Upload continuous evaluation checkpoints, batch 7

.gitattributes

@@ -68,3 +68,7 @@ continuous/qwen25/mc2_b32_384/finetune/tokenizer.json filter=lfs diff=lfs merge=
 continuous/qwen25/mc2_g14_224/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 continuous/qwen25/mc2_g14_378/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 continuous/qwen25/mc2_h14_378/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+continuous/qwen25/mc2_l14_224/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+continuous/qwen25/mc2_m16_224/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+continuous/qwen25/mc2_m16_224_mt5/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+continuous/qwen25/mc2_m16_384/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text

continuous/qwen25/mc2_l14_224/finetune/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:870d61c9babb0227f192947bf1c51503965d312d8ca36b859b914fdb28f50560
+size 4177931704

continuous/qwen25/mc2_l14_224/finetune/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_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|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

continuous/qwen25/mc2_l14_224/finetune/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

continuous/qwen25/mc2_l14_224/finetune/tokenizer_config.json

@@ -0,0 +1,209 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_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|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# 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>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\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\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 8192,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

continuous/qwen25/mc2_l14_224/finetune/trainer_state.json

@@ -0,0 +1,1856 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999037443449803,
+  "eval_steps": 500,
+  "global_step": 2597,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003850226200789296,
+      "grad_norm": 15.114603996276855,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 1.9052,
+      "step": 10
+    },
+    {
+      "epoch": 0.007700452401578592,
+      "grad_norm": 5.379757404327393,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 1.4923,
+      "step": 20
+    },
+    {
+      "epoch": 0.01155067860236789,
+      "grad_norm": 3.456979990005493,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.0956,
+      "step": 30
+    },
+    {
+      "epoch": 0.015400904803157185,
+      "grad_norm": 0.6332670450210571,
+      "learning_rate": 5e-06,
+      "loss": 0.9267,
+      "step": 40
+    },
+    {
+      "epoch": 0.019251131003946483,
+      "grad_norm": 0.3476366102695465,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 0.8956,
+      "step": 50
+    },
+    {
+      "epoch": 0.02310135720473578,
+      "grad_norm": 0.265860378742218,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 0.8475,
+      "step": 60
+    },
+    {
+      "epoch": 0.026951583405525073,
+      "grad_norm": 0.22542931139469147,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.8215,
+      "step": 70
+    },
+    {
+      "epoch": 0.03080180960631437,
+      "grad_norm": 0.22437433898448944,
+      "learning_rate": 9.999996111488723e-06,
+      "loss": 0.8293,
+      "step": 80
+    },
+    {
+      "epoch": 0.034652035807103666,
+      "grad_norm": 0.2079779952764511,
+      "learning_rate": 9.999529497453782e-06,
+      "loss": 0.8111,
+      "step": 90
+    },
+    {
+      "epoch": 0.038502262007892966,
+      "grad_norm": 0.22076578438282013,
+      "learning_rate": 9.998285264324193e-06,
+      "loss": 0.7898,
+      "step": 100
+    },
+    {
+      "epoch": 0.04235248820868226,
+      "grad_norm": 0.19754542410373688,
+      "learning_rate": 9.996263605626052e-06,
+      "loss": 0.7818,
+      "step": 110
+    },
+    {
+      "epoch": 0.04620271440947156,
+      "grad_norm": 0.2224540412425995,
+      "learning_rate": 9.99346483580503e-06,
+      "loss": 0.7941,
+      "step": 120
+    },
+    {
+      "epoch": 0.05005294061026085,
+      "grad_norm": 0.19936759769916534,
+      "learning_rate": 9.989889390177467e-06,
+      "loss": 0.7821,
+      "step": 130
+    },
+    {
+      "epoch": 0.053903166811050146,
+      "grad_norm": 0.22732502222061157,
+      "learning_rate": 9.985537824862647e-06,
+      "loss": 0.7854,
+      "step": 140
+    },
+    {
+      "epoch": 0.057753393011839446,
+      "grad_norm": 0.22679132223129272,
+      "learning_rate": 9.980410816696317e-06,
+      "loss": 0.7663,
+      "step": 150
+    },
+    {
+      "epoch": 0.06160361921262874,
+      "grad_norm": 0.2128051221370697,
+      "learning_rate": 9.97450916312541e-06,
+      "loss": 0.7775,
+      "step": 160
+    },
+    {
+      "epoch": 0.06545384541341803,
+      "grad_norm": 0.19375000894069672,
+      "learning_rate": 9.967833782084004e-06,
+      "loss": 0.778,
+      "step": 170
+    },
+    {
+      "epoch": 0.06930407161420733,
+      "grad_norm": 0.22488422691822052,
+      "learning_rate": 9.960385711850554e-06,
+      "loss": 0.8017,
+      "step": 180
+    },
+    {
+      "epoch": 0.07315429781499663,
+      "grad_norm": 0.21468931436538696,
+      "learning_rate": 9.952166110886394e-06,
+      "loss": 0.795,
+      "step": 190
+    },
+    {
+      "epoch": 0.07700452401578593,
+      "grad_norm": 0.2400956153869629,
+      "learning_rate": 9.943176257655567e-06,
+      "loss": 0.7853,
+      "step": 200
+    },
+    {
+      "epoch": 0.08085475021657522,
+      "grad_norm": 0.2095334380865097,
+      "learning_rate": 9.933417550425952e-06,
+      "loss": 0.7817,
+      "step": 210
+    },
+    {
+      "epoch": 0.08470497641736452,
+      "grad_norm": 0.20858250558376312,
+      "learning_rate": 9.9228915070518e-06,
+      "loss": 0.7655,
+      "step": 220
+    },
+    {
+      "epoch": 0.08855520261815382,
+      "grad_norm": 0.21185994148254395,
+      "learning_rate": 9.911599764737632e-06,
+      "loss": 0.7688,
+      "step": 230
+    },
+    {
+      "epoch": 0.09240542881894312,
+      "grad_norm": 0.21817010641098022,
+      "learning_rate": 9.899544079783606e-06,
+      "loss": 0.7625,
+      "step": 240
+    },
+    {
+      "epoch": 0.0962556550197324,
+      "grad_norm": 0.21825969219207764,
+      "learning_rate": 9.886726327312338e-06,
+      "loss": 0.7791,
+      "step": 250
+    },
+    {
+      "epoch": 0.1001058812205217,
+      "grad_norm": 0.2143014818429947,
+      "learning_rate": 9.87314850097724e-06,
+      "loss": 0.7758,
+      "step": 260
+    },
+    {
+      "epoch": 0.103956107421311,
+      "grad_norm": 0.20040369033813477,
+      "learning_rate": 9.858812712652452e-06,
+      "loss": 0.7656,
+      "step": 270
+    },
+    {
+      "epoch": 0.10780633362210029,
+      "grad_norm": 0.19657689332962036,
+      "learning_rate": 9.843721192104341e-06,
+      "loss": 0.7632,
+      "step": 280
+    },
+    {
+      "epoch": 0.11165655982288959,
+      "grad_norm": 0.2087303251028061,
+      "learning_rate": 9.827876286644698e-06,
+      "loss": 0.785,
+      "step": 290
+    },
+    {
+      "epoch": 0.11550678602367889,
+      "grad_norm": 0.20141665637493134,
+      "learning_rate": 9.811280460765643e-06,
+      "loss": 0.7482,
+      "step": 300
+    },
+    {
+      "epoch": 0.11935701222446819,
+      "grad_norm": 0.2114259898662567,
+      "learning_rate": 9.793936295756292e-06,
+      "loss": 0.7557,
+      "step": 310
+    },
+    {
+      "epoch": 0.12320723842525748,
+      "grad_norm": 0.21333669126033783,
+      "learning_rate": 9.775846489301279e-06,
+      "loss": 0.7603,
+      "step": 320
+    },
+    {
+      "epoch": 0.12705746462604678,
+      "grad_norm": 0.2160813808441162,
+      "learning_rate": 9.757013855061157e-06,
+      "loss": 0.7583,
+      "step": 330
+    },
+    {
+      "epoch": 0.13090769082683607,
+      "grad_norm": 0.21442899107933044,
+      "learning_rate": 9.737441322234758e-06,
+      "loss": 0.7724,
+      "step": 340
+    },
+    {
+      "epoch": 0.13475791702762538,
+      "grad_norm": 0.21706406772136688,
+      "learning_rate": 9.717131935103605e-06,
+      "loss": 0.7712,
+      "step": 350
+    },
+    {
+      "epoch": 0.13860814322841467,
+      "grad_norm": 0.21671874821186066,
+      "learning_rate": 9.696088852558398e-06,
+      "loss": 0.758,
+      "step": 360
+    },
+    {
+      "epoch": 0.14245836942920398,
+      "grad_norm": 0.2087402641773224,
+      "learning_rate": 9.674315347607686e-06,
+      "loss": 0.7837,
+      "step": 370
+    },
+    {
+      "epoch": 0.14630859562999327,
+      "grad_norm": 0.21003581583499908,
+      "learning_rate": 9.651814806868796e-06,
+      "loss": 0.7489,
+      "step": 380
+    },
+    {
+      "epoch": 0.15015882183078255,
+      "grad_norm": 0.23703910410404205,
+      "learning_rate": 9.62859073004107e-06,
+      "loss": 0.7508,
+      "step": 390
+    },
+    {
+      "epoch": 0.15400904803157187,
+      "grad_norm": 0.18968799710273743,
+      "learning_rate": 9.604646729361547e-06,
+      "loss": 0.7553,
+      "step": 400
+    },
+    {
+      "epoch": 0.15785927423236115,
+      "grad_norm": 0.20380418002605438,
+      "learning_rate": 9.579986529043099e-06,
+      "loss": 0.7464,
+      "step": 410
+    },
+    {
+      "epoch": 0.16170950043315044,
+      "grad_norm": 0.2087632417678833,
+      "learning_rate": 9.554613964695189e-06,
+      "loss": 0.7634,
+      "step": 420
+    },
+    {
+      "epoch": 0.16555972663393975,
+      "grad_norm": 0.2162550687789917,
+      "learning_rate": 9.528532982727281e-06,
+      "loss": 0.756,
+      "step": 430
+    },
+    {
+      "epoch": 0.16940995283472904,
+      "grad_norm": 0.20226193964481354,
+      "learning_rate": 9.501747639735018e-06,
+      "loss": 0.7524,
+      "step": 440
+    },
+    {
+      "epoch": 0.17326017903551832,
+      "grad_norm": 0.20702673494815826,
+      "learning_rate": 9.474262101869272e-06,
+      "loss": 0.7411,
+      "step": 450
+    },
+    {
+      "epoch": 0.17711040523630764,
+      "grad_norm": 0.21235325932502747,
+      "learning_rate": 9.446080644188141e-06,
+      "loss": 0.7479,
+      "step": 460
+    },
+    {
+      "epoch": 0.18096063143709692,
+      "grad_norm": 0.2024451047182083,
+      "learning_rate": 9.417207649992017e-06,
+      "loss": 0.7528,
+      "step": 470
+    },
+    {
+      "epoch": 0.18481085763788624,
+      "grad_norm": 0.21309994161128998,
+      "learning_rate": 9.387647610141816e-06,
+      "loss": 0.7512,
+      "step": 480
+    },
+    {
+      "epoch": 0.18866108383867553,
+      "grad_norm": 0.18987809121608734,
+      "learning_rate": 9.35740512236047e-06,
+      "loss": 0.7384,
+      "step": 490
+    },
+    {
+      "epoch": 0.1925113100394648,
+      "grad_norm": 0.1991511583328247,
+      "learning_rate": 9.326484890517809e-06,
+      "loss": 0.7515,
+      "step": 500
+    },
+    {
+      "epoch": 0.19636153624025413,
+      "grad_norm": 0.22268106043338776,
+      "learning_rate": 9.294891723898926e-06,
+      "loss": 0.7405,
+      "step": 510
+    },
+    {
+      "epoch": 0.2002117624410434,
+      "grad_norm": 0.2213958501815796,
+      "learning_rate": 9.262630536456153e-06,
+      "loss": 0.7697,
+      "step": 520
+    },
+    {
+      "epoch": 0.2040619886418327,
+      "grad_norm": 0.21954451501369476,
+      "learning_rate": 9.229706346044749e-06,
+      "loss": 0.7441,
+      "step": 530
+    },
+    {
+      "epoch": 0.207912214842622,
+      "grad_norm": 0.2248673290014267,
+      "learning_rate": 9.196124273642422e-06,
+      "loss": 0.756,
+      "step": 540
+    },
+    {
+      "epoch": 0.2117624410434113,
+      "grad_norm": 0.21131162345409393,
+      "learning_rate": 9.161889542552842e-06,
+      "loss": 0.7612,
+      "step": 550
+    },
+    {
+      "epoch": 0.21561266724420058,
+      "grad_norm": 0.19974969327449799,
+      "learning_rate": 9.127007477593193e-06,
+      "loss": 0.7618,
+      "step": 560
+    },
+    {
+      "epoch": 0.2194628934449899,
+      "grad_norm": 0.22532647848129272,
+      "learning_rate": 9.091483504265974e-06,
+      "loss": 0.7687,
+      "step": 570
+    },
+    {
+      "epoch": 0.22331311964577918,
+      "grad_norm": 0.21998955309391022,
+      "learning_rate": 9.055323147915116e-06,
+      "loss": 0.7498,
+      "step": 580
+    },
+    {
+      "epoch": 0.2271633458465685,
+      "grad_norm": 0.21207381784915924,
+      "learning_rate": 9.018532032866585e-06,
+      "loss": 0.7594,
+      "step": 590
+    },
+    {
+      "epoch": 0.23101357204735778,
+      "grad_norm": 0.19901323318481445,
+      "learning_rate": 8.981115881553579e-06,
+      "loss": 0.7518,
+      "step": 600
+    },
+    {
+      "epoch": 0.23486379824814707,
+      "grad_norm": 0.22377325594425201,
+      "learning_rate": 8.943080513626476e-06,
+      "loss": 0.7735,
+      "step": 610
+    },
+    {
+      "epoch": 0.23871402444893638,
+      "grad_norm": 0.21403208374977112,
+      "learning_rate": 8.904431845047641e-06,
+      "loss": 0.7576,
+      "step": 620
+    },
+    {
+      "epoch": 0.24256425064972567,
+      "grad_norm": 0.21642917394638062,
+      "learning_rate": 8.86517588717128e-06,
+      "loss": 0.7727,
+      "step": 630
+    },
+    {
+      "epoch": 0.24641447685051496,
+      "grad_norm": 0.2034228891134262,
+      "learning_rate": 8.82531874580844e-06,
+      "loss": 0.7665,
+      "step": 640
+    },
+    {
+      "epoch": 0.25026470305130427,
+      "grad_norm": 0.1914915144443512,
+      "learning_rate": 8.784866620277314e-06,
+      "loss": 0.723,
+      "step": 650
+    },
+    {
+      "epoch": 0.25411492925209356,
+      "grad_norm": 0.22455662488937378,
+      "learning_rate": 8.743825802439018e-06,
+      "loss": 0.7533,
+      "step": 660
+    },
+    {
+      "epoch": 0.25796515545288284,
+      "grad_norm": 0.23121367394924164,
+      "learning_rate": 8.702202675718956e-06,
+      "loss": 0.7644,
+      "step": 670
+    },
+    {
+      "epoch": 0.26181538165367213,
+      "grad_norm": 0.19596794247627258,
+      "learning_rate": 8.660003714113959e-06,
+      "loss": 0.7315,
+      "step": 680
+    },
+    {
+      "epoch": 0.26566560785446147,
+      "grad_norm": 0.21284624934196472,
+      "learning_rate": 8.617235481185323e-06,
+      "loss": 0.753,
+      "step": 690
+    },
+    {
+      "epoch": 0.26951583405525076,
+      "grad_norm": 0.21604591608047485,
+      "learning_rate": 8.573904629037931e-06,
+      "loss": 0.7596,
+      "step": 700
+    },
+    {
+      "epoch": 0.27336606025604004,
+      "grad_norm": 0.20862695574760437,
+      "learning_rate": 8.530017897285585e-06,
+      "loss": 0.7426,
+      "step": 710
+    },
+    {
+      "epoch": 0.27721628645682933,
+      "grad_norm": 0.1971656084060669,
+      "learning_rate": 8.485582112002746e-06,
+      "loss": 0.7475,
+      "step": 720
+    },
+    {
+      "epoch": 0.2810665126576186,
+      "grad_norm": 0.2059982866048813,
+      "learning_rate": 8.440604184662811e-06,
+      "loss": 0.7492,
+      "step": 730
+    },
+    {
+      "epoch": 0.28491673885840796,
+      "grad_norm": 0.21329063177108765,
+      "learning_rate": 8.395091111063118e-06,
+      "loss": 0.7392,
+      "step": 740
+    },
+    {
+      "epoch": 0.28876696505919724,
+      "grad_norm": 0.1931198686361313,
+      "learning_rate": 8.349049970236822e-06,
+      "loss": 0.7418,
+      "step": 750
+    },
+    {
+      "epoch": 0.29261719125998653,
+      "grad_norm": 0.2042660415172577,
+      "learning_rate": 8.302487923351842e-06,
+      "loss": 0.7312,
+      "step": 760
+    },
+    {
+      "epoch": 0.2964674174607758,
+      "grad_norm": 0.20536047220230103,
+      "learning_rate": 8.255412212597021e-06,
+      "loss": 0.7493,
+      "step": 770
+    },
+    {
+      "epoch": 0.3003176436615651,
+      "grad_norm": 0.1983938068151474,
+      "learning_rate": 8.207830160055688e-06,
+      "loss": 0.7457,
+      "step": 780
+    },
+    {
+      "epoch": 0.3041678698623544,
+      "grad_norm": 0.21412931382656097,
+      "learning_rate": 8.159749166566786e-06,
+      "loss": 0.7321,
+      "step": 790
+    },
+    {
+      "epoch": 0.30801809606314373,
+      "grad_norm": 0.1993480920791626,
+      "learning_rate": 8.11117671057377e-06,
+      "loss": 0.758,
+      "step": 800
+    },
+    {
+      "epoch": 0.311868322263933,
+      "grad_norm": 0.21769478917121887,
