grail_v2op_traj_b64_sft_e3_ck640

.gitattributes

@@ -85,3 +85,4 @@ grail_v2op_fcurlong_b64_sft_e3_ck640/tokenizer.json filter=lfs diff=lfs merge=lf
 grail_v2op_fcur_frombase_b64_sft_e0_ck640/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 grail_v2op_fcur_lora64_b64_sft_e3_merged/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 grail_v1_traj_b64_sft_e3_ck300/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+grail_v2op_traj_b64_sft_e3_ck640/tokenizer.json filter=lfs diff=lfs merge=lfs -text

grail_v2op_traj_b64_sft_e3_ck640/all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 1.2167300380228137,
+    "eval_loss": 0.02630898542702198,
+    "eval_runtime": 126.8539,
+    "eval_samples_per_second": 13.969,
+    "eval_steps_per_second": 3.492,
+    "total_flos": 263393572028416.0,
+    "train_loss": 0.08146042495500297,
+    "train_runtime": 7467.1868,
+    "train_samples_per_second": 5.485,
+    "train_steps_per_second": 0.086
+}

grail_v2op_traj_b64_sft_e3_ck640/chat_template.jinja

@@ -0,0 +1,120 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {%- if messages[0].content is string %}
+            {{- messages[0].content }}
+        {%- else %}
+            {%- for content in messages[0].content %}
+                {%- if 'text' in content %}
+                    {{- content.text }}
+                {%- endif %}
+            {%- endfor %}
+        {%- endif %}
+        {{- '\n\n' }}
+    {%- endif %}
+    {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {{- '<|im_start|>system\n' }}
+        {%- if messages[0].content is string %}
+            {{- messages[0].content }}
+        {%- else %}
+            {%- for content in messages[0].content %}
+                {%- if 'text' in content %}
+                    {{- content.text }}
+                {%- endif %}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- for message in messages %}
+    {%- if message.role == "user" %}
+        {{- '<|im_start|>' + message.role + '\n' }}
+        {%- if message.content is string %}
+            {{- message.content }}
+        {%- else %}
+            {%- for content in message.content %}
+                {%- if content.type == 'image' or 'image' in content or 'image_url' in content %}
+                    {%- set image_count.value = image_count.value + 1 %}
+                    {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%}
+                    <|vision_start|><|image_pad|><|vision_end|>
+                {%- elif content.type == 'video' or 'video' in content %}
+                    {%- set video_count.value = video_count.value + 1 %}
+                    {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%}
+                    <|vision_start|><|video_pad|><|vision_end|>
+                {%- elif 'text' in content %}
+                    {{- content.text }}
+                {%- endif %}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role + '\n' }}
+        {%- if message.content is string %}
+            {{- message.content }}
+        {%- else %}
+            {%- for content_item in message.content %}
+                {%- if 'text' in content_item %}
+                    {{- content_item.text }}
+                {%- endif %}
+            {%- endfor %}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and message.content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {%- if message.content is string %}
+            {{- message.content }}
+        {%- else %}
+            {%- for content in message.content %}
+                {%- if content.type == 'image' or 'image' in content or 'image_url' in content %}
+                    {%- set image_count.value = image_count.value + 1 %}
+                    {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%}
+                    <|vision_start|><|image_pad|><|vision_end|>
+                {%- elif content.type == 'video' or 'video' in content %}
+                    {%- set video_count.value = video_count.value + 1 %}
+                    {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%}
+                    <|vision_start|><|video_pad|><|vision_end|>
+                {%- elif 'text' in content %}
+                    {{- content.text }}
+                {%- endif %}
+            {%- endfor %}
+        {%- endif %}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}

grail_v2op_traj_b64_sft_e3_ck640/config.json

@@ -0,0 +1,70 @@
+{
+  "architectures": [
+    "Qwen3VLForConditionalGeneration"
+  ],
+  "bos_token_id": null,
+  "dtype": "bfloat16",
+  "eos_token_id": 151645,
+  "hidden_size": 4096,
+  "image_token_id": 151655,
+  "model_type": "qwen3_vl",
+  "pad_token_id": 151643,
+  "text_config": {
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "bos_token_id": 151643,
+    "dtype": "bfloat16",
+    "eos_token_id": 151645,
+    "head_dim": 128,
+    "hidden_act": "silu",
+    "hidden_size": 4096,
+    "initializer_range": 0.02,
+    "intermediate_size": 12288,
+    "max_position_embeddings": 262144,
+    "model_type": "qwen3_vl_text",
+    "num_attention_heads": 32,
+    "num_hidden_layers": 36,
+    "num_key_value_heads": 8,
+    "pad_token_id": null,
+    "rms_norm_eps": 1e-06,
+    "rope_parameters": {
+      "mrope_interleaved": true,
+      "mrope_section": [
+        24,
+        20,
+        20
+      ],
+      "rope_theta": 5000000,
+      "rope_type": "default"
+    },
+    "use_cache": false,
+    "vocab_size": 151936
+  },
+  "tie_word_embeddings": false,
+  "transformers_version": "5.8.0",
+  "use_cache": false,
+  "video_token_id": 151656,
+  "vision_config": {
+    "deepstack_visual_indexes": [
+      8,
+      16,
+      24
+    ],
+    "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_vl_vision",
+    "num_heads": 16,
+    "num_position_embeddings": 2304,
+    "out_hidden_size": 4096,
+    "patch_size": 16,
+    "spatial_merge_size": 2,
+    "temporal_patch_size": 2
+  },
+  "vision_end_token_id": 151653,
+  "vision_start_token_id": 151652
+}

grail_v2op_traj_b64_sft_e3_ck640/eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 1.2167300380228137,
+    "eval_loss": 0.02630898542702198,
+    "eval_runtime": 126.8539,
+    "eval_samples_per_second": 13.969,
+    "eval_steps_per_second": 3.492
+}

grail_v2op_traj_b64_sft_e3_ck640/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.0,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "5.8.0"
+}

grail_v2op_traj_b64_sft_e3_ck640/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb6a23694039e531a5713377c2bca831efe61802c85698840fac8e964df15719
+size 17534340584

grail_v2op_traj_b64_sft_e3_ck640/processor_config.json

@@ -0,0 +1,60 @@
+{
+  "image_processor": {
+    "do_convert_rgb": true,
+    "do_normalize": true,
+    "do_rescale": true,
+    "do_resize": true,
+    "image_mean": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "image_processor_type": "Qwen2VLImageProcessor",
+    "image_std": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "merge_size": 2,
+    "patch_size": 16,
+    "resample": 3,
+    "rescale_factor": 0.00392156862745098,
+    "size": {
+      "longest_edge": 16777216,
+      "shortest_edge": 65536
+    },
+    "temporal_patch_size": 2
+  },
+  "processor_class": "Qwen3VLProcessor",
+  "video_processor": {
+    "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,
+    "return_metadata": false,
+    "size": {
+      "longest_edge": 25165824,
+      "shortest_edge": 4096
+    },
+    "temporal_patch_size": 2,
+    "video_processor_type": "Qwen3VLVideoProcessor"
+  }
+}

grail_v2op_traj_b64_sft_e3_ck640/sft.yaml

@@ -0,0 +1,58 @@
+# LLaMA-Factory full-parameter SFT of Qwen3-VL-8B-Instruct on one HER-final dataset.
+# Four L40S (46 GB), one run at a time: DeepSpeed ZeRO-3 with optimizer offload to host
+# RAM (~130 GB per run, so two runs do not fit in 372 GB); the vision tower and projector
+# stay frozen (LLaMA-Factory's default for VLMs), the LLM trains.
+# Filled per run by launch_sft.sh: grail_v2op_traj_sft_e3, exps/sft/grail_v2op_traj_sft_e3_k1536, /scratch/kwc5354/viewagent/viewagent_private/exps/sft_runs/grail_v2op_traj_b64_sft_e3.
+### model
+model_name_or_path: /scratch/kwc5354/viewagent/viewagent_private/exps/models/graph_nemi_sft_e2
+template: qwen3_vl_nothink
+trust_remote_code: true
+image_max_pixels: 65536          # frames are 256x256 already; never upscale
+image_min_pixels: 4096
+
+### method
+stage: sft
+do_train: true
+finetuning_type: full
+freeze_vision_tower: true
+freeze_multi_modal_projector: true
+freeze_language_model: false
+deepspeed: /scratch/kwc5354/viewagent/viewagent_private/LLaMA-Factory/examples/deepspeed/ds_z3_config.json
+
+### dataset
+dataset: grail_v2op_traj_sft_e3
+dataset_dir: exps/sft/grail_v2op_traj_sft_e3_k1536
+media_dir: /scratch/kwc5354/viewagent/viewagent_private
+cutoff_len: 8192
+overwrite_cache: true
+preprocessing_num_workers: 16
+dataloader_num_workers: 4
+
+### output
+output_dir: /scratch/kwc5354/viewagent/viewagent_private/exps/sft_runs/grail_v2op_traj_b64_sft_e3
+logging_steps: 10
+save_strategy: steps
+save_steps: 640
+save_total_limit: 100
+save_only_model: true
+plot_loss: true
+overwrite_output_dir: true
+report_to: wandb
+
+### train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 16
+max_steps: 640
+learning_rate: 1.0e-5
+lr_scheduler_type: cosine
+warmup_ratio: 0.03
+weight_decay: 0.01
+bf16: true
+gradient_checkpointing: true
+ddp_timeout: 180000000
+seed: 0
+val_size: 0.05
+eval_strategy: steps
+eval_steps: 640
+per_device_eval_batch_size: 1
+run_name: grail_v2op_traj_b64_sft_e3

grail_v2op_traj_b64_sft_e3_ck640/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
+size 11422650

grail_v2op_traj_b64_sft_e3_ck640/tokenizer_config.json

@@ -0,0 +1,32 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": true,
+  "local_files_only": false,
+  "model_max_length": 262144,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "processor_class": "Qwen3VLProcessor",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

grail_v2op_traj_b64_sft_e3_ck640/train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 1.2167300380228137,
+    "total_flos": 263393572028416.0,
+    "train_loss": 0.08146042495500297,
+    "train_runtime": 7467.1868,
+    "train_samples_per_second": 5.485,
+    "train_steps_per_second": 0.086
+}

grail_v2op_traj_b64_sft_e3_ck640/trainer_state.json

@@ -0,0 +1,499 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.2167300380228137,
+  "eval_steps": 640,
+  "global_step": 640,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.019011406844106463,
+      "grad_norm": 6.833532627858379,
+      "learning_rate": 4.5e-06,
+      "loss": 1.1798740386962892,
+      "step": 10
+    },
+    {
+      "epoch": 0.03802281368821293,
+      "grad_norm": 0.8675944227650964,
+      "learning_rate": 9.5e-06,
+      "loss": 0.36087474822998045,
+      "step": 20
+    },
+    {
+      "epoch": 0.057034220532319393,
+      "grad_norm": 0.7583905914291712,
+      "learning_rate": 9.994801642711945e-06,
+      "loss": 0.19843592643737792,
+      "step": 30
+    },
+    {
+      "epoch": 0.07604562737642585,
+      "grad_norm": 0.6964028245001298,
+      "learning_rate": 9.976845889479778e-06,
+      "loss": 0.17371606826782227,
+      "step": 40
+    },
+    {
+      "epoch": 0.09505703422053231,
+      "grad_norm": 0.5235682627317625,
+      "learning_rate": 9.946114644227765e-06,
+      "loss": 0.15696249008178711,
+      "step": 50
+    },
+    {
+      "epoch": 0.11406844106463879,
+      "grad_norm": 0.6361116135640386,
+      "learning_rate": 9.90268679362172e-06,
+      "loss": 0.14409921169281006,
+      "step": 60
+    },
+    {
+      "epoch": 0.13307984790874525,
+      "grad_norm": 0.6338464377511862,
+      "learning_rate": 9.846673816331297e-06,
+      "loss": 0.14039350748062135,
+      "step": 70
+    },
+    {
+      "epoch": 0.1520912547528517,
+      "grad_norm": 0.598179796278668,
+      "learning_rate": 9.778219496865864e-06,
+      "loss": 0.123904287815094,
+      "step": 80
+    },
+    {
+      "epoch": 0.17110266159695817,
+      "grad_norm": 0.5788919647747465,
+      "learning_rate": 9.6974995564817e-06,
+      "loss": 0.12066190242767334,
+      "step": 90
+    },
+    {
+      "epoch": 0.19011406844106463,
+      "grad_norm": 0.6779205740174066,
+      "learning_rate": 9.604721202108084e-06,
+      "loss": 0.1133304238319397,
+      "step": 100
+    },
+    {
+      "epoch": 0.20912547528517111,
+      "grad_norm": 0.6436771238979078,
+      "learning_rate": 9.500122594450056e-06,
+      "loss": 0.10433969497680665,
+      "step": 110
+    },
+    {
+      "epoch": 0.22813688212927757,
+      "grad_norm": 0.7915766239312892,
+      "learning_rate": 9.383972236633308e-06,
+      "loss": 0.10093061923980713,
+      "step": 120
+    },
+    {
+      "epoch": 0.24714828897338403,
+      "grad_norm": 0.5862935298482929,
+      "learning_rate": 9.256568284960513e-06,
+      "loss": 0.09776815176010131,
+      "step": 130
+    },
+    {
+      "epoch": 0.2661596958174905,
+      "grad_norm": 0.6461671702588686,
+      "learning_rate": 9.118237783548376e-06,
+      "loss": 0.09077458381652832,
+      "step": 140
+    },
+    {
+      "epoch": 0.28517110266159695,
+      "grad_norm": 0.7633650662685545,
+      "learning_rate": 8.969335824810075e-06,
+      "loss": 0.08747783899307252,
+      "step": 150
+    },
+    {
+      "epoch": 0.3041825095057034,
+      "grad_norm": 0.6223130214887873,
+      "learning_rate": 8.810244637938117e-06,
+      "loss": 0.08635616302490234,
+      "step": 160
+    },
+    {
+      "epoch": 0.3231939163498099,
+      "grad_norm": 0.59046762476648,
+      "learning_rate": 8.641372607727503e-06,
+      "loss": 0.08176609873771667,
+      "step": 170
+    },
+    {
+      "epoch": 0.34220532319391633,
+      "grad_norm": 0.6292093448637252,
+      "learning_rate": 8.463153226257805e-06,
+      "loss": 0.0785952091217041,
+      "step": 180
+    },
+    {
+      "epoch": 0.3612167300380228,
+      "grad_norm": 0.7709470523083949,
+      "learning_rate": 8.276043980125213e-06,
+      "loss": 0.07542359828948975,
+      "step": 190
+    },
+    {
+      "epoch": 0.38022813688212925,
+      "grad_norm": 0.676026518054413,
+      "learning_rate": 8.080525176081006e-06,
+      "loss": 0.06985285282135009,
+      "step": 200
+    },
+    {
+      "epoch": 0.39923954372623577,
+      "grad_norm": 0.6709319883269886,
+      "learning_rate": 7.877098708091017e-06,
+      "loss": 0.06881477832794189,
+      "step": 210
+    },
+    {
+      "epoch": 0.41825095057034223,
+      "grad_norm": 0.689634666914331,
+      "learning_rate": 7.666286768981012e-06,
+      "loss": 0.0663729727268219,
+      "step": 220
+    },
+    {
+      "epoch": 0.4372623574144487,
+      "grad_norm": 0.6872124123166982,
+      "learning_rate": 7.448630509975199e-06,
+      "loss": 0.0639960527420044,
+      "step": 230
+    },
+    {
+      "epoch": 0.45627376425855515,
+      "grad_norm": 0.635100494186257,
+      "learning_rate": 7.224688651568797e-06,
+      "loss": 0.060294950008392335,
+      "step": 240
+    },
+    {
+      "epoch": 0.4752851711026616,
+      "grad_norm": 0.6089469882541673,
+      "learning_rate": 6.995036049300539e-06,
+      "loss": 0.059751582145690915,
+      "step": 250
+    },
+    {
+      "epoch": 0.49429657794676807,
+      "grad_norm": 0.6479737152230922,
+      "learning_rate": 6.7602622181067425e-06,
+      "loss": 0.05485899448394775,
+      "step": 260
+    },
+    {
+      "epoch": 0.5133079847908745,
+      "grad_norm": 0.6833861102040915,
+      "learning_rate": 6.520969819044933e-06,
+      "loss": 0.055822563171386716,
+      "step": 270
+    },
+    {
+      "epoch": 0.532319391634981,
+      "grad_norm": 0.6023016958618561,
+      "learning_rate": 6.277773112271545e-06,
+      "loss": 0.052158784866333005,
+      "step": 280
+    },
+    {
+      "epoch": 0.5513307984790875,
+      "grad_norm": 0.6371911068922275,
+      "learning_rate": 6.031296380244915e-06,
+      "loss": 0.04927125871181488,
+      "step": 290
+    },
+    {
+      "epoch": 0.5703422053231939,
+      "grad_norm": 0.6784439648867187,
+      "learning_rate": 5.782172325201155e-06,
+      "loss": 0.04956686794757843,
+      "step": 300
+    },
+    {
+      "epoch": 0.5893536121673004,
+      "grad_norm": 0.7099300377069291,
+      "learning_rate": 5.531040445016604e-06,
+      "loss": 0.04914776384830475,
+      "step": 310
+    },
+    {
+      "epoch": 0.6083650190114068,
+      "grad_norm": 0.8253831804698116,
+      "learning_rate": 5.278545391625931e-06,
+      "loss": 0.047037053108215335,
+      "step": 320
+    },
+    {
+      "epoch": 0.6273764258555133,
+      "grad_norm": 0.6492888915423779,
+      "learning_rate": 5.025335316209883e-06,
+      "loss": 0.04610362648963928,
+      "step": 330
+    },
+    {
+      "epoch": 0.6463878326996197,
+      "grad_norm": 0.623969074455222,
+      "learning_rate": 4.7720602054004865e-06,
+      "loss": 0.04243798553943634,
+      "step": 340
+    },
+    {
+      "epoch": 0.6653992395437263,
+      "grad_norm": 0.6769755602417163,
+      "learning_rate": 4.51937021277469e-06,
+      "loss": 0.04274915456771851,
+      "step": 350
+    },
+    {
+      "epoch": 0.6844106463878327,
+      "grad_norm": 0.638653090576167,
+      "learning_rate": 4.267913989919457e-06,
+      "loss": 0.044309118390083314,
+      "step": 360
+    },
+    {
+      "epoch": 0.7034220532319392,
+      "grad_norm": 0.6424778479790612,
+      "learning_rate": 4.018337021352447e-06,
+      "loss": 0.034453320503234866,
+      "step": 370
+    },
+    {
+      "epoch": 0.7224334600760456,
+      "grad_norm": 0.7765984940970069,
+      "learning_rate": 3.7712799675725105e-06,
+      "loss": 0.04009739756584167,
+      "step": 380
+    },
+    {
+      "epoch": 0.7414448669201521,
+      "grad_norm": 0.590225448366126,
+      "learning_rate": 3.5273770204933814e-06,
+      "loss": 0.04006598591804504,
+      "step": 390
+    },
+    {
+      "epoch": 0.7604562737642585,
+      "grad_norm": 0.5890444076012925,
+      "learning_rate": 3.2872542754821334e-06,
+      "loss": 0.0395596444606781,
+      "step": 400
+    },
+    {
+      "epoch": 0.779467680608365,
+      "grad_norm": 0.5101376718808112,
+      "learning_rate": 3.051528124181342e-06,
+      "loss": 0.03626045882701874,
+      "step": 410
+    },
+    {
+      "epoch": 0.7984790874524715,
+      "grad_norm": 0.7548574110894845,
+      "learning_rate": 2.8208036722405895e-06,
+      "loss": 0.03296552896499634,
+      "step": 420
+    },
+    {
+      "epoch": 0.8174904942965779,
+      "grad_norm": 0.5399807177143271,
+      "learning_rate": 2.595673186018929e-06,
+      "loss": 0.03616112172603607,
+      "step": 430
+    },
+    {
+      "epoch": 0.8365019011406845,
+      "grad_norm": 0.49704464549734645,
+      "learning_rate": 2.376714572245651e-06,
+      "loss": 0.037315943837165834,
+      "step": 440
+    },
+    {
+      "epoch": 0.8555133079847909,
+      "grad_norm": 1.0906599758285942,
+      "learning_rate": 2.1644898945419814e-06,
+      "loss": 0.032875528931617735,
+      "step": 450
+    },
+    {
+      "epoch": 0.8745247148288974,
+      "grad_norm": 0.6319206502516904,
+      "learning_rate": 1.959543930611835e-06,
+      "loss": 0.030941855907440186,
+      "step": 460
+    },
+    {
+      "epoch": 0.8935361216730038,
+      "grad_norm": 0.6314112516264699,
+      "learning_rate": 1.7624027738052669e-06,
+      "loss": 0.030404505133628846,
+      "step": 470
+    },
+    {
+      "epoch": 0.9125475285171103,
+      "grad_norm": 0.6027593855195658,
+      "learning_rate": 1.5735724826443788e-06,
+      "loss": 0.029952824115753174,
+      "step": 480
+    },
+    {
+      "epoch": 0.9315589353612167,
+      "grad_norm": 0.5307440008863118,
+      "learning_rate": 1.3935377817783585e-06,
+      "loss": 0.029353293776512145,
+      "step": 490
+    },
+    {
+      "epoch": 0.9505703422053232,
+      "grad_norm": 0.4499056558638885,
+      "learning_rate": 1.2227608177022522e-06,
+      "loss": 0.028975510597229005,
+      "step": 500
+    },
+    {
+      "epoch": 0.9695817490494296,
+      "grad_norm": 0.6906605138662051,
+      "learning_rate": 1.0616799724335387e-06,
+      "loss": 0.030493971705436707,
+      "step": 510
+    },
+    {
+      "epoch": 0.9885931558935361,
+      "grad_norm": 0.7091902771093809,
+      "learning_rate": 9.107087381917878e-07,
+      "loss": 0.029538416862487794,
+      "step": 520
+    },
+    {
+      "epoch": 1.0076045627376427,
+      "grad_norm": 0.36437080328785,
+      "learning_rate": 7.702346559700502e-07,
+      "loss": 0.023704057931900023,
+      "step": 530
+    },
+    {
+      "epoch": 1.026615969581749,
+      "grad_norm": 0.7128556834288021,
+      "learning_rate": 6.406183207226973e-07,
+      "loss": 0.019738171994686127,
+      "step": 540
+    },
+    {
+      "epoch": 1.0456273764258555,
+      "grad_norm": 0.48032089571607406,
+      "learning_rate": 5.22192455723336e-07,
+      "loss": 0.01951468288898468,
+      "step": 550
+    },
+    {
+      "epoch": 1.064638783269962,
+      "grad_norm": 0.5000082229759188,
+      "learning_rate": 4.152610584689443e-07,
+      "loss": 0.021027599275112153,
+      "step": 560
+    },
+    {
+      "epoch": 1.0836501901140685,
+      "grad_norm": 0.6809401395526675,
+      "learning_rate": 3.200986203226536e-07,
+      "loss": 0.018143051862716676,
+      "step": 570
+    },
+    {
+      "epoch": 1.102661596958175,
+      "grad_norm": 0.5654617627527649,
+      "learning_rate": 2.3694942189835423e-07,
+      "loss": 0.018912801146507265,
+      "step": 580
+    },
+    {
+      "epoch": 1.1216730038022813,
+      "grad_norm": 0.5618698717171875,
+      "learning_rate": 1.6602690599586747e-07,
+      "loss": 0.01971060335636139,
+      "step": 590
+    },
+    {
+      "epoch": 1.1406844106463878,
+      "grad_norm": 0.5999058275608201,
+      "learning_rate": 1.0751312969633665e-07,
+      "loss": 0.017468313872814178,
+      "step": 600
+    },
+    {
+      "epoch": 1.1596958174904943,
+      "grad_norm": 0.3849724194613528,
+      "learning_rate": 6.15582970243117e-08,
+      "loss": 0.018527662754058837,
+      "step": 610
+    },
+    {
+      "epoch": 1.1787072243346008,
+      "grad_norm": 0.636445498431748,
+      "learning_rate": 2.828037337617273e-08,
+      "loss": 0.0197751522064209,
+      "step": 620
+    },
+    {
+      "epoch": 1.1977186311787071,
+      "grad_norm": 0.5907779687253281,
+      "learning_rate": 7.764782704649532e-09,
+      "loss": 0.01971716284751892,
+      "step": 630
+    },
+    {
+      "epoch": 1.2167300380228137,
+      "grad_norm": 0.6453465285927367,
+      "learning_rate": 6.418823675857777e-11,
+      "loss": 0.019581705331802368,
+      "step": 640
+    },
+    {
+      "epoch": 1.2167300380228137,
+      "eval_loss": 0.02630898542702198,
+      "eval_runtime": 127.4298,
+      "eval_samples_per_second": 13.906,
+      "eval_steps_per_second": 3.476,
+      "step": 640
+    },
+    {
+      "epoch": 1.2167300380228137,
+      "step": 640,
+      "total_flos": 263393572028416.0,
+      "train_loss": 0.08146042495500297,
+      "train_runtime": 7467.1868,
+      "train_samples_per_second": 5.485,
+      "train_steps_per_second": 0.086
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 640,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 640,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 263393572028416.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}