,
+ "loss": 0.505037784576416,
+ "step": 266
+ },
+ {
+ "epoch": 0.14008394543546696,
+ "grad_norm": 0.2257998691620493,
+ "learning_rate": 4.7856081503325875e-05,
+ "loss": 0.5027364492416382,
+ "step": 267
+ },
+ {
+ "epoch": 0.14060860440713535,
+ "grad_norm": 0.2602576139887493,
+ "learning_rate": 4.784025457735675e-05,
+ "loss": 0.510755717754364,
+ "step": 268
+ },
+ {
+ "epoch": 0.14113326337880377,
+ "grad_norm": 0.252596538962873,
+ "learning_rate": 4.7824372391469286e-05,
+ "loss": 0.5040655732154846,
+ "step": 269
+ },
+ {
+ "epoch": 0.1416579223504722,
+ "grad_norm": 0.2088630112867918,
+ "learning_rate": 4.780843498881185e-05,
+ "loss": 0.4826688766479492,
+ "step": 270
+ },
+ {
+ "epoch": 0.14218258132214062,
+ "grad_norm": 0.2451388718936559,
+ "learning_rate": 4.779244241268279e-05,
+ "loss": 0.49606287479400635,
+ "step": 271
+ },
+ {
+ "epoch": 0.14270724029380902,
+ "grad_norm": 0.21842578127387305,
+ "learning_rate": 4.777639470653035e-05,
+ "loss": 0.4882327914237976,
+ "step": 272
+ },
+ {
+ "epoch": 0.14323189926547744,
+ "grad_norm": 0.20889116915462674,
+ "learning_rate": 4.7760291913952554e-05,
+ "loss": 0.48267611861228943,
+ "step": 273
+ },
+ {
+ "epoch": 0.14375655823714587,
+ "grad_norm": 0.20361079261167978,
+ "learning_rate": 4.77441340786971e-05,
+ "loss": 0.4839272201061249,
+ "step": 274
+ },
+ {
+ "epoch": 0.14428121720881426,
+ "grad_norm": 0.24265346582567346,
+ "learning_rate": 4.7727921244661196e-05,
+ "loss": 0.5002747774124146,
+ "step": 275
+ },
+ {
+ "epoch": 0.14480587618048268,
+ "grad_norm": 0.21550771493613938,
+ "learning_rate": 4.771165345589148e-05,
+ "loss": 0.4920514225959778,
+ "step": 276
+ },
+ {
+ "epoch": 0.1453305351521511,
+ "grad_norm": 0.2110925316802284,
+ "learning_rate": 4.76953307565839e-05,
+ "loss": 0.4966445565223694,
+ "step": 277
+ },
+ {
+ "epoch": 0.1458551941238195,
+ "grad_norm": 0.22447923307166887,
+ "learning_rate": 4.767895319108358e-05,
+ "loss": 0.4997980296611786,
+ "step": 278
+ },
+ {
+ "epoch": 0.14637985309548793,
+ "grad_norm": 0.23215569266848568,
+ "learning_rate": 4.7662520803884684e-05,
+ "loss": 0.4889225661754608,
+ "step": 279
+ },
+ {
+ "epoch": 0.14690451206715635,
+ "grad_norm": 0.20805333548773317,
+ "learning_rate": 4.764603363963035e-05,
+ "loss": 0.4929523468017578,
+ "step": 280
+ },
+ {
+ "epoch": 0.14742917103882477,
+ "grad_norm": 0.21144835822202035,
+ "learning_rate": 4.7629491743112485e-05,
+ "loss": 0.4972894787788391,
+ "step": 281
+ },
+ {
+ "epoch": 0.14795383001049317,
+ "grad_norm": 0.20475807515777222,
+ "learning_rate": 4.7612895159271744e-05,
+ "loss": 0.485750287771225,
+ "step": 282
+ },
+ {
+ "epoch": 0.1484784889821616,
+ "grad_norm": 0.20111772833692326,
+ "learning_rate": 4.759624393319732e-05,
+ "loss": 0.4783627390861511,
+ "step": 283
+ },
+ {
+ "epoch": 0.14900314795383002,
+ "grad_norm": 0.1865822060873931,
+ "learning_rate": 4.757953811012688e-05,
+ "loss": 0.4868030250072479,
+ "step": 284
+ },
+ {
+ "epoch": 0.1495278069254984,
+ "grad_norm": 0.20237500636134131,
+ "learning_rate": 4.7562777735446374e-05,
+ "loss": 0.4796566963195801,
+ "step": 285
+ },
+ {
+ "epoch": 0.15005246589716684,
+ "grad_norm": 0.20143822492012056,
+ "learning_rate": 4.7545962854690014e-05,
+ "loss": 0.48732325434684753,
+ "step": 286
+ },
+ {
+ "epoch": 0.15057712486883526,
+ "grad_norm": 0.20004714827487116,
+ "learning_rate": 4.752909351354006e-05,
+ "loss": 0.48853757977485657,
+ "step": 287
+ },
+ {
+ "epoch": 0.15110178384050368,
+ "grad_norm": 0.1976470703620628,
+ "learning_rate": 4.7512169757826735e-05,
+ "loss": 0.48465389013290405,
+ "step": 288
+ },
+ {
+ "epoch": 0.15162644281217208,
+ "grad_norm": 0.20043605391798197,
+ "learning_rate": 4.7495191633528096e-05,
+ "loss": 0.48968803882598877,
+ "step": 289
+ },
+ {
+ "epoch": 0.1521511017838405,
+ "grad_norm": 0.1960330993311956,
+ "learning_rate": 4.747815918676991e-05,
+ "loss": 0.49025195837020874,
+ "step": 290
+ },
+ {
+ "epoch": 0.15267576075550893,
+ "grad_norm": 0.21900091811742575,
+ "learning_rate": 4.7461072463825507e-05,
+ "loss": 0.5043517351150513,
+ "step": 291
+ },
+ {
+ "epoch": 0.15320041972717732,
+ "grad_norm": 0.19788639202925754,
+ "learning_rate": 4.74439315111157e-05,
+ "loss": 0.49339932203292847,
+ "step": 292
+ },
+ {
+ "epoch": 0.15372507869884575,
+ "grad_norm": 0.21152466070201445,
+ "learning_rate": 4.742673637520862e-05,
+ "loss": 0.5017908215522766,
+ "step": 293
+ },
+ {
+ "epoch": 0.15424973767051417,
+ "grad_norm": 0.20904752908869453,
+ "learning_rate": 4.740948710281961e-05,
+ "loss": 0.5105178356170654,
+ "step": 294
+ },
+ {
+ "epoch": 0.1547743966421826,
+ "grad_norm": 0.22111503611422464,
+ "learning_rate": 4.7392183740811055e-05,
+ "loss": 0.5002360939979553,
+ "step": 295
+ },
+ {
+ "epoch": 0.155299055613851,
+ "grad_norm": 0.2049330244999369,
+ "learning_rate": 4.737482633619235e-05,
+ "loss": 0.4924122095108032,
+ "step": 296
+ },
+ {
+ "epoch": 0.1558237145855194,
+ "grad_norm": 0.19795733631000484,
+ "learning_rate": 4.735741493611966e-05,
+ "loss": 0.5041211843490601,
+ "step": 297
+ },
+ {
+ "epoch": 0.15634837355718784,
+ "grad_norm": 0.2154458207780425,
+ "learning_rate": 4.7339949587895846e-05,
+ "loss": 0.48981621861457825,
+ "step": 298
+ },
+ {
+ "epoch": 0.15687303252885623,
+ "grad_norm": 0.20407512819227694,
+ "learning_rate": 4.732243033897038e-05,
+ "loss": 0.48598241806030273,
+ "step": 299
+ },
+ {
+ "epoch": 0.15739769150052466,
+ "grad_norm": 0.20135634283286236,
+ "learning_rate": 4.730485723693912e-05,
+ "loss": 0.5038883686065674,
+ "step": 300
+ },
+ {
+ "epoch": 0.15792235047219308,
+ "grad_norm": 0.21276990576284707,
+ "learning_rate": 4.7287230329544266e-05,
+ "loss": 0.4827083349227905,
+ "step": 301
+ },
+ {
+ "epoch": 0.1584470094438615,
+ "grad_norm": 0.2141775797258927,
+ "learning_rate": 4.726954966467415e-05,
+ "loss": 0.49595269560813904,
+ "step": 302
+ },
+ {
+ "epoch": 0.1589716684155299,
+ "grad_norm": 0.20197523230197592,
+ "learning_rate": 4.7251815290363214e-05,
+ "loss": 0.49274614453315735,
+ "step": 303
+ },
+ {
+ "epoch": 0.15949632738719832,
+ "grad_norm": 0.20654827342690685,
+ "learning_rate": 4.723402725479176e-05,
+ "loss": 0.4889271557331085,
+ "step": 304
+ },
+ {
+ "epoch": 0.16002098635886675,
+ "grad_norm": 0.18717180310640436,
+ "learning_rate": 4.721618560628591e-05,
+ "loss": 0.4777800440788269,
+ "step": 305
+ },
+ {
+ "epoch": 0.16054564533053514,
+ "grad_norm": 0.20306910861588276,
+ "learning_rate": 4.7198290393317436e-05,
+ "loss": 0.4965876638889313,
+ "step": 306
+ },
+ {
+ "epoch": 0.16107030430220357,
+ "grad_norm": 0.20602310369943766,
+ "learning_rate": 4.71803416645036e-05,
+ "loss": 0.48937660455703735,
+ "step": 307
+ },
+ {
+ "epoch": 0.161594963273872,
+ "grad_norm": 0.18879654253536324,
+ "learning_rate": 4.7162339468607094e-05,
+ "loss": 0.46815598011016846,
+ "step": 308
+ },
+ {
+ "epoch": 0.1621196222455404,
+ "grad_norm": 0.21234717675548817,
+ "learning_rate": 4.714428385453587e-05,
+ "loss": 0.4768769443035126,
+ "step": 309
+ },
+ {
+ "epoch": 0.1626442812172088,
+ "grad_norm": 0.21426369960607713,
+ "learning_rate": 4.7126174871342974e-05,
+ "loss": 0.481756329536438,
+ "step": 310
+ },
+ {
+ "epoch": 0.16316894018887723,
+ "grad_norm": 0.20884490820339016,
+ "learning_rate": 4.710801256822646e-05,
+ "loss": 0.4855930507183075,
+ "step": 311
+ },
+ {
+ "epoch": 0.16369359916054566,
+ "grad_norm": 0.20991205135866156,
+ "learning_rate": 4.7089796994529234e-05,
+ "loss": 0.4742157459259033,
+ "step": 312
+ },
+ {
+ "epoch": 0.16421825813221405,
+ "grad_norm": 0.19810536500544093,
+ "learning_rate": 4.707152819973894e-05,
+ "loss": 0.4848288893699646,
+ "step": 313
+ },
+ {
+ "epoch": 0.16474291710388247,
+ "grad_norm": 0.20394512395283337,
+ "learning_rate": 4.7053206233487805e-05,
+ "loss": 0.48789310455322266,
+ "step": 314
+ },
+ {
+ "epoch": 0.1652675760755509,
+ "grad_norm": 0.1971541018498332,
+ "learning_rate": 4.703483114555251e-05,
+ "loss": 0.4921473264694214,
+ "step": 315
+ },
+ {
+ "epoch": 0.1657922350472193,
+ "grad_norm": 0.19793628207240357,
+ "learning_rate": 4.701640298585405e-05,
+ "loss": 0.4894387722015381,
+ "step": 316
+ },
+ {
+ "epoch": 0.16631689401888772,
+ "grad_norm": 0.20445254694916395,
+ "learning_rate": 4.699792180445761e-05,
+ "loss": 0.5037619471549988,
+ "step": 317
+ },
+ {
+ "epoch": 0.16684155299055614,
+ "grad_norm": 0.20226836031933246,
+ "learning_rate": 4.6979387651572435e-05,
+ "loss": 0.4965515434741974,
+ "step": 318
+ },
+ {
+ "epoch": 0.16736621196222456,
+ "grad_norm": 0.2039065174051651,
+ "learning_rate": 4.696080057755165e-05,
+ "loss": 0.49192219972610474,
+ "step": 319
+ },
+ {
+ "epoch": 0.16789087093389296,
+ "grad_norm": 0.22103322022956556,
+ "learning_rate": 4.6942160632892186e-05,
+ "loss": 0.5071749687194824,
+ "step": 320
+ },
+ {
+ "epoch": 0.16841552990556138,
+ "grad_norm": 0.21564730973684285,
+ "learning_rate": 4.69234678682346e-05,
+ "loss": 0.47806400060653687,
+ "step": 321
+ },
+ {
+ "epoch": 0.1689401888772298,
+ "grad_norm": 0.20601617921603987,
+ "learning_rate": 4.690472233436294e-05,
+ "loss": 0.482598215341568,
+ "step": 322
+ },
+ {
+ "epoch": 0.1694648478488982,
+ "grad_norm": 0.2387711418230785,
+ "learning_rate": 4.688592408220463e-05,
+ "loss": 0.4898022413253784,
+ "step": 323
+ },
+ {
+ "epoch": 0.16998950682056663,
+ "grad_norm": 0.21268103353554485,
+ "learning_rate": 4.6867073162830324e-05,
+ "loss": 0.4905577301979065,
+ "step": 324
+ },
+ {
+ "epoch": 0.17051416579223505,
+ "grad_norm": 0.19715742151976873,
+ "learning_rate": 4.684816962745374e-05,
+ "loss": 0.4740475118160248,
+ "step": 325
+ },
+ {
+ "epoch": 0.17103882476390347,
+ "grad_norm": 0.23117030350312417,
+ "learning_rate": 4.682921352743155e-05,
+ "loss": 0.4988901615142822,
+ "step": 326
+ },
+ {
+ "epoch": 0.17156348373557187,
+ "grad_norm": 0.19156692418028007,
+ "learning_rate": 4.681020491426324e-05,
+ "loss": 0.48951083421707153,
+ "step": 327
+ },
+ {
+ "epoch": 0.1720881427072403,
+ "grad_norm": 0.20204310760555502,
+ "learning_rate": 4.6791143839590954e-05,
+ "loss": 0.49523672461509705,
+ "step": 328
+ },
+ {
+ "epoch": 0.17261280167890872,
+ "grad_norm": 0.20546781421488164,
+ "learning_rate": 4.677203035519935e-05,
+ "loss": 0.4931063652038574,
+ "step": 329
+ },
+ {
+ "epoch": 0.1731374606505771,
+ "grad_norm": 0.1974199465251975,
+ "learning_rate": 4.675286451301551e-05,
+ "loss": 0.49605268239974976,
+ "step": 330
+ },
+ {
+ "epoch": 0.17366211962224554,
+ "grad_norm": 0.19521163088447754,
+ "learning_rate": 4.673364636510874e-05,
+ "loss": 0.4679698646068573,
+ "step": 331
+ },
+ {
+ "epoch": 0.17418677859391396,
+ "grad_norm": 0.2288885699536713,
+ "learning_rate": 4.671437596369042e-05,
+ "loss": 0.5009484887123108,
+ "step": 332
+ },
+ {
+ "epoch": 0.17471143756558238,
+ "grad_norm": 0.19843654017688225,
+ "learning_rate": 4.6695053361113934e-05,
+ "loss": 0.48696738481521606,
+ "step": 333
+ },
+ {
+ "epoch": 0.17523609653725078,
+ "grad_norm": 0.20369700635699925,
+ "learning_rate": 4.667567860987446e-05,
+ "loss": 0.503460168838501,
+ "step": 334
+ },
+ {
+ "epoch": 0.1757607555089192,
+ "grad_norm": 0.19467856800365996,
+ "learning_rate": 4.6656251762608864e-05,
+ "loss": 0.5148646831512451,
+ "step": 335
+ },
+ {
+ "epoch": 0.17628541448058763,
+ "grad_norm": 0.20183959160051948,
+ "learning_rate": 4.663677287209554e-05,
+ "loss": 0.4726570248603821,
+ "step": 336
+ },
+ {
+ "epoch": 0.17681007345225602,
+ "grad_norm": 0.19428625377872036,
+ "learning_rate": 4.6617241991254276e-05,
+ "loss": 0.5030290484428406,
+ "step": 337
+ },
+ {
+ "epoch": 0.17733473242392445,
+ "grad_norm": 0.1987355254080962,
+ "learning_rate": 4.6597659173146095e-05,
+ "loss": 0.49913081526756287,
+ "step": 338
+ },
+ {
+ "epoch": 0.17785939139559287,
+ "grad_norm": 0.19404813469718887,
+ "learning_rate": 4.657802447097315e-05,
+ "loss": 0.4883667528629303,
+ "step": 339
+ },
+ {
+ "epoch": 0.1783840503672613,
+ "grad_norm": 0.20184751335688705,
+ "learning_rate": 4.6558337938078505e-05,
+ "loss": 0.48960620164871216,
+ "step": 340
+ },
+ {
+ "epoch": 0.1789087093389297,
+ "grad_norm": 0.19883905060437146,
+ "learning_rate": 4.653859962794608e-05,
+ "loss": 0.502848744392395,
+ "step": 341
+ },
+ {
+ "epoch": 0.1794333683105981,
+ "grad_norm": 0.1891918859597783,
+ "learning_rate": 4.651880959420043e-05,
+ "loss": 0.47708186507225037,
+ "step": 342
+ },
+ {
+ "epoch": 0.17995802728226654,
+ "grad_norm": 0.19844251269020877,
+ "learning_rate": 4.649896789060667e-05,
+ "loss": 0.4847305417060852,
+ "step": 343
+ },
+ {
+ "epoch": 0.18048268625393493,
+ "grad_norm": 0.19920278249907905,
+ "learning_rate": 4.647907457107025e-05,
+ "loss": 0.4991101324558258,
+ "step": 344
+ },
+ {
+ "epoch": 0.18100734522560336,
+ "grad_norm": 0.19392889835203725,
+ "learning_rate": 4.645912968963687e-05,
+ "loss": 0.48793840408325195,
+ "step": 345
+ },
+ {
+ "epoch": 0.18153200419727178,
+ "grad_norm": 0.21489774082743496,
+ "learning_rate": 4.64391333004923e-05,
+ "loss": 0.5058388113975525,
+ "step": 346
+ },
+ {
+ "epoch": 0.1820566631689402,
+ "grad_norm": 0.20538704516632084,
+ "learning_rate": 4.6419085457962264e-05,
+ "loss": 0.5034661889076233,
+ "step": 347
+ },
+ {
+ "epoch": 0.1825813221406086,
+ "grad_norm": 0.19673110960930915,
+ "learning_rate": 4.639898621651226e-05,
+ "loss": 0.4912847876548767,
+ "step": 348
+ },
+ {
+ "epoch": 0.18310598111227702,
+ "grad_norm": 0.2017511592646111,
+ "learning_rate": 4.637883563074743e-05,
+ "loss": 0.48043256998062134,
+ "step": 349
+ },
+ {
+ "epoch": 0.18363064008394545,
+ "grad_norm": 0.20093163028321342,
+ "learning_rate": 4.635863375541239e-05,
+ "loss": 0.49590930342674255,
+ "step": 350
+ },
+ {
+ "epoch": 0.18415529905561384,
+ "grad_norm": 0.21350487224525785,
+ "learning_rate": 4.6338380645391125e-05,
+ "loss": 0.48712003231048584,
+ "step": 351
+ },
+ {
+ "epoch": 0.18467995802728226,
+ "grad_norm": 0.19661734920471458,
+ "learning_rate": 4.6318076355706804e-05,
+ "loss": 0.4726638197898865,
+ "step": 352
+ },
+ {
+ "epoch": 0.1852046169989507,
+ "grad_norm": 0.18305587152256153,
+ "learning_rate": 4.629772094152164e-05,
+ "loss": 0.46838825941085815,
+ "step": 353
+ },
+ {
+ "epoch": 0.18572927597061908,
+ "grad_norm": 0.21807773993637447,
+ "learning_rate": 4.627731445813672e-05,
+ "loss": 0.4992654025554657,
+ "step": 354
+ },
+ {
+ "epoch": 0.1862539349422875,
+ "grad_norm": 0.19654836688620947,
+ "learning_rate": 4.62568569609919e-05,
+ "loss": 0.5152627825737,
+ "step": 355
+ },
+ {
+ "epoch": 0.18677859391395593,
+ "grad_norm": 0.2207743593785138,
+ "learning_rate": 4.623634850566562e-05,
+ "loss": 0.4884442389011383,
+ "step": 356
+ },
+ {
+ "epoch": 0.18730325288562436,
+ "grad_norm": 0.2074611972773911,
+ "learning_rate": 4.6215789147874763e-05,
+ "loss": 0.4881437420845032,
+ "step": 357
+ },
+ {
+ "epoch": 0.18782791185729275,
+ "grad_norm": 0.20735292885978085,
+ "learning_rate": 4.619517894347451e-05,
+ "loss": 0.4888407588005066,
+ "step": 358
+ },
+ {
+ "epoch": 0.18835257082896117,
+ "grad_norm": 0.2176780014155129,
+ "learning_rate": 4.617451794845816e-05,
+ "loss": 0.4760252833366394,
+ "step": 359
+ },
+ {
+ "epoch": 0.1888772298006296,
+ "grad_norm": 0.21377246451124693,
+ "learning_rate": 4.6153806218957e-05,
+ "loss": 0.49749091267585754,
+ "step": 360
+ },
+ {
+ "epoch": 0.189401888772298,
+ "grad_norm": 0.21037393449486458,
+ "learning_rate": 4.613304381124019e-05,
+ "loss": 0.49243104457855225,
+ "step": 361
+ },
+ {
+ "epoch": 0.18992654774396642,
+ "grad_norm": 0.21728961765853474,
+ "learning_rate": 4.6112230781714534e-05,
+ "loss": 0.4972265362739563,
+ "step": 362
+ },
+ {
+ "epoch": 0.19045120671563484,
+ "grad_norm": 0.21711913658007095,
+ "learning_rate": 4.609136718692437e-05,
+ "loss": 0.4787655174732208,
+ "step": 363
+ },
+ {
+ "epoch": 0.19097586568730326,
+ "grad_norm": 0.1980053551148303,
+ "learning_rate": 4.6070453083551415e-05,
+ "loss": 0.49180781841278076,
+ "step": 364
+ },
+ {
+ "epoch": 0.19150052465897166,
+ "grad_norm": 0.2128862573273591,
+ "learning_rate": 4.60494885284146e-05,
+ "loss": 0.491728812456131,
+ "step": 365
+ },
+ {
+ "epoch": 0.19202518363064008,
+ "grad_norm": 0.21736774782963791,
+ "learning_rate": 4.602847357846994e-05,
+ "loss": 0.4938160181045532,
+ "step": 366
+ },
+ {
+ "epoch": 0.1925498426023085,
+ "grad_norm": 0.19449397649953404,
+ "learning_rate": 4.600740829081034e-05,
+ "loss": 0.4858305752277374,
+ "step": 367
+ },
+ {
+ "epoch": 0.1930745015739769,
+ "grad_norm": 0.21388852002557526,
+ "learning_rate": 4.5986292722665494e-05,
+ "loss": 0.48994073271751404,
+ "step": 368
+ },
+ {
+ "epoch": 0.19359916054564533,
+ "grad_norm": 0.19577830188970796,
+ "learning_rate": 4.596512693140164e-05,
+ "loss": 0.482921302318573,
+ "step": 369
+ },
+ {
+ "epoch": 0.19412381951731375,
+ "grad_norm": 0.19951728266464225,
+ "learning_rate": 4.594391097452151e-05,
+ "loss": 0.4711063504219055,
+ "step": 370
+ },
+ {
+ "epoch": 0.19464847848898217,
+ "grad_norm": 0.2025906777494991,
+ "learning_rate": 4.592264490966411e-05,
+ "loss": 0.5043122172355652,
+ "step": 371
+ },
+ {
+ "epoch": 0.19517313746065057,
+ "grad_norm": 0.2034983871315486,
+ "learning_rate": 4.590132879460458e-05,
+ "loss": 0.4876527190208435,
+ "step": 372
+ },
+ {
+ "epoch": 0.195697796432319,
+ "grad_norm": 0.23162768822995744,
+ "learning_rate": 4.587996268725402e-05,
+ "loss": 0.48991626501083374,
+ "step": 373
+ },
+ {
+ "epoch": 0.19622245540398742,
+ "grad_norm": 0.1886112619543641,
+ "learning_rate": 4.585854664565939e-05,
+ "loss": 0.5020086169242859,
+ "step": 374
+ },
+ {
+ "epoch": 0.1967471143756558,
+ "grad_norm": 0.2097288447878384,
+ "learning_rate": 4.583708072800325e-05,
+ "loss": 0.5055171847343445,
+ "step": 375
+ },
+ {
+ "epoch": 0.19727177334732424,
+ "grad_norm": 0.2018286380770378,
+ "learning_rate": 4.58155649926037e-05,
+ "loss": 0.5072212219238281,
+ "step": 376
+ },
+ {
+ "epoch": 0.19779643231899266,
+ "grad_norm": 0.1955970746997133,
+ "learning_rate": 4.579399949791419e-05,
+ "loss": 0.48105719685554504,
+ "step": 377
+ },
+ {
+ "epoch": 0.19832109129066108,
+ "grad_norm": 0.20673536008034943,
+ "learning_rate": 4.5772384302523325e-05,
+ "loss": 0.5041084289550781,
+ "step": 378
+ },
+ {
+ "epoch": 0.19884575026232948,
+ "grad_norm": 0.20536034147238338,
+ "learning_rate": 4.575071946515477e-05,
+ "loss": 0.4688987731933594,
+ "step": 379
+ },
+ {
+ "epoch": 0.1993704092339979,
+ "grad_norm": 0.20272239051808535,
+ "learning_rate": 4.572900504466702e-05,
+ "loss": 0.4640740752220154,
+ "step": 380
+ },
+ {
+ "epoch": 0.19989506820566633,
+ "grad_norm": 0.25058140305501153,
+ "learning_rate": 4.57072411000533e-05,
+ "loss": 0.480777382850647,
+ "step": 381
+ },
+ {
+ "epoch": 0.20041972717733472,
+ "grad_norm": 0.21137024512187075,
+ "learning_rate": 4.5685427690441385e-05,
+ "loss": 0.4951503872871399,
+ "step": 382
+ },
+ {
+ "epoch": 0.20094438614900315,
+ "grad_norm": 0.20543066953545808,
+ "learning_rate": 4.566356487509341e-05,
+ "loss": 0.49060487747192383,
+ "step": 383
+ },
+ {
+ "epoch": 0.20146904512067157,
+ "grad_norm": 0.19053043556097235,
+ "learning_rate": 4.564165271340577e-05,
+ "loss": 0.48229914903640747,
+ "step": 384
+ },
+ {
+ "epoch": 0.20199370409233997,
+ "grad_norm": 0.2415409691663337,
+ "learning_rate": 4.561969126490888e-05,
+ "loss": 0.5014164447784424,
+ "step": 385
+ },
+ {
+ "epoch": 0.2025183630640084,
+ "grad_norm": 0.19321490977259959,
+ "learning_rate": 4.55976805892671e-05,
+ "loss": 0.49548542499542236,
+ "step": 386
+ },
+ {
+ "epoch": 0.2030430220356768,
+ "grad_norm": 0.19888025519816874,
+ "learning_rate": 4.557562074627849e-05,
+ "loss": 0.49908390641212463,
+ "step": 387
+ },
+ {
+ "epoch": 0.20356768100734524,
+ "grad_norm": 0.19045907984518948,
+ "learning_rate": 4.555351179587471e-05,
+ "loss": 0.5011584162712097,
+ "step": 388
+ },
+ {
+ "epoch": 0.20409233997901363,
+ "grad_norm": 0.2034373140106159,
+ "learning_rate": 4.5531353798120854e-05,
+ "loss": 0.4809190034866333,
+ "step": 389
+ },
+ {
+ "epoch": 0.20461699895068206,
+ "grad_norm": 0.21362046688037958,
+ "learning_rate": 4.5509146813215204e-05,
+ "loss": 0.5129217505455017,
+ "step": 390
+ },
+ {
+ "epoch": 0.20514165792235048,
+ "grad_norm": 0.2005531452611148,
+ "learning_rate": 4.548689090148919e-05,
+ "loss": 0.4901890456676483,
+ "step": 391
+ },
+ {
+ "epoch": 0.20566631689401887,
+ "grad_norm": 0.23789259327109627,
+ "learning_rate": 4.546458612340713e-05,
+ "loss": 0.5273652076721191,
+ "step": 392
+ },
+ {
+ "epoch": 0.2061909758656873,
+ "grad_norm": 0.21125301565528945,
+ "learning_rate": 4.5442232539566124e-05,
+ "loss": 0.48382890224456787,
+ "step": 393
+ },
+ {
+ "epoch": 0.20671563483735572,
+ "grad_norm": 0.20984681657728893,
+ "learning_rate": 4.5419830210695845e-05,
+ "loss": 0.4895797669887543,
+ "step": 394
+ },
+ {
+ "epoch": 0.20724029380902415,
+ "grad_norm": 0.19952182387837472,
+ "learning_rate": 4.539737919765842e-05,
+ "loss": 0.4788748025894165,
+ "step": 395
+ },
+ {
+ "epoch": 0.20776495278069254,
+ "grad_norm": 0.1930005781516943,
+ "learning_rate": 4.537487956144821e-05,
+ "loss": 0.47772473096847534,
+ "step": 396
+ },
+ {
+ "epoch": 0.20828961175236096,
+ "grad_norm": 0.232101495398028,
+ "learning_rate": 4.535233136319169e-05,
+ "loss": 0.5068112015724182,
+ "step": 397
+ },
+ {
+ "epoch": 0.2088142707240294,
+ "grad_norm": 0.21056643176194195,
+ "learning_rate": 4.5329734664147286e-05,
+ "loss": 0.5078309774398804,
+ "step": 398
+ },
+ {
+ "epoch": 0.20933892969569778,
+ "grad_norm": 0.2274874586448457,
+ "learning_rate": 4.530708952570515e-05,
+ "loss": 0.4837700426578522,
+ "step": 399
+ },
+ {
+ "epoch": 0.2098635886673662,
+ "grad_norm": 0.20773883357516573,
+ "learning_rate": 4.528439600938706e-05,
+ "loss": 0.4915201961994171,
+ "step": 400
+ },
+ {
+ "epoch": 0.21038824763903463,
+ "grad_norm": 0.21292775120746008,
+ "learning_rate": 4.5261654176846224e-05,
+ "loss": 0.5043562650680542,
+ "step": 401
+ },
+ {
+ "epoch": 0.21091290661070305,
+ "grad_norm": 0.21547961984597777,
+ "learning_rate": 4.52388640898671e-05,
+ "loss": 0.49059492349624634,
+ "step": 402
+ },
+ {
+ "epoch": 0.21143756558237145,
+ "grad_norm": 0.1894071498395343,
+ "learning_rate": 4.521602581036527e-05,
+ "loss": 0.4886820912361145,
+ "step": 403
+ },
+ {
+ "epoch": 0.21196222455403987,
+ "grad_norm": 0.20536390063908713,
+ "learning_rate": 4.519313940038721e-05,
+ "loss": 0.4859333634376526,
+ "step": 404
+ },
+ {
+ "epoch": 0.2124868835257083,
+ "grad_norm": 0.19248861110769996,
+ "learning_rate": 4.5170204922110174e-05,
+ "loss": 0.4755397140979767,
+ "step": 405
+ },
+ {
+ "epoch": 0.2130115424973767,
+ "grad_norm": 0.19632639204982114,
+ "learning_rate": 4.5147222437842024e-05,
+ "loss": 0.48414891958236694,
+ "step": 406
+ },
+ {
+ "epoch": 0.21353620146904512,
+ "grad_norm": 0.19335035210986967,
+ "learning_rate": 4.5124192010021005e-05,
+ "loss": 0.48097655177116394,
+ "step": 407
+ },
+ {
+ "epoch": 0.21406086044071354,
+ "grad_norm": 0.1956907617136089,
+ "learning_rate": 4.510111370121564e-05,
+ "loss": 0.48582807183265686,
+ "step": 408
+ },
+ {
+ "epoch": 0.21458551941238196,
+ "grad_norm": 0.19740074807757949,
+ "learning_rate": 4.5077987574124524e-05,
+ "loss": 0.5069482326507568,
+ "step": 409
+ },
+ {
+ "epoch": 0.21511017838405036,
+ "grad_norm": 0.19522492404537745,
+ "learning_rate": 4.505481369157617e-05,
+ "loss": 0.47752654552459717,
+ "step": 410
+ },
+ {
+ "epoch": 0.21563483735571878,
+ "grad_norm": 0.20078806911355473,
+ "learning_rate": 4.503159211652884e-05,
+ "loss": 0.49054259061813354,
+ "step": 411
+ },
+ {
+ "epoch": 0.2161594963273872,
+ "grad_norm": 0.1922916020853349,
+ "learning_rate": 4.500832291207033e-05,
+ "loss": 0.47457975149154663,
+ "step": 412
+ },
+ {
+ "epoch": 0.2166841552990556,
+ "grad_norm": 0.18906406634138712,
+ "learning_rate": 4.49850061414179e-05,
+ "loss": 0.5094400644302368,
+ "step": 413
+ },
+ {
+ "epoch": 0.21720881427072403,
+ "grad_norm": 0.19486985008167207,
+ "learning_rate": 4.496164186791795e-05,
+ "loss": 0.4839191734790802,
+ "step": 414
+ },
+ {
+ "epoch": 0.21773347324239245,
+ "grad_norm": 0.20869563884726042,
+ "learning_rate": 4.4938230155046005e-05,
+ "loss": 0.5022982954978943,
+ "step": 415
+ },
+ {
+ "epoch": 0.21825813221406087,
+ "grad_norm": 0.20115353521050966,
+ "learning_rate": 4.4914771066406454e-05,
+ "loss": 0.49222704768180847,
+ "step": 416
+ },
+ {
+ "epoch": 0.21878279118572927,
+ "grad_norm": 0.21789297317266892,
+ "learning_rate": 4.4891264665732365e-05,
+ "loss": 0.4971074163913727,
+ "step": 417
+ },
+ {
+ "epoch": 0.2193074501573977,
+ "grad_norm": 0.2031937995150951,
+ "learning_rate": 4.4867711016885386e-05,
+ "loss": 0.49928638339042664,
+ "step": 418
+ },
+ {
+ "epoch": 0.21983210912906612,
+ "grad_norm": 0.21043700159975373,
+ "learning_rate": 4.484411018385551e-05,
+ "loss": 0.48748651146888733,
+ "step": 419
+ },
+ {
+ "epoch": 0.2203567681007345,
+ "grad_norm": 0.20148226383642173,
+ "learning_rate": 4.482046223076091e-05,
+ "loss": 0.4856518805027008,
+ "step": 420
+ },
+ {
+ "epoch": 0.22088142707240294,
+ "grad_norm": 0.1951352394773881,
+ "learning_rate": 4.479676722184778e-05,
+ "loss": 0.4749642312526703,
+ "step": 421
+ },
+ {
+ "epoch": 0.22140608604407136,
+ "grad_norm": 0.1948457701162797,
+ "learning_rate": 4.477302522149016e-05,
+ "loss": 0.49277958273887634,
+ "step": 422
+ },
+ {
+ "epoch": 0.22193074501573976,
+ "grad_norm": 0.19769333670396683,
+ "learning_rate": 4.474923629418974e-05,
+ "loss": 0.46302592754364014,
+ "step": 423
+ },
+ {
+ "epoch": 0.22245540398740818,
+ "grad_norm": 0.20865373408852095,
+ "learning_rate": 4.472540050457572e-05,
+ "loss": 0.4929874837398529,
+ "step": 424
+ },
+ {
+ "epoch": 0.2229800629590766,
+ "grad_norm": 0.19449471200427154,
+ "learning_rate": 4.470151791740459e-05,
+ "loss": 0.5030213594436646,
+ "step": 425
+ },
+ {
+ "epoch": 0.22350472193074503,
+ "grad_norm": 0.19636212249634782,
+ "learning_rate": 4.467758859756002e-05,
+ "loss": 0.47791340947151184,
+ "step": 426
+ },
+ {
+ "epoch": 0.22402938090241342,
+ "grad_norm": 0.2125527348630816,
+ "learning_rate": 4.46536126100526e-05,
+ "loss": 0.5006207227706909,
+ "step": 427
+ },
+ {
+ "epoch": 0.22455403987408185,
+ "grad_norm": 0.1934909467765078,
+ "learning_rate": 4.462959002001971e-05,
+ "loss": 0.4889323115348816,
+ "step": 428
+ },
+ {
+ "epoch": 0.22507869884575027,
+ "grad_norm": 0.18918797496722067,
+ "learning_rate": 4.460552089272536e-05,
+ "loss": 0.4747370779514313,
+ "step": 429
+ },
+ {
+ "epoch": 0.22560335781741866,
+ "grad_norm": 0.1920955645668967,
+ "learning_rate": 4.458140529355998e-05,
+ "loss": 0.49264299869537354,
+ "step": 430
+ },
+ {
+ "epoch": 0.2261280167890871,
+ "grad_norm": 0.1781250367672512,
+ "learning_rate": 4.455724328804026e-05,
+ "loss": 0.4665558934211731,
+ "step": 431
+ },
+ {
+ "epoch": 0.2266526757607555,
+ "grad_norm": 0.1853281136078876,
+ "learning_rate": 4.4533034941808956e-05,
+ "loss": 0.46890392899513245,
+ "step": 432
+ },
+ {
+ "epoch": 0.22717733473242394,
+ "grad_norm": 0.19411101834670585,
+ "learning_rate": 4.450878032063472e-05,
+ "loss": 0.4868405759334564,
+ "step": 433
+ },
+ {
+ "epoch": 0.22770199370409233,
+ "grad_norm": 0.18603153581391266,
+ "learning_rate": 4.448447949041194e-05,
+ "loss": 0.4699792265892029,
+ "step": 434
+ },
+ {
+ "epoch": 0.22822665267576075,
+ "grad_norm": 0.19934522400817525,
+ "learning_rate": 4.446013251716051e-05,
+ "loss": 0.48219117522239685,
+ "step": 435
+ },
+ {
+ "epoch": 0.22875131164742918,
+ "grad_norm": 0.21201904919578815,
+ "learning_rate": 4.443573946702573e-05,
+ "loss": 0.4985590875148773,
+ "step": 436
+ },
+ {
+ "epoch": 0.22927597061909757,
+ "grad_norm": 0.22340402335613319,
+ "learning_rate": 4.441130040627804e-05,
+ "loss": 0.5099130868911743,
+ "step": 437
+ },
+ {
+ "epoch": 0.229800629590766,
+ "grad_norm": 0.22796376691421188,
+ "learning_rate": 4.4386815401312915e-05,
+ "loss": 0.48687899112701416,
+ "step": 438
+ },
+ {
+ "epoch": 0.23032528856243442,
+ "grad_norm": 0.19515518112968896,
+ "learning_rate": 4.436228451865061e-05,
+ "loss": 0.4919915795326233,
+ "step": 439
+ },
+ {
+ "epoch": 0.23084994753410285,
+ "grad_norm": 0.20888204295786816,
+ "learning_rate": 4.433770782493606e-05,
+ "loss": 0.5035967826843262,
+ "step": 440
+ },
+ {
+ "epoch": 0.23137460650577124,
+ "grad_norm": 0.19957691207438416,
+ "learning_rate": 4.431308538693862e-05,
+ "loss": 0.4963775873184204,
+ "step": 441
+ },
+ {
+ "epoch": 0.23189926547743966,
+ "grad_norm": 0.20713157708679947,
+ "learning_rate": 4.4288417271551974e-05,
+ "loss": 0.494832843542099,
+ "step": 442
+ },
+ {
+ "epoch": 0.2324239244491081,
+ "grad_norm": 0.2082730788778929,
+ "learning_rate": 4.426370354579384e-05,
+ "loss": 0.49278348684310913,
+ "step": 443
+ },
+ {
+ "epoch": 0.23294858342077648,
+ "grad_norm": 0.22920179495341536,
+ "learning_rate": 4.4238944276805896e-05,
+ "loss": 0.47680842876434326,
+ "step": 444
+ },
+ {
+ "epoch": 0.2334732423924449,
+ "grad_norm": 0.18679233772124446,
+ "learning_rate": 4.421413953185351e-05,
+ "loss": 0.5002862811088562,
+ "step": 445
+ },
+ {
+ "epoch": 0.23399790136411333,
+ "grad_norm": 0.23175479918934758,
+ "learning_rate": 4.418928937832565e-05,
+ "loss": 0.5007957220077515,
+ "step": 446
+ },
+ {
+ "epoch": 0.23452256033578175,
+ "grad_norm": 0.2097609824358498,
+ "learning_rate": 4.41643938837346e-05,
+ "loss": 0.5131531357765198,
+ "step": 447
+ },
+ {
+ "epoch": 0.23504721930745015,
+ "grad_norm": 0.21048463332579867,
+ "learning_rate": 4.413945311571586e-05,
+ "loss": 0.4840421974658966,
+ "step": 448
+ },
+ {
+ "epoch": 0.23557187827911857,
+ "grad_norm": 0.22127179437429526,
+ "learning_rate": 4.4114467142027905e-05,
+ "loss": 0.4967547357082367,
+ "step": 449
+ },
+ {
+ "epoch": 0.236096537250787,
+ "grad_norm": 0.19834993446742008,
+ "learning_rate": 4.4089436030552035e-05,
+ "loss": 0.4757929742336273,
+ "step": 450
+ },
+ {
+ "epoch": 0.2366211962224554,
+ "grad_norm": 0.21065798784728926,
+ "learning_rate": 4.4064359849292166e-05,
+ "loss": 0.4977768659591675,
+ "step": 451
+ },
+ {
+ "epoch": 0.23714585519412382,
+ "grad_norm": 0.2002922630027721,
+ "learning_rate": 4.403923866637469e-05,
+ "loss": 0.45635008811950684,
+ "step": 452
+ },
+ {
+ "epoch": 0.23767051416579224,
+ "grad_norm": 0.20198999164868367,
+ "learning_rate": 4.4014072550048227e-05,
+ "loss": 0.4840908944606781,
+ "step": 453
+ },
+ {
+ "epoch": 0.23819517313746066,
+ "grad_norm": 0.19919284698933104,
+ "learning_rate": 4.398886156868347e-05,
+ "loss": 0.4777500033378601,
+ "step": 454
+ },
+ {
+ "epoch": 0.23871983210912906,
+ "grad_norm": 0.3002196602346047,
+ "learning_rate": 4.3963605790773026e-05,
+ "loss": 0.49407732486724854,
+ "step": 455
+ },
+ {
+ "epoch": 0.23924449108079748,
+ "grad_norm": 0.1843790565631538,
+ "learning_rate": 4.393830528493118e-05,
+ "loss": 0.4797516465187073,
+ "step": 456
+ },
+ {
+ "epoch": 0.2397691500524659,
+ "grad_norm": 0.2316267764426789,
+ "learning_rate": 4.3912960119893755e-05,
+ "loss": 0.4796208441257477,
+ "step": 457
+ },
+ {
+ "epoch": 0.2402938090241343,
+ "grad_norm": 0.22658430570217925,
+ "learning_rate": 4.3887570364517885e-05,
+ "loss": 0.481270968914032,
+ "step": 458
+ },
+ {
+ "epoch": 0.24081846799580273,
+ "grad_norm": 0.19338831323124744,
+ "learning_rate": 4.3862136087781846e-05,
+ "loss": 0.48800894618034363,
+ "step": 459
+ },
+ {
+ "epoch": 0.24134312696747115,
+ "grad_norm": 0.19363741267291593,
+ "learning_rate": 4.3836657358784876e-05,
+ "loss": 0.49191194772720337,
+ "step": 460
+ },
+ {
+ "epoch": 0.24186778593913955,
+ "grad_norm": 0.2066675796348607,
+ "learning_rate": 4.381113424674699e-05,
+ "loss": 0.4997405707836151,
+ "step": 461
+ },
+ {
+ "epoch": 0.24239244491080797,
+ "grad_norm": 0.19639676711322537,
+ "learning_rate": 4.378556682100876e-05,
+ "loss": 0.4781111776828766,
+ "step": 462
+ },
+ {
+ "epoch": 0.2429171038824764,
+ "grad_norm": 0.18397664561272883,
+ "learning_rate": 4.375995515103114e-05,
+ "loss": 0.4825688600540161,
+ "step": 463
+ },
+ {
+ "epoch": 0.24344176285414482,
+ "grad_norm": 0.2056538092643436,
+ "learning_rate": 4.373429930639534e-05,
+ "loss": 0.4947561025619507,
+ "step": 464
+ },
+ {
+ "epoch": 0.2439664218258132,
+ "grad_norm": 0.1809360096304517,
+ "learning_rate": 4.370859935680252e-05,
+ "loss": 0.48019152879714966,
+ "step": 465
+ },
+ {
+ "epoch": 0.24449108079748164,
+ "grad_norm": 0.1972433864757527,
+ "learning_rate": 4.368285537207371e-05,
+ "loss": 0.48355674743652344,
+ "step": 466
+ },
+ {
+ "epoch": 0.24501573976915006,
+ "grad_norm": 0.19761941269394526,
+ "learning_rate": 4.365706742214953e-05,
+ "loss": 0.491325318813324,
+ "step": 467
+ },
+ {
+ "epoch": 0.24554039874081846,
+ "grad_norm": 0.1897877301879071,
+ "learning_rate": 4.3631235577090085e-05,
+ "loss": 0.4849388599395752,
+ "step": 468
+ },
+ {
+ "epoch": 0.24606505771248688,
+ "grad_norm": 0.19698394331583843,
+ "learning_rate": 4.3605359907074704e-05,
+ "loss": 0.48221641778945923,
+ "step": 469
+ },
+ {
+ "epoch": 0.2465897166841553,
+ "grad_norm": 0.1847208738538242,
+ "learning_rate": 4.3579440482401805e-05,
+ "loss": 0.48547977209091187,
+ "step": 470
+ },
+ {
+ "epoch": 0.24711437565582373,
+ "grad_norm": 0.20103873148388024,
+ "learning_rate": 4.355347737348864e-05,
+ "loss": 0.4920230507850647,
+ "step": 471
+ },
+ {
+ "epoch": 0.24763903462749212,
+ "grad_norm": 0.19361216779355914,
+ "learning_rate": 4.352747065087119e-05,
+ "loss": 0.47582632303237915,
+ "step": 472
+ },
+ {
+ "epoch": 0.24816369359916055,
+ "grad_norm": 0.18642424818234848,
+ "learning_rate": 4.3501420385203884e-05,
+ "loss": 0.4738594889640808,
+ "step": 473
+ },
+ {
+ "epoch": 0.24868835257082897,
+ "grad_norm": 0.19631285880492338,
+ "learning_rate": 4.347532664725946e-05,
+ "loss": 0.490800678730011,
+ "step": 474
+ },
+ {
+ "epoch": 0.24921301154249736,
+ "grad_norm": 0.18508072382838214,
+ "learning_rate": 4.3449189507928776e-05,
+ "loss": 0.4833751320838928,
+ "step": 475
+ },
+ {
+ "epoch": 0.2497376705141658,
+ "grad_norm": 0.1995899934010246,
+ "learning_rate": 4.3423009038220565e-05,
+ "loss": 0.48172321915626526,
+ "step": 476
+ },
+ {
+ "epoch": 0.2502623294858342,
+ "grad_norm": 0.17992358256391341,
+ "learning_rate": 4.339678530926132e-05,
+ "loss": 0.4740462303161621,
+ "step": 477
+ },
+ {
+ "epoch": 0.25078698845750264,
+ "grad_norm": 0.18932927399162522,
+ "learning_rate": 4.337051839229504e-05,
+ "loss": 0.47711318731307983,
+ "step": 478
+ },
+ {
+ "epoch": 0.25131164742917106,
+ "grad_norm": 0.19048957223864676,
+ "learning_rate": 4.3344208358683047e-05,
+ "loss": 0.48262083530426025,
+ "step": 479
+ },
+ {
+ "epoch": 0.2518363064008394,
+ "grad_norm": 0.19267359824618632,
+ "learning_rate": 4.331785527990384e-05,
+ "loss": 0.47504326701164246,
+ "step": 480
+ },
+ {
+ "epoch": 0.25236096537250785,
+ "grad_norm": 0.20127175925725735,
+ "learning_rate": 4.329145922755281e-05,
+ "loss": 0.47620856761932373,
+ "step": 481
+ },
+ {
+ "epoch": 0.2528856243441763,
+ "grad_norm": 0.19093478454057455,
+ "learning_rate": 4.326502027334213e-05,
+ "loss": 0.48457974195480347,
+ "step": 482
+ },
+ {
+ "epoch": 0.2534102833158447,
+ "grad_norm": 0.20457208623404696,
+ "learning_rate": 4.3238538489100514e-05,
+ "loss": 0.49174126982688904,
+ "step": 483
+ },
+ {
+ "epoch": 0.2539349422875131,
+ "grad_norm": 0.19715906998524946,
+ "learning_rate": 4.321201394677304e-05,
+ "loss": 0.4941791296005249,
+ "step": 484
+ },
+ {
+ "epoch": 0.25445960125918154,
+ "grad_norm": 0.19726199111118628,
+ "learning_rate": 4.3185446718420976e-05,
+ "loss": 0.49088531732559204,
+ "step": 485
+ },
+ {
+ "epoch": 0.25498426023084997,
+ "grad_norm": 0.20546733535813932,
+ "learning_rate": 4.31588368762215e-05,
+ "loss": 0.4862939715385437,
+ "step": 486
+ },
+ {
+ "epoch": 0.25550891920251834,
+ "grad_norm": 0.1891768308679286,
+ "learning_rate": 4.313218449246763e-05,
+ "loss": 0.47618573904037476,
+ "step": 487
+ },
+ {
+ "epoch": 0.25603357817418676,
+ "grad_norm": 0.20036177448461404,
+ "learning_rate": 4.3105489639567906e-05,
+ "loss": 0.46958842873573303,
+ "step": 488
+ },
+ {
+ "epoch": 0.2565582371458552,
+ "grad_norm": 0.19729242845497308,
+ "learning_rate": 4.307875239004626e-05,
+ "loss": 0.4836847186088562,
+ "step": 489
+ },
+ {
+ "epoch": 0.2570828961175236,
+ "grad_norm": 0.21655150877973078,
+ "learning_rate": 4.305197281654183e-05,
+ "loss": 0.4758085310459137,
+ "step": 490
+ },
+ {
+ "epoch": 0.25760755508919203,
+ "grad_norm": 0.18700542975974652,
+ "learning_rate": 4.302515099180872e-05,
+ "loss": 0.48014035820961,
+ "step": 491
+ },
+ {
+ "epoch": 0.25813221406086045,
+ "grad_norm": 0.2048485994863198,
+ "learning_rate": 4.299828698871581e-05,
+ "loss": 0.49596071243286133,
+ "step": 492
+ },
+ {
+ "epoch": 0.2586568730325289,
+ "grad_norm": 0.19673652760478702,
+ "learning_rate": 4.29713808802466e-05,
+ "loss": 0.47907233238220215,
+ "step": 493
+ },
+ {
+ "epoch": 0.25918153200419725,
+ "grad_norm": 0.19711305451180386,
+ "learning_rate": 4.294443273949895e-05,
+ "loss": 0.47071751952171326,
+ "step": 494
+ },
+ {
+ "epoch": 0.25970619097586567,
+ "grad_norm": 0.19982536801201056,
+ "learning_rate": 4.291744263968494e-05,
+ "loss": 0.47717398405075073,
+ "step": 495
+ },
+ {
+ "epoch": 0.2602308499475341,
+ "grad_norm": 0.20221441757437503,
+ "learning_rate": 4.289041065413062e-05,
+ "loss": 0.50141441822052,
+ "step": 496
+ },
+ {
+ "epoch": 0.2607555089192025,
+ "grad_norm": 0.1911546037476864,
+ "learning_rate": 4.2863336856275844e-05,
+ "loss": 0.466622531414032,
+ "step": 497
+ },
+ {
+ "epoch": 0.26128016789087094,
+ "grad_norm": 0.18591508233820475,
+ "learning_rate": 4.2836221319674055e-05,
+ "loss": 0.4638373851776123,
+ "step": 498
+ },
+ {
+ "epoch": 0.26180482686253936,
+ "grad_norm": 0.19572770104268478,
+ "learning_rate": 4.28090641179921e-05,
+ "loss": 0.4795790910720825,
+ "step": 499
+ },
+ {
+ "epoch": 0.2623294858342078,
+ "grad_norm": 0.1909010655134962,
+ "learning_rate": 4.278186532501003e-05,
+ "loss": 0.4681857228279114,
+ "step": 500
+ },
+ {
+ "epoch": 0.26285414480587616,
+ "grad_norm": 0.1976655350910736,
+ "learning_rate": 4.275462501462087e-05,
+ "loss": 0.4821563959121704,
+ "step": 501
+ },
+ {
+ "epoch": 0.2633788037775446,
+ "grad_norm": 0.2001578759674193,
+ "learning_rate": 4.272734326083042e-05,
+ "loss": 0.5014116168022156,
+ "step": 502
+ },
+ {
+ "epoch": 0.263903462749213,
+ "grad_norm": 0.19029704860176525,
+ "learning_rate": 4.2700020137757125e-05,
+ "loss": 0.4792998135089874,
+ "step": 503
+ },
+ {
+ "epoch": 0.2644281217208814,
+ "grad_norm": 0.19769506689523492,
+ "learning_rate": 4.2672655719631785e-05,
+ "loss": 0.5015027523040771,
+ "step": 504
+ },
+ {
+ "epoch": 0.26495278069254985,
+ "grad_norm": 0.1981483122419844,
+ "learning_rate": 4.264525008079741e-05,
+ "loss": 0.49372148513793945,
+ "step": 505
+ },
+ {
+ "epoch": 0.2654774396642183,
+ "grad_norm": 0.18524860498238543,
+ "learning_rate": 4.261780329570897e-05,
+ "loss": 0.4777575731277466,
+ "step": 506
+ },
+ {
+ "epoch": 0.2660020986358867,
+ "grad_norm": 0.19914310403457625,
+ "learning_rate": 4.259031543893323e-05,
+ "loss": 0.49828243255615234,
+ "step": 507
+ },
+ {
+ "epoch": 0.26652675760755506,
+ "grad_norm": 0.1829259151106153,
+ "learning_rate": 4.2562786585148575e-05,
+ "loss": 0.47125187516212463,
+ "step": 508
+ },
+ {
+ "epoch": 0.2670514165792235,
+ "grad_norm": 0.17929086425332166,
+ "learning_rate": 4.2535216809144705e-05,
+ "loss": 0.49947690963745117,
+ "step": 509
+ },
+ {
+ "epoch": 0.2675760755508919,
+ "grad_norm": 0.18535028148655724,
+ "learning_rate": 4.250760618582255e-05,
+ "loss": 0.47412362694740295,
+ "step": 510
+ },
+ {
+ "epoch": 0.26810073452256034,
+ "grad_norm": 0.18025075382576644,
+ "learning_rate": 4.2479954790193974e-05,
+ "loss": 0.48555561900138855,
+ "step": 511
+ },
+ {
+ "epoch": 0.26862539349422876,
+ "grad_norm": 0.18536039309609798,
+ "learning_rate": 4.2452262697381656e-05,
+ "loss": 0.4787901043891907,
+ "step": 512
+ },
+ {
+ "epoch": 0.2691500524658972,
+ "grad_norm": 0.1898064970749832,
+ "learning_rate": 4.242452998261879e-05,
+ "loss": 0.4829568862915039,
+ "step": 513
+ },
+ {
+ "epoch": 0.2696747114375656,
+ "grad_norm": 0.18469792856530137,
+ "learning_rate": 4.239675672124896e-05,
+ "loss": 0.46647170186042786,
+ "step": 514
+ },
+ {
+ "epoch": 0.270199370409234,
+ "grad_norm": 0.1909466529292699,
+ "learning_rate": 4.236894298872591e-05,
+ "loss": 0.49455198645591736,
+ "step": 515
+ },
+ {
+ "epoch": 0.2707240293809024,
+ "grad_norm": 0.18628690740837425,
+ "learning_rate": 4.2341088860613305e-05,
+ "loss": 0.46410971879959106,
+ "step": 516
+ },
+ {
+ "epoch": 0.2712486883525708,
+ "grad_norm": 0.19201282567978442,
+ "learning_rate": 4.231319441258459e-05,
+ "loss": 0.5083948969841003,
+ "step": 517
+ },
+ {
+ "epoch": 0.27177334732423925,
+ "grad_norm": 0.19823761982256113,
+ "learning_rate": 4.2285259720422726e-05,
+ "loss": 0.48851656913757324,
+ "step": 518
+ },
+ {
+ "epoch": 0.27229800629590767,
+ "grad_norm": 0.20289491576277816,
+ "learning_rate": 4.225728486002004e-05,
+ "loss": 0.4843178987503052,
+ "step": 519
+ },
+ {
+ "epoch": 0.2728226652675761,
+ "grad_norm": 0.1813367715900217,
+ "learning_rate": 4.2229269907377933e-05,
+ "loss": 0.46726176142692566,
+ "step": 520
+ },
+ {
+ "epoch": 0.2733473242392445,
+ "grad_norm": 0.20031212853886943,
+ "learning_rate": 4.220121493860678e-05,
+ "loss": 0.5005958080291748,
+ "step": 521
+ },
+ {
+ "epoch": 0.2738719832109129,
+ "grad_norm": 0.18334650156283827,
+ "learning_rate": 4.217312002992565e-05,
+ "loss": 0.47628265619277954,
+ "step": 522
+ },
+ {
+ "epoch": 0.2743966421825813,
+ "grad_norm": 0.19550837239251015,
+ "learning_rate": 4.2144985257662126e-05,
+ "loss": 0.4686697721481323,
+ "step": 523
+ },
+ {
+ "epoch": 0.27492130115424973,
+ "grad_norm": 0.17926260767716534,
+ "learning_rate": 4.211681069825207e-05,
+ "loss": 0.4543974697589874,
+ "step": 524
+ },
+ {
+ "epoch": 0.27544596012591815,
+ "grad_norm": 0.21021464381226687,
+ "learning_rate": 4.2088596428239466e-05,
+ "loss": 0.5026243925094604,
+ "step": 525
+ },
+ {
+ "epoch": 0.2759706190975866,
+ "grad_norm": 0.1862432311147689,
+ "learning_rate": 4.206034252427616e-05,
+ "loss": 0.4828398823738098,
+ "step": 526
+ },
+ {
+ "epoch": 0.276495278069255,
+ "grad_norm": 0.20349427012469418,
+ "learning_rate": 4.2032049063121696e-05,
+ "loss": 0.47698354721069336,
+ "step": 527
+ },
+ {
+ "epoch": 0.2770199370409234,
+ "grad_norm": 0.19141725963537087,
+ "learning_rate": 4.2003716121643055e-05,
+ "loss": 0.4852374494075775,
+ "step": 528
+ },
+ {
+ "epoch": 0.2775445960125918,
+ "grad_norm": 0.18922366835728505,
+ "learning_rate": 4.197534377681452e-05,
+ "loss": 0.47030404210090637,
+ "step": 529
+ },
+ {
+ "epoch": 0.2780692549842602,
+ "grad_norm": 0.206175328344501,
+ "learning_rate": 4.194693210571738e-05,
+ "loss": 0.4711211621761322,
+ "step": 530
+ },
+ {
+ "epoch": 0.27859391395592864,
+ "grad_norm": 0.1916539730173641,
+ "learning_rate": 4.19184811855398e-05,
+ "loss": 0.4880971908569336,
+ "step": 531
+ },
+ {
+ "epoch": 0.27911857292759706,
+ "grad_norm": 0.19900636884109554,
+ "learning_rate": 4.188999109357655e-05,
+ "loss": 0.49334031343460083,
+ "step": 532
+ },
+ {
+ "epoch": 0.2796432318992655,
+ "grad_norm": 0.18404893731783803,
+ "learning_rate": 4.186146190722883e-05,
+ "loss": 0.48461294174194336,
+ "step": 533
+ },
+ {
+ "epoch": 0.2801678908709339,
+ "grad_norm": 0.19301712559137277,
+ "learning_rate": 4.183289370400407e-05,
+ "loss": 0.47637492418289185,
+ "step": 534
+ },
+ {
+ "epoch": 0.28069254984260233,
+ "grad_norm": 0.19465652334673852,
+ "learning_rate": 4.180428656151567e-05,
+ "loss": 0.4746769070625305,
+ "step": 535
+ },
+ {
+ "epoch": 0.2812172088142707,
+ "grad_norm": 0.18036833457926044,
+ "learning_rate": 4.177564055748284e-05,
+ "loss": 0.46309196949005127,
+ "step": 536
+ },
+ {
+ "epoch": 0.2817418677859391,
+ "grad_norm": 0.18662944746848986,
+ "learning_rate": 4.174695576973036e-05,
+ "loss": 0.4916302561759949,
+ "step": 537
+ },
+ {
+ "epoch": 0.28226652675760755,
+ "grad_norm": 0.18636157100153936,
+ "learning_rate": 4.171823227618836e-05,
+ "loss": 0.4789227843284607,
+ "step": 538
+ },
+ {
+ "epoch": 0.282791185729276,
+ "grad_norm": 0.1843430677763694,
+ "learning_rate": 4.168947015489215e-05,
+ "loss": 0.48610541224479675,
+ "step": 539
+ },
+ {
+ "epoch": 0.2833158447009444,
+ "grad_norm": 0.17970176908343424,
+ "learning_rate": 4.166066948398199e-05,
+ "loss": 0.4731542468070984,
+ "step": 540
+ },
+ {
+ "epoch": 0.2838405036726128,
+ "grad_norm": 0.18982786153102205,
+ "learning_rate": 4.163183034170284e-05,
+ "loss": 0.49426978826522827,
+ "step": 541
+ },
+ {
+ "epoch": 0.28436516264428124,
+ "grad_norm": 0.18262832530431292,
+ "learning_rate": 4.1602952806404195e-05,
+ "loss": 0.47087955474853516,
+ "step": 542
+ },
+ {
+ "epoch": 0.2848898216159496,
+ "grad_norm": 0.18082132267228873,
+ "learning_rate": 4.1574036956539864e-05,
+ "loss": 0.47580409049987793,
+ "step": 543
+ },
+ {
+ "epoch": 0.28541448058761804,
+ "grad_norm": 0.19544404992693376,
+ "learning_rate": 4.154508287066772e-05,
+ "loss": 0.4801163971424103,
+ "step": 544
+ },
+ {
+ "epoch": 0.28593913955928646,
+ "grad_norm": 0.18071890849291686,
+ "learning_rate": 4.151609062744956e-05,
+ "loss": 0.4923740029335022,
+ "step": 545
+ },
+ {
+ "epoch": 0.2864637985309549,
+ "grad_norm": 0.1991746069706558,
+ "learning_rate": 4.148706030565079e-05,
+ "loss": 0.4868098497390747,
+ "step": 546
+ },
+ {
+ "epoch": 0.2869884575026233,
+ "grad_norm": 0.17707720674550964,
+ "learning_rate": 4.145799198414031e-05,
+ "loss": 0.459911584854126,
+ "step": 547
+ },
+ {
+ "epoch": 0.28751311647429173,
+ "grad_norm": 0.20416240852956843,
+ "learning_rate": 4.1428885741890244e-05,
+ "loss": 0.4870449900627136,
+ "step": 548
+ },
+ {
+ "epoch": 0.28803777544596015,
+ "grad_norm": 0.1797988029496339,
+ "learning_rate": 4.1399741657975724e-05,
+ "loss": 0.47509634494781494,
+ "step": 549
+ },
+ {
+ "epoch": 0.2885624344176285,
+ "grad_norm": 0.18038401737178636,
+ "learning_rate": 4.1370559811574726e-05,
+ "loss": 0.465260773897171,
+ "step": 550
+ },
+ {
+ "epoch": 0.28908709338929695,
+ "grad_norm": 0.18967415639647642,
+ "learning_rate": 4.134134028196777e-05,
+ "loss": 0.4796587824821472,
+ "step": 551
+ },
+ {
+ "epoch": 0.28961175236096537,
+ "grad_norm": 0.17423992719302386,
+ "learning_rate": 4.131208314853778e-05,
+ "loss": 0.4601815342903137,
+ "step": 552
+ },
+ {
+ "epoch": 0.2901364113326338,
+ "grad_norm": 0.17378377507331835,
+ "learning_rate": 4.1282788490769844e-05,
+ "loss": 0.46536317467689514,
+ "step": 553
+ },
+ {
+ "epoch": 0.2906610703043022,
+ "grad_norm": 0.19107939923258302,
+ "learning_rate": 4.1253456388250996e-05,
+ "loss": 0.48128658533096313,
+ "step": 554
+ },
+ {
+ "epoch": 0.29118572927597064,
+ "grad_norm": 0.181023799398351,
+ "learning_rate": 4.1224086920669986e-05,
+ "loss": 0.489425390958786,
+ "step": 555
+ },
+ {
+ "epoch": 0.291710388247639,
+ "grad_norm": 0.18318124679332567,
+ "learning_rate": 4.1194680167817074e-05,
+ "loss": 0.48763418197631836,
+ "step": 556
+ },
+ {
+ "epoch": 0.29223504721930743,
+ "grad_norm": 0.184903676814236,
+ "learning_rate": 4.116523620958384e-05,
+ "loss": 0.48576080799102783,
+ "step": 557
+ },
+ {
+ "epoch": 0.29275970619097585,
+ "grad_norm": 0.1878104056371087,
+ "learning_rate": 4.113575512596291e-05,
+ "loss": 0.47175419330596924,
+ "step": 558
+ },
+ {
+ "epoch": 0.2932843651626443,
+ "grad_norm": 0.18665604649109063,
+ "learning_rate": 4.11062369970478e-05,
+ "loss": 0.46720027923583984,
+ "step": 559
+ },
+ {
+ "epoch": 0.2938090241343127,
+ "grad_norm": 0.18682959716574435,
+ "learning_rate": 4.107668190303266e-05,
+ "loss": 0.4827365279197693,
+ "step": 560
+ },
+ {
+ "epoch": 0.2943336831059811,
+ "grad_norm": 0.18993292154980318,
+ "learning_rate": 4.1047089924212054e-05,
+ "loss": 0.48860135674476624,
+ "step": 561
+ },
+ {
+ "epoch": 0.29485834207764955,
+ "grad_norm": 0.1898197495285551,
+ "learning_rate": 4.1017461140980785e-05,
+ "loss": 0.4663747251033783,
+ "step": 562
+ },
+ {
+ "epoch": 0.2953830010493179,
+ "grad_norm": 0.18103066640629287,
+ "learning_rate": 4.09877956338336e-05,
+ "loss": 0.4649704396724701,
+ "step": 563
+ },
+ {
+ "epoch": 0.29590766002098634,
+ "grad_norm": 0.21700021489548446,
+ "learning_rate": 4.0958093483365055e-05,
+ "loss": 0.4909958839416504,
+ "step": 564
+ },
+ {
+ "epoch": 0.29643231899265476,
+ "grad_norm": 0.18294152749883993,
+ "learning_rate": 4.092835477026924e-05,
+ "loss": 0.4720481038093567,
+ "step": 565
+ },
+ {
+ "epoch": 0.2969569779643232,
+ "grad_norm": 0.19728437261973147,
+ "learning_rate": 4.089857957533957e-05,
+ "loss": 0.4950045049190521,
+ "step": 566
+ },
+ {
+ "epoch": 0.2974816369359916,
+ "grad_norm": 0.1959773716418508,
+ "learning_rate": 4.086876797946861e-05,
+ "loss": 0.47431886196136475,
+ "step": 567
+ },
+ {
+ "epoch": 0.29800629590766003,
+ "grad_norm": 0.18563709182351354,
+ "learning_rate": 4.083892006364777e-05,
+ "loss": 0.46619534492492676,
+ "step": 568
+ },
+ {
+ "epoch": 0.29853095487932846,
+ "grad_norm": 0.1892395016668497,
+ "learning_rate": 4.080903590896716e-05,
+ "loss": 0.48647207021713257,
+ "step": 569
+ },
+ {
+ "epoch": 0.2990556138509968,
+ "grad_norm": 0.19254597286971845,
+ "learning_rate": 4.0779115596615346e-05,
+ "loss": 0.4681573510169983,
+ "step": 570
+ },
+ {
+ "epoch": 0.29958027282266525,
+ "grad_norm": 0.18898908724918076,
+ "learning_rate": 4.074915920787912e-05,
+ "loss": 0.4771970510482788,
+ "step": 571
+ },
+ {
+ "epoch": 0.3001049317943337,
+ "grad_norm": 0.18527925057789466,
+ "learning_rate": 4.071916682414327e-05,
+ "loss": 0.49014243483543396,
+ "step": 572
+ },
+ {
+ "epoch": 0.3006295907660021,
+ "grad_norm": 0.19231827116273723,
+ "learning_rate": 4.0689138526890394e-05,
+ "loss": 0.48276934027671814,
+ "step": 573
+ },
+ {
+ "epoch": 0.3011542497376705,
+ "grad_norm": 0.18578460788549242,
+ "learning_rate": 4.065907439770066e-05,
+ "loss": 0.48563656210899353,
+ "step": 574
+ },
+ {
+ "epoch": 0.30167890870933894,
+ "grad_norm": 0.19338139569047225,
+ "learning_rate": 4.0628974518251575e-05,
+ "loss": 0.4691210389137268,
+ "step": 575
+ },
+ {
+ "epoch": 0.30220356768100737,
+ "grad_norm": 0.18848623641848672,
+ "learning_rate": 4.0598838970317765e-05,
+ "loss": 0.481357604265213,
+ "step": 576
+ },
+ {
+ "epoch": 0.30272822665267574,
+ "grad_norm": 0.1982288241100154,
+ "learning_rate": 4.0568667835770783e-05,
+ "loss": 0.470198392868042,
+ "step": 577
+ },
+ {
+ "epoch": 0.30325288562434416,
+ "grad_norm": 0.20275299445938652,
+ "learning_rate": 4.0538461196578835e-05,
+ "loss": 0.4942198395729065,
+ "step": 578
+ },
+ {
+ "epoch": 0.3037775445960126,
+ "grad_norm": 0.19435056916343918,
+ "learning_rate": 4.05082191348066e-05,
+ "loss": 0.4693480134010315,
+ "step": 579
+ },
+ {
+ "epoch": 0.304302203567681,
+ "grad_norm": 0.18958878206006355,
+ "learning_rate": 4.047794173261499e-05,
+ "loss": 0.4832199811935425,
+ "step": 580
+ },
+ {
+ "epoch": 0.30482686253934943,
+ "grad_norm": 0.18486502647006745,
+ "learning_rate": 4.044762907226093e-05,
+ "loss": 0.4676910638809204,
+ "step": 581
+ },
+ {
+ "epoch": 0.30535152151101785,
+ "grad_norm": 0.19326574060114263,
+ "learning_rate": 4.041728123609714e-05,
+ "loss": 0.49034518003463745,
+ "step": 582
+ },
+ {
+ "epoch": 0.3058761804826863,
+ "grad_norm": 0.18012391893837743,
+ "learning_rate": 4.038689830657187e-05,
+ "loss": 0.4750268757343292,
+ "step": 583
+ },
+ {
+ "epoch": 0.30640083945435465,
+ "grad_norm": 0.18450728254563814,
+ "learning_rate": 4.035648036622879e-05,
+ "loss": 0.4865539073944092,
+ "step": 584
+ },
+ {
+ "epoch": 0.30692549842602307,
+ "grad_norm": 0.18838506093398688,
+ "learning_rate": 4.032602749770658e-05,
+ "loss": 0.48115286231040955,
+ "step": 585
+ },
+ {
+ "epoch": 0.3074501573976915,
+ "grad_norm": 0.18115296003186768,
+ "learning_rate": 4.029553978373889e-05,
+ "loss": 0.46923011541366577,
+ "step": 586
+ },
+ {
+ "epoch": 0.3079748163693599,
+ "grad_norm": 0.18290085829721717,
+ "learning_rate": 4.026501730715401e-05,
+ "loss": 0.47181451320648193,
+ "step": 587
+ },
+ {
+ "epoch": 0.30849947534102834,
+ "grad_norm": 0.19513117762825424,
+ "learning_rate": 4.023446015087469e-05,
+ "loss": 0.4860653281211853,
+ "step": 588
+ },
+ {
+ "epoch": 0.30902413431269676,
+ "grad_norm": 0.18581581250210708,
+ "learning_rate": 4.020386839791786e-05,
+ "loss": 0.4601367115974426,
+ "step": 589
+ },
+ {
+ "epoch": 0.3095487932843652,
+ "grad_norm": 0.19898478659379668,
+ "learning_rate": 4.017324213139448e-05,
+ "loss": 0.48900914192199707,
+ "step": 590
+ },
+ {
+ "epoch": 0.31007345225603355,
+ "grad_norm": 0.17813270247026924,
+ "learning_rate": 4.014258143450926e-05,
+ "loss": 0.4816719889640808,
+ "step": 591
+ },
+ {
+ "epoch": 0.310598111227702,
+ "grad_norm": 0.18676381614678764,
+ "learning_rate": 4.011188639056044e-05,
+ "loss": 0.4765278697013855,
+ "step": 592
+ },
+ {
+ "epoch": 0.3111227701993704,
+ "grad_norm": 0.19217728375384824,
+ "learning_rate": 4.00811570829396e-05,
+ "loss": 0.468919575214386,
+ "step": 593
+ },
+ {
+ "epoch": 0.3116474291710388,
+ "grad_norm": 0.1905097118122076,
+ "learning_rate": 4.0050393595131366e-05,
+ "loss": 0.47396987676620483,
+ "step": 594
+ },
+ {
+ "epoch": 0.31217208814270725,
+ "grad_norm": 0.20101371047625896,
+ "learning_rate": 4.001959601071327e-05,
+ "loss": 0.47947967052459717,
+ "step": 595
+ },
+ {
+ "epoch": 0.3126967471143757,
+ "grad_norm": 0.19931836156324098,
+ "learning_rate": 3.998876441335544e-05,
+ "loss": 0.47473740577697754,
+ "step": 596
+ },
+ {
+ "epoch": 0.3132214060860441,
+ "grad_norm": 0.18935675800322405,
+ "learning_rate": 3.995789888682043e-05,
+ "loss": 0.4774348735809326,
+ "step": 597
+ },
+ {
+ "epoch": 0.31374606505771246,
+ "grad_norm": 0.18144723598187917,
+ "learning_rate": 3.9926999514962974e-05,
+ "loss": 0.47736242413520813,
+ "step": 598
+ },
+ {
+ "epoch": 0.3142707240293809,
+ "grad_norm": 0.18704474216590725,
+ "learning_rate": 3.989606638172972e-05,
+ "loss": 0.4696865379810333,
+ "step": 599
+ },
+ {
+ "epoch": 0.3147953830010493,
+ "grad_norm": 0.1963206607978136,
+ "learning_rate": 3.9865099571159096e-05,
+ "loss": 0.4707121253013611,
+ "step": 600
+ },
+ {
+ "epoch": 0.31532004197271774,
+ "grad_norm": 0.19042032319199087,
+ "learning_rate": 3.9834099167380984e-05,
+ "loss": 0.48166126012802124,
+ "step": 601
+ },
+ {
+ "epoch": 0.31584470094438616,
+ "grad_norm": 0.19667274077040733,
+ "learning_rate": 3.980306525461655e-05,
+ "loss": 0.4597858488559723,
+ "step": 602
+ },
+ {
+ "epoch": 0.3163693599160546,
+ "grad_norm": 0.18296501511671764,
+ "learning_rate": 3.9771997917177965e-05,
+ "loss": 0.4749276041984558,
+ "step": 603
+ },
+ {
+ "epoch": 0.316894018887723,
+ "grad_norm": 0.20424750040165998,
+ "learning_rate": 3.974089723946826e-05,
+ "loss": 0.48608389496803284,
+ "step": 604
+ },
+ {
+ "epoch": 0.3174186778593914,
+ "grad_norm": 0.179164516760055,
+ "learning_rate": 3.970976330598098e-05,
+ "loss": 0.4613687992095947,
+ "step": 605
+ },
+ {
+ "epoch": 0.3179433368310598,
+ "grad_norm": 0.1971675487899417,
+ "learning_rate": 3.967859620130008e-05,
+ "loss": 0.48067009449005127,
+ "step": 606
+ },
+ {
+ "epoch": 0.3184679958027282,
+ "grad_norm": 0.1984353794116141,
+ "learning_rate": 3.96473960100996e-05,
+ "loss": 0.46984297037124634,
+ "step": 607
+ },
+ {
+ "epoch": 0.31899265477439664,
+ "grad_norm": 0.19052277949743157,
+ "learning_rate": 3.961616281714346e-05,
+ "loss": 0.46637284755706787,
+ "step": 608
+ },
+ {
+ "epoch": 0.31951731374606507,
+ "grad_norm": 0.1948466211954685,
+ "learning_rate": 3.958489670728529e-05,
+ "loss": 0.45997026562690735,
+ "step": 609
+ },
+ {
+ "epoch": 0.3200419727177335,
+ "grad_norm": 0.1914611009126952,
+ "learning_rate": 3.9553597765468064e-05,
+ "loss": 0.46965113282203674,
+ "step": 610
+ },
+ {
+ "epoch": 0.3205666316894019,
+ "grad_norm": 0.18838407148180245,
+ "learning_rate": 3.9522266076724034e-05,
+ "loss": 0.46311426162719727,
+ "step": 611
+ },
+ {
+ "epoch": 0.3210912906610703,
+ "grad_norm": 0.18320616024247421,
+ "learning_rate": 3.949090172617437e-05,
+ "loss": 0.47325724363327026,
+ "step": 612
+ },
+ {
+ "epoch": 0.3216159496327387,
+ "grad_norm": 0.18208055375614307,
+ "learning_rate": 3.9459504799028996e-05,
+ "loss": 0.4793781638145447,
+ "step": 613
+ },
+ {
+ "epoch": 0.32214060860440713,
+ "grad_norm": 0.1779623681390907,
+ "learning_rate": 3.942807538058634e-05,
+ "loss": 0.49063172936439514,
+ "step": 614
+ },
+ {
+ "epoch": 0.32266526757607555,
+ "grad_norm": 0.1705509396899137,
+ "learning_rate": 3.9396613556233076e-05,
+ "loss": 0.46445411443710327,
+ "step": 615
+ },
+ {
+ "epoch": 0.323189926547744,
+ "grad_norm": 0.18638331049160017,
+ "learning_rate": 3.936511941144397e-05,
+ "loss": 0.4820844531059265,
+ "step": 616
+ },
+ {
+ "epoch": 0.3237145855194124,
+ "grad_norm": 0.1742031551459907,
+ "learning_rate": 3.933359303178152e-05,
+ "loss": 0.48277541995048523,
+ "step": 617
+ },
+ {
+ "epoch": 0.3242392444910808,
+ "grad_norm": 0.19402936613519242,
+ "learning_rate": 3.930203450289588e-05,
+ "loss": 0.48094409704208374,
+ "step": 618
+ },
+ {
+ "epoch": 0.3247639034627492,
+ "grad_norm": 0.17056373983083972,
+ "learning_rate": 3.9270443910524504e-05,
+ "loss": 0.47330591082572937,
+ "step": 619
+ },
+ {
+ "epoch": 0.3252885624344176,
+ "grad_norm": 0.19590242038608663,
+ "learning_rate": 3.9238821340491936e-05,
+ "loss": 0.463482528924942,
+ "step": 620
+ },
+ {
+ "epoch": 0.32581322140608604,
+ "grad_norm": 0.1773174424661309,
+ "learning_rate": 3.920716687870964e-05,
+ "loss": 0.4714035391807556,
+ "step": 621
+ },
+ {
+ "epoch": 0.32633788037775446,
+ "grad_norm": 0.21309078461219066,
+ "learning_rate": 3.9175480611175705e-05,
+ "loss": 0.48028913140296936,
+ "step": 622
+ },
+ {
+ "epoch": 0.3268625393494229,
+ "grad_norm": 0.18051754700054698,
+ "learning_rate": 3.91437626239746e-05,
+ "loss": 0.47057634592056274,
+ "step": 623
+ },
+ {
+ "epoch": 0.3273871983210913,
+ "grad_norm": 0.20308178439324137,
+ "learning_rate": 3.9112013003277025e-05,
+ "loss": 0.4846119284629822,
+ "step": 624
+ },
+ {
+ "epoch": 0.3279118572927597,
+ "grad_norm": 0.18763948014394452,
+ "learning_rate": 3.908023183533958e-05,
+ "loss": 0.4895526170730591,
+ "step": 625
+ },
+ {
+ "epoch": 0.3284365162644281,
+ "grad_norm": 0.20089961653784144,
+ "learning_rate": 3.904841920650458e-05,
+ "loss": 0.47670379281044006,
+ "step": 626
+ },
+ {
+ "epoch": 0.3289611752360965,
+ "grad_norm": 0.19361688884357725,
+ "learning_rate": 3.9016575203199836e-05,
+ "loss": 0.48227179050445557,
+ "step": 627
+ },
+ {
+ "epoch": 0.32948583420776495,
+ "grad_norm": 0.18811840125964013,
+ "learning_rate": 3.898469991193835e-05,
+ "loss": 0.48012083768844604,
+ "step": 628
+ },
+ {
+ "epoch": 0.3300104931794334,
+ "grad_norm": 0.20588907108781773,
+ "learning_rate": 3.895279341931816e-05,
+ "loss": 0.4617851972579956,
+ "step": 629
+ },
+ {
+ "epoch": 0.3305351521511018,
+ "grad_norm": 0.1824162959270584,
+ "learning_rate": 3.892085581202208e-05,
+ "loss": 0.4913630485534668,
+ "step": 630
+ },
+ {
+ "epoch": 0.3310598111227702,
+ "grad_norm": 0.21864095229522051,
+ "learning_rate": 3.888888717681742e-05,
+ "loss": 0.4805128872394562,
+ "step": 631
+ },
+ {
+ "epoch": 0.3315844700944386,
+ "grad_norm": 0.20737303366539847,
+ "learning_rate": 3.8856887600555814e-05,
+ "loss": 0.47446706891059875,
+ "step": 632
+ },
+ {
+ "epoch": 0.332109129066107,
+ "grad_norm": 0.1919647540172787,
+ "learning_rate": 3.882485717017294e-05,
+ "loss": 0.4690658152103424,
+ "step": 633
+ },
+ {
+ "epoch": 0.33263378803777544,
+ "grad_norm": 0.21875432305516018,
+ "learning_rate": 3.879279597268832e-05,
+ "loss": 0.47906386852264404,
+ "step": 634
+ },
+ {
+ "epoch": 0.33315844700944386,
+ "grad_norm": 0.1892578809878526,
+ "learning_rate": 3.8760704095205034e-05,
+ "loss": 0.49976128339767456,
+ "step": 635
+ },
+ {
+ "epoch": 0.3336831059811123,
+ "grad_norm": 0.21644492122330394,
+ "learning_rate": 3.8728581624909557e-05,
+ "loss": 0.4707983434200287,
+ "step": 636
+ },
+ {
+ "epoch": 0.3342077649527807,
+ "grad_norm": 0.19174191036050525,
+ "learning_rate": 3.869642864907142e-05,
+ "loss": 0.4768666625022888,
+ "step": 637
+ },
+ {
+ "epoch": 0.33473242392444913,
+ "grad_norm": 0.2024752056109731,
+ "learning_rate": 3.866424525504307e-05,
+ "loss": 0.4523438811302185,
+ "step": 638
+ },
+ {
+ "epoch": 0.3352570828961175,
+ "grad_norm": 0.2303129129798601,
+ "learning_rate": 3.8632031530259596e-05,
+ "loss": 0.4779718220233917,
+ "step": 639
+ },
+ {
+ "epoch": 0.3357817418677859,
+ "grad_norm": 0.1921588821226795,
+ "learning_rate": 3.8599787562238455e-05,
+ "loss": 0.48594534397125244,
+ "step": 640
+ },
+ {
+ "epoch": 0.33630640083945434,
+ "grad_norm": 0.19501707802067358,
+ "learning_rate": 3.85675134385793e-05,
+ "loss": 0.4620492458343506,
+ "step": 641
+ },
+ {
+ "epoch": 0.33683105981112277,
+ "grad_norm": 0.2101264181752891,
+ "learning_rate": 3.853520924696372e-05,
+ "loss": 0.4693989157676697,
+ "step": 642
+ },
+ {
+ "epoch": 0.3373557187827912,
+ "grad_norm": 0.18848815814943715,
+ "learning_rate": 3.8502875075154935e-05,
+ "loss": 0.4820001721382141,
+ "step": 643
+ },
+ {
+ "epoch": 0.3378803777544596,
+ "grad_norm": 0.19485532962923965,
+ "learning_rate": 3.8470511010997687e-05,
+ "loss": 0.465011864900589,
+ "step": 644
+ },
+ {
+ "epoch": 0.33840503672612804,
+ "grad_norm": 0.19299597479807248,
+ "learning_rate": 3.8438117142417864e-05,
+ "loss": 0.4724501371383667,
+ "step": 645
+ },
+ {
+ "epoch": 0.3389296956977964,
+ "grad_norm": 0.1973619476486774,
+ "learning_rate": 3.840569355742237e-05,
+ "loss": 0.4736691117286682,
+ "step": 646
+ },
+ {
+ "epoch": 0.33945435466946483,
+ "grad_norm": 0.19514678127770438,
+ "learning_rate": 3.8373240344098824e-05,
+ "loss": 0.48075538873672485,
+ "step": 647
+ },
+ {
+ "epoch": 0.33997901364113325,
+ "grad_norm": 0.18859654613839305,
+ "learning_rate": 3.834075759061535e-05,
+ "loss": 0.4810548722743988,
+ "step": 648
+ },
+ {
+ "epoch": 0.3405036726128017,
+ "grad_norm": 0.1804310916374026,
+ "learning_rate": 3.830824538522031e-05,
+ "loss": 0.47870394587516785,
+ "step": 649
+ },
+ {
+ "epoch": 0.3410283315844701,
+ "grad_norm": 0.2065277336294725,
+ "learning_rate": 3.8275703816242085e-05,
+ "loss": 0.4836372137069702,
+ "step": 650
+ },
+ {
+ "epoch": 0.3415529905561385,
+ "grad_norm": 0.18345549713407117,
+ "learning_rate": 3.824313297208883e-05,
+ "loss": 0.48482322692871094,
+ "step": 651
+ },
+ {
+ "epoch": 0.34207764952780695,
+ "grad_norm": 0.1988436545531611,
+ "learning_rate": 3.8210532941248256e-05,
+ "loss": 0.49088552594184875,
+ "step": 652
+ },
+ {
+ "epoch": 0.3426023084994753,
+ "grad_norm": 0.19102516535098402,
+ "learning_rate": 3.8177903812287324e-05,
+ "loss": 0.4875457286834717,
+ "step": 653
+ },
+ {
+ "epoch": 0.34312696747114374,
+ "grad_norm": 0.18898452432771068,
+ "learning_rate": 3.81452456738521e-05,
+ "loss": 0.47810572385787964,
+ "step": 654
+ },
+ {
+ "epoch": 0.34365162644281216,
+ "grad_norm": 0.184714343677509,
+ "learning_rate": 3.811255861466741e-05,
+ "loss": 0.4739091396331787,
+ "step": 655
+ },
+ {
+ "epoch": 0.3441762854144806,
+ "grad_norm": 0.18737269677842625,
+ "learning_rate": 3.8079842723536684e-05,
+ "loss": 0.470388263463974,
+ "step": 656
+ },
+ {
+ "epoch": 0.344700944386149,
+ "grad_norm": 0.18302901684492437,
+ "learning_rate": 3.804709808934169e-05,
+ "loss": 0.48345351219177246,
+ "step": 657
+ },
+ {
+ "epoch": 0.34522560335781743,
+ "grad_norm": 0.19169089239810186,
+ "learning_rate": 3.801432480104225e-05,
+ "loss": 0.47023966908454895,
+ "step": 658
+ },
+ {
+ "epoch": 0.34575026232948586,
+ "grad_norm": 0.20280394871510182,
+ "learning_rate": 3.798152294767606e-05,
+ "loss": 0.487832248210907,
+ "step": 659
+ },
+ {
+ "epoch": 0.3462749213011542,
+ "grad_norm": 0.17659046418887644,
+ "learning_rate": 3.794869261835842e-05,
+ "loss": 0.4812365472316742,
+ "step": 660
+ },
+ {
+ "epoch": 0.34679958027282265,
+ "grad_norm": 0.19591485346258952,
+ "learning_rate": 3.791583390228197e-05,
+ "loss": 0.46557363867759705,
+ "step": 661
+ },
+ {
+ "epoch": 0.3473242392444911,
+ "grad_norm": 0.17486393651428034,
+ "learning_rate": 3.7882946888716484e-05,
+ "loss": 0.4659643769264221,
+ "step": 662
+ },
+ {
+ "epoch": 0.3478488982161595,
+ "grad_norm": 0.1963689773822065,
+ "learning_rate": 3.785003166700863e-05,
+ "loss": 0.4977255165576935,
+ "step": 663
+ },
+ {
+ "epoch": 0.3483735571878279,
+ "grad_norm": 0.18700519459285708,
+ "learning_rate": 3.781708832658169e-05,
+ "loss": 0.4631659984588623,
+ "step": 664
+ },
+ {
+ "epoch": 0.34889821615949634,
+ "grad_norm": 0.18894931466314774,
+ "learning_rate": 3.778411695693535e-05,
+ "loss": 0.45918798446655273,
+ "step": 665
+ },
+ {
+ "epoch": 0.34942287513116477,
+ "grad_norm": 0.18880152953171528,
+ "learning_rate": 3.775111764764543e-05,
+ "loss": 0.46911996603012085,
+ "step": 666
+ },
+ {
+ "epoch": 0.34994753410283314,
+ "grad_norm": 0.19160839987122374,
+ "learning_rate": 3.771809048836368e-05,
+ "loss": 0.460246205329895,
+ "step": 667
+ },
+ {
+ "epoch": 0.35047219307450156,
+ "grad_norm": 0.18613118424903888,
+ "learning_rate": 3.7685035568817486e-05,
+ "loss": 0.48542192578315735,
+ "step": 668
+ },
+ {
+ "epoch": 0.35099685204617,
+ "grad_norm": 0.19397545367897925,
+ "learning_rate": 3.765195297880967e-05,
+ "loss": 0.4696044325828552,
+ "step": 669
+ },
+ {
+ "epoch": 0.3515215110178384,
+ "grad_norm": 0.20084948640831596,
+ "learning_rate": 3.761884280821823e-05,
+ "loss": 0.48373615741729736,
+ "step": 670
+ },
+ {
+ "epoch": 0.35204616998950683,
+ "grad_norm": 0.19537748883131847,
+ "learning_rate": 3.758570514699608e-05,
+ "loss": 0.48010650277137756,
+ "step": 671
+ },
+ {
+ "epoch": 0.35257082896117525,
+ "grad_norm": 0.2132579743116352,
+ "learning_rate": 3.755254008517082e-05,
+ "loss": 0.47235727310180664,
+ "step": 672
+ },
+ {
+ "epoch": 0.3530954879328437,
+ "grad_norm": 0.18822908863970023,
+ "learning_rate": 3.751934771284452e-05,
+ "loss": 0.4659617245197296,
+ "step": 673
+ },
+ {
+ "epoch": 0.35362014690451204,
+ "grad_norm": 0.19862262493500743,
+ "learning_rate": 3.7486128120193396e-05,
+ "loss": 0.46733272075653076,
+ "step": 674
+ },
+ {
+ "epoch": 0.35414480587618047,
+ "grad_norm": 0.19237282620477125,
+ "learning_rate": 3.745288139746767e-05,
+ "loss": 0.4723363518714905,
+ "step": 675
+ },
+ {
+ "epoch": 0.3546694648478489,
+ "grad_norm": 0.19521980909605932,
+ "learning_rate": 3.7419607634991235e-05,
+ "loss": 0.4794524610042572,
+ "step": 676
+ },
+ {
+ "epoch": 0.3551941238195173,
+ "grad_norm": 0.1924269412551977,
+ "learning_rate": 3.738630692316146e-05,
+ "loss": 0.4653412997722626,
+ "step": 677
+ },
+ {
+ "epoch": 0.35571878279118574,
+ "grad_norm": 0.20267833717192388,
+ "learning_rate": 3.73529793524489e-05,
+ "loss": 0.46611469984054565,
+ "step": 678
+ },
+ {
+ "epoch": 0.35624344176285416,
+ "grad_norm": 0.19496180396727691,
+ "learning_rate": 3.731962501339714e-05,
+ "loss": 0.4949174225330353,
+ "step": 679
+ },
+ {
+ "epoch": 0.3567681007345226,
+ "grad_norm": 0.19351166305358677,
+ "learning_rate": 3.7286243996622435e-05,
+ "loss": 0.4892667233943939,
+ "step": 680
+ },
+ {
+ "epoch": 0.35729275970619095,
+ "grad_norm": 0.18192561299969984,
+ "learning_rate": 3.725283639281354e-05,
+ "loss": 0.482984721660614,
+ "step": 681
+ },
+ {
+ "epoch": 0.3578174186778594,
+ "grad_norm": 0.1962335214892454,
+ "learning_rate": 3.7219402292731426e-05,
+ "loss": 0.4856143295764923,
+ "step": 682
+ },
+ {
+ "epoch": 0.3583420776495278,
+ "grad_norm": 0.18770397209908057,
+ "learning_rate": 3.718594178720907e-05,
+ "loss": 0.4512186646461487,
+ "step": 683
+ },
+ {
+ "epoch": 0.3588667366211962,
+ "grad_norm": 0.17502501408253013,
+ "learning_rate": 3.715245496715116e-05,
+ "loss": 0.46332547068595886,
+ "step": 684
+ },
+ {
+ "epoch": 0.35939139559286465,
+ "grad_norm": 0.17939698357560485,
+ "learning_rate": 3.71189419235339e-05,
+ "loss": 0.4803802967071533,
+ "step": 685
+ },
+ {
+ "epoch": 0.35991605456453307,
+ "grad_norm": 0.16822088302873006,
+ "learning_rate": 3.708540274740473e-05,
+ "loss": 0.4618965983390808,
+ "step": 686
+ },
+ {
+ "epoch": 0.3604407135362015,
+ "grad_norm": 0.16625239168850683,
+ "learning_rate": 3.7051837529882085e-05,
+ "loss": 0.4509458541870117,
+ "step": 687
+ },
+ {
+ "epoch": 0.36096537250786986,
+ "grad_norm": 0.18162962626025703,
+ "learning_rate": 3.7018246362155133e-05,
+ "loss": 0.49511533975601196,
+ "step": 688
+ },
+ {
+ "epoch": 0.3614900314795383,
+ "grad_norm": 0.1742780021476734,
+ "learning_rate": 3.698462933548358e-05,
+ "loss": 0.47176942229270935,
+ "step": 689
+ },
+ {
+ "epoch": 0.3620146904512067,
+ "grad_norm": 0.1810888382469795,
+ "learning_rate": 3.6950986541197344e-05,
+ "loss": 0.47207826375961304,
+ "step": 690
+ },
+ {
+ "epoch": 0.36253934942287513,
+ "grad_norm": 0.17997894837922476,
+ "learning_rate": 3.691731807069636e-05,
+ "loss": 0.46649712324142456,
+ "step": 691
+ },
+ {
+ "epoch": 0.36306400839454356,
+ "grad_norm": 0.18913883779426133,
+ "learning_rate": 3.688362401545034e-05,
+ "loss": 0.4813411235809326,
+ "step": 692
+ },
+ {
+ "epoch": 0.363588667366212,
+ "grad_norm": 0.1795309829744703,
+ "learning_rate": 3.684990446699849e-05,
+ "loss": 0.49149495363235474,
+ "step": 693
+ },
+ {
+ "epoch": 0.3641133263378804,
+ "grad_norm": 0.1776284682782097,
+ "learning_rate": 3.681615951694926e-05,
+ "loss": 0.46758022904396057,
+ "step": 694
+ },
+ {
+ "epoch": 0.3646379853095488,
+ "grad_norm": 0.19188148744936434,
+ "learning_rate": 3.6782389256980145e-05,
+ "loss": 0.49666500091552734,
+ "step": 695
+ },
+ {
+ "epoch": 0.3651626442812172,
+ "grad_norm": 0.1785908089372554,
+ "learning_rate": 3.6748593778837366e-05,
+ "loss": 0.473566472530365,
+ "step": 696
+ },
+ {
+ "epoch": 0.3656873032528856,
+ "grad_norm": 0.17057576472240485,
+ "learning_rate": 3.671477317433567e-05,
+ "loss": 0.44567710161209106,
+ "step": 697
+ },
+ {
+ "epoch": 0.36621196222455404,
+ "grad_norm": 0.1733038192323539,
+ "learning_rate": 3.66809275353581e-05,
+ "loss": 0.46948713064193726,
+ "step": 698
+ },
+ {
+ "epoch": 0.36673662119622247,
+ "grad_norm": 0.1809290560000329,
+ "learning_rate": 3.664705695385562e-05,
+ "loss": 0.46283137798309326,
+ "step": 699
+ },
+ {
+ "epoch": 0.3672612801678909,
+ "grad_norm": 0.1797433700402037,
+ "learning_rate": 3.6613161521847075e-05,
+ "loss": 0.47661077976226807,
+ "step": 700
+ },
+ {
+ "epoch": 0.36778593913955926,
+ "grad_norm": 0.19739139956644097,
+ "learning_rate": 3.6579241331418746e-05,
+ "loss": 0.48989033699035645,
+ "step": 701
+ },
+ {
+ "epoch": 0.3683105981112277,
+ "grad_norm": 0.17434870321164336,
+ "learning_rate": 3.654529647472419e-05,
+ "loss": 0.46496671438217163,
+ "step": 702
+ },
+ {
+ "epoch": 0.3688352570828961,
+ "grad_norm": 0.1897108745458232,
+ "learning_rate": 3.651132704398399e-05,
+ "loss": 0.4777913987636566,
+ "step": 703
+ },
+ {
+ "epoch": 0.36935991605456453,
+ "grad_norm": 0.1694937550246897,
+ "learning_rate": 3.647733313148548e-05,
+ "loss": 0.4681360125541687,
+ "step": 704
+ },
+ {
+ "epoch": 0.36988457502623295,
+ "grad_norm": 0.19024406770648913,
+ "learning_rate": 3.644331482958252e-05,
+ "loss": 0.47455257177352905,
+ "step": 705
+ },
+ {
+ "epoch": 0.3704092339979014,
+ "grad_norm": 0.17617259328958404,
+ "learning_rate": 3.640927223069521e-05,
+ "loss": 0.4734068512916565,
+ "step": 706
+ },
+ {
+ "epoch": 0.3709338929695698,
+ "grad_norm": 0.18956238546722706,
+ "learning_rate": 3.637520542730968e-05,
+ "loss": 0.46937721967697144,
+ "step": 707
+ },
+ {
+ "epoch": 0.37145855194123817,
+ "grad_norm": 0.17525932440626427,
+ "learning_rate": 3.634111451197781e-05,
+ "loss": 0.47101178765296936,
+ "step": 708
+ },
+ {
+ "epoch": 0.3719832109129066,
+ "grad_norm": 0.18644176814870617,
+ "learning_rate": 3.630699957731696e-05,
+ "loss": 0.4720165729522705,
+ "step": 709
+ },
+ {
+ "epoch": 0.372507869884575,
+ "grad_norm": 0.17261128054772287,
+ "learning_rate": 3.627286071600981e-05,
+ "loss": 0.4692431688308716,
+ "step": 710
+ },
+ {
+ "epoch": 0.37303252885624344,
+ "grad_norm": 0.19948374376925587,
+ "learning_rate": 3.6238698020803966e-05,
+ "loss": 0.4788331389427185,
+ "step": 711
+ },
+ {
+ "epoch": 0.37355718782791186,
+ "grad_norm": 0.17223872964081752,
+ "learning_rate": 3.620451158451186e-05,
+ "loss": 0.4711341857910156,
+ "step": 712
+ },
+ {
+ "epoch": 0.3740818467995803,
+ "grad_norm": 0.20785601498072281,
+ "learning_rate": 3.6170301500010357e-05,
+ "loss": 0.48784637451171875,
+ "step": 713
+ },
+ {
+ "epoch": 0.3746065057712487,
+ "grad_norm": 0.1692027736628384,
+ "learning_rate": 3.613606786024061e-05,
+ "loss": 0.47366487979888916,
+ "step": 714
+ },
+ {
+ "epoch": 0.3751311647429171,
+ "grad_norm": 0.19692035233814445,
+ "learning_rate": 3.610181075820777e-05,
+ "loss": 0.5048215985298157,
+ "step": 715
+ },
+ {
+ "epoch": 0.3756558237145855,
+ "grad_norm": 0.1890028394125962,
+ "learning_rate": 3.606753028698067e-05,
+ "loss": 0.4747781753540039,
+ "step": 716
+ },
+ {
+ "epoch": 0.3761804826862539,
+ "grad_norm": 0.190934478768137,
+ "learning_rate": 3.603322653969172e-05,
+ "loss": 0.46920639276504517,
+ "step": 717
+ },
+ {
+ "epoch": 0.37670514165792235,
+ "grad_norm": 0.18208605796741986,
+ "learning_rate": 3.5998899609536506e-05,
+ "loss": 0.4760962724685669,
+ "step": 718
+ },
+ {
+ "epoch": 0.3772298006295908,
+ "grad_norm": 0.19069420563455455,
+ "learning_rate": 3.596454958977361e-05,
+ "loss": 0.4782423675060272,
+ "step": 719
+ },
+ {
+ "epoch": 0.3777544596012592,
+ "grad_norm": 0.19209837648058897,
+ "learning_rate": 3.5930176573724355e-05,
+ "loss": 0.47083091735839844,
+ "step": 720
+ },
+ {
+ "epoch": 0.3782791185729276,
+ "grad_norm": 0.18576517620810123,
+ "learning_rate": 3.589578065477252e-05,
+ "loss": 0.46874189376831055,
+ "step": 721
+ },
+ {
+ "epoch": 0.378803777544596,
+ "grad_norm": 0.18193394451341524,
+ "learning_rate": 3.586136192636413e-05,
+ "loss": 0.48555469512939453,
+ "step": 722
+ },
+ {
+ "epoch": 0.3793284365162644,
+ "grad_norm": 0.18443943112028222,
+ "learning_rate": 3.582692048200715e-05,
+ "loss": 0.4774631857872009,
+ "step": 723
+ },
+ {
+ "epoch": 0.37985309548793283,
+ "grad_norm": 0.18982995478401762,
+ "learning_rate": 3.5792456415271285e-05,
+ "loss": 0.4939538538455963,
+ "step": 724
+ },
+ {
+ "epoch": 0.38037775445960126,
+ "grad_norm": 0.17873207526952814,
+ "learning_rate": 3.5757969819787686e-05,
+ "loss": 0.47975707054138184,
+ "step": 725
+ },
+ {
+ "epoch": 0.3809024134312697,
+ "grad_norm": 0.1902036698282205,
+ "learning_rate": 3.5723460789248714e-05,
+ "loss": 0.46293818950653076,
+ "step": 726
+ },
+ {
+ "epoch": 0.3814270724029381,
+ "grad_norm": 0.17383025942930141,
+ "learning_rate": 3.5688929417407674e-05,
+ "loss": 0.48262521624565125,
+ "step": 727
+ },
+ {
+ "epoch": 0.38195173137460653,
+ "grad_norm": 0.17553539715723387,
+ "learning_rate": 3.565437579807858e-05,
+ "loss": 0.47784456610679626,
+ "step": 728
+ },
+ {
+ "epoch": 0.3824763903462749,
+ "grad_norm": 0.17506043495839824,
+ "learning_rate": 3.561980002513588e-05,
+ "loss": 0.45129501819610596,
+ "step": 729
+ },
+ {
+ "epoch": 0.3830010493179433,
+ "grad_norm": 0.19300536460320428,
+ "learning_rate": 3.55852021925142e-05,
+ "loss": 0.4777262210845947,
+ "step": 730
+ },
+ {
+ "epoch": 0.38352570828961174,
+ "grad_norm": 0.16586668482715058,
+ "learning_rate": 3.5550582394208126e-05,
+ "loss": 0.4725654125213623,
+ "step": 731
+ },
+ {
+ "epoch": 0.38405036726128017,
+ "grad_norm": 0.18443313868804215,
+ "learning_rate": 3.551594072427188e-05,
+ "loss": 0.46294188499450684,
+ "step": 732
+ },
+ {
+ "epoch": 0.3845750262329486,
+ "grad_norm": 0.17638641942568364,
+ "learning_rate": 3.5481277276819135e-05,
+ "loss": 0.4640953540802002,
+ "step": 733
+ },
+ {
+ "epoch": 0.385099685204617,
+ "grad_norm": 0.18546263638205782,
+ "learning_rate": 3.5446592146022705e-05,
+ "loss": 0.47433286905288696,
+ "step": 734
+ },
+ {
+ "epoch": 0.38562434417628544,
+ "grad_norm": 0.18378876457852503,
+ "learning_rate": 3.541188542611435e-05,
+ "loss": 0.4626409113407135,
+ "step": 735
+ },
+ {
+ "epoch": 0.3861490031479538,
+ "grad_norm": 0.1840855442861905,
+ "learning_rate": 3.537715721138444e-05,
+ "loss": 0.4584702253341675,
+ "step": 736
+ },
+ {
+ "epoch": 0.38667366211962223,
+ "grad_norm": 0.17431868598237935,
+ "learning_rate": 3.5342407596181756e-05,
+ "loss": 0.46623867750167847,
+ "step": 737
+ },
+ {
+ "epoch": 0.38719832109129065,
+ "grad_norm": 0.18676696402145893,
+ "learning_rate": 3.5307636674913236e-05,
+ "loss": 0.47735559940338135,
+ "step": 738
+ },
+ {
+ "epoch": 0.3877229800629591,
+ "grad_norm": 0.17350534499406192,
+ "learning_rate": 3.5272844542043684e-05,
+ "loss": 0.46928340196609497,
+ "step": 739
+ },
+ {
+ "epoch": 0.3882476390346275,
+ "grad_norm": 0.18522239767382523,
+ "learning_rate": 3.523803129209554e-05,
+ "loss": 0.4713566303253174,
+ "step": 740
+ },
+ {
+ "epoch": 0.3887722980062959,
+ "grad_norm": 0.1758276789151429,
+ "learning_rate": 3.520319701964861e-05,
+ "loss": 0.4708835780620575,
+ "step": 741
+ },
+ {
+ "epoch": 0.38929695697796435,
+ "grad_norm": 0.1807634512537148,
+ "learning_rate": 3.5168341819339814e-05,
+ "loss": 0.4628238081932068,
+ "step": 742
+ },
+ {
+ "epoch": 0.3898216159496327,
+ "grad_norm": 0.18330480185134132,
+ "learning_rate": 3.513346578586293e-05,
+ "loss": 0.4731083810329437,
+ "step": 743
+ },
+ {
+ "epoch": 0.39034627492130114,
+ "grad_norm": 0.17794661224360395,
+ "learning_rate": 3.509856901396831e-05,
+ "loss": 0.46726107597351074,
+ "step": 744
+ },
+ {
+ "epoch": 0.39087093389296956,
+ "grad_norm": 0.1920009277299435,
+ "learning_rate": 3.506365159846271e-05,
+ "loss": 0.46715307235717773,
+ "step": 745
+ },
+ {
+ "epoch": 0.391395592864638,
+ "grad_norm": 0.1747159001305072,
+ "learning_rate": 3.5028713634208914e-05,
+ "loss": 0.4766237139701843,
+ "step": 746
+ },
+ {
+ "epoch": 0.3919202518363064,
+ "grad_norm": 0.20209319046342394,
+ "learning_rate": 3.499375521612554e-05,
+ "loss": 0.4678901433944702,
+ "step": 747
+ },
+ {
+ "epoch": 0.39244491080797483,
+ "grad_norm": 0.17507182935416063,
+ "learning_rate": 3.4958776439186796e-05,
+ "loss": 0.46668827533721924,
+ "step": 748
+ },
+ {
+ "epoch": 0.39296956977964326,
+ "grad_norm": 0.20197990142261657,
+ "learning_rate": 3.492377739842219e-05,
+ "loss": 0.4900662899017334,
+ "step": 749
+ },
+ {
+ "epoch": 0.3934942287513116,
+ "grad_norm": 0.17482655657075183,
+ "learning_rate": 3.4888758188916274e-05,
+ "loss": 0.4739720821380615,
+ "step": 750
+ },
+ {
+ "epoch": 0.39401888772298005,
+ "grad_norm": 0.18871692146823568,
+ "learning_rate": 3.4853718905808405e-05,
+ "loss": 0.46060967445373535,
+ "step": 751
+ },
+ {
+ "epoch": 0.3945435466946485,
+ "grad_norm": 0.17453187843806395,
+ "learning_rate": 3.4818659644292474e-05,
+ "loss": 0.46890175342559814,
+ "step": 752
+ },
+ {
+ "epoch": 0.3950682056663169,
+ "grad_norm": 0.19130839632557065,
+ "learning_rate": 3.478358049961663e-05,
+ "loss": 0.46784400939941406,
+ "step": 753
+ },
+ {
+ "epoch": 0.3955928646379853,
+ "grad_norm": 0.17440785596043434,
+ "learning_rate": 3.474848156708308e-05,
+ "loss": 0.4774072766304016,
+ "step": 754
+ },
+ {
+ "epoch": 0.39611752360965374,
+ "grad_norm": 0.18014440318099736,
+ "learning_rate": 3.4713362942047745e-05,
+ "loss": 0.47180938720703125,
+ "step": 755
+ },
+ {
+ "epoch": 0.39664218258132217,
+ "grad_norm": 0.18307376044235557,
+ "learning_rate": 3.4678224719920083e-05,
+ "loss": 0.4629283845424652,
+ "step": 756
+ },
+ {
+ "epoch": 0.39716684155299053,
+ "grad_norm": 0.17565693913236313,
+ "learning_rate": 3.464306699616277e-05,
+ "loss": 0.47490352392196655,
+ "step": 757
+ },
+ {
+ "epoch": 0.39769150052465896,
+ "grad_norm": 0.1929290633352642,
+ "learning_rate": 3.460788986629147e-05,
+ "loss": 0.4784941077232361,
+ "step": 758
+ },
+ {
+ "epoch": 0.3982161594963274,
+ "grad_norm": 0.17509419605212645,
+ "learning_rate": 3.457269342587458e-05,
+ "loss": 0.4653838574886322,
+ "step": 759
+ },
+ {
+ "epoch": 0.3987408184679958,
+ "grad_norm": 0.18128390992721163,
+ "learning_rate": 3.453747777053293e-05,
+ "loss": 0.4671784043312073,
+ "step": 760
+ },
+ {
+ "epoch": 0.39926547743966423,
+ "grad_norm": 0.1791616728645312,
+ "learning_rate": 3.4502242995939596e-05,
+ "loss": 0.4871884882450104,
+ "step": 761
+ },
+ {
+ "epoch": 0.39979013641133265,
+ "grad_norm": 0.19074658173660747,
+ "learning_rate": 3.446698919781956e-05,
+ "loss": 0.44543662667274475,
+ "step": 762
+ },
+ {
+ "epoch": 0.4003147953830011,
+ "grad_norm": 0.20126643497694777,
+ "learning_rate": 3.443171647194951e-05,
+ "loss": 0.46859604120254517,
+ "step": 763
+ },
+ {
+ "epoch": 0.40083945435466944,
+ "grad_norm": 0.19094202368491892,
+ "learning_rate": 3.4396424914157524e-05,
+ "loss": 0.480099081993103,
+ "step": 764
+ },
+ {
+ "epoch": 0.40136411332633787,
+ "grad_norm": 0.19366222488694126,
+ "learning_rate": 3.4361114620322875e-05,
+ "loss": 0.47904855012893677,
+ "step": 765
+ },
+ {
+ "epoch": 0.4018887722980063,
+ "grad_norm": 0.19702291884184286,
+ "learning_rate": 3.432578568637574e-05,
+ "loss": 0.4629756808280945,
+ "step": 766
+ },
+ {
+ "epoch": 0.4024134312696747,
+ "grad_norm": 0.16691251936787443,
+ "learning_rate": 3.429043820829691e-05,
+ "loss": 0.45320895314216614,
+ "step": 767
+ },
+ {
+ "epoch": 0.40293809024134314,
+ "grad_norm": 0.19622688405826869,
+ "learning_rate": 3.425507228211757e-05,
+ "loss": 0.4798918664455414,
+ "step": 768
+ },
+ {
+ "epoch": 0.40346274921301156,
+ "grad_norm": 0.1782912454285934,
+ "learning_rate": 3.421968800391903e-05,
+ "loss": 0.46745413541793823,
+ "step": 769
+ },
+ {
+ "epoch": 0.40398740818467993,
+ "grad_norm": 0.1900756854637891,
+ "learning_rate": 3.418428546983245e-05,
+ "loss": 0.47500404715538025,
+ "step": 770
+ },
+ {
+ "epoch": 0.40451206715634835,
+ "grad_norm": 0.17832290645561236,
+ "learning_rate": 3.414886477603859e-05,
+ "loss": 0.45256009697914124,
+ "step": 771
+ },
+ {
+ "epoch": 0.4050367261280168,
+ "grad_norm": 0.18392866667597935,
+ "learning_rate": 3.411342601876754e-05,
+ "loss": 0.4751660227775574,
+ "step": 772
+ },
+ {
+ "epoch": 0.4055613850996852,
+ "grad_norm": 0.1882109158311612,
+ "learning_rate": 3.407796929429849e-05,
+ "loss": 0.46648120880126953,
+ "step": 773
+ },
+ {
+ "epoch": 0.4060860440713536,
+ "grad_norm": 0.18951718125370368,
+ "learning_rate": 3.40424946989594e-05,
+ "loss": 0.46590569615364075,
+ "step": 774
+ },
+ {
+ "epoch": 0.40661070304302205,
+ "grad_norm": 0.19849977165588717,
+ "learning_rate": 3.400700232912681e-05,
+ "loss": 0.47594356536865234,
+ "step": 775
+ },
+ {
+ "epoch": 0.40713536201469047,
+ "grad_norm": 0.18782224015917673,
+ "learning_rate": 3.397149228122555e-05,
+ "loss": 0.4912794828414917,
+ "step": 776
+ },
+ {
+ "epoch": 0.40766002098635884,
+ "grad_norm": 0.1931513109313898,
+ "learning_rate": 3.393596465172848e-05,
+ "loss": 0.46562087535858154,
+ "step": 777
+ },
+ {
+ "epoch": 0.40818467995802726,
+ "grad_norm": 0.17925977530765735,
+ "learning_rate": 3.3900419537156186e-05,
+ "loss": 0.46635520458221436,
+ "step": 778
+ },
+ {
+ "epoch": 0.4087093389296957,
+ "grad_norm": 0.17825481237355673,
+ "learning_rate": 3.386485703407682e-05,
+ "loss": 0.46161317825317383,
+ "step": 779
+ },
+ {
+ "epoch": 0.4092339979013641,
+ "grad_norm": 0.1760886807555647,
+ "learning_rate": 3.3829277239105714e-05,
+ "loss": 0.4768933355808258,
+ "step": 780
+ },
+ {
+ "epoch": 0.40975865687303253,
+ "grad_norm": 0.17886191066012203,
+ "learning_rate": 3.3793680248905226e-05,
+ "loss": 0.47968024015426636,
+ "step": 781
+ },
+ {
+ "epoch": 0.41028331584470096,
+ "grad_norm": 0.17960501055090095,
+ "learning_rate": 3.3758066160184405e-05,
+ "loss": 0.4786861538887024,
+ "step": 782
+ },
+ {
+ "epoch": 0.4108079748163694,
+ "grad_norm": 0.1764747678311139,
+ "learning_rate": 3.3722435069698746e-05,
+ "loss": 0.4682663083076477,
+ "step": 783
+ },
+ {
+ "epoch": 0.41133263378803775,
+ "grad_norm": 0.17865265007610734,
+ "learning_rate": 3.368678707424996e-05,
+ "loss": 0.4737027883529663,
+ "step": 784
+ },
+ {
+ "epoch": 0.4118572927597062,
+ "grad_norm": 0.18295805209998312,
+ "learning_rate": 3.365112227068566e-05,
+ "loss": 0.4658524990081787,
+ "step": 785
+ },
+ {
+ "epoch": 0.4123819517313746,
+ "grad_norm": 0.176320646726516,
+ "learning_rate": 3.3615440755899154e-05,
+ "loss": 0.48119238018989563,
+ "step": 786
+ },
+ {
+ "epoch": 0.412906610703043,
+ "grad_norm": 0.1875070556670073,
+ "learning_rate": 3.3579742626829095e-05,
+ "loss": 0.47216686606407166,
+ "step": 787
+ },
+ {
+ "epoch": 0.41343126967471144,
+ "grad_norm": 0.1706565484233721,
+ "learning_rate": 3.354402798045935e-05,
+ "loss": 0.44997406005859375,
+ "step": 788
+ },
+ {
+ "epoch": 0.41395592864637987,
+ "grad_norm": 0.18512519460732668,
+ "learning_rate": 3.350829691381858e-05,
+ "loss": 0.47532424330711365,
+ "step": 789
+ },
+ {
+ "epoch": 0.4144805876180483,
+ "grad_norm": 0.17883995824220275,
+ "learning_rate": 3.347254952398011e-05,
+ "loss": 0.45862317085266113,
+ "step": 790
+ },
+ {
+ "epoch": 0.41500524658971666,
+ "grad_norm": 0.18026583479320943,
+ "learning_rate": 3.343678590806161e-05,
+ "loss": 0.4655536413192749,
+ "step": 791
+ },
+ {
+ "epoch": 0.4155299055613851,
+ "grad_norm": 0.19136184328481412,
+ "learning_rate": 3.340100616322479e-05,
+ "loss": 0.4819777011871338,
+ "step": 792
+ },
+ {
+ "epoch": 0.4160545645330535,
+ "grad_norm": 0.1778448857429592,
+ "learning_rate": 3.336521038667523e-05,
+ "loss": 0.45306986570358276,
+ "step": 793
+ },
+ {
+ "epoch": 0.41657922350472193,
+ "grad_norm": 0.21144395575441685,
+ "learning_rate": 3.332939867566203e-05,
+ "loss": 0.4786582589149475,
+ "step": 794
+ },
+ {
+ "epoch": 0.41710388247639035,
+ "grad_norm": 0.16947561406287748,
+ "learning_rate": 3.329357112747759e-05,
+ "loss": 0.45980104804039,
+ "step": 795
+ },
+ {
+ "epoch": 0.4176285414480588,
+ "grad_norm": 0.19916426912928414,
+ "learning_rate": 3.3257727839457355e-05,
+ "loss": 0.4771812856197357,
+ "step": 796
+ },
+ {
+ "epoch": 0.4181532004197272,
+ "grad_norm": 0.18135502182361274,
+ "learning_rate": 3.322186890897949e-05,
+ "loss": 0.4651404917240143,
+ "step": 797
+ },
+ {
+ "epoch": 0.41867785939139557,
+ "grad_norm": 0.18446053787470598,
+ "learning_rate": 3.318599443346471e-05,
+ "loss": 0.4591749608516693,
+ "step": 798
+ },
+ {
+ "epoch": 0.419202518363064,
+ "grad_norm": 0.1854243496640156,
+ "learning_rate": 3.3150104510375905e-05,
+ "loss": 0.4756122827529907,
+ "step": 799
+ },
+ {
+ "epoch": 0.4197271773347324,
+ "grad_norm": 0.17863210960014014,
+ "learning_rate": 3.311419923721798e-05,
+ "loss": 0.45419955253601074,
+ "step": 800
+ },
+ {
+ "epoch": 0.42025183630640084,
+ "grad_norm": 0.2012694989641943,
+ "learning_rate": 3.307827871153752e-05,
+ "loss": 0.4819367229938507,
+ "step": 801
+ },
+ {
+ "epoch": 0.42077649527806926,
+ "grad_norm": 0.1693683871587881,
+ "learning_rate": 3.3042343030922555e-05,
+ "loss": 0.4602414667606354,
+ "step": 802
+ },
+ {
+ "epoch": 0.4213011542497377,
+ "grad_norm": 0.19774070264511331,
+ "learning_rate": 3.3006392293002284e-05,
+ "loss": 0.46831682324409485,
+ "step": 803
+ },
+ {
+ "epoch": 0.4218258132214061,
+ "grad_norm": 0.17724479264201887,
+ "learning_rate": 3.2970426595446804e-05,
+ "loss": 0.4741601347923279,
+ "step": 804
+ },
+ {
+ "epoch": 0.4223504721930745,
+ "grad_norm": 0.18551771187740332,
+ "learning_rate": 3.293444603596688e-05,
+ "loss": 0.4742834270000458,
+ "step": 805
+ },
+ {
+ "epoch": 0.4228751311647429,
+ "grad_norm": 0.1696027485701005,
+ "learning_rate": 3.2898450712313624e-05,
+ "loss": 0.4645093083381653,
+ "step": 806
+ },
+ {
+ "epoch": 0.4233997901364113,
+ "grad_norm": 0.17349448260552724,
+ "learning_rate": 3.2862440722278284e-05,
+ "loss": 0.4489970803260803,
+ "step": 807
+ },
+ {
+ "epoch": 0.42392444910807975,
+ "grad_norm": 0.19295574973895221,
+ "learning_rate": 3.282641616369193e-05,
+ "loss": 0.46199095249176025,
+ "step": 808
+ },
+ {
+ "epoch": 0.42444910807974817,
+ "grad_norm": 0.17967818065949828,
+ "learning_rate": 3.279037713442523e-05,
+ "loss": 0.47169333696365356,
+ "step": 809
+ },
+ {
+ "epoch": 0.4249737670514166,
+ "grad_norm": 0.20163103402616964,
+ "learning_rate": 3.275432373238815e-05,
+ "loss": 0.45942437648773193,
+ "step": 810
+ },
+ {
+ "epoch": 0.425498426023085,
+ "grad_norm": 0.18358785705656622,
+ "learning_rate": 3.271825605552973e-05,
+ "loss": 0.4602029323577881,
+ "step": 811
+ },
+ {
+ "epoch": 0.4260230849947534,
+ "grad_norm": 0.18112866528723698,
+ "learning_rate": 3.268217420183777e-05,
+ "loss": 0.45105835795402527,
+ "step": 812
+ },
+ {
+ "epoch": 0.4265477439664218,
+ "grad_norm": 0.20571671712289333,
+ "learning_rate": 3.264607826933857e-05,
+ "loss": 0.46072322130203247,
+ "step": 813
+ },
+ {
+ "epoch": 0.42707240293809023,
+ "grad_norm": 0.17963891336766089,
+ "learning_rate": 3.260996835609672e-05,
+ "loss": 0.4416537880897522,
+ "step": 814
+ },
+ {
+ "epoch": 0.42759706190975866,
+ "grad_norm": 0.17172650763122463,
+ "learning_rate": 3.2573844560214754e-05,
+ "loss": 0.4714724123477936,
+ "step": 815
+ },
+ {
+ "epoch": 0.4281217208814271,
+ "grad_norm": 0.1833239618528263,
+ "learning_rate": 3.2537706979832956e-05,
+ "loss": 0.4733719229698181,
+ "step": 816
+ },
+ {
+ "epoch": 0.4286463798530955,
+ "grad_norm": 0.17758216623120027,
+ "learning_rate": 3.250155571312903e-05,
+ "loss": 0.4746835231781006,
+ "step": 817
+ },
+ {
+ "epoch": 0.42917103882476393,
+ "grad_norm": 0.1746032400447865,
+ "learning_rate": 3.246539085831787e-05,
+ "loss": 0.45553964376449585,
+ "step": 818
+ },
+ {
+ "epoch": 0.4296956977964323,
+ "grad_norm": 0.1871343685963966,
+ "learning_rate": 3.2429212513651314e-05,
+ "loss": 0.4555765986442566,
+ "step": 819
+ },
+ {
+ "epoch": 0.4302203567681007,
+ "grad_norm": 0.1792760090402548,
+ "learning_rate": 3.23930207774178e-05,
+ "loss": 0.49340271949768066,
+ "step": 820
+ },
+ {
+ "epoch": 0.43074501573976914,
+ "grad_norm": 0.18249133498413006,
+ "learning_rate": 3.235681574794219e-05,
+ "loss": 0.4710668623447418,
+ "step": 821
+ },
+ {
+ "epoch": 0.43126967471143757,
+ "grad_norm": 0.17672806714478967,
+ "learning_rate": 3.232059752358543e-05,
+ "loss": 0.47569704055786133,
+ "step": 822
+ },
+ {
+ "epoch": 0.431794333683106,
+ "grad_norm": 0.18641432224245422,
+ "learning_rate": 3.2284366202744335e-05,
+ "loss": 0.4691632091999054,
+ "step": 823
+ },
+ {
+ "epoch": 0.4323189926547744,
+ "grad_norm": 0.1719261720825053,
+ "learning_rate": 3.224812188385129e-05,
+ "loss": 0.45778918266296387,
+ "step": 824
+ },
+ {
+ "epoch": 0.43284365162644284,
+ "grad_norm": 0.18683887926081086,
+ "learning_rate": 3.221186466537402e-05,
+ "loss": 0.4613558351993561,
+ "step": 825
+ },
+ {
+ "epoch": 0.4333683105981112,
+ "grad_norm": 0.16969741876720854,
+ "learning_rate": 3.217559464581523e-05,
+ "loss": 0.4585855007171631,
+ "step": 826
+ },
+ {
+ "epoch": 0.43389296956977963,
+ "grad_norm": 0.18472092831216783,
+ "learning_rate": 3.213931192371248e-05,
+ "loss": 0.4576684832572937,
+ "step": 827
+ },
+ {
+ "epoch": 0.43441762854144805,
+ "grad_norm": 0.17154369054810312,
+ "learning_rate": 3.210301659763777e-05,
+ "loss": 0.45437556505203247,
+ "step": 828
+ },
+ {
+ "epoch": 0.4349422875131165,
+ "grad_norm": 0.18042389014075988,
+ "learning_rate": 3.2066708766197406e-05,
+ "loss": 0.4654255211353302,
+ "step": 829
+ },
+ {
+ "epoch": 0.4354669464847849,
+ "grad_norm": 0.17308036910647218,
+ "learning_rate": 3.203038852803162e-05,
+ "loss": 0.46544164419174194,
+ "step": 830
+ },
+ {
+ "epoch": 0.4359916054564533,
+ "grad_norm": 0.1818956612485934,
+ "learning_rate": 3.199405598181438e-05,
+ "loss": 0.46769338846206665,
+ "step": 831
+ },
+ {
+ "epoch": 0.43651626442812175,
+ "grad_norm": 0.17181156893900013,
+ "learning_rate": 3.1957711226253075e-05,
+ "loss": 0.46617552638053894,
+ "step": 832
+ },
+ {
+ "epoch": 0.4370409233997901,
+ "grad_norm": 0.18045912652510682,
+ "learning_rate": 3.192135436008827e-05,
+ "loss": 0.476098895072937,
+ "step": 833
+ },
+ {
+ "epoch": 0.43756558237145854,
+ "grad_norm": 0.1779320299090377,
+ "learning_rate": 3.1884985482093426e-05,
+ "loss": 0.4711330831050873,
+ "step": 834
+ },
+ {
+ "epoch": 0.43809024134312696,
+ "grad_norm": 0.1750785521590116,
+ "learning_rate": 3.1848604691074655e-05,
+ "loss": 0.46946778893470764,
+ "step": 835
+ },
+ {
+ "epoch": 0.4386149003147954,
+ "grad_norm": 0.17698182449515382,
+ "learning_rate": 3.181221208587041e-05,
+ "loss": 0.4675105810165405,
+ "step": 836
+ },
+ {
+ "epoch": 0.4391395592864638,
+ "grad_norm": 0.16809336680661224,
+ "learning_rate": 3.1775807765351264e-05,
+ "loss": 0.4664367735385895,
+ "step": 837
+ },
+ {
+ "epoch": 0.43966421825813223,
+ "grad_norm": 0.1779284662856733,
+ "learning_rate": 3.173939182841959e-05,
+ "loss": 0.4691658616065979,
+ "step": 838
+ },
+ {
+ "epoch": 0.44018887722980066,
+ "grad_norm": 0.18158282038148463,
+ "learning_rate": 3.1702964374009356e-05,
+ "loss": 0.47681736946105957,
+ "step": 839
+ },
+ {
+ "epoch": 0.440713536201469,
+ "grad_norm": 0.17424121130172063,
+ "learning_rate": 3.16665255010858e-05,
+ "loss": 0.4832586646080017,
+ "step": 840
+ },
+ {
+ "epoch": 0.44123819517313745,
+ "grad_norm": 0.1946839178805279,
+ "learning_rate": 3.163007530864515e-05,
+ "loss": 0.4796105623245239,
+ "step": 841
+ },
+ {
+ "epoch": 0.44176285414480587,
+ "grad_norm": 0.1729676950978342,
+ "learning_rate": 3.159361389571446e-05,
+ "loss": 0.45837289094924927,
+ "step": 842
+ },
+ {
+ "epoch": 0.4422875131164743,
+ "grad_norm": 0.1815936391211383,
+ "learning_rate": 3.155714136135121e-05,
+ "loss": 0.45617926120758057,
+ "step": 843
+ },
+ {
+ "epoch": 0.4428121720881427,
+ "grad_norm": 0.18492332666214983,
+ "learning_rate": 3.152065780464311e-05,
+ "loss": 0.4543980658054352,
+ "step": 844
+ },
+ {
+ "epoch": 0.44333683105981114,
+ "grad_norm": 0.17705825903751285,
+ "learning_rate": 3.148416332470782e-05,
+ "loss": 0.4591902494430542,
+ "step": 845
+ },
+ {
+ "epoch": 0.4438614900314795,
+ "grad_norm": 0.18334931746091654,
+ "learning_rate": 3.144765802069267e-05,
+ "loss": 0.46418845653533936,
+ "step": 846
+ },
+ {
+ "epoch": 0.44438614900314793,
+ "grad_norm": 0.1764131192946885,
+ "learning_rate": 3.14111419917744e-05,
+ "loss": 0.45251819491386414,
+ "step": 847
+ },
+ {
+ "epoch": 0.44491080797481636,
+ "grad_norm": 0.17092125726313687,
+ "learning_rate": 3.137461533715887e-05,
+ "loss": 0.4768788814544678,
+ "step": 848
+ },
+ {
+ "epoch": 0.4454354669464848,
+ "grad_norm": 0.172371784237127,
+ "learning_rate": 3.1338078156080854e-05,
+ "loss": 0.45612800121307373,
+ "step": 849
+ },
+ {
+ "epoch": 0.4459601259181532,
+ "grad_norm": 0.18155942681350976,
+ "learning_rate": 3.130153054780368e-05,
+ "loss": 0.47879016399383545,
+ "step": 850
+ },
+ {
+ "epoch": 0.44648478488982163,
+ "grad_norm": 0.1765744108320689,
+ "learning_rate": 3.1264972611619e-05,
+ "loss": 0.46737438440322876,
+ "step": 851
+ },
+ {
+ "epoch": 0.44700944386149005,
+ "grad_norm": 0.17937790491025424,
+ "learning_rate": 3.122840444684658e-05,
+ "loss": 0.46918386220932007,
+ "step": 852
+ },
+ {
+ "epoch": 0.4475341028331584,
+ "grad_norm": 0.16893607826790863,
+ "learning_rate": 3.119182615283391e-05,
+ "loss": 0.45767197012901306,
+ "step": 853
+ },
+ {
+ "epoch": 0.44805876180482684,
+ "grad_norm": 0.17790762342811886,
+ "learning_rate": 3.1155237828956035e-05,
+ "loss": 0.46405214071273804,
+ "step": 854
+ },
+ {
+ "epoch": 0.44858342077649527,
+ "grad_norm": 0.18130690095111512,
+ "learning_rate": 3.1118639574615236e-05,
+ "loss": 0.47842586040496826,
+ "step": 855
+ },
+ {
+ "epoch": 0.4491080797481637,
+ "grad_norm": 0.17084443519974984,
+ "learning_rate": 3.108203148924078e-05,
+ "loss": 0.4425484538078308,
+ "step": 856
+ },
+ {
+ "epoch": 0.4496327387198321,
+ "grad_norm": 0.1656553911213869,
+ "learning_rate": 3.1045413672288636e-05,
+ "loss": 0.4654726982116699,
+ "step": 857
+ },
+ {
+ "epoch": 0.45015739769150054,
+ "grad_norm": 0.16901395208369688,
+ "learning_rate": 3.100878622324123e-05,
+ "loss": 0.4691373109817505,
+ "step": 858
+ },
+ {
+ "epoch": 0.45068205666316896,
+ "grad_norm": 0.16932810685899333,
+ "learning_rate": 3.097214924160712e-05,
+ "loss": 0.4631810784339905,
+ "step": 859
+ },
+ {
+ "epoch": 0.45120671563483733,
+ "grad_norm": 0.17289591787117292,
+ "learning_rate": 3.093550282692081e-05,
+ "loss": 0.47004759311676025,
+ "step": 860
+ },
+ {
+ "epoch": 0.45173137460650575,
+ "grad_norm": 0.18267781169263053,
+ "learning_rate": 3.089884707874237e-05,
+ "loss": 0.47008734941482544,
+ "step": 861
+ },
+ {
+ "epoch": 0.4522560335781742,
+ "grad_norm": 0.19731984788376258,
+ "learning_rate": 3.0862182096657296e-05,
+ "loss": 0.4688064455986023,
+ "step": 862
+ },
+ {
+ "epoch": 0.4527806925498426,
+ "grad_norm": 0.17030427209496843,
+ "learning_rate": 3.082550798027612e-05,
+ "loss": 0.46062836050987244,
+ "step": 863
+ },
+ {
+ "epoch": 0.453305351521511,
+ "grad_norm": 0.17802260101618392,
+ "learning_rate": 3.07888248292342e-05,
+ "loss": 0.4703964591026306,
+ "step": 864
+ },
+ {
+ "epoch": 0.45383001049317945,
+ "grad_norm": 0.1792813037794071,
+ "learning_rate": 3.0752132743191466e-05,
+ "loss": 0.48758721351623535,
+ "step": 865
+ },
+ {
+ "epoch": 0.45435466946484787,
+ "grad_norm": 0.178956981513226,
+ "learning_rate": 3.0715431821832076e-05,
+ "loss": 0.4744793772697449,
+ "step": 866
+ },
+ {
+ "epoch": 0.45487932843651624,
+ "grad_norm": 0.174949760942311,
+ "learning_rate": 3.067872216486425e-05,
+ "loss": 0.48484891653060913,
+ "step": 867
+ },
+ {
+ "epoch": 0.45540398740818466,
+ "grad_norm": 0.18137397131920793,
+ "learning_rate": 3.064200387201987e-05,
+ "loss": 0.4631873369216919,
+ "step": 868
+ },
+ {
+ "epoch": 0.4559286463798531,
+ "grad_norm": 0.1732496277596729,
+ "learning_rate": 3.060527704305435e-05,
+ "loss": 0.465931236743927,
+ "step": 869
+ },
+ {
+ "epoch": 0.4564533053515215,
+ "grad_norm": 0.18674581439596288,
+ "learning_rate": 3.056854177774625e-05,
+ "loss": 0.4731084704399109,
+ "step": 870
+ },
+ {
+ "epoch": 0.45697796432318993,
+ "grad_norm": 0.1705183528405742,
+ "learning_rate": 3.053179817589705e-05,
+ "loss": 0.4759066104888916,
+ "step": 871
+ },
+ {
+ "epoch": 0.45750262329485836,
+ "grad_norm": 0.18333960418663817,
+ "learning_rate": 3.049504633733092e-05,
+ "loss": 0.4720202088356018,
+ "step": 872
+ },
+ {
+ "epoch": 0.4580272822665268,
+ "grad_norm": 0.1721271589969461,
+ "learning_rate": 3.0458286361894357e-05,
+ "loss": 0.4494917690753937,
+ "step": 873
+ },
+ {
+ "epoch": 0.45855194123819515,
+ "grad_norm": 0.179622679302864,
+ "learning_rate": 3.042151834945599e-05,
+ "loss": 0.48819422721862793,
+ "step": 874
+ },
+ {
+ "epoch": 0.4590766002098636,
+ "grad_norm": 0.17079486351495243,
+ "learning_rate": 3.0384742399906273e-05,
+ "loss": 0.4689975678920746,
+ "step": 875
+ },
+ {
+ "epoch": 0.459601259181532,
+ "grad_norm": 0.17997285458346546,
+ "learning_rate": 3.034795861315723e-05,
+ "loss": 0.45560094714164734,
+ "step": 876
+ },
+ {
+ "epoch": 0.4601259181532004,
+ "grad_norm": 0.1726957000314432,
+ "learning_rate": 3.031116708914219e-05,
+ "loss": 0.46888601779937744,
+ "step": 877
+ },
+ {
+ "epoch": 0.46065057712486884,
+ "grad_norm": 0.17681806153988336,
+ "learning_rate": 3.0274367927815468e-05,
+ "loss": 0.46470487117767334,
+ "step": 878
+ },
+ {
+ "epoch": 0.46117523609653727,
+ "grad_norm": 0.1773327553526153,
+ "learning_rate": 3.023756122915215e-05,
+ "loss": 0.47452306747436523,
+ "step": 879
+ },
+ {
+ "epoch": 0.4616998950682057,
+ "grad_norm": 0.17963476535942613,
+ "learning_rate": 3.020074709314781e-05,
+ "loss": 0.4480860233306885,
+ "step": 880
+ },
+ {
+ "epoch": 0.46222455403987406,
+ "grad_norm": 0.17799413072105774,
+ "learning_rate": 3.0163925619818205e-05,
+ "loss": 0.45236068964004517,
+ "step": 881
+ },
+ {
+ "epoch": 0.4627492130115425,
+ "grad_norm": 0.17193695597882394,
+ "learning_rate": 3.0127096909199037e-05,
+ "loss": 0.4458444118499756,
+ "step": 882
+ },
+ {
+ "epoch": 0.4632738719832109,
+ "grad_norm": 0.17467538561637588,
+ "learning_rate": 3.009026106134567e-05,
+ "loss": 0.44784224033355713,
+ "step": 883
+ },
+ {
+ "epoch": 0.46379853095487933,
+ "grad_norm": 0.16749678930092807,
+ "learning_rate": 3.0053418176332844e-05,
+ "loss": 0.45795536041259766,
+ "step": 884
+ },
+ {
+ "epoch": 0.46432318992654775,
+ "grad_norm": 0.17299069140182316,
+ "learning_rate": 3.001656835425445e-05,
+ "loss": 0.44555261731147766,
+ "step": 885
+ },
+ {
+ "epoch": 0.4648478488982162,
+ "grad_norm": 0.17491097835310385,
+ "learning_rate": 2.9979711695223183e-05,
+ "loss": 0.461821049451828,
+ "step": 886
+ },
+ {
+ "epoch": 0.4653725078698846,
+ "grad_norm": 0.17793031337455156,
+ "learning_rate": 2.9942848299370363e-05,
+ "loss": 0.48487529158592224,
+ "step": 887
+ },
+ {
+ "epoch": 0.46589716684155297,
+ "grad_norm": 0.17185716106368243,
+ "learning_rate": 2.9905978266845564e-05,
+ "loss": 0.4482880234718323,
+ "step": 888
+ },
+ {
+ "epoch": 0.4664218258132214,
+ "grad_norm": 0.17592478369985728,
+ "learning_rate": 2.9869101697816416e-05,
+ "loss": 0.46595093607902527,
+ "step": 889
+ },
+ {
+ "epoch": 0.4669464847848898,
+ "grad_norm": 0.1705955811915332,
+ "learning_rate": 2.9832218692468294e-05,
+ "loss": 0.4602831304073334,
+ "step": 890
+ },
+ {
+ "epoch": 0.46747114375655824,
+ "grad_norm": 0.1828910384017392,
+ "learning_rate": 2.9795329351004082e-05,
+ "loss": 0.46365973353385925,
+ "step": 891
+ },
+ {
+ "epoch": 0.46799580272822666,
+ "grad_norm": 0.16932062798080852,
+ "learning_rate": 2.975843377364385e-05,
+ "loss": 0.4667619466781616,
+ "step": 892
+ },
+ {
+ "epoch": 0.4685204616998951,
+ "grad_norm": 0.17835717431399314,
+ "learning_rate": 2.972153206062463e-05,
+ "loss": 0.4726160764694214,
+ "step": 893
+ },
+ {
+ "epoch": 0.4690451206715635,
+ "grad_norm": 0.1584662133261623,
+ "learning_rate": 2.968462431220011e-05,
+ "loss": 0.4468635320663452,
+ "step": 894
+ },
+ {
+ "epoch": 0.4695697796432319,
+ "grad_norm": 0.1759588909434276,
+ "learning_rate": 2.9647710628640384e-05,
+ "loss": 0.48429811000823975,
+ "step": 895
+ },
+ {
+ "epoch": 0.4700944386149003,
+ "grad_norm": 0.1673210075807756,
+ "learning_rate": 2.9610791110231668e-05,
+ "loss": 0.47632431983947754,
+ "step": 896
+ },
+ {
+ "epoch": 0.4706190975865687,
+ "grad_norm": 0.17043437023204477,
+ "learning_rate": 2.9573865857276028e-05,
+ "loss": 0.4587501287460327,
+ "step": 897
+ },
+ {
+ "epoch": 0.47114375655823715,
+ "grad_norm": 0.1646548157692475,
+ "learning_rate": 2.9536934970091108e-05,
+ "loss": 0.4642453193664551,
+ "step": 898
+ },
+ {
+ "epoch": 0.47166841552990557,
+ "grad_norm": 0.15940980831419563,
+ "learning_rate": 2.949999854900985e-05,
+ "loss": 0.4391498863697052,
+ "step": 899
+ },
+ {
+ "epoch": 0.472193074501574,
+ "grad_norm": 0.1684031143485564,
+ "learning_rate": 2.9463056694380265e-05,
+ "loss": 0.46176138520240784,
+ "step": 900
+ },
+ {
+ "epoch": 0.4727177334732424,
+ "grad_norm": 0.16858641414424183,
+ "learning_rate": 2.942610950656509e-05,
+ "loss": 0.45706647634506226,
+ "step": 901
+ },
+ {
+ "epoch": 0.4732423924449108,
+ "grad_norm": 0.16800791620216,
+ "learning_rate": 2.938915708594156e-05,
+ "loss": 0.4599732756614685,
+ "step": 902
+ },
+ {
+ "epoch": 0.4737670514165792,
+ "grad_norm": 0.17121821936362464,
+ "learning_rate": 2.9352199532901143e-05,
+ "loss": 0.4753255844116211,
+ "step": 903
+ },
+ {
+ "epoch": 0.47429171038824763,
+ "grad_norm": 0.16597472801521954,
+ "learning_rate": 2.9315236947849213e-05,
+ "loss": 0.45036259293556213,
+ "step": 904
+ },
+ {
+ "epoch": 0.47481636935991606,
+ "grad_norm": 0.16948740028977227,
+ "learning_rate": 2.9278269431204862e-05,
+ "loss": 0.4506021738052368,
+ "step": 905
+ },
+ {
+ "epoch": 0.4753410283315845,
+ "grad_norm": 0.16623273345325396,
+ "learning_rate": 2.9241297083400554e-05,
+ "loss": 0.4460686445236206,
+ "step": 906
+ },
+ {
+ "epoch": 0.4758656873032529,
+ "grad_norm": 0.16895551138978052,
+ "learning_rate": 2.9204320004881863e-05,
+ "loss": 0.4550820291042328,
+ "step": 907
+ },
+ {
+ "epoch": 0.47639034627492133,
+ "grad_norm": 0.16609221147963488,
+ "learning_rate": 2.916733829610726e-05,
+ "loss": 0.4568508267402649,
+ "step": 908
+ },
+ {
+ "epoch": 0.4769150052465897,
+ "grad_norm": 0.1761776165947237,
+ "learning_rate": 2.913035205754775e-05,
+ "loss": 0.47477489709854126,
+ "step": 909
+ },
+ {
+ "epoch": 0.4774396642182581,
+ "grad_norm": 0.1720771870966164,
+ "learning_rate": 2.9093361389686674e-05,
+ "loss": 0.45254939794540405,
+ "step": 910
+ },
+ {
+ "epoch": 0.47796432318992654,
+ "grad_norm": 0.16803097453010016,
+ "learning_rate": 2.905636639301939e-05,
+ "loss": 0.4596205949783325,
+ "step": 911
+ },
+ {
+ "epoch": 0.47848898216159497,
+ "grad_norm": 0.17187206195790292,
+ "learning_rate": 2.901936716805302e-05,
+ "loss": 0.4835141897201538,
+ "step": 912
+ },
+ {
+ "epoch": 0.4790136411332634,
+ "grad_norm": 0.17138230203736526,
+ "learning_rate": 2.898236381530618e-05,
+ "loss": 0.4552057981491089,
+ "step": 913
+ },
+ {
+ "epoch": 0.4795383001049318,
+ "grad_norm": 0.16684081343053822,
+ "learning_rate": 2.8945356435308696e-05,
+ "loss": 0.45461946725845337,
+ "step": 914
+ },
+ {
+ "epoch": 0.48006295907660024,
+ "grad_norm": 0.17268002595144114,
+ "learning_rate": 2.890834512860134e-05,
+ "loss": 0.47163042426109314,
+ "step": 915
+ },
+ {
+ "epoch": 0.4805876180482686,
+ "grad_norm": 0.16874650464609184,
+ "learning_rate": 2.887132999573554e-05,
+ "loss": 0.4635821580886841,
+ "step": 916
+ },
+ {
+ "epoch": 0.48111227701993703,
+ "grad_norm": 0.16861058639601054,
+ "learning_rate": 2.8834311137273117e-05,
+ "loss": 0.4604582190513611,
+ "step": 917
+ },
+ {
+ "epoch": 0.48163693599160545,
+ "grad_norm": 0.1666087546902995,
+ "learning_rate": 2.8797288653786043e-05,
+ "loss": 0.4603103995323181,
+ "step": 918
+ },
+ {
+ "epoch": 0.4821615949632739,
+ "grad_norm": 0.17374719993517845,
+ "learning_rate": 2.8760262645856096e-05,
+ "loss": 0.46453070640563965,
+ "step": 919
+ },
+ {
+ "epoch": 0.4826862539349423,
+ "grad_norm": 0.17116073729656917,
+ "learning_rate": 2.8723233214074675e-05,
+ "loss": 0.46432310342788696,
+ "step": 920
+ },
+ {
+ "epoch": 0.4832109129066107,
+ "grad_norm": 0.1618534380792579,
+ "learning_rate": 2.8686200459042438e-05,
+ "loss": 0.44625040888786316,
+ "step": 921
+ },
+ {
+ "epoch": 0.4837355718782791,
+ "grad_norm": 0.18484190074651213,
+ "learning_rate": 2.86491644813691e-05,
+ "loss": 0.46191471815109253,
+ "step": 922
+ },
+ {
+ "epoch": 0.4842602308499475,
+ "grad_norm": 0.16324184235136427,
+ "learning_rate": 2.8612125381673112e-05,
+ "loss": 0.4640767276287079,
+ "step": 923
+ },
+ {
+ "epoch": 0.48478488982161594,
+ "grad_norm": 0.18269463317294554,
+ "learning_rate": 2.857508326058142e-05,
+ "loss": 0.4676060080528259,
+ "step": 924
+ },
+ {
+ "epoch": 0.48530954879328436,
+ "grad_norm": 0.1616079538284595,
+ "learning_rate": 2.8538038218729187e-05,
+ "loss": 0.4431381821632385,
+ "step": 925
+ },
+ {
+ "epoch": 0.4858342077649528,
+ "grad_norm": 0.1746601662445674,
+ "learning_rate": 2.8500990356759476e-05,
+ "loss": 0.4572705328464508,
+ "step": 926
+ },
+ {
+ "epoch": 0.4863588667366212,
+ "grad_norm": 0.1745623763362984,
+ "learning_rate": 2.8463939775323057e-05,
+ "loss": 0.4578236937522888,
+ "step": 927
+ },
+ {
+ "epoch": 0.48688352570828963,
+ "grad_norm": 0.17181307571126248,
+ "learning_rate": 2.842688657507805e-05,
+ "loss": 0.4509286880493164,
+ "step": 928
+ },
+ {
+ "epoch": 0.487408184679958,
+ "grad_norm": 0.17725373882628745,
+ "learning_rate": 2.8389830856689715e-05,
+ "loss": 0.46443605422973633,
+ "step": 929
+ },
+ {
+ "epoch": 0.4879328436516264,
+ "grad_norm": 0.17093469695389601,
+ "learning_rate": 2.8352772720830138e-05,
+ "loss": 0.45976129174232483,
+ "step": 930
+ },
+ {
+ "epoch": 0.48845750262329485,
+ "grad_norm": 0.1735988134522573,
+ "learning_rate": 2.8315712268177987e-05,
+ "loss": 0.45718681812286377,
+ "step": 931
+ },
+ {
+ "epoch": 0.48898216159496327,
+ "grad_norm": 0.17145609849005775,
+ "learning_rate": 2.8278649599418194e-05,
+ "loss": 0.47013211250305176,
+ "step": 932
+ },
+ {
+ "epoch": 0.4895068205666317,
+ "grad_norm": 0.18295274946907686,
+ "learning_rate": 2.8241584815241755e-05,
+ "loss": 0.4669569730758667,
+ "step": 933
+ },
+ {
+ "epoch": 0.4900314795383001,
+ "grad_norm": 0.1674773049076024,
+ "learning_rate": 2.8204518016345388e-05,
+ "loss": 0.4614064693450928,
+ "step": 934
+ },
+ {
+ "epoch": 0.49055613850996854,
+ "grad_norm": 0.17064185980244923,
+ "learning_rate": 2.8167449303431276e-05,
+ "loss": 0.46194010972976685,
+ "step": 935
+ },
+ {
+ "epoch": 0.4910807974816369,
+ "grad_norm": 0.15983150053137107,
+ "learning_rate": 2.8130378777206828e-05,
+ "loss": 0.4550577402114868,
+ "step": 936
+ },
+ {
+ "epoch": 0.49160545645330533,
+ "grad_norm": 0.17703751111643015,
+ "learning_rate": 2.8093306538384357e-05,
+ "loss": 0.4712820053100586,
+ "step": 937
+ },
+ {
+ "epoch": 0.49213011542497376,
+ "grad_norm": 0.17388586598408573,
+ "learning_rate": 2.8056232687680834e-05,
+ "loss": 0.46234333515167236,
+ "step": 938
+ },
+ {
+ "epoch": 0.4926547743966422,
+ "grad_norm": 0.17188541123455303,
+ "learning_rate": 2.8019157325817623e-05,
+ "loss": 0.4513767957687378,
+ "step": 939
+ },
+ {
+ "epoch": 0.4931794333683106,
+ "grad_norm": 0.1793942730427001,
+ "learning_rate": 2.7982080553520167e-05,
+ "loss": 0.4611365497112274,
+ "step": 940
+ },
+ {
+ "epoch": 0.49370409233997903,
+ "grad_norm": 0.16890499968477882,
+ "learning_rate": 2.7945002471517773e-05,
+ "loss": 0.4438607394695282,
+ "step": 941
+ },
+ {
+ "epoch": 0.49422875131164745,
+ "grad_norm": 0.1645393636692259,
+ "learning_rate": 2.7907923180543277e-05,
+ "loss": 0.4503021836280823,
+ "step": 942
+ },
+ {
+ "epoch": 0.4947534102833158,
+ "grad_norm": 0.1685242276058161,
+ "learning_rate": 2.787084278133283e-05,
+ "loss": 0.46390002965927124,
+ "step": 943
+ },
+ {
+ "epoch": 0.49527806925498424,
+ "grad_norm": 0.16427997039461814,
+ "learning_rate": 2.7833761374625543e-05,
+ "loss": 0.4478822350502014,
+ "step": 944
+ },
+ {
+ "epoch": 0.49580272822665267,
+ "grad_norm": 0.17005794828810356,
+ "learning_rate": 2.7796679061163335e-05,
+ "loss": 0.45072460174560547,
+ "step": 945
+ },
+ {
+ "epoch": 0.4963273871983211,
+ "grad_norm": 0.16429463559042642,
+ "learning_rate": 2.7759595941690526e-05,
+ "loss": 0.46297261118888855,
+ "step": 946
+ },
+ {
+ "epoch": 0.4968520461699895,
+ "grad_norm": 0.1686355306041516,
+ "learning_rate": 2.772251211695365e-05,
+ "loss": 0.48136377334594727,
+ "step": 947
+ },
+ {
+ "epoch": 0.49737670514165794,
+ "grad_norm": 0.1584642448830141,
+ "learning_rate": 2.7685427687701172e-05,
+ "loss": 0.4592211842536926,
+ "step": 948
+ },
+ {
+ "epoch": 0.49790136411332636,
+ "grad_norm": 0.16302353022861613,
+ "learning_rate": 2.764834275468317e-05,
+ "loss": 0.4561595618724823,
+ "step": 949
+ },
+ {
+ "epoch": 0.49842602308499473,
+ "grad_norm": 0.1658881804860426,
+ "learning_rate": 2.7611257418651116e-05,
+ "loss": 0.4650879502296448,
+ "step": 950
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 1906,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 1,
+ "save_steps": 50,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 2.5145154180481024e+16,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}