+      "learning_rate": 8.062120346961404e-06,
+      "loss": 0.7502,
+      "step": 810
+    },
+    {
+      "epoch": 0.3157185484647223,
+      "grad_norm": 0.1956753432750702,
+      "learning_rate": 8.012587705880705e-06,
+      "loss": 0.7456,
+      "step": 820
+    },
+    {
+      "epoch": 0.3195687746655116,
+      "grad_norm": 0.21049824357032776,
+      "learning_rate": 7.962586491562146e-06,
+      "loss": 0.7305,
+      "step": 830
+    },
+    {
+      "epoch": 0.3234190008663009,
+      "grad_norm": 0.21689215302467346,
+      "learning_rate": 7.912124481117357e-06,
+      "loss": 0.7366,
+      "step": 840
+    },
+    {
+      "epoch": 0.3272692270670902,
+      "grad_norm": 0.20401524007320404,
+      "learning_rate": 7.861209523329494e-06,
+      "loss": 0.7496,
+      "step": 850
+    },
+    {
+      "epoch": 0.3311194532678795,
+      "grad_norm": 0.20821674168109894,
+      "learning_rate": 7.809849537432432e-06,
+      "loss": 0.732,
+      "step": 860
+    },
+    {
+      "epoch": 0.3349696794686688,
+      "grad_norm": 0.2078704684972763,
+      "learning_rate": 7.758052511879035e-06,
+      "loss": 0.7426,
+      "step": 870
+    },
+    {
+      "epoch": 0.3388199056694581,
+      "grad_norm": 0.19707965850830078,
+      "learning_rate": 7.705826503098642e-06,
+      "loss": 0.7385,
+      "step": 880
+    },
+    {
+      "epoch": 0.34267013187024736,
+      "grad_norm": 0.20079182088375092,
+      "learning_rate": 7.653179634243979e-06,
+      "loss": 0.738,
+      "step": 890
+    },
+    {
+      "epoch": 0.34652035807103665,
+      "grad_norm": 0.1860584169626236,
+      "learning_rate": 7.600120093927696e-06,
+      "loss": 0.7392,
+      "step": 900
+    },
+    {
+      "epoch": 0.350370584271826,
+      "grad_norm": 0.21798209846019745,
+      "learning_rate": 7.546656134948729e-06,
+      "loss": 0.7248,
+      "step": 910
+    },
+    {
+      "epoch": 0.3542208104726153,
+      "grad_norm": 0.21079829335212708,
+      "learning_rate": 7.492796073008661e-06,
+      "loss": 0.7301,
+      "step": 920
+    },
+    {
+      "epoch": 0.35807103667340456,
+      "grad_norm": 0.22210615873336792,
+      "learning_rate": 7.438548285418326e-06,
+      "loss": 0.7174,
+      "step": 930
+    },
+    {
+      "epoch": 0.36192126287419385,
+      "grad_norm": 0.23038989305496216,
+      "learning_rate": 7.3839212097948e-06,
+      "loss": 0.7526,
+      "step": 940
+    },
+    {
+      "epoch": 0.36577148907498314,
+      "grad_norm": 0.2096896469593048,
+      "learning_rate": 7.328923342749039e-06,
+      "loss": 0.7658,
+      "step": 950
+    },
+    {
+      "epoch": 0.3696217152757725,
+      "grad_norm": 0.21523195505142212,
+      "learning_rate": 7.273563238564319e-06,
+      "loss": 0.7516,
+      "step": 960
+    },
+    {
+      "epoch": 0.37347194147656176,
+      "grad_norm": 0.21206583082675934,
+      "learning_rate": 7.217849507865724e-06,
+      "loss": 0.7291,
+      "step": 970
+    },
+    {
+      "epoch": 0.37732216767735105,
+      "grad_norm": 0.2058362513780594,
+      "learning_rate": 7.161790816280858e-06,
+      "loss": 0.7378,
+      "step": 980
+    },
+    {
+      "epoch": 0.38117239387814034,
+      "grad_norm": 0.21002115309238434,
+      "learning_rate": 7.1053958830920125e-06,
+      "loss": 0.7175,
+      "step": 990
+    },
+    {
+      "epoch": 0.3850226200789296,
+      "grad_norm": 0.20490789413452148,
+      "learning_rate": 7.0486734798799715e-06,
+      "loss": 0.7379,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3888728462797189,
+      "grad_norm": 0.20403194427490234,
+      "learning_rate": 6.991632429159708e-06,
+      "loss": 0.761,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39272307248050825,
+      "grad_norm": 0.22212830185890198,
+      "learning_rate": 6.934281603008132e-06,
+      "loss": 0.7521,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39657329868129754,
+      "grad_norm": 0.19919133186340332,
+      "learning_rate": 6.876629921684143e-06,
+      "loss": 0.7498,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4004235248820868,
+      "grad_norm": 0.21023645997047424,
+      "learning_rate": 6.818686352241199e-06,
+      "loss": 0.7401,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4042737510828761,
+      "grad_norm": 0.19032973051071167,
+      "learning_rate": 6.760459907132573e-06,
+      "loss": 0.736,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4081239772836654,
+      "grad_norm": 0.1914728581905365,
+      "learning_rate": 6.701959642809589e-06,
+      "loss": 0.7412,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41197420348445474,
+      "grad_norm": 0.2087353616952896,
+      "learning_rate": 6.643194658312987e-06,
+      "loss": 0.7321,
+      "step": 1070
+    },
+    {
+      "epoch": 0.415824429685244,
+      "grad_norm": 0.19705414772033691,
+      "learning_rate": 6.584174093857676e-06,
+      "loss": 0.7558,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4196746558860333,
+      "grad_norm": 0.22641058266162872,
+      "learning_rate": 6.5249071294110735e-06,
+      "loss": 0.7405,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4235248820868226,
+      "grad_norm": 0.21289943158626556,
+      "learning_rate": 6.465402983265274e-06,
+      "loss": 0.7546,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4273751082876119,
+      "grad_norm": 0.21463654935359955,
+      "learning_rate": 6.405670910603243e-06,
+      "loss": 0.7495,
+      "step": 1110
+    },
+    {
+      "epoch": 0.43122533448840117,
+      "grad_norm": 0.2237464040517807,
+      "learning_rate": 6.345720202059291e-06,
+      "loss": 0.7461,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4350755606891905,
+      "grad_norm": 0.20567022264003754,
+      "learning_rate": 6.285560182274008e-06,
+      "loss": 0.7233,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4389257868899798,
+      "grad_norm": 0.20194895565509796,
+      "learning_rate": 6.225200208443936e-06,
+      "loss": 0.7461,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4427760130907691,
+      "grad_norm": 0.2019462287425995,
+      "learning_rate": 6.16464966886616e-06,
+      "loss": 0.7354,
+      "step": 1150
+    },
+    {
+      "epoch": 0.44662623929155837,
+      "grad_norm": 0.24128669500350952,
+      "learning_rate": 6.103917981478065e-06,
+      "loss": 0.7603,
+      "step": 1160
+    },
+    {
+      "epoch": 0.45047646549234766,
+      "grad_norm": 0.2029421180486679,
+      "learning_rate": 6.043014592392489e-06,
+      "loss": 0.7716,
+      "step": 1170
+    },
+    {
+      "epoch": 0.454326691693137,
+      "grad_norm": 0.19456329941749573,
+      "learning_rate": 5.981948974428485e-06,
+      "loss": 0.7447,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4581769178939263,
+      "grad_norm": 0.2101643830537796,
+      "learning_rate": 5.920730625637934e-06,
+      "loss": 0.7223,
+      "step": 1190
+    },
+    {
+      "epoch": 0.46202714409471557,
+      "grad_norm": 0.2161208987236023,
+      "learning_rate": 5.8593690678282355e-06,
+      "loss": 0.7367,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46587737029550486,
+      "grad_norm": 0.21383459866046906,
+      "learning_rate": 5.797873845081301e-06,
+      "loss": 0.7396,
+      "step": 1210
+    },
+    {
+      "epoch": 0.46972759649629414,
+      "grad_norm": 0.20989835262298584,
+      "learning_rate": 5.736254522269075e-06,
+      "loss": 0.7556,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47357782269708343,
+      "grad_norm": 0.20152361690998077,
+      "learning_rate": 5.6745206835658386e-06,
+      "loss": 0.7615,
+      "step": 1230
+    },
+    {
+      "epoch": 0.47742804889787277,
+      "grad_norm": 0.1987544447183609,
+      "learning_rate": 5.612681930957495e-06,
+      "loss": 0.7288,
+      "step": 1240
+    },
+    {
+      "epoch": 0.48127827509866206,
+      "grad_norm": 0.19656851887702942,
+      "learning_rate": 5.550747882748095e-06,
+      "loss": 0.7255,
+      "step": 1250
+    },
+    {
+      "epoch": 0.48512850129945134,
+      "grad_norm": 0.21723157167434692,
+      "learning_rate": 5.4887281720638195e-06,
+      "loss": 0.7365,
+      "step": 1260
+    },
+    {
+      "epoch": 0.48897872750024063,
+      "grad_norm": 0.23111577332019806,
+      "learning_rate": 5.42663244535466e-06,
+      "loss": 0.7166,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4928289537010299,
+      "grad_norm": 0.20663127303123474,
+      "learning_rate": 5.3644703608940205e-06,
+      "loss": 0.7284,
+      "step": 1280
+    },
+    {
+      "epoch": 0.49667917990181926,
+      "grad_norm": 0.20647293329238892,
+      "learning_rate": 5.302251587276491e-06,
+      "loss": 0.7371,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5005294061026085,
+      "grad_norm": 0.19643406569957733,
+      "learning_rate": 5.2399858019140005e-06,
+      "loss": 0.7396,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5043796323033978,
+      "grad_norm": 0.19850556552410126,
+      "learning_rate": 5.177682689530615e-06,
+      "loss": 0.7532,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5082298585041871,
+      "grad_norm": 0.21492964029312134,
+      "learning_rate": 5.11535194065619e-06,
+      "loss": 0.7223,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5120800847049765,
+      "grad_norm": 0.22107228636741638,
+      "learning_rate": 5.053003250119113e-06,
+      "loss": 0.7302,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5159303109057657,
+      "grad_norm": 0.2126016914844513,
+      "learning_rate": 4.9906463155383925e-06,
+      "loss": 0.7369,
+      "step": 1340
+    },
+    {
+      "epoch": 0.519780537106555,
+      "grad_norm": 0.20797733962535858,
+      "learning_rate": 4.9282908358153046e-06,
+      "loss": 0.7333,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5236307633073443,
+      "grad_norm": 0.23392651975154877,
+      "learning_rate": 4.8659465096248345e-06,
+      "loss": 0.748,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5274809895081336,
+      "grad_norm": 0.21066848933696747,
+      "learning_rate": 4.803623033907169e-06,
+      "loss": 0.7435,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5313312157089229,
+      "grad_norm": 0.1987970620393753,
+      "learning_rate": 4.741330102359441e-06,
+      "loss": 0.7523,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5351814419097122,
+      "grad_norm": 0.2203347086906433,
+      "learning_rate": 4.679077403927996e-06,
+      "loss": 0.7306,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5390316681105015,
+      "grad_norm": 0.20622745156288147,
+      "learning_rate": 4.616874621301373e-06,
+      "loss": 0.7336,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5428818943112907,
+      "grad_norm": 0.2118251919746399,
+      "learning_rate": 4.554731429404293e-06,
+      "loss": 0.7466,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5467321205120801,
+      "grad_norm": 0.21767885982990265,
+      "learning_rate": 4.492657493892822e-06,
+      "loss": 0.7306,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5505823467128694,
+      "grad_norm": 0.22103926539421082,
+      "learning_rate": 4.430662469650999e-06,
+      "loss": 0.7465,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5544325729136587,
+      "grad_norm": 0.21318268775939941,
+      "learning_rate": 4.368755999289118e-06,
+      "loss": 0.752,
+      "step": 1440
+    },
+    {
+      "epoch": 0.558282799114448,
+      "grad_norm": 0.21599628031253815,
+      "learning_rate": 4.306947711643949e-06,
+      "loss": 0.7312,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5621330253152372,
+      "grad_norm": 0.21296727657318115,
+      "learning_rate": 4.245247220281068e-06,
+      "loss": 0.7374,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5659832515160266,
+      "grad_norm": 0.20392480492591858,
+      "learning_rate": 4.183664121999587e-06,
+      "loss": 0.7322,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5698334777168159,
+      "grad_norm": 0.22679561376571655,
+      "learning_rate": 4.122207995339474e-06,
+      "loss": 0.7318,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5736837039176051,
+      "grad_norm": 0.1984763741493225,
+      "learning_rate": 4.060888399091734e-06,
+      "loss": 0.7349,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5775339301183945,
+      "grad_norm": 0.19852308928966522,
+      "learning_rate": 3.999714870811639e-06,
+      "loss": 0.7474,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5813841563191837,
+      "grad_norm": 0.2058991640806198,
+      "learning_rate": 3.938696925335275e-06,
+      "loss": 0.7397,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5852343825199731,
+      "grad_norm": 0.20268718898296356,
+      "learning_rate": 3.87784405329962e-06,
+      "loss": 0.7474,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5890846087207623,
+      "grad_norm": 0.20448251068592072,
+      "learning_rate": 3.817165719666382e-06,
+      "loss": 0.7473,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5929348349215516,
+      "grad_norm": 0.20428811013698578,
+      "learning_rate": 3.756671362249841e-06,
+      "loss": 0.7542,
+      "step": 1540
+    },
+    {
+      "epoch": 0.596785061122341,
+      "grad_norm": 0.215225487947464,
+      "learning_rate": 3.6963703902488907e-06,
+      "loss": 0.7401,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6006352873231302,
+      "grad_norm": 0.20664937794208527,
+      "learning_rate": 3.6362721827835655e-06,
+      "loss": 0.7527,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6044855135239195,
+      "grad_norm": 0.2130601853132248,
+      "learning_rate": 3.576386087436212e-06,
+      "loss": 0.7519,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6083357397247088,
+      "grad_norm": 0.2195250689983368,
+      "learning_rate": 3.5167214187975913e-06,
+      "loss": 0.7408,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6121859659254981,
+      "grad_norm": 0.1836192011833191,
+      "learning_rate": 3.457287457018086e-06,
+      "loss": 0.7432,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6160361921262875,
+      "grad_norm": 0.194973886013031,
+      "learning_rate": 3.3980934463643003e-06,
+      "loss": 0.7312,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6198864183270767,
+      "grad_norm": 0.19340261816978455,
+      "learning_rate": 3.3391485937812053e-06,
+      "loss": 0.7126,
+      "step": 1610
+    },
+    {
+      "epoch": 0.623736644527866,
+      "grad_norm": 0.21016700565814972,
+      "learning_rate": 3.280462067460115e-06,
+      "loss": 0.7122,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6275868707286553,
+      "grad_norm": 0.203949972987175,
+      "learning_rate": 3.222042995412669e-06,
+      "loss": 0.7452,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6314370969294446,
+      "grad_norm": 0.18483348190784454,
+      "learning_rate": 3.163900464051089e-06,
+      "loss": 0.7335,
+      "step": 1640
+    },
+    {
+      "epoch": 0.635287323130234,
+      "grad_norm": 0.2075316607952118,
+      "learning_rate": 3.1060435167748826e-06,
+      "loss": 0.7449,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6391375493310232,
+      "grad_norm": 0.2047448456287384,
+      "learning_rate": 3.048481152564253e-06,
+      "loss": 0.7251,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6429877755318125,
+      "grad_norm": 0.2189546525478363,
+      "learning_rate": 2.9912223245804017e-06,
+      "loss": 0.7458,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6468380017326018,
+      "grad_norm": 0.21480692923069,
+      "learning_rate": 2.934275938772981e-06,
+      "loss": 0.739,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6506882279333911,
+      "grad_norm": 0.20565976202487946,
+      "learning_rate": 2.8776508524948676e-06,
+      "loss": 0.7298,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6545384541341804,
+      "grad_norm": 0.1837766170501709,
+      "learning_rate": 2.821355873124503e-06,
+      "loss": 0.7283,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6583886803349697,
+      "grad_norm": 0.20932051539421082,
+      "learning_rate": 2.7653997566960144e-06,
+      "loss": 0.7424,
+      "step": 1710
+    },
+    {
+      "epoch": 0.662238906535759,
+      "grad_norm": 0.20843954384326935,
+      "learning_rate": 2.709791206537309e-06,
+      "loss": 0.7321,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6660891327365482,
+      "grad_norm": 0.21107088029384613,
+      "learning_rate": 2.6545388719163747e-06,
+      "loss": 0.7329,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6699393589373376,
+      "grad_norm": 0.21113397181034088,
+      "learning_rate": 2.599651346695979e-06,
+      "loss": 0.7357,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6737895851381268,
+      "grad_norm": 0.20257095992565155,
+      "learning_rate": 2.545137167997007e-06,
+      "loss": 0.7352,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6776398113389162,
+      "grad_norm": 0.19663965702056885,
+      "learning_rate": 2.4910048148705972e-06,
+      "loss": 0.7218,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6814900375397055,
+      "grad_norm": 0.2160496562719345,
+      "learning_rate": 2.4372627069793327e-06,
+      "loss": 0.7681,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6853402637404947,
+      "grad_norm": 0.22264611721038818,
+      "learning_rate": 2.3839192032876486e-06,
+      "loss": 0.7506,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6891904899412841,
+      "grad_norm": 0.20332592725753784,
+      "learning_rate": 2.3309826007617047e-06,
+      "loss": 0.7611,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6930407161420733,
+      "grad_norm": 0.2058313935995102,
+      "learning_rate": 2.278461133078881e-06,
+      "loss": 0.7256,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6968909423428626,
+      "grad_norm": 0.21930497884750366,
+      "learning_rate": 2.2263629693471258e-06,
+      "loss": 0.717,
+      "step": 1810
+    },
+    {
+      "epoch": 0.700741168543652,
+      "grad_norm": 0.19607724249362946,
+      "learning_rate": 2.1746962128343414e-06,
+      "loss": 0.7389,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7045913947444412,
+      "grad_norm": 0.2033325582742691,
+      "learning_rate": 2.1234688997080183e-06,
+      "loss": 0.7293,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7084416209452306,
+      "grad_norm": 0.22264441847801208,
+      "learning_rate": 2.0726889977853053e-06,
+      "loss": 0.7358,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7122918471460198,
+      "grad_norm": 0.22031886875629425,
+      "learning_rate": 2.022364405293703e-06,
+      "loss": 0.7475,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7161420733468091,
+      "grad_norm": 0.20421592891216278,
+      "learning_rate": 1.9725029496425814e-06,
+      "loss": 0.7418,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7199922995475985,
+      "grad_norm": 0.19128301739692688,
+      "learning_rate": 1.923112386205727e-06,
+      "loss": 0.7365,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7238425257483877,
+      "grad_norm": 0.1990060955286026,
+      "learning_rate": 1.874200397115079e-06,
+      "loss": 0.7456,
+      "step": 1880
+    },
+    {
+      "epoch": 0.727692751949177,
+      "grad_norm": 0.2213444858789444,
+      "learning_rate": 1.825774590065857e-06,
+      "loss": 0.728,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7315429781499663,
+      "grad_norm": 0.21128568053245544,
+      "learning_rate": 1.7778424971332813e-06,
+      "loss": 0.7383,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7353932043507556,
+      "grad_norm": 0.19812627136707306,
+      "learning_rate": 1.7304115736010407e-06,
+      "loss": 0.7344,
+      "step": 1910
+    },
+    {
+      "epoch": 0.739243430551545,
+      "grad_norm": 0.20407681167125702,
+      "learning_rate": 1.6834891968017108e-06,
+      "loss": 0.73,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7430936567523342,
+      "grad_norm": 0.2137129008769989,
+      "learning_rate": 1.6370826649692884e-06,
+      "loss": 0.7431,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7469438829531235,
+      "grad_norm": 0.2247190624475479,
+      "learning_rate": 1.5911991961040463e-06,
+      "loss": 0.752,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7507941091539128,
+      "grad_norm": 0.21254469454288483,
+      "learning_rate": 1.545845926849851e-06,
+      "loss": 0.7407,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7546443353547021,
+      "grad_norm": 0.2218170315027237,
+      "learning_rate": 1.5010299113841397e-06,
+      "loss": 0.7566,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7584945615554913,
+      "grad_norm": 0.20940083265304565,
+      "learning_rate": 1.4567581203207238e-06,
+      "loss": 0.7468,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7623447877562807,
+      "grad_norm": 0.21801768243312836,
+      "learning_rate": 1.413037439625593e-06,
+      "loss": 0.7451,
+      "step": 1980
+    },
+    {
+      "epoch": 0.76619501395707,
+      "grad_norm": 0.21765030920505524,
+      "learning_rate": 1.3698746695458847e-06,
+      "loss": 0.7393,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7700452401578592,
+      "grad_norm": 0.20398850739002228,
+      "learning_rate": 1.3272765235521745e-06,
+      "loss": 0.7316,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7738954663586486,
+      "grad_norm": 0.20528504252433777,
+      "learning_rate": 1.2852496272942877e-06,
+      "loss": 0.7425,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7777456925594378,
+      "grad_norm": 0.19537873566150665,
+      "learning_rate": 1.2438005175707396e-06,
+      "loss": 0.7275,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7815959187602272,
+      "grad_norm": 0.20100075006484985,
+      "learning_rate": 1.2029356413120297e-06,
+      "loss": 0.7122,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7854461449610165,
+      "grad_norm": 0.20554952323436737,
+      "learning_rate": 1.1626613545778786e-06,
+      "loss": 0.7287,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7892963711618057,
+      "grad_norm": 0.20241102576255798,
+      "learning_rate": 1.1229839215686294e-06,
+      "loss": 0.7331,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7931465973625951,
+      "grad_norm": 0.19400659203529358,
+      "learning_rate": 1.0839095136509188e-06,
+      "loss": 0.7164,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7969968235633843,
+      "grad_norm": 0.21605609357357025,
+      "learning_rate": 1.045444208397791e-06,
+      "loss": 0.738,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8008470497641736,
+      "grad_norm": 0.19360129535198212,
+      "learning_rate": 1.007593988643402e-06,
+      "loss": 0.7231,
+      "step": 2080
+    },
+    {
+      "epoch": 0.804697275964963,
+      "grad_norm": 0.19576580822467804,
+      "learning_rate": 9.703647415524613e-07,
+      "loss": 0.7441,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8085475021657522,
+      "grad_norm": 0.20411916077136993,
+      "learning_rate": 9.337622577045485e-07,
+      "loss": 0.7315,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8123977283665416,
+      "grad_norm": 0.20931822061538696,
+      "learning_rate": 8.977922301934588e-07,
+      "loss": 0.743,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8162479545673308,
+      "grad_norm": 0.20833367109298706,
+      "learning_rate": 8.624602537417009e-07,
+      "loss": 0.7439,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8200981807681201,
+      "grad_norm": 0.22011397778987885,
+      "learning_rate": 8.277718238303123e-07,
+      "loss": 0.7287,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8239484069689095,
+      "grad_norm": 0.20507727563381195,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.7328,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8277986331696987,
+      "grad_norm": 0.2230200171470642,
+      "learning_rate": 7.603470842324223e-07,
+      "loss": 0.724,
+      "step": 2150
+    },
+    {
+      "epoch": 0.831648859370488,
+      "grad_norm": 0.20112238824367523,
+      "learning_rate": 7.276212616857575e-07,
+      "loss": 0.7436,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8354990855712773,
+      "grad_norm": 0.20401395857334137,
+      "learning_rate": 6.955599583279837e-07,
+      "loss": 0.7422,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8393493117720666,
+      "grad_norm": 0.20610256493091583,
+      "learning_rate": 6.641681609246981e-07,
+      "loss": 0.7348,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8431995379728559,
+      "grad_norm": 0.19775360822677612,
+      "learning_rate": 6.334507521075711e-07,
+      "loss": 0.7407,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8470497641736452,
+      "grad_norm": 0.21478189527988434,
+      "learning_rate": 6.034125096149179e-07,
+      "loss": 0.7373,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8508999903744345,
+      "grad_norm": 0.21097274124622345,
+      "learning_rate": 5.740581055485722e-07,
+      "loss": 0.7072,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8547502165752238,
+      "grad_norm": 0.19041503965854645,
+      "learning_rate": 5.453921056471961e-07,
+      "loss": 0.7394,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8586004427760131,
+      "grad_norm": 0.21016176044940948,
+      "learning_rate": 5.174189685761288e-07,
+      "loss": 0.7365,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8624506689768023,
+      "grad_norm": 0.19182974100112915,
+      "learning_rate": 4.901430452338979e-07,
+      "loss": 0.7472,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8663008951775917,
+      "grad_norm": 0.1939581036567688,
+      "learning_rate": 4.6356857807548286e-07,
+      "loss": 0.7304,
+      "step": 2250
+    },
+    {
+      "epoch": 0.870151121378381,
+      "grad_norm": 0.20594309270381927,
+      "learning_rate": 4.376997004524541e-07,
+      "loss": 0.7207,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8740013475791703,
+      "grad_norm": 0.20231139659881592,
+      "learning_rate": 4.12540435970073e-07,
+      "loss": 0.7268,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8778515737799596,
+      "grad_norm": 0.20554253458976746,
+      "learning_rate": 3.880946978614703e-07,
+      "loss": 0.7409,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8817017999807488,
+      "grad_norm": 0.2030980885028839,
+      "learning_rate": 3.643662883789878e-07,
+      "loss": 0.7412,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8855520261815382,
+      "grad_norm": 0.19294650852680206,
+      "learning_rate": 3.41358898202781e-07,
+      "loss": 0.7219,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8894022523823275,
+      "grad_norm": 0.19389821588993073,
+      "learning_rate": 3.190761058667741e-07,
+      "loss": 0.7332,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8932524785831167,
+      "grad_norm": 0.203164741396904,
+      "learning_rate": 2.9752137720206463e-07,
+      "loss": 0.737,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8971027047839061,
+      "grad_norm": 0.20090574026107788,
+      "learning_rate": 2.7669806479785067e-07,
+      "loss": 0.7427,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9009529309846953,
+      "grad_norm": 0.20257596671581268,
+      "learning_rate": 2.566094074799763e-07,
+      "loss": 0.7353,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9048031571854847,
+      "grad_norm": 0.20320729911327362,
+      "learning_rate": 2.3725852980716557e-07,
+      "loss": 0.7218,
+      "step": 2350
+    },
+    {
+      "epoch": 0.908653383386274,
+      "grad_norm": 0.19884338974952698,
+      "learning_rate": 2.1864844158503983e-07,
+      "loss": 0.7316,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9125036095870632,
+      "grad_norm": 0.2189999222755432,
+      "learning_rate": 2.007820373979702e-07,
+      "loss": 0.7251,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9163538357878526,
+      "grad_norm": 0.20163306593894958,
+      "learning_rate": 1.8366209615886532e-07,
+      "loss": 0.7345,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9202040619886418,
+      "grad_norm": 0.20421792566776276,
+      "learning_rate": 1.6729128067693613e-07,
+      "loss": 0.7474,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9240542881894311,
+      "grad_norm": 0.22121217846870422,
+      "learning_rate": 1.5167213724353426e-07,
+      "loss": 0.7461,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9279045143902205,
+      "grad_norm": 0.20139385759830475,
+      "learning_rate": 1.3680709523610058e-07,
+      "loss": 0.727,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9317547405910097,
+      "grad_norm": 0.2055376023054123,
+      "learning_rate": 1.2269846674030418e-07,
+      "loss": 0.75,
+      "step": 2420
+    },
+    {
+      "epoch": 0.935604966791799,
+      "grad_norm": 0.21648965775966644,
+      "learning_rate": 1.0934844619042595e-07,
+      "loss": 0.7453,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9394551929925883,
+      "grad_norm": 0.21709614992141724,
+      "learning_rate": 9.675911002803961e-08,
+      "loss": 0.734,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9433054191933776,
+      "grad_norm": 0.2211645096540451,
+      "learning_rate": 8.493241637904381e-08,
+      "loss": 0.7308,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9471556453941669,
+      "grad_norm": 0.20659662783145905,
+      "learning_rate": 7.387020474909689e-08,
+      "loss": 0.7508,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9510058715949562,
+      "grad_norm": 0.22171872854232788,
+      "learning_rate": 6.357419573750856e-08,
+      "loss": 0.7466,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9548560977957455,
+      "grad_norm": 0.20153294503688812,
+      "learning_rate": 5.4045990769613365e-08,
+      "loss": 0.7272,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9587063239965348,
+      "grad_norm": 0.2043210119009018,
+      "learning_rate": 4.5287071847694434e-08,
+      "loss": 0.7344,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9625565501973241,
+      "grad_norm": 0.20310825109481812,
+      "learning_rate": 3.7298801320467817e-08,
+      "loss": 0.732,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9664067763981133,
+      "grad_norm": 0.20677487552165985,
+      "learning_rate": 3.0082421671192576e-08,
+      "loss": 0.7259,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9702570025989027,
+      "grad_norm": 0.20481035113334656,
+      "learning_rate": 2.3639055324410355e-08,
+      "loss": 0.7458,
+      "step": 2520
+    },
+    {
+      "epoch": 0.974107228799692,
+      "grad_norm": 0.20366270840168,
+      "learning_rate": 1.7969704471369477e-08,
+      "loss": 0.7298,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9779574550004813,
+      "grad_norm": 0.19724921882152557,
+      "learning_rate": 1.3075250914142413e-08,
+      "loss": 0.7316,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9818076812012706,
+      "grad_norm": 0.19427776336669922,
+      "learning_rate": 8.956455928476049e-09,
+      "loss": 0.7402,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9856579074020598,
+      "grad_norm": 0.20086972415447235,
+      "learning_rate": 5.6139601453786275e-09,
+      "loss": 0.7328,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9895081336028492,
+      "grad_norm": 0.20216654241085052,
+      "learning_rate": 3.0482834514805602e-09,
+      "loss": 0.7231,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9933583598036385,
+      "grad_norm": 0.19881804287433624,
+      "learning_rate": 1.2598249081713366e-09,
+      "loss": 0.7137,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9972085860044277,
+      "grad_norm": 0.20754384994506836,
+      "learning_rate": 2.488626895302826e-10,
+      "loss": 0.7487,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9999037443449803,
+      "step": 2597,
+      "total_flos": 4.066856054212788e+18,
+      "train_loss": 0.7553673464378854,
+      "train_runtime": 8619.1113,
+      "train_samples_per_second": 77.143,
+      "train_steps_per_second": 0.301
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2597,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.066856054212788e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen25/mc2_l14_224/finetune/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea0db9a8aec619b6af07d21e8d0e7b15193cfb792e2538ee83296c3bf96c1f76
+size 6136

continuous/qwen25/mc2_l14_224/pretrain/config.json

@@ -0,0 +1,58 @@
+{
+  "_attn_implementation_autoset": true,
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 21,
+  "mm_hidden_size": 1024,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "flat",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_tunable_parts": "mm_mlp_adapter",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "open_clip_hub:ViT-L-14",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen2",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 32768,
+  "tie_word_embeddings": true,
+  "tokenizer_model_max_length": 8192,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.3",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_tower_pretrained": "/cache/ckpt/download/tokenizer/continuous/mc2/mc2_l14_224.pt",
+  "vocab_size": 151936
+}

continuous/qwen25/mc2_l14_224/pretrain/mm_projector.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2532bd4b26b781c8c7886dd7f6a46b25bb402c8d5fd3824598939a49a2ca04ab
+size 15743032

continuous/qwen25/mc2_l14_224/pretrain/trainer_state.json

@@ -0,0 +1,421 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 545,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01834862385321101,
+      "grad_norm": 1.5605828762054443,
+      "learning_rate": 0.00032142857142857147,
+      "loss": 4.7922,
+      "step": 10
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 0.32338252663612366,
+      "learning_rate": 0.0006785714285714287,
+      "loss": 3.5007,
+      "step": 20
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 0.10640215128660202,
+      "learning_rate": 0.0009999907688176173,
+      "loss": 3.2576,
+      "step": 30
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 0.11233284324407578,
+      "learning_rate": 0.0009988834393115766,
+      "loss": 3.1714,
+      "step": 40
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 0.143575057387352,
+      "learning_rate": 0.0009959345572940772,
+      "loss": 3.1056,
+      "step": 50
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 0.13603857159614563,
+      "learning_rate": 0.0009911550081155983,
+      "loss": 3.0313,
+      "step": 60
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.13515746593475342,
+      "learning_rate": 0.0009845624347566433,
+      "loss": 2.9743,
+      "step": 70
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 0.11585989594459534,
+      "learning_rate": 0.0009761811727013547,
+      "loss": 2.926,
+      "step": 80
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 0.21629685163497925,
+      "learning_rate": 0.0009660421601067666,
+      "loss": 2.8706,
+      "step": 90
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 0.21672171354293823,
+      "learning_rate": 0.0009541828235992926,
+      "loss": 2.8452,
+      "step": 100
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 0.1830170452594757,
+      "learning_rate": 0.0009406469401200151,
+      "loss": 2.8094,
+      "step": 110
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 0.16867105662822723,
+      "learning_rate": 0.0009254844753287492,
+      "loss": 2.8078,
+      "step": 120
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 0.13484440743923187,
+      "learning_rate": 0.0009087513991633924,
+      "loss": 2.7745,
+      "step": 130
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 0.10492187738418579,
+      "learning_rate": 0.0008905094792353917,
+      "loss": 2.7475,
+      "step": 140
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 0.12424638867378235,
+      "learning_rate": 0.0008708260528239789,
+      "loss": 2.7333,
+      "step": 150
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.10864352434873581,
+      "learning_rate": 0.0008497737783108237,
+      "loss": 2.7026,
+      "step": 160
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 0.14068374037742615,
+      "learning_rate": 0.0008274303669726426,
+      "loss": 2.682,
+      "step": 170
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 0.11920050531625748,
+      "learning_rate": 0.0008038782961218137,
+      "loss": 2.6487,
+      "step": 180
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.13926160335540771,
+      "learning_rate": 0.000779204504653894,
+      "loss": 2.6246,
+      "step": 190
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 0.13258130848407745,
+      "learning_rate": 0.0007535000721258791,
+      "loss": 2.5879,
+      "step": 200
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 0.12856905162334442,
+      "learning_rate": 0.00072685988254984,
+      "loss": 2.5728,
+      "step": 210
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 0.12240317463874817,
+      "learning_rate": 0.0006993822741429906,
+      "loss": 2.5739,
+      "step": 220
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.11876916885375977,
+      "learning_rate": 0.0006711686763270818,
+      "loss": 2.5613,
+      "step": 230
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 0.12407710403203964,
+      "learning_rate": 0.0006423232353170831,
+      "loss": 2.5448,
+      "step": 240
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 0.1333903968334198,
+      "learning_rate": 0.0006129524296812335,
+      "loss": 2.5407,
+      "step": 250
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 0.11863861978054047,
+      "learning_rate": 0.0005831646772915651,
+      "loss": 2.541,
+      "step": 260
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.14252015948295593,
+      "learning_rate": 0.0005530699351157809,
+      "loss": 2.5206,
+      "step": 270
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 0.13033610582351685,
+      "learning_rate": 0.0005227792933277943,
+      "loss": 2.5219,
+      "step": 280
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 0.12918302416801453,
+      "learning_rate": 0.0004924045652352092,
+      "loss": 2.5133,
+      "step": 290
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 0.11928386241197586,
+      "learning_rate": 0.0004620578745374633,
+      "loss": 2.5079,
+      "step": 300
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 0.1346321702003479,
+      "learning_rate": 0.0004318512414382059,
+      "loss": 2.5143,
+      "step": 310
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 0.11825684458017349,
+      "learning_rate": 0.0004018961691397148,
+      "loss": 2.5085,
+      "step": 320
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 0.1431489735841751,
+      "learning_rate": 0.0003723032322457458,
+      "loss": 2.4943,
+      "step": 330
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.10054672509431839,
+      "learning_rate": 0.0003431816685921625,
+      "loss": 2.4985,
+      "step": 340
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 0.13861466944217682,
+      "learning_rate": 0.0003146389760120416,
+      "loss": 2.4951,
+      "step": 350
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 0.11972185969352722,
+      "learning_rate": 0.00028678051552373485,
+      "loss": 2.4978,
+      "step": 360
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 0.10848913341760635,
+      "learning_rate": 0.0002597091224066581,
+      "loss": 2.4895,
+      "step": 370
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 0.12226279824972153,
+      "learning_rate": 0.0002335247266004629,
+      "loss": 2.4971,
+      "step": 380
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 0.12621009349822998,
+      "learning_rate": 0.000208323983828828,
+      "loss": 2.487,
+      "step": 390
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.1062069684267044,
+      "learning_rate": 0.0001841999188095224,
+      "loss": 2.4804,
+      "step": 400
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 0.1101030558347702,
+      "learning_rate": 0.00016124158186777676,
+      "loss": 2.482,
+      "step": 410
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 0.10745077580213547,
+      "learning_rate": 0.0001395337202205211,
+      "loss": 2.4786,
+      "step": 420
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 0.10573945194482803,
+      "learning_rate": 0.00011915646514489292,
+      "loss": 2.4771,
+      "step": 430
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 0.11785120517015457,
+      "learning_rate": 0.00010018503618578818,
+      "loss": 2.4861,
+      "step": 440
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.11175502091646194,
+      "learning_rate": 8.268946349432583e-05,
+      "loss": 2.4776,
+      "step": 450
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 0.10645473003387451,
+      "learning_rate": 6.673432932217117e-05,
+      "loss": 2.4725,
+      "step": 460
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.11124389618635178,
+      "learning_rate": 5.2378529625954684e-05,
+      "loss": 2.4627,
+      "step": 470
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.11082123965024948,
+      "learning_rate": 3.9675056661785556e-05,
+      "loss": 2.4652,
+      "step": 480
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.10290926694869995,
+      "learning_rate": 2.8670803372379538e-05,
+      "loss": 2.4739,
+      "step": 490
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 0.10559625178575516,
+      "learning_rate": 1.9406390288876586e-05,
+      "loss": 2.4799,
+      "step": 500
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 0.1101013645529747,
+      "learning_rate": 1.191601558631461e-05,
+      "loss": 2.4721,
+      "step": 510
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 0.10641390830278397,
+      "learning_rate": 6.2273288462559305e-06,
+      "loss": 2.4688,
+      "step": 520
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 0.10214255005121231,
+      "learning_rate": 2.361328992552314e-06,
+      "loss": 2.4698,
+      "step": 530
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 0.10081050544977188,
+      "learning_rate": 3.322867770044202e-07,
+      "loss": 2.4668,
+      "step": 540
+    },
+    {
+      "epoch": 1.0,
+      "step": 545,
+      "total_flos": 3.2179952270966784e+17,
+      "train_loss": 2.6759143199395696,
+      "train_runtime": 5301.6675,
+      "train_samples_per_second": 105.274,
+      "train_steps_per_second": 0.103
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 545,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.2179952270966784e+17,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen25/mc2_m16_224/finetune/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

continuous/qwen25/mc2_m16_224/finetune/config.json

@@ -0,0 +1,57 @@
+{
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 21,
+  "mm_hidden_size": 512,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "flat",
+  "mm_projector_lr": 0.0001,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_tunable_parts": "mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "open_clip_hub:ViT-M-16",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen2",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 32768,
+  "tie_word_embeddings": true,
+  "tokenizer_model_max_length": 8192,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.3",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_tower_pretrained": "/cache/ckpt/download/tokenizer/continuous/mc2/mc2_m16_224.pt",
+  "vocab_size": 151936
+}

continuous/qwen25/mc2_m16_224/finetune/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.1,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.51.3"
+}

continuous/qwen25/mc2_m16_224/finetune/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41954a40088bb4dee18d9c120e1b2f444df366156477b938b644ac34351b44a3
+size 3644007016

continuous/qwen25/mc2_m16_224/finetune/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_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|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

continuous/qwen25/mc2_m16_224/finetune/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

continuous/qwen25/mc2_m16_224/finetune/tokenizer_config.json

@@ -0,0 +1,209 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_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|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# 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>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\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\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 8192,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

continuous/qwen25/mc2_m16_224/finetune/trainer_state.json

@@ -0,0 +1,1856 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999037443449803,
+  "eval_steps": 500,
+  "global_step": 2597,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003850226200789296,
+      "grad_norm": 16.288795471191406,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 2.1761,
+      "step": 10
+    },
+    {
+      "epoch": 0.007700452401578592,
+      "grad_norm": 11.511601448059082,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 1.8153,
+      "step": 20
+    },
+    {
+      "epoch": 0.01155067860236789,
+      "grad_norm": 3.9250426292419434,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.2468,
+      "step": 30
+    },
+    {
+      "epoch": 0.015400904803157185,
+      "grad_norm": 0.8525418043136597,
+      "learning_rate": 5e-06,
+      "loss": 0.9727,
+      "step": 40
+    },
+    {
+      "epoch": 0.019251131003946483,
+      "grad_norm": 0.440680593252182,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 0.9193,
+      "step": 50
+    },
+    {
+      "epoch": 0.02310135720473578,
+      "grad_norm": 0.2900700271129608,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 0.8691,
+      "step": 60
+    },
+    {
+      "epoch": 0.026951583405525073,
+      "grad_norm": 0.2321665734052658,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.8413,
+      "step": 70
+    },
+    {
+      "epoch": 0.03080180960631437,
+      "grad_norm": 0.23111803829669952,
+      "learning_rate": 9.999996111488723e-06,
+      "loss": 0.8488,
+      "step": 80
+    },
+    {
+      "epoch": 0.034652035807103666,
+      "grad_norm": 0.21423622965812683,
+      "learning_rate": 9.999529497453782e-06,
+      "loss": 0.8292,
+      "step": 90
+    },
+    {
+      "epoch": 0.038502262007892966,
+      "grad_norm": 0.2186075747013092,
+      "learning_rate": 9.998285264324193e-06,
+      "loss": 0.8084,
+      "step": 100
+    },
+    {
+      "epoch": 0.04235248820868226,
+      "grad_norm": 0.1999548077583313,
+      "learning_rate": 9.996263605626052e-06,
+      "loss": 0.7989,
+      "step": 110
+    },
+    {
+      "epoch": 0.04620271440947156,
+      "grad_norm": 0.2237752228975296,
+      "learning_rate": 9.99346483580503e-06,
+      "loss": 0.8102,
+      "step": 120
+    },
+    {
+      "epoch": 0.05005294061026085,
+      "grad_norm": 0.20301717519760132,
+      "learning_rate": 9.989889390177467e-06,
+      "loss": 0.7991,
+      "step": 130
+    },
+    {
+      "epoch": 0.053903166811050146,
+      "grad_norm": 0.23235280811786652,
+      "learning_rate": 9.985537824862647e-06,
+      "loss": 0.8014,
+      "step": 140
+    },
+    {
+      "epoch": 0.057753393011839446,
+      "grad_norm": 0.23169316351413727,
+      "learning_rate": 9.980410816696317e-06,
+      "loss": 0.7828,
+      "step": 150
+    },
+    {
+      "epoch": 0.06160361921262874,
+      "grad_norm": 0.21564103662967682,
+      "learning_rate": 9.97450916312541e-06,
+      "loss": 0.7924,
+      "step": 160
+    },
+    {
+      "epoch": 0.06545384541341803,
+      "grad_norm": 0.19768257439136505,
+      "learning_rate": 9.967833782084004e-06,
+      "loss": 0.793,
+      "step": 170
+    },
+    {
+      "epoch": 0.06930407161420733,
+      "grad_norm": 0.23142103850841522,
+      "learning_rate": 9.960385711850554e-06,
+      "loss": 0.8177,
+      "step": 180
+    },
+    {
+      "epoch": 0.07315429781499663,
+      "grad_norm": 0.22244900465011597,
+      "learning_rate": 9.952166110886394e-06,
+      "loss": 0.8105,
+      "step": 190
+    },
+    {
+      "epoch": 0.07700452401578593,
+      "grad_norm": 0.2380964457988739,
+      "learning_rate": 9.943176257655567e-06,
+      "loss": 0.8017,
+      "step": 200
+    },
+    {
+      "epoch": 0.08085475021657522,
+      "grad_norm": 0.21020683646202087,
+      "learning_rate": 9.933417550425952e-06,
+      "loss": 0.7954,
+      "step": 210
+    },
+    {
+      "epoch": 0.08470497641736452,
+      "grad_norm": 0.21145296096801758,
+      "learning_rate": 9.9228915070518e-06,
+      "loss": 0.7813,
+      "step": 220
+    },
+    {
+      "epoch": 0.08855520261815382,
+      "grad_norm": 0.2158515751361847,
+      "learning_rate": 9.911599764737632e-06,
+      "loss": 0.7848,
+      "step": 230
+    },
+    {
+      "epoch": 0.09240542881894312,
+      "grad_norm": 0.22187089920043945,
+      "learning_rate": 9.899544079783606e-06,
+      "loss": 0.7774,
+      "step": 240
+    },
+    {
+      "epoch": 0.0962556550197324,
+      "grad_norm": 0.22201354801654816,
+      "learning_rate": 9.886726327312338e-06,
+      "loss": 0.7943,
+      "step": 250
+    },
+    {
+      "epoch": 0.1001058812205217,
+      "grad_norm": 0.21093527972698212,
+      "learning_rate": 9.87314850097724e-06,
+      "loss": 0.7916,
+      "step": 260
+    },
+    {
+      "epoch": 0.103956107421311,
+      "grad_norm": 0.2035299390554428,
+      "learning_rate": 9.858812712652452e-06,
+      "loss": 0.7819,
+      "step": 270
+    },
+    {
+      "epoch": 0.10780633362210029,
+      "grad_norm": 0.20283955335617065,
+      "learning_rate": 9.843721192104341e-06,
+      "loss": 0.7805,
+      "step": 280
+    },
+    {
+      "epoch": 0.11165655982288959,
+      "grad_norm": 0.2071700245141983,
+      "learning_rate": 9.827876286644698e-06,
+      "loss": 0.8011,
+      "step": 290
+    },
+    {
+      "epoch": 0.11550678602367889,
+      "grad_norm": 0.1991768330335617,
+      "learning_rate": 9.811280460765643e-06,
+      "loss": 0.7628,
+      "step": 300
+    },
+    {
+      "epoch": 0.11935701222446819,
+      "grad_norm": 0.20873670279979706,
+      "learning_rate": 9.793936295756292e-06,
+      "loss": 0.7717,
+      "step": 310
+    },
+    {
+      "epoch": 0.12320723842525748,
+      "grad_norm": 0.21808387339115143,
+      "learning_rate": 9.775846489301279e-06,
+      "loss": 0.7767,
+      "step": 320
+    },
+    {
+      "epoch": 0.12705746462604678,
+      "grad_norm": 0.21633084118366241,
+      "learning_rate": 9.757013855061157e-06,
+      "loss": 0.7741,
+      "step": 330
+    },
+    {
+      "epoch": 0.13090769082683607,
+      "grad_norm": 0.211192324757576,
+      "learning_rate": 9.737441322234758e-06,
+      "loss": 0.7878,
+      "step": 340
+    },
+    {
+      "epoch": 0.13475791702762538,
+      "grad_norm": 0.22047904133796692,
+      "learning_rate": 9.717131935103605e-06,
+      "loss": 0.7867,
+      "step": 350
+    },
+    {
+      "epoch": 0.13860814322841467,
+      "grad_norm": 0.21744397282600403,
+      "learning_rate": 9.696088852558398e-06,
+      "loss": 0.7726,
+      "step": 360
+    },
+    {
+      "epoch": 0.14245836942920398,
+      "grad_norm": 0.20912770926952362,
+      "learning_rate": 9.674315347607686e-06,
+      "loss": 0.7988,
+      "step": 370
+    },
+    {
+      "epoch": 0.14630859562999327,
+      "grad_norm": 0.2104756236076355,
+      "learning_rate": 9.651814806868796e-06,
+      "loss": 0.7646,
+      "step": 380
+    },
+    {
+      "epoch": 0.15015882183078255,
+      "grad_norm": 0.22212260961532593,
+      "learning_rate": 9.62859073004107e-06,
+      "loss": 0.7661,
+      "step": 390
+    },
+    {
+      "epoch": 0.15400904803157187,
+      "grad_norm": 0.20146134495735168,
+      "learning_rate": 9.604646729361547e-06,
+      "loss": 0.7714,
+      "step": 400
+    },
+    {
+      "epoch": 0.15785927423236115,
+      "grad_norm": 0.2030630260705948,
+      "learning_rate": 9.579986529043099e-06,
+      "loss": 0.762,
+      "step": 410
+    },
+    {
+      "epoch": 0.16170950043315044,
+      "grad_norm": 0.21657682955265045,
+      "learning_rate": 9.554613964695189e-06,
+      "loss": 0.7784,
+      "step": 420
+    },
+    {
+      "epoch": 0.16555972663393975,
+      "grad_norm": 0.21809349954128265,
+      "learning_rate": 9.528532982727281e-06,
+      "loss": 0.771,
+      "step": 430
+    },
+    {
+      "epoch": 0.16940995283472904,
+      "grad_norm": 0.19967098534107208,
+      "learning_rate": 9.501747639735018e-06,
+      "loss": 0.7673,
+      "step": 440
+    },
+    {
+      "epoch": 0.17326017903551832,
+      "grad_norm": 0.2111966907978058,
+      "learning_rate": 9.474262101869272e-06,
+      "loss": 0.7562,
+      "step": 450
+    },
+    {
+      "epoch": 0.17711040523630764,
+      "grad_norm": 0.2139514833688736,
+      "learning_rate": 9.446080644188141e-06,
+      "loss": 0.763,
+      "step": 460
+    },
+    {
+      "epoch": 0.18096063143709692,
+      "grad_norm": 0.20092251896858215,
+      "learning_rate": 9.417207649992017e-06,
+      "loss": 0.7683,
+      "step": 470
+    },
+    {
+      "epoch": 0.18481085763788624,
+      "grad_norm": 0.20846688747406006,
+      "learning_rate": 9.387647610141816e-06,
+      "loss": 0.7673,
+      "step": 480
+    },
+    {
+      "epoch": 0.18866108383867553,
+      "grad_norm": 0.19388310611248016,
+      "learning_rate": 9.35740512236047e-06,
+      "loss": 0.7534,
+      "step": 490
+    },
+    {
+      "epoch": 0.1925113100394648,
+      "grad_norm": 0.1973726749420166,
+      "learning_rate": 9.326484890517809e-06,
+      "loss": 0.7658,
+      "step": 500
+    },
+    {
+      "epoch": 0.19636153624025413,
+      "grad_norm": 0.2133578658103943,
+      "learning_rate": 9.294891723898926e-06,
+      "loss": 0.7551,
+      "step": 510
+    },
+    {
+      "epoch": 0.2002117624410434,
+      "grad_norm": 0.2236829549074173,
+      "learning_rate": 9.262630536456153e-06,
+      "loss": 0.7846,
+      "step": 520
+    },
+    {
+      "epoch": 0.2040619886418327,
+      "grad_norm": 0.22007854282855988,
+      "learning_rate": 9.229706346044749e-06,
+      "loss": 0.7587,
+      "step": 530
+    },
+    {
+      "epoch": 0.207912214842622,
+      "grad_norm": 0.22696907818317413,
+      "learning_rate": 9.196124273642422e-06,
+      "loss": 0.7703,
+      "step": 540
+    },
+    {
+      "epoch": 0.2117624410434113,
+      "grad_norm": 0.20491227507591248,
+      "learning_rate": 9.161889542552842e-06,
+      "loss": 0.7755,
+      "step": 550
+    },
+    {
+      "epoch": 0.21561266724420058,
+      "grad_norm": 0.20329317450523376,
+      "learning_rate": 9.127007477593193e-06,
+      "loss": 0.7763,
+      "step": 560
+    },
+    {
+      "epoch": 0.2194628934449899,
+      "grad_norm": 0.22808435559272766,
+      "learning_rate": 9.091483504265974e-06,
+      "loss": 0.7857,
+      "step": 570
+    },
+    {
+      "epoch": 0.22331311964577918,
+      "grad_norm": 0.21905148029327393,
+      "learning_rate": 9.055323147915116e-06,
+      "loss": 0.7658,
+      "step": 580
+    },
+    {
+      "epoch": 0.2271633458465685,
+      "grad_norm": 0.21039220690727234,
+      "learning_rate": 9.018532032866585e-06,
+      "loss": 0.7755,
+      "step": 590
+    },
+    {
+      "epoch": 0.23101357204735778,
+      "grad_norm": 0.1985267549753189,
+      "learning_rate": 8.981115881553579e-06,
+      "loss": 0.7679,
+      "step": 600
+    },
+    {
+      "epoch": 0.23486379824814707,
+      "grad_norm": 0.22400768101215363,
+      "learning_rate": 8.943080513626476e-06,
+      "loss": 0.7886,
+      "step": 610
+    },
+    {
+      "epoch": 0.23871402444893638,
+      "grad_norm": 0.21560271084308624,
+      "learning_rate": 8.904431845047641e-06,
+      "loss": 0.7725,
+      "step": 620
+    },
+    {
+      "epoch": 0.24256425064972567,
+      "grad_norm": 0.22577184438705444,
+      "learning_rate": 8.86517588717128e-06,
+      "loss": 0.7889,
+      "step": 630
+    },
+    {
+      "epoch": 0.24641447685051496,
+      "grad_norm": 0.2039911448955536,
+      "learning_rate": 8.82531874580844e-06,
+      "loss": 0.7807,
+      "step": 640
+    },
+    {
+      "epoch": 0.25026470305130427,
+      "grad_norm": 0.18994522094726562,
+      "learning_rate": 8.784866620277314e-06,
+      "loss": 0.7385,
+      "step": 650
+    },
+    {
+      "epoch": 0.25411492925209356,
+      "grad_norm": 0.22563709318637848,
+      "learning_rate": 8.743825802439018e-06,
+      "loss": 0.7701,
+      "step": 660
+    },
+    {
+      "epoch": 0.25796515545288284,
+      "grad_norm": 0.23138193786144257,
+      "learning_rate": 8.702202675718956e-06,
+      "loss": 0.7785,
+      "step": 670
+    },
+    {
+      "epoch": 0.26181538165367213,
+      "grad_norm": 0.20282208919525146,
+      "learning_rate": 8.660003714113959e-06,
+      "loss": 0.7468,
+      "step": 680
+    },
+    {
+      "epoch": 0.26566560785446147,
+      "grad_norm": 0.21072335541248322,
+      "learning_rate": 8.617235481185323e-06,
+      "loss": 0.7684,
+      "step": 690
+    },
+    {
+      "epoch": 0.26951583405525076,
+      "grad_norm": 0.2123950868844986,
+      "learning_rate": 8.573904629037931e-06,
+      "loss": 0.7752,
+      "step": 700
+    },
+    {
+      "epoch": 0.27336606025604004,
+      "grad_norm": 0.20947079360485077,
+      "learning_rate": 8.530017897285585e-06,
+      "loss": 0.7582,
+      "step": 710
+    },
+    {
+      "epoch": 0.27721628645682933,
+      "grad_norm": 0.1925039291381836,
+      "learning_rate": 8.485582112002746e-06,
+      "loss": 0.763,
+      "step": 720
+    },
+    {
+      "epoch": 0.2810665126576186,
+      "grad_norm": 0.20791536569595337,
+      "learning_rate": 8.440604184662811e-06,
+      "loss": 0.7654,
+      "step": 730
+    },
+    {
+      "epoch": 0.28491673885840796,
+      "grad_norm": 0.2156745046377182,
+      "learning_rate": 8.395091111063118e-06,
+      "loss": 0.756,
+      "step": 740
+    },
+    {
+      "epoch": 0.28876696505919724,
+      "grad_norm": 0.19460701942443848,
+      "learning_rate": 8.349049970236822e-06,
+      "loss": 0.7545,
+      "step": 750
+    },
+    {
+      "epoch": 0.29261719125998653,
+      "grad_norm": 0.20894278585910797,
+      "learning_rate": 8.302487923351842e-06,
+      "loss": 0.7475,
+      "step": 760
+    },
+    {
+      "epoch": 0.2964674174607758,
+      "grad_norm": 0.20247431099414825,
+      "learning_rate": 8.255412212597021e-06,
+      "loss": 0.7651,
+      "step": 770
+    },
+    {
+      "epoch": 0.3003176436615651,
+      "grad_norm": 0.2028290033340454,
+      "learning_rate": 8.207830160055688e-06,
+      "loss": 0.7627,
+      "step": 780
+    },
+    {
+      "epoch": 0.3041678698623544,
+      "grad_norm": 0.21736489236354828,
+      "learning_rate": 8.159749166566786e-06,
+      "loss": 0.7482,
+      "step": 790
+    },
+    {
+      "epoch": 0.30801809606314373,
+      "grad_norm": 0.20517832040786743,
+      "learning_rate": 8.11117671057377e-06,
+      "loss": 0.7758,
+      "step": 800
+    },
+    {
+      "epoch": 0.311868322263933,
+      "grad_norm": 0.219059020280838,
+      "learning_rate": 8.062120346961404e-06,
+      "loss": 0.7666,
+      "step": 810
+    },
+    {
+      "epoch": 0.3157185484647223,
+      "grad_norm": 0.19764654338359833,
+      "learning_rate": 8.012587705880705e-06,
+      "loss": 0.7623,
+      "step": 820
+    },
+    {
+      "epoch": 0.3195687746655116,
+      "grad_norm": 0.2051871120929718,
+      "learning_rate": 7.962586491562146e-06,
+      "loss": 0.746,
+      "step": 830
+    },
+    {
+      "epoch": 0.3234190008663009,
+      "grad_norm": 0.22835248708724976,
+      "learning_rate": 7.912124481117357e-06,
+      "loss": 0.7524,
+      "step": 840
+    },
+    {
+      "epoch": 0.3272692270670902,
+      "grad_norm": 0.20835348963737488,
+      "learning_rate": 7.861209523329494e-06,
+      "loss": 0.7646,
+      "step": 850
+    },
+    {
+      "epoch": 0.3311194532678795,
+      "grad_norm": 0.21009822189807892,
+      "learning_rate": 7.809849537432432e-06,
+      "loss": 0.7481,
+      "step": 860
+    },
+    {
+      "epoch": 0.3349696794686688,
+      "grad_norm": 0.20456050336360931,
+      "learning_rate": 7.758052511879035e-06,
+      "loss": 0.7591,
+      "step": 870
+    },
+    {
+      "epoch": 0.3388199056694581,
+      "grad_norm": 0.19596132636070251,
+      "learning_rate": 7.705826503098642e-06,
+      "loss": 0.7531,
+      "step": 880
+    },
+    {
+      "epoch": 0.34267013187024736,
+      "grad_norm": 0.19885829091072083,
+      "learning_rate": 7.653179634243979e-06,
+      "loss": 0.7531,
+      "step": 890
+    },
+    {
+      "epoch": 0.34652035807103665,
+      "grad_norm": 0.18923085927963257,
+      "learning_rate": 7.600120093927696e-06,
+      "loss": 0.7567,
+      "step": 900
+    },
+    {
+      "epoch": 0.350370584271826,
+      "grad_norm": 0.21729962527751923,
+      "learning_rate": 7.546656134948729e-06,
+      "loss": 0.7418,
+      "step": 910
+    },
+    {
+      "epoch": 0.3542208104726153,
+      "grad_norm": 0.21061204373836517,
+      "learning_rate": 7.492796073008661e-06,
+      "loss": 0.7452,
+      "step": 920
+    },
+    {
+      "epoch": 0.35807103667340456,
+      "grad_norm": 0.21927441656589508,
+      "learning_rate": 7.438548285418326e-06,
+      "loss": 0.7339,
+      "step": 930
+    },
+    {
+      "epoch": 0.36192126287419385,
+      "grad_norm": 0.22907552123069763,
+      "learning_rate": 7.3839212097948e-06,
+      "loss": 0.7667,
+      "step": 940
+    },
+    {
+      "epoch": 0.36577148907498314,
+      "grad_norm": 0.21692071855068207,
+      "learning_rate": 7.328923342749039e-06,
+      "loss": 0.7814,
+      "step": 950
+    },
+    {
+      "epoch": 0.3696217152757725,
+      "grad_norm": 0.22222675383090973,
+      "learning_rate": 7.273563238564319e-06,
+      "loss": 0.7681,
+      "step": 960
+    },
+    {
+      "epoch": 0.37347194147656176,
+      "grad_norm": 0.206553652882576,
+      "learning_rate": 7.217849507865724e-06,
+      "loss": 0.7464,
+      "step": 970
+    },
+    {
+      "epoch": 0.37732216767735105,
+      "grad_norm": 0.20847830176353455,
+      "learning_rate": 7.161790816280858e-06,
+      "loss": 0.7528,
+      "step": 980
+    },
+    {
+      "epoch": 0.38117239387814034,
+      "grad_norm": 0.2141447514295578,
+      "learning_rate": 7.1053958830920125e-06,
+      "loss": 0.7339,
+      "step": 990
+    },
+    {
+      "epoch": 0.3850226200789296,
+      "grad_norm": 0.20676593482494354,
+      "learning_rate": 7.0486734798799715e-06,
+      "loss": 0.7525,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3888728462797189,
+      "grad_norm": 0.2014734148979187,
+      "learning_rate": 6.991632429159708e-06,
+      "loss": 0.7767,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39272307248050825,
+      "grad_norm": 0.22567221522331238,
+      "learning_rate": 6.934281603008132e-06,
+      "loss": 0.7686,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39657329868129754,
+      "grad_norm": 0.2050096094608307,
+      "learning_rate": 6.876629921684143e-06,
+      "loss": 0.7659,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4004235248820868,
+      "grad_norm": 0.21002596616744995,
+      "learning_rate": 6.818686352241199e-06,
+      "loss": 0.7566,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4042737510828761,
+      "grad_norm": 0.1931946724653244,
+      "learning_rate": 6.760459907132573e-06,
+      "loss": 0.753,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4081239772836654,
+      "grad_norm": 0.19158756732940674,
+      "learning_rate": 6.701959642809589e-06,
+      "loss": 0.7581,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41197420348445474,
+      "grad_norm": 0.21262875199317932,
+      "learning_rate": 6.643194658312987e-06,
+      "loss": 0.748,
+      "step": 1070
+    },
+    {
+      "epoch": 0.415824429685244,
+      "grad_norm": 0.2017115205526352,
+      "learning_rate": 6.584174093857676e-06,
+      "loss": 0.7722,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4196746558860333,
+      "grad_norm": 0.2277713268995285,
+      "learning_rate": 6.5249071294110735e-06,
+      "loss": 0.7571,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4235248820868226,
+      "grad_norm": 0.21170935034751892,
+      "learning_rate": 6.465402983265274e-06,
+      "loss": 0.7717,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4273751082876119,
+      "grad_norm": 0.21431288123130798,
+      "learning_rate": 6.405670910603243e-06,
+      "loss": 0.767,
+      "step": 1110
+    },
+    {
+      "epoch": 0.43122533448840117,
+      "grad_norm": 0.22440817952156067,
+      "learning_rate": 6.345720202059291e-06,
+      "loss": 0.7622,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4350755606891905,
+      "grad_norm": 0.20508740842342377,
+      "learning_rate": 6.285560182274008e-06,
+      "loss": 0.7392,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4389257868899798,
+      "grad_norm": 0.2058320939540863,
+      "learning_rate": 6.225200208443936e-06,
+      "loss": 0.7612,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4427760130907691,
+      "grad_norm": 0.20070785284042358,
+      "learning_rate": 6.16464966886616e-06,
+      "loss": 0.7524,
+      "step": 1150
+    },
+    {
+      "epoch": 0.44662623929155837,
+      "grad_norm": 0.24046966433525085,
+      "learning_rate": 6.103917981478065e-06,
+      "loss": 0.7768,
+      "step": 1160
+    },
+    {
+      "epoch": 0.45047646549234766,
+      "grad_norm": 0.20911650359630585,
+      "learning_rate": 6.043014592392489e-06,
+      "loss": 0.7871,
+      "step": 1170
+    },
+    {
+      "epoch": 0.454326691693137,
+      "grad_norm": 0.19310633838176727,
+      "learning_rate": 5.981948974428485e-06,
+      "loss": 0.7603,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4581769178939263,
+      "grad_norm": 0.21135659515857697,
+      "learning_rate": 5.920730625637934e-06,
+      "loss": 0.7394,
+      "step": 1190
+    },
+    {
+      "epoch": 0.46202714409471557,
+      "grad_norm": 0.21752288937568665,
+      "learning_rate": 5.8593690678282355e-06,
+      "loss": 0.7534,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46587737029550486,
+      "grad_norm": 0.21521420776844025,
+      "learning_rate": 5.797873845081301e-06,
+      "loss": 0.7557,
+      "step": 1210
+    },
+    {
+      "epoch": 0.46972759649629414,
+      "grad_norm": 0.21382513642311096,
+      "learning_rate": 5.736254522269075e-06,
+      "loss": 0.7716,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47357782269708343,
+      "grad_norm": 0.20670993626117706,
+      "learning_rate": 5.6745206835658386e-06,
+      "loss": 0.7758,
+      "step": 1230
+    },
+    {
+      "epoch": 0.47742804889787277,
+      "grad_norm": 0.19914288818836212,
+      "learning_rate": 5.612681930957495e-06,
+      "loss": 0.7446,
+      "step": 1240
+    },
+    {
+      "epoch": 0.48127827509866206,
+      "grad_norm": 0.1923430860042572,
+      "learning_rate": 5.550747882748095e-06,
+      "loss": 0.7424,
+      "step": 1250
+    },
+    {
+      "epoch": 0.48512850129945134,
+      "grad_norm": 0.2222111076116562,
+      "learning_rate": 5.4887281720638195e-06,
+      "loss": 0.7524,
+      "step": 1260
+    },
+    {
+      "epoch": 0.48897872750024063,
+      "grad_norm": 0.23338721692562103,
+      "learning_rate": 5.42663244535466e-06,
+      "loss": 0.7337,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4928289537010299,
+      "grad_norm": 0.21165338158607483,
+      "learning_rate": 5.3644703608940205e-06,
+      "loss": 0.7433,
+      "step": 1280
+    },
+    {
+      "epoch": 0.49667917990181926,
+      "grad_norm": 0.2127082496881485,
+      "learning_rate": 5.302251587276491e-06,
+      "loss": 0.7538,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5005294061026085,
+      "grad_norm": 0.19939866662025452,
+      "learning_rate": 5.2399858019140005e-06,
+      "loss": 0.7554,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5043796323033978,
+      "grad_norm": 0.2039617896080017,
+      "learning_rate": 5.177682689530615e-06,
+      "loss": 0.767,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5082298585041871,
+      "grad_norm": 0.218671515583992,
+      "learning_rate": 5.11535194065619e-06,
+      "loss": 0.7391,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5120800847049765,
+      "grad_norm": 0.2246018499135971,
+      "learning_rate": 5.053003250119113e-06,
+      "loss": 0.7463,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5159303109057657,
+      "grad_norm": 0.21361508965492249,
+      "learning_rate": 4.9906463155383925e-06,
+      "loss": 0.7538,
+      "step": 1340
+    },
+    {
+      "epoch": 0.519780537106555,
+      "grad_norm": 0.2089388072490692,
+      "learning_rate": 4.9282908358153046e-06,
+      "loss": 0.7486,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5236307633073443,
+      "grad_norm": 0.23477107286453247,
+      "learning_rate": 4.8659465096248345e-06,
+      "loss": 0.764,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5274809895081336,
+      "grad_norm": 0.21296969056129456,
+      "learning_rate": 4.803623033907169e-06,
+      "loss": 0.7607,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5313312157089229,
+      "grad_norm": 0.1997753083705902,
+      "learning_rate": 4.741330102359441e-06,
+      "loss": 0.7689,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5351814419097122,
+      "grad_norm": 0.22383525967597961,
+      "learning_rate": 4.679077403927996e-06,
+      "loss": 0.7483,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5390316681105015,
+      "grad_norm": 0.21045644581317902,
+      "learning_rate": 4.616874621301373e-06,
+      "loss": 0.7492,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5428818943112907,
+      "grad_norm": 0.21369719505310059,
+      "learning_rate": 4.554731429404293e-06,
+      "loss": 0.7631,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5467321205120801,
+      "grad_norm": 0.22261664271354675,
+      "learning_rate": 4.492657493892822e-06,
+      "loss": 0.7465,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5505823467128694,
+      "grad_norm": 0.2256411910057068,
+      "learning_rate": 4.430662469650999e-06,
+      "loss": 0.763,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5544325729136587,
+      "grad_norm": 0.21729658544063568,
+      "learning_rate": 4.368755999289118e-06,
+      "loss": 0.7675,
+      "step": 1440
+    },
+    {
+      "epoch": 0.558282799114448,
+      "grad_norm": 0.2179233431816101,
+      "learning_rate": 4.306947711643949e-06,
+      "loss": 0.747,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5621330253152372,
+      "grad_norm": 0.21981894969940186,
+      "learning_rate": 4.245247220281068e-06,
+      "loss": 0.7541,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5659832515160266,
+      "grad_norm": 0.20603090524673462,
+      "learning_rate": 4.183664121999587e-06,
+      "loss": 0.7493,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5698334777168159,
+      "grad_norm": 0.22008609771728516,
+      "learning_rate": 4.122207995339474e-06,
+      "loss": 0.7469,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5736837039176051,
+      "grad_norm": 0.2020481675863266,
+      "learning_rate": 4.060888399091734e-06,
+      "loss": 0.7521,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5775339301183945,
+      "grad_norm": 0.2006344199180603,
+      "learning_rate": 3.999714870811639e-06,
+      "loss": 0.7628,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5813841563191837,
+      "grad_norm": 0.2121267169713974,
+      "learning_rate": 3.938696925335275e-06,
+      "loss": 0.7553,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5852343825199731,
+      "grad_norm": 0.2053990364074707,
+      "learning_rate": 3.87784405329962e-06,
+      "loss": 0.7637,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5890846087207623,
+      "grad_norm": 0.21236294507980347,
+      "learning_rate": 3.817165719666382e-06,
+      "loss": 0.764,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5929348349215516,
+      "grad_norm": 0.20262780785560608,
+      "learning_rate": 3.756671362249841e-06,
+      "loss": 0.7698,
+      "step": 1540
+    },
+    {
+      "epoch": 0.596785061122341,
+      "grad_norm": 0.2086733877658844,
+      "learning_rate": 3.6963703902488907e-06,
+      "loss": 0.756,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6006352873231302,
+      "grad_norm": 0.2092708796262741,
+      "learning_rate": 3.6362721827835655e-06,
+      "loss": 0.7702,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6044855135239195,
+      "grad_norm": 0.21840761601924896,
+      "learning_rate": 3.576386087436212e-06,
+      "loss": 0.7673,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6083357397247088,
+      "grad_norm": 0.22421573102474213,
+      "learning_rate": 3.5167214187975913e-06,
+      "loss": 0.7577,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6121859659254981,
+      "grad_norm": 0.18610623478889465,
+      "learning_rate": 3.457287457018086e-06,
+      "loss": 0.7592,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6160361921262875,
+      "grad_norm": 0.1981368362903595,
+      "learning_rate": 3.3980934463643003e-06,
+      "loss": 0.7481,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6198864183270767,
+      "grad_norm": 0.1967846155166626,
+      "learning_rate": 3.3391485937812053e-06,
+      "loss": 0.7303,
+      "step": 1610
+    },
+    {
+      "epoch": 0.623736644527866,
+      "grad_norm": 0.2125241458415985,
+      "learning_rate": 3.280462067460115e-06,
+      "loss": 0.7286,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6275868707286553,
+      "grad_norm": 0.2055991142988205,
+      "learning_rate": 3.222042995412669e-06,
+      "loss": 0.7607,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6314370969294446,
+      "grad_norm": 0.18595252931118011,
+      "learning_rate": 3.163900464051089e-06,
+      "loss": 0.7491,
+      "step": 1640
+    },
+    {
+      "epoch": 0.635287323130234,
+      "grad_norm": 0.20913900434970856,
+      "learning_rate": 3.1060435167748826e-06,
+      "loss": 0.7607,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6391375493310232,
+      "grad_norm": 0.2088804543018341,
+      "learning_rate": 3.048481152564253e-06,
+      "loss": 0.7408,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6429877755318125,
+      "grad_norm": 0.22024321556091309,
+      "learning_rate": 2.9912223245804017e-06,
+      "loss": 0.7619,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6468380017326018,
+      "grad_norm": 0.2176813930273056,
+      "learning_rate": 2.934275938772981e-06,
+      "loss": 0.7555,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6506882279333911,
+      "grad_norm": 0.22187024354934692,
+      "learning_rate": 2.8776508524948676e-06,
+      "loss": 0.7469,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6545384541341804,
+      "grad_norm": 0.18795916438102722,
+      "learning_rate": 2.821355873124503e-06,
+      "loss": 0.7448,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6583886803349697,
+      "grad_norm": 0.2104453295469284,
+      "learning_rate": 2.7653997566960144e-06,
+      "loss": 0.7585,
+      "step": 1710
+    },
+    {
+      "epoch": 0.662238906535759,
+      "grad_norm": 0.2132103592157364,
+      "learning_rate": 2.709791206537309e-06,
+      "loss": 0.7474,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6660891327365482,
+      "grad_norm": 0.21366068720817566,
+      "learning_rate": 2.6545388719163747e-06,
+      "loss": 0.7519,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6699393589373376,
+      "grad_norm": 0.21275867521762848,
+      "learning_rate": 2.599651346695979e-06,
+      "loss": 0.7527,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6737895851381268,
+      "grad_norm": 0.19982588291168213,
+      "learning_rate": 2.545137167997007e-06,
+      "loss": 0.7512,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6776398113389162,
+      "grad_norm": 0.19792386889457703,
+      "learning_rate": 2.4910048148705972e-06,
+      "loss": 0.7389,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6814900375397055,
+      "grad_norm": 0.21640290319919586,
+      "learning_rate": 2.4372627069793327e-06,
+      "loss": 0.7836,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6853402637404947,
+      "grad_norm": 0.22064004838466644,
+      "learning_rate": 2.3839192032876486e-06,
+      "loss": 0.7673,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6891904899412841,
+      "grad_norm": 0.20353646576404572,
+      "learning_rate": 2.3309826007617047e-06,
+      "loss": 0.779,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6930407161420733,
+      "grad_norm": 0.20568087697029114,
+      "learning_rate": 2.278461133078881e-06,
+      "loss": 0.7414,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6968909423428626,
+      "grad_norm": 0.2179674655199051,
+      "learning_rate": 2.2263629693471258e-06,
+      "loss": 0.7348,
+      "step": 1810
+    },
+    {
+      "epoch": 0.700741168543652,
+      "grad_norm": 0.20492053031921387,
+      "learning_rate": 2.1746962128343414e-06,
+      "loss": 0.755,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7045913947444412,
+      "grad_norm": 0.20481650531291962,
+      "learning_rate": 2.1234688997080183e-06,
+      "loss": 0.7459,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7084416209452306,
+      "grad_norm": 0.2210734486579895,
+      "learning_rate": 2.0726889977853053e-06,
+      "loss": 0.7526,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7122918471460198,
+      "grad_norm": 0.22121140360832214,
+      "learning_rate": 2.022364405293703e-06,
+      "loss": 0.763,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7161420733468091,
+      "grad_norm": 0.2089855670928955,
+      "learning_rate": 1.9725029496425814e-06,
+      "loss": 0.7582,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7199922995475985,
+      "grad_norm": 0.1928728073835373,
+      "learning_rate": 1.923112386205727e-06,
+      "loss": 0.753,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7238425257483877,
+      "grad_norm": 0.2032562494277954,
+      "learning_rate": 1.874200397115079e-06,
+      "loss": 0.7609,
+      "step": 1880
+    },
+    {
+      "epoch": 0.727692751949177,
+      "grad_norm": 0.219020277261734,
+      "learning_rate": 1.825774590065857e-06,
+      "loss": 0.7437,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7315429781499663,
+      "grad_norm": 0.22221846878528595,
+      "learning_rate": 1.7778424971332813e-06,
+      "loss": 0.7551,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7353932043507556,
+      "grad_norm": 0.19995063543319702,
+      "learning_rate": 1.7304115736010407e-06,
+      "loss": 0.7508,
+      "step": 1910
+    },
+    {
+      "epoch": 0.739243430551545,
+      "grad_norm": 0.2064635157585144,
+      "learning_rate": 1.6834891968017108e-06,
+      "loss": 0.7469,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7430936567523342,
+      "grad_norm": 0.2150709480047226,
+      "learning_rate": 1.6370826649692884e-06,
+      "loss": 0.7606,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7469438829531235,
+      "grad_norm": 0.22300992906093597,
+      "learning_rate": 1.5911991961040463e-06,
+      "loss": 0.7675,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7507941091539128,
+      "grad_norm": 0.21358713507652283,
+      "learning_rate": 1.545845926849851e-06,
+      "loss": 0.7554,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7546443353547021,
+      "grad_norm": 0.2264849990606308,
+      "learning_rate": 1.5010299113841397e-06,
+      "loss": 0.7727,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7584945615554913,
+      "grad_norm": 0.2113042175769806,
+      "learning_rate": 1.4567581203207238e-06,
+      "loss": 0.7633,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7623447877562807,
+      "grad_norm": 0.21814097464084625,
+      "learning_rate": 1.413037439625593e-06,
+      "loss": 0.7617,
+      "step": 1980
+    },
+    {
+      "epoch": 0.76619501395707,
+      "grad_norm": 0.22122417390346527,
+      "learning_rate": 1.3698746695458847e-06,
+      "loss": 0.7556,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7700452401578592,
+      "grad_norm": 0.20687873661518097,
+      "learning_rate": 1.3272765235521745e-06,
+      "loss": 0.7482,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7738954663586486,
+      "grad_norm": 0.20615042746067047,
+      "learning_rate": 1.2852496272942877e-06,
+      "loss": 0.7573,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7777456925594378,
+      "grad_norm": 0.1958792805671692,
+      "learning_rate": 1.2438005175707396e-06,
+      "loss": 0.7436,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7815959187602272,
+      "grad_norm": 0.20327842235565186,
+      "learning_rate": 1.2029356413120297e-06,
+      "loss": 0.7275,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7854461449610165,
+      "grad_norm": 0.20761027932167053,
+      "learning_rate": 1.1626613545778786e-06,
+      "loss": 0.7453,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7892963711618057,
+      "grad_norm": 0.19965431094169617,
+      "learning_rate": 1.1229839215686294e-06,
+      "loss": 0.7504,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7931465973625951,
+      "grad_norm": 0.1919877976179123,
+      "learning_rate": 1.0839095136509188e-06,
+      "loss": 0.7324,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7969968235633843,
+      "grad_norm": 0.2224104106426239,
+      "learning_rate": 1.045444208397791e-06,
+      "loss": 0.7548,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8008470497641736,
+      "grad_norm": 0.19875513017177582,
+      "learning_rate": 1.007593988643402e-06,
+      "loss": 0.74,
+      "step": 2080
+    },
+    {
+      "epoch": 0.804697275964963,
+      "grad_norm": 0.2057286500930786,
+      "learning_rate": 9.703647415524613e-07,
+      "loss": 0.7607,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8085475021657522,
+      "grad_norm": 0.20953015983104706,
+      "learning_rate": 9.337622577045485e-07,
+      "loss": 0.7489,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8123977283665416,
+      "grad_norm": 0.2101055383682251,
+      "learning_rate": 8.977922301934588e-07,
+      "loss": 0.7588,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8162479545673308,
+      "grad_norm": 0.20772676169872284,
+      "learning_rate": 8.624602537417009e-07,
+      "loss": 0.7596,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8200981807681201,
+      "grad_norm": 0.22364068031311035,
+      "learning_rate": 8.277718238303123e-07,
+      "loss": 0.7457,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8239484069689095,
+      "grad_norm": 0.21129044890403748,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.7503,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8277986331696987,
+      "grad_norm": 0.2233978509902954,
+      "learning_rate": 7.603470842324223e-07,
+      "loss": 0.7408,
+      "step": 2150
+    },
+    {
+      "epoch": 0.831648859370488,
+      "grad_norm": 0.20042496919631958,
+      "learning_rate": 7.276212616857575e-07,
+      "loss": 0.7589,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8354990855712773,
+      "grad_norm": 0.2062404602766037,
+      "learning_rate": 6.955599583279837e-07,
+      "loss": 0.7586,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8393493117720666,
+      "grad_norm": 0.2032700479030609,
+      "learning_rate": 6.641681609246981e-07,
+      "loss": 0.7502,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8431995379728559,
+      "grad_norm": 0.2017485648393631,
+      "learning_rate": 6.334507521075711e-07,
+      "loss": 0.7576,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8470497641736452,
+      "grad_norm": 0.21377256512641907,
+      "learning_rate": 6.034125096149179e-07,
+      "loss": 0.7537,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8508999903744345,
+      "grad_norm": 0.2112908959388733,
+      "learning_rate": 5.740581055485722e-07,
+      "loss": 0.7233,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8547502165752238,
+      "grad_norm": 0.19390273094177246,
+      "learning_rate": 5.453921056471961e-07,
+      "loss": 0.7563,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8586004427760131,
+      "grad_norm": 0.20678476989269257,
+      "learning_rate": 5.174189685761288e-07,
+      "loss": 0.7534,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8624506689768023,
+      "grad_norm": 0.1947387158870697,
+      "learning_rate": 4.901430452338979e-07,
+      "loss": 0.7632,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8663008951775917,
+      "grad_norm": 0.19909805059432983,
+      "learning_rate": 4.6356857807548286e-07,
+      "loss": 0.7475,
+      "step": 2250
+    },
+    {
+      "epoch": 0.870151121378381,
+      "grad_norm": 0.20846129953861237,
+      "learning_rate": 4.376997004524541e-07,
+      "loss": 0.7371,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8740013475791703,
+      "grad_norm": 0.20840491354465485,
+      "learning_rate": 4.12540435970073e-07,
+      "loss": 0.7443,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8778515737799596,
+      "grad_norm": 0.2092905193567276,
+      "learning_rate": 3.880946978614703e-07,
+      "loss": 0.7564,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8817017999807488,
+      "grad_norm": 0.20847876369953156,
+      "learning_rate": 3.643662883789878e-07,
+      "loss": 0.7587,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8855520261815382,
+      "grad_norm": 0.1944025158882141,
+      "learning_rate": 3.41358898202781e-07,
+      "loss": 0.7389,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8894022523823275,
+      "grad_norm": 0.19681933522224426,
+      "learning_rate": 3.190761058667741e-07,
+      "loss": 0.7509,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8932524785831167,
+      "grad_norm": 0.20260721445083618,
+      "learning_rate": 2.9752137720206463e-07,
+      "loss": 0.7538,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8971027047839061,
+      "grad_norm": 0.20376621186733246,
+      "learning_rate": 2.7669806479785067e-07,
+      "loss": 0.7592,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9009529309846953,
+      "grad_norm": 0.20533882081508636,
+      "learning_rate": 2.566094074799763e-07,
+      "loss": 0.7525,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9048031571854847,
+      "grad_norm": 0.20630250871181488,
+      "learning_rate": 2.3725852980716557e-07,
+      "loss": 0.7379,
+      "step": 2350
+    },
+    {
+      "epoch": 0.908653383386274,
+      "grad_norm": 0.20250754058361053,
+      "learning_rate": 2.1864844158503983e-07,
+      "loss": 0.748,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9125036095870632,
+      "grad_norm": 0.22006921470165253,
+      "learning_rate": 2.007820373979702e-07,
+      "loss": 0.743,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9163538357878526,
+      "grad_norm": 0.20922455191612244,
+      "learning_rate": 1.8366209615886532e-07,
+      "loss": 0.7514,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9202040619886418,
+      "grad_norm": 0.20728087425231934,
+      "learning_rate": 1.6729128067693613e-07,
+      "loss": 0.7634,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9240542881894311,
+      "grad_norm": 0.2228490561246872,
+      "learning_rate": 1.5167213724353426e-07,
+      "loss": 0.7628,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9279045143902205,
+      "grad_norm": 0.2038961797952652,
+      "learning_rate": 1.3680709523610058e-07,
+      "loss": 0.7427,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9317547405910097,
+      "grad_norm": 0.20545853674411774,
+      "learning_rate": 1.2269846674030418e-07,
+      "loss": 0.7671,
+      "step": 2420
+    },
+    {
+      "epoch": 0.935604966791799,
+      "grad_norm": 0.2172468900680542,
+      "learning_rate": 1.0934844619042595e-07,
+      "loss": 0.7624,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9394551929925883,
+      "grad_norm": 0.21919716894626617,
+      "learning_rate": 9.675911002803961e-08,
+      "loss": 0.7506,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9433054191933776,
+      "grad_norm": 0.22320537269115448,
+      "learning_rate": 8.493241637904381e-08,
+      "loss": 0.7477,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9471556453941669,
+      "grad_norm": 0.2103501707315445,
+      "learning_rate": 7.387020474909689e-08,
+      "loss": 0.7675,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9510058715949562,
+      "grad_norm": 0.22292308509349823,
+      "learning_rate": 6.357419573750856e-08,
+      "loss": 0.7631,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9548560977957455,
+      "grad_norm": 0.20595529675483704,
+      "learning_rate": 5.4045990769613365e-08,
+      "loss": 0.7441,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9587063239965348,
+      "grad_norm": 0.20607131719589233,
+      "learning_rate": 4.5287071847694434e-08,
+      "loss": 0.7507,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9625565501973241,
+      "grad_norm": 0.20815503597259521,
+      "learning_rate": 3.7298801320467817e-08,
+      "loss": 0.7483,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9664067763981133,
+      "grad_norm": 0.20976288616657257,
+      "learning_rate": 3.0082421671192576e-08,
+      "loss": 0.7416,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9702570025989027,
+      "grad_norm": 0.20461627840995789,
+      "learning_rate": 2.3639055324410355e-08,
+      "loss": 0.7619,
+      "step": 2520
+    },
+    {
+      "epoch": 0.974107228799692,
+      "grad_norm": 0.209821879863739,
+      "learning_rate": 1.7969704471369477e-08,
+      "loss": 0.7465,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9779574550004813,
+      "grad_norm": 0.19958749413490295,
+      "learning_rate": 1.3075250914142413e-08,
+      "loss": 0.7481,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9818076812012706,
+      "grad_norm": 0.20056262612342834,
+      "learning_rate": 8.956455928476049e-09,
+      "loss": 0.7568,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9856579074020598,
+      "grad_norm": 0.20076054334640503,
+      "learning_rate": 5.6139601453786275e-09,
+      "loss": 0.7494,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9895081336028492,
+      "grad_norm": 0.2042563110589981,
+      "learning_rate": 3.0482834514805602e-09,
+      "loss": 0.7393,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9933583598036385,
+      "grad_norm": 0.20023392140865326,
+      "learning_rate": 1.2598249081713366e-09,
+      "loss": 0.7312,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9972085860044277,
+      "grad_norm": 0.2173144817352295,
+      "learning_rate": 2.488626895302826e-10,
+      "loss": 0.7643,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9999037443449803,
+      "step": 2597,
+      "total_flos": 3.3979476276590674e+18,
+      "train_loss": 0.7743726096155096,
+      "train_runtime": 8177.4213,
+      "train_samples_per_second": 81.31,
+      "train_steps_per_second": 0.318
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2597,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.3979476276590674e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen25/mc2_m16_224/finetune/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77fe1a9a63c172bab929df0b458645f6bcd6964fbf4c09d5207e186508794e16
+size 6136

continuous/qwen25/mc2_m16_224/pretrain/config.json

@@ -0,0 +1,58 @@
+{
+  "_attn_implementation_autoset": true,
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 21,
+  "mm_hidden_size": 512,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "flat",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_tunable_parts": "mm_mlp_adapter",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "open_clip_hub:ViT-M-16",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen2",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 32768,
+  "tie_word_embeddings": true,
+  "tokenizer_model_max_length": 8192,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.3",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_tower_pretrained": "/cache/ckpt/download/tokenizer/continuous/mc2/mc2_m16_224.pt",
+  "vocab_size": 151936
+}

continuous/qwen25/mc2_m16_224/pretrain/mm_projector.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94c29412e906ac5cd09e7c71a594ab81e5c5afca7c35a020dcfe5b1b98e2d057
+size 12597304

continuous/qwen25/mc2_m16_224/pretrain/trainer_state.json

@@ -0,0 +1,421 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 545,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01834862385321101,
+      "grad_norm": 3.271164655685425,
+      "learning_rate": 0.00032142857142857147,
+      "loss": 4.6396,
+      "step": 10
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 0.38319090008735657,
+      "learning_rate": 0.0006785714285714287,
+      "loss": 3.5046,
+      "step": 20
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 0.09917092323303223,
+      "learning_rate": 0.0009999907688176173,
+      "loss": 3.2819,
+      "step": 30
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 0.09828156977891922,
+      "learning_rate": 0.0009988834393115766,
+      "loss": 3.218,
+      "step": 40
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 0.10038913786411285,
+      "learning_rate": 0.0009959345572940772,
+      "loss": 3.1853,
+      "step": 50
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 0.07526467740535736,
+      "learning_rate": 0.0009911550081155983,
+      "loss": 3.1294,
+      "step": 60
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.08796021342277527,
+      "learning_rate": 0.0009845624347566433,
+      "loss": 3.0839,
+      "step": 70
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 0.1084410548210144,
+      "learning_rate": 0.0009761811727013547,
+      "loss": 3.0415,
+      "step": 80
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 0.1288289576768875,
+      "learning_rate": 0.0009660421601067666,
+      "loss": 2.9928,
+      "step": 90
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 0.12765370309352875,
+      "learning_rate": 0.0009541828235992926,
+      "loss": 2.9656,
+      "step": 100
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 0.13520975410938263,
+      "learning_rate": 0.0009406469401200151,
+      "loss": 2.934,
+      "step": 110
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 0.1450054794549942,
+      "learning_rate": 0.0009254844753287492,
+      "loss": 2.9201,
+      "step": 120
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 0.1835455596446991,
+      "learning_rate": 0.0009087513991633924,
+      "loss": 2.8867,
+      "step": 130
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 0.11476298421621323,
+      "learning_rate": 0.0008905094792353917,
+      "loss": 2.866,
+      "step": 140
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 0.12783539295196533,
+      "learning_rate": 0.0008708260528239789,
+      "loss": 2.8581,
+      "step": 150
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.13325990736484528,
+      "learning_rate": 0.0008497737783108237,
+      "loss": 2.8254,
+      "step": 160
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 0.13115906715393066,
+      "learning_rate": 0.0008274303669726426,
+      "loss": 2.8012,
+      "step": 170
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 0.10847330838441849,
+      "learning_rate": 0.0008038782961218137,
+      "loss": 2.7747,
+      "step": 180
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.11461023986339569,
+      "learning_rate": 0.000779204504653894,
+      "loss": 2.7579,
+      "step": 190
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 0.09740302711725235,
+      "learning_rate": 0.0007535000721258791,
+      "loss": 2.7267,
+      "step": 200
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 0.10305294394493103,
+      "learning_rate": 0.00072685988254984,
+      "loss": 2.7162,
+      "step": 210
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 0.12250901758670807,
+      "learning_rate": 0.0006993822741429906,
+      "loss": 2.723,
+      "step": 220
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.1224115639925003,
+      "learning_rate": 0.0006711686763270818,
+      "loss": 2.7114,
+      "step": 230
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 0.10863951593637466,
+      "learning_rate": 0.0006423232353170831,
+      "loss": 2.6951,
+      "step": 240
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 0.11608681827783585,
+      "learning_rate": 0.0006129524296812335,
+      "loss": 2.6898,
+      "step": 250
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 0.09788592904806137,
+      "learning_rate": 0.0005831646772915651,
+      "loss": 2.697,
+      "step": 260
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.12143095582723618,
+      "learning_rate": 0.0005530699351157809,
+      "loss": 2.6743,
+      "step": 270
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 0.13118606805801392,
+      "learning_rate": 0.0005227792933277943,
+      "loss": 2.6754,
+      "step": 280
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 0.12023335695266724,
+      "learning_rate": 0.0004924045652352092,
+      "loss": 2.6614,
+      "step": 290
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 0.10531511902809143,
+      "learning_rate": 0.0004620578745374633,
+      "loss": 2.6619,
+      "step": 300
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 0.10306553542613983,
+      "learning_rate": 0.0004318512414382059,
+      "loss": 2.666,
+      "step": 310
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 0.12332956492900848,
+      "learning_rate": 0.0004018961691397148,
+      "loss": 2.6619,
+      "step": 320
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 0.09926410019397736,
+      "learning_rate": 0.0003723032322457458,
+      "loss": 2.6468,
+      "step": 330
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.12872529029846191,
+      "learning_rate": 0.0003431816685921625,
+      "loss": 2.6488,
+      "step": 340
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 0.11775332689285278,
+      "learning_rate": 0.0003146389760120416,
+      "loss": 2.6468,
+      "step": 350
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 0.1280543953180313,
+      "learning_rate": 0.00028678051552373485,
+      "loss": 2.6512,
+      "step": 360
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 0.10719691962003708,
+      "learning_rate": 0.0002597091224066581,
+      "loss": 2.6448,
+      "step": 370
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 0.09066632390022278,
+      "learning_rate": 0.0002335247266004629,
+      "loss": 2.6515,
+      "step": 380
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 0.1206979900598526,
+      "learning_rate": 0.000208323983828828,
+      "loss": 2.6393,
+      "step": 390
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.1239546611905098,
+      "learning_rate": 0.0001841999188095224,
+      "loss": 2.6345,
+      "step": 400
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 0.09086289256811142,
+      "learning_rate": 0.00016124158186777676,
+      "loss": 2.6376,
+      "step": 410
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 0.08489479869604111,
+      "learning_rate": 0.0001395337202205211,
+      "loss": 2.6331,
+      "step": 420
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 0.08497538417577744,
+      "learning_rate": 0.00011915646514489292,
+      "loss": 2.6302,
+      "step": 430
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 0.12431782484054565,
+      "learning_rate": 0.00010018503618578818,
+      "loss": 2.6411,
+      "step": 440
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.08686698228120804,
+      "learning_rate": 8.268946349432583e-05,
+      "loss": 2.6342,
+      "step": 450
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 0.08690136671066284,
+      "learning_rate": 6.673432932217117e-05,
+      "loss": 2.6274,
+      "step": 460
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.09988857060670853,
+      "learning_rate": 5.2378529625954684e-05,
+      "loss": 2.6189,
+      "step": 470
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.09928096830844879,
+      "learning_rate": 3.9675056661785556e-05,
+      "loss": 2.6226,
+      "step": 480
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.11798673868179321,
+      "learning_rate": 2.8670803372379538e-05,
+      "loss": 2.6319,
+      "step": 490
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 0.09891407936811447,
+      "learning_rate": 1.9406390288876586e-05,
+      "loss": 2.6375,
+      "step": 500
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 0.09769132733345032,
+      "learning_rate": 1.191601558631461e-05,
+      "loss": 2.6255,
+      "step": 510
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 0.08942878246307373,
+      "learning_rate": 6.2273288462559305e-06,
+      "loss": 2.6253,
+      "step": 520
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 0.08646480739116669,
+      "learning_rate": 2.361328992552314e-06,
+      "loss": 2.626,
+      "step": 530
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 0.09041832387447357,
+      "learning_rate": 3.322867770044202e-07,
+      "loss": 2.6228,
+      "step": 540
+    },
+    {
+      "epoch": 1.0,
+      "step": 545,
+      "total_flos": 2.688705702783877e+17,
+      "train_loss": 2.806120305542552,
+      "train_runtime": 5246.0172,
+      "train_samples_per_second": 106.391,
+      "train_steps_per_second": 0.104
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 545,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.688705702783877e+17,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen25/mc2_m16_224_mt5/finetune/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

continuous/qwen25/mc2_m16_224_mt5/finetune/config.json

@@ -0,0 +1,57 @@
+{
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 21,
+  "mm_hidden_size": 512,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "flat",
+  "mm_projector_lr": 0.0001,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_tunable_parts": "mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "open_clip_hub:ViT-M-16",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen2",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 32768,
+  "tie_word_embeddings": true,
+  "tokenizer_model_max_length": 8192,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.3",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_tower_pretrained": "/cache/ckpt/download/tokenizer/continuous/mc2/mc2_m16_224_mt5.pt",
+  "vocab_size": 151936
+}

continuous/qwen25/mc2_m16_224_mt5/finetune/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.1,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.51.3"
+}

continuous/qwen25/mc2_m16_224_mt5/finetune/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5600f444f7661dd945f2dc46cbb314c10eb1dc22b953ead7a270a0e7a527a63d
+size 3644007016

continuous/qwen25/mc2_m16_224_mt5/finetune/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_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|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

continuous/qwen25/mc2_m16_224_mt5/finetune/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

continuous/qwen25/mc2_m16_224_mt5/finetune/tokenizer_config.json

@@ -0,0 +1,209 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_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|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# 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>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\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\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 8192,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

continuous/qwen25/mc2_m16_224_mt5/finetune/trainer_state.json

@@ -0,0 +1,1856 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999037443449803,
+  "eval_steps": 500,
+  "global_step": 2597,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003850226200789296,
+      "grad_norm": 17.946489334106445,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 2.019,
+      "step": 10
+    },
+    {
+      "epoch": 0.007700452401578592,
+      "grad_norm": 10.400087356567383,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 1.6443,
+      "step": 20
+    },
+    {
+      "epoch": 0.01155067860236789,
+      "grad_norm": 3.3501930236816406,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.1167,
+      "step": 30
+    },
+    {
+      "epoch": 0.015400904803157185,
+      "grad_norm": 0.757749617099762,
+      "learning_rate": 5e-06,
+      "loss": 0.9455,
+      "step": 40
+    },
+    {
+      "epoch": 0.019251131003946483,
+      "grad_norm": 0.3714425563812256,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 0.9117,
+      "step": 50
+    },
+    {
+      "epoch": 0.02310135720473578,
+      "grad_norm": 0.2789915204048157,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 0.8646,
+      "step": 60
+    },
+    {
+      "epoch": 0.026951583405525073,
+      "grad_norm": 0.22899337112903595,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.8384,
+      "step": 70
+    },
+    {
+      "epoch": 0.03080180960631437,
+      "grad_norm": 0.23750527203083038,
+      "learning_rate": 9.999996111488723e-06,
+      "loss": 0.8474,
+      "step": 80
+    },
+    {
+      "epoch": 0.034652035807103666,
+      "grad_norm": 0.21610580384731293,
+      "learning_rate": 9.999529497453782e-06,
+      "loss": 0.8281,
+      "step": 90
+    },
+    {
+      "epoch": 0.038502262007892966,
+      "grad_norm": 0.2217676043510437,
+      "learning_rate": 9.998285264324193e-06,
+      "loss": 0.8077,
+      "step": 100
+    },
+    {
+      "epoch": 0.04235248820868226,
+      "grad_norm": 0.19842316210269928,
+      "learning_rate": 9.996263605626052e-06,
+      "loss": 0.798,
+      "step": 110
+    },
+    {
+      "epoch": 0.04620271440947156,
+      "grad_norm": 0.22483044862747192,
+      "learning_rate": 9.99346483580503e-06,
+      "loss": 0.8096,
+      "step": 120
+    },
+    {
+      "epoch": 0.05005294061026085,
+      "grad_norm": 0.20166398584842682,
+      "learning_rate": 9.989889390177467e-06,
+      "loss": 0.7992,
+      "step": 130
+    },
+    {
+      "epoch": 0.053903166811050146,
+      "grad_norm": 0.23341001570224762,
+      "learning_rate": 9.985537824862647e-06,
+      "loss": 0.8015,
+      "step": 140
+    },
+    {
+      "epoch": 0.057753393011839446,
+      "grad_norm": 0.23562155663967133,
+      "learning_rate": 9.980410816696317e-06,
+      "loss": 0.783,
+      "step": 150
+    },
+    {
+      "epoch": 0.06160361921262874,
+      "grad_norm": 0.21823126077651978,
+      "learning_rate": 9.97450916312541e-06,
+      "loss": 0.7922,
+      "step": 160
+    },
+    {
+      "epoch": 0.06545384541341803,
+      "grad_norm": 0.19712361693382263,
+      "learning_rate": 9.967833782084004e-06,
+      "loss": 0.7929,
+      "step": 170
+    },
+    {
+      "epoch": 0.06930407161420733,
+      "grad_norm": 0.2300766408443451,
+      "learning_rate": 9.960385711850554e-06,
+      "loss": 0.8179,
+      "step": 180
+    },
+    {
+      "epoch": 0.07315429781499663,
+      "grad_norm": 0.22057217359542847,
+      "learning_rate": 9.952166110886394e-06,
+      "loss": 0.8109,
+      "step": 190
+    },
+    {
+      "epoch": 0.07700452401578593,
+      "grad_norm": 0.250495582818985,
+      "learning_rate": 9.943176257655567e-06,
+      "loss": 0.8019,
+      "step": 200
+    },
+    {
+      "epoch": 0.08085475021657522,
+      "grad_norm": 0.21307650208473206,
+      "learning_rate": 9.933417550425952e-06,
+      "loss": 0.7959,
+      "step": 210
+    },
+    {
+      "epoch": 0.08470497641736452,
+      "grad_norm": 0.21728742122650146,
+      "learning_rate": 9.9228915070518e-06,
+      "loss": 0.7812,
+      "step": 220
+    },
+    {
+      "epoch": 0.08855520261815382,
+      "grad_norm": 0.215750053524971,
+      "learning_rate": 9.911599764737632e-06,
+      "loss": 0.7841,
+      "step": 230
+    },
+    {
+      "epoch": 0.09240542881894312,
+      "grad_norm": 0.2224174588918686,
+      "learning_rate": 9.899544079783606e-06,
+      "loss": 0.7778,
+      "step": 240
+    },
+    {
+      "epoch": 0.0962556550197324,
+      "grad_norm": 0.2246825397014618,
+      "learning_rate": 9.886726327312338e-06,
+      "loss": 0.7944,
+      "step": 250
+    },
+    {
+      "epoch": 0.1001058812205217,
+      "grad_norm": 0.21317213773727417,
+      "learning_rate": 9.87314850097724e-06,
+      "loss": 0.7919,
+      "step": 260
+    },
+    {
+      "epoch": 0.103956107421311,
+      "grad_norm": 0.20239098370075226,
+      "learning_rate": 9.858812712652452e-06,
+      "loss": 0.7818,
+      "step": 270
+    },
+    {
+      "epoch": 0.10780633362210029,
+      "grad_norm": 0.19969305396080017,
+      "learning_rate": 9.843721192104341e-06,
+      "loss": 0.7802,
+      "step": 280
+    },
+    {
+      "epoch": 0.11165655982288959,
+      "grad_norm": 0.21302467584609985,
+      "learning_rate": 9.827876286644698e-06,
+      "loss": 0.8015,
+      "step": 290
+    },
+    {
+      "epoch": 0.11550678602367889,
+      "grad_norm": 0.19669854640960693,
+      "learning_rate": 9.811280460765643e-06,
+      "loss": 0.7627,
+      "step": 300
+    },
+    {
+      "epoch": 0.11935701222446819,
+      "grad_norm": 0.20913861691951752,
+      "learning_rate": 9.793936295756292e-06,
+      "loss": 0.7723,
+      "step": 310
+    },
+    {
+      "epoch": 0.12320723842525748,
+      "grad_norm": 0.2178550660610199,
+      "learning_rate": 9.775846489301279e-06,
+      "loss": 0.7764,
+      "step": 320
+    },
+    {
+      "epoch": 0.12705746462604678,
+      "grad_norm": 0.2206895351409912,
+      "learning_rate": 9.757013855061157e-06,
+      "loss": 0.7741,
+      "step": 330
+    },
+    {
+      "epoch": 0.13090769082683607,
+      "grad_norm": 0.21030789613723755,
+      "learning_rate": 9.737441322234758e-06,
+      "loss": 0.7877,
+      "step": 340
+    },
+    {
+      "epoch": 0.13475791702762538,
+      "grad_norm": 0.22136712074279785,
+      "learning_rate": 9.717131935103605e-06,
+      "loss": 0.7872,
+      "step": 350
+    },
+    {
+      "epoch": 0.13860814322841467,
+      "grad_norm": 0.21178635954856873,
+      "learning_rate": 9.696088852558398e-06,
+      "loss": 0.7732,
+      "step": 360
+    },
+    {
+      "epoch": 0.14245836942920398,
+      "grad_norm": 0.2112646847963333,
+      "learning_rate": 9.674315347607686e-06,
+      "loss": 0.799,
+      "step": 370
+    },
+    {
+      "epoch": 0.14630859562999327,
+      "grad_norm": 0.21052142977714539,
+      "learning_rate": 9.651814806868796e-06,
+      "loss": 0.7645,
+      "step": 380
+    },
+    {
+      "epoch": 0.15015882183078255,
+      "grad_norm": 0.21341727674007416,
+      "learning_rate": 9.62859073004107e-06,
+      "loss": 0.766,
+      "step": 390
+    },
+    {
+      "epoch": 0.15400904803157187,
+      "grad_norm": 0.20068255066871643,
+      "learning_rate": 9.604646729361547e-06,
+      "loss": 0.7717,
+      "step": 400
+    },
+    {
+      "epoch": 0.15785927423236115,
+      "grad_norm": 0.20311903953552246,
+      "learning_rate": 9.579986529043099e-06,
+      "loss": 0.762,
+      "step": 410
+    },
+    {
+      "epoch": 0.16170950043315044,
+      "grad_norm": 0.21248885989189148,
+      "learning_rate": 9.554613964695189e-06,
+      "loss": 0.7785,
+      "step": 420
+    },
+    {
+      "epoch": 0.16555972663393975,
+      "grad_norm": 0.21587100625038147,
+      "learning_rate": 9.528532982727281e-06,
+      "loss": 0.771,
+      "step": 430
+    },
+    {
+      "epoch": 0.16940995283472904,
+      "grad_norm": 0.20114363729953766,
+      "learning_rate": 9.501747639735018e-06,
+      "loss": 0.7681,
+      "step": 440
+    },
+    {
+      "epoch": 0.17326017903551832,
+      "grad_norm": 0.21528469026088715,
+      "learning_rate": 9.474262101869272e-06,
+      "loss": 0.7565,
+      "step": 450
+    },
+    {
+      "epoch": 0.17711040523630764,
+      "grad_norm": 0.21150706708431244,
+      "learning_rate": 9.446080644188141e-06,
+      "loss": 0.7633,
+      "step": 460
+    },
+    {
+      "epoch": 0.18096063143709692,
+      "grad_norm": 0.2017327845096588,
+      "learning_rate": 9.417207649992017e-06,
+      "loss": 0.7681,
+      "step": 470
+    },
+    {
+      "epoch": 0.18481085763788624,
+      "grad_norm": 0.20973379909992218,
+      "learning_rate": 9.387647610141816e-06,
+      "loss": 0.7674,
+      "step": 480
+    },
+    {
+      "epoch": 0.18866108383867553,
+      "grad_norm": 0.19291938841342926,
+      "learning_rate": 9.35740512236047e-06,
+      "loss": 0.7541,
+      "step": 490
+    },
+    {
+      "epoch": 0.1925113100394648,
+      "grad_norm": 0.1934889853000641,
+      "learning_rate": 9.326484890517809e-06,
+      "loss": 0.7656,
+      "step": 500
+    },
+    {
+      "epoch": 0.19636153624025413,
+      "grad_norm": 0.21811489760875702,
+      "learning_rate": 9.294891723898926e-06,
+      "loss": 0.755,
+      "step": 510
+    },
+    {
+      "epoch": 0.2002117624410434,
+      "grad_norm": 0.22245967388153076,
+      "learning_rate": 9.262630536456153e-06,
+      "loss": 0.7853,
+      "step": 520
+    },
+    {
+      "epoch": 0.2040619886418327,
+      "grad_norm": 0.21980620920658112,
+      "learning_rate": 9.229706346044749e-06,
+      "loss": 0.759,
+      "step": 530
+    },
+    {
+      "epoch": 0.207912214842622,
+      "grad_norm": 0.2326648086309433,
+      "learning_rate": 9.196124273642422e-06,
+      "loss": 0.771,
+      "step": 540
+    },
+    {
+      "epoch": 0.2117624410434113,
+      "grad_norm": 0.20531702041625977,
+      "learning_rate": 9.161889542552842e-06,
+      "loss": 0.7756,
+      "step": 550
+    },
+    {
+      "epoch": 0.21561266724420058,
+      "grad_norm": 0.20525620877742767,
+      "learning_rate": 9.127007477593193e-06,
+      "loss": 0.7766,
+      "step": 560
+    },
+    {
+      "epoch": 0.2194628934449899,
+      "grad_norm": 0.226873517036438,
+      "learning_rate": 9.091483504265974e-06,
+      "loss": 0.7856,
+      "step": 570
+    },
+    {
+      "epoch": 0.22331311964577918,
+      "grad_norm": 0.22353024780750275,
+      "learning_rate": 9.055323147915116e-06,
+      "loss": 0.7656,
+      "step": 580
+    },
+    {
+      "epoch": 0.2271633458465685,
+      "grad_norm": 0.2133764922618866,
+      "learning_rate": 9.018532032866585e-06,
+      "loss": 0.7755,
+      "step": 590
+    },
+    {
+      "epoch": 0.23101357204735778,
+      "grad_norm": 0.19909435510635376,
+      "learning_rate": 8.981115881553579e-06,
+      "loss": 0.7678,
+      "step": 600
+    },
+    {
+      "epoch": 0.23486379824814707,
+      "grad_norm": 0.23409725725650787,
+      "learning_rate": 8.943080513626476e-06,
+      "loss": 0.7887,
+      "step": 610
+    },
+    {
+      "epoch": 0.23871402444893638,
+      "grad_norm": 0.21527846157550812,
+      "learning_rate": 8.904431845047641e-06,
+      "loss": 0.7726,
+      "step": 620
+    },
+    {
+      "epoch": 0.24256425064972567,
+      "grad_norm": 0.22071720659732819,
+      "learning_rate": 8.86517588717128e-06,
+      "loss": 0.7882,
+      "step": 630
+    },
+    {
+      "epoch": 0.24641447685051496,
+      "grad_norm": 0.20500795543193817,
+      "learning_rate": 8.82531874580844e-06,
+      "loss": 0.7806,
+      "step": 640
+    },
+    {
+      "epoch": 0.25026470305130427,
+      "grad_norm": 0.19545669853687286,
+      "learning_rate": 8.784866620277314e-06,
+      "loss": 0.7385,
+      "step": 650
+    },
+    {
+      "epoch": 0.25411492925209356,
+      "grad_norm": 0.22581326961517334,
+      "learning_rate": 8.743825802439018e-06,
+      "loss": 0.7702,
+      "step": 660
+    },
+    {
+      "epoch": 0.25796515545288284,
+      "grad_norm": 0.2226068526506424,
+      "learning_rate": 8.702202675718956e-06,
+      "loss": 0.7785,
+      "step": 670
+    },
+    {
+      "epoch": 0.26181538165367213,
+      "grad_norm": 0.20091012120246887,
+      "learning_rate": 8.660003714113959e-06,
+      "loss": 0.7467,
+      "step": 680
+    },
+    {
+      "epoch": 0.26566560785446147,
+      "grad_norm": 0.21000726521015167,
+      "learning_rate": 8.617235481185323e-06,
+      "loss": 0.7687,
+      "step": 690
+    },
+    {
+      "epoch": 0.26951583405525076,
+      "grad_norm": 0.21369947493076324,
+      "learning_rate": 8.573904629037931e-06,
+      "loss": 0.775,
+      "step": 700
+    },
+    {
+      "epoch": 0.27336606025604004,
+      "grad_norm": 0.21368445456027985,
+      "learning_rate": 8.530017897285585e-06,
+      "loss": 0.7582,
+      "step": 710
+    },
+    {
+      "epoch": 0.27721628645682933,
+      "grad_norm": 0.19397543370723724,
+      "learning_rate": 8.485582112002746e-06,
+      "loss": 0.7632,
+      "step": 720
+    },
+    {
+      "epoch": 0.2810665126576186,
+      "grad_norm": 0.2081318199634552,
+      "learning_rate": 8.440604184662811e-06,
+      "loss": 0.7656,
+      "step": 730
+    },
+    {
+      "epoch": 0.28491673885840796,
+      "grad_norm": 0.21429389715194702,
+      "learning_rate": 8.395091111063118e-06,
+      "loss": 0.7554,
+      "step": 740
+    },
+    {
+      "epoch": 0.28876696505919724,
+      "grad_norm": 0.1985684186220169,
+      "learning_rate": 8.349049970236822e-06,
+      "loss": 0.7549,
+      "step": 750
+    },
+    {
+      "epoch": 0.29261719125998653,
+      "grad_norm": 0.20911246538162231,
+      "learning_rate": 8.302487923351842e-06,
+      "loss": 0.7475,
+      "step": 760
+    },
+    {
+      "epoch": 0.2964674174607758,
+      "grad_norm": 0.20631681382656097,
+      "learning_rate": 8.255412212597021e-06,
+      "loss": 0.7652,
+      "step": 770
+    },
+    {
+      "epoch": 0.3003176436615651,
+      "grad_norm": 0.19921927154064178,
+      "learning_rate": 8.207830160055688e-06,
+      "loss": 0.7626,
+      "step": 780
+    },
+    {
+      "epoch": 0.3041678698623544,
+      "grad_norm": 0.21662472188472748,
+      "learning_rate": 8.159749166566786e-06,
+      "loss": 0.7484,
+      "step": 790
+    },
+    {
+      "epoch": 0.30801809606314373,
+      "grad_norm": 0.20460261404514313,
+      "learning_rate": 8.11117671057377e-06,
+      "loss": 0.7757,
+      "step": 800
+    },
+    {
+      "epoch": 0.311868322263933,
+      "grad_norm": 0.2161598652601242,
+      "learning_rate": 8.062120346961404e-06,
+      "loss": 0.7672,
+      "step": 810
+    },
+    {
+      "epoch": 0.3157185484647223,
+      "grad_norm": 0.20158053934574127,
+      "learning_rate": 8.012587705880705e-06,
+      "loss": 0.7618,
+      "step": 820
+    },
+    {
+      "epoch": 0.3195687746655116,
+      "grad_norm": 0.20826669037342072,
+      "learning_rate": 7.962586491562146e-06,
+      "loss": 0.7469,
+      "step": 830
+    },
+    {
+      "epoch": 0.3234190008663009,
+      "grad_norm": 0.2278338372707367,
+      "learning_rate": 7.912124481117357e-06,
+      "loss": 0.7529,
+      "step": 840
+    },
+    {
+      "epoch": 0.3272692270670902,
+      "grad_norm": 0.2085474580526352,
+      "learning_rate": 7.861209523329494e-06,
+      "loss": 0.7649,
+      "step": 850
+    },
+    {
+      "epoch": 0.3311194532678795,
+      "grad_norm": 0.2084922045469284,
+      "learning_rate": 7.809849537432432e-06,
+      "loss": 0.7482,
+      "step": 860
+    },
+    {
+      "epoch": 0.3349696794686688,
+      "grad_norm": 0.2052047997713089,
+      "learning_rate": 7.758052511879035e-06,
+      "loss": 0.7588,
+      "step": 870
+    },
+    {
+      "epoch": 0.3388199056694581,
+      "grad_norm": 0.19877785444259644,
+      "learning_rate": 7.705826503098642e-06,
+      "loss": 0.7538,
+      "step": 880
+    },
+    {
+      "epoch": 0.34267013187024736,
+      "grad_norm": 0.20094160735607147,
+      "learning_rate": 7.653179634243979e-06,
+      "loss": 0.7536,
+      "step": 890
+    },
+    {
+      "epoch": 0.34652035807103665,
+      "grad_norm": 0.1901484876871109,
+      "learning_rate": 7.600120093927696e-06,
+      "loss": 0.7566,
+      "step": 900
+    },
+    {
+      "epoch": 0.350370584271826,
+      "grad_norm": 0.21791775524616241,
+      "learning_rate": 7.546656134948729e-06,
+      "loss": 0.7416,
+      "step": 910
+    },
+    {
+      "epoch": 0.3542208104726153,
+      "grad_norm": 0.2077992856502533,
+      "learning_rate": 7.492796073008661e-06,
+      "loss": 0.7459,
+      "step": 920
+    },
+    {
+      "epoch": 0.35807103667340456,
+      "grad_norm": 0.22126300632953644,
+      "learning_rate": 7.438548285418326e-06,
+      "loss": 0.7342,
+      "step": 930
+    },
+    {
+      "epoch": 0.36192126287419385,
+      "grad_norm": 0.23142030835151672,
+      "learning_rate": 7.3839212097948e-06,
+      "loss": 0.7673,
+      "step": 940
+    },
+    {
+      "epoch": 0.36577148907498314,
+      "grad_norm": 0.21628747880458832,
+      "learning_rate": 7.328923342749039e-06,
+      "loss": 0.7804,
+      "step": 950
+    },
+    {
+      "epoch": 0.3696217152757725,
+      "grad_norm": 0.2196776270866394,
+      "learning_rate": 7.273563238564319e-06,
+      "loss": 0.7678,
+      "step": 960
+    },
+    {
+      "epoch": 0.37347194147656176,
+      "grad_norm": 0.21377627551555634,
+      "learning_rate": 7.217849507865724e-06,
+      "loss": 0.7462,
+      "step": 970
+    },
+    {
+      "epoch": 0.37732216767735105,
+      "grad_norm": 0.20526368916034698,
+      "learning_rate": 7.161790816280858e-06,
+      "loss": 0.7532,
+      "step": 980
+    },
+    {
+      "epoch": 0.38117239387814034,
+      "grad_norm": 0.2162705510854721,
+      "learning_rate": 7.1053958830920125e-06,
+      "loss": 0.735,
+      "step": 990
+    },
+    {
+      "epoch": 0.3850226200789296,
+      "grad_norm": 0.20701883733272552,
+      "learning_rate": 7.0486734798799715e-06,
+      "loss": 0.7527,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3888728462797189,
+      "grad_norm": 0.20308168232440948,
+      "learning_rate": 6.991632429159708e-06,
+      "loss": 0.7765,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39272307248050825,
+      "grad_norm": 0.22281944751739502,
+      "learning_rate": 6.934281603008132e-06,
+      "loss": 0.7688,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39657329868129754,
+      "grad_norm": 0.2043282389640808,
+      "learning_rate": 6.876629921684143e-06,
+      "loss": 0.766,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4004235248820868,
+      "grad_norm": 0.21408423781394958,
+      "learning_rate": 6.818686352241199e-06,
+      "loss": 0.756,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4042737510828761,
+      "grad_norm": 0.19420672953128815,
+      "learning_rate": 6.760459907132573e-06,
+      "loss": 0.7529,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4081239772836654,
+      "grad_norm": 0.1940777599811554,
+      "learning_rate": 6.701959642809589e-06,
+      "loss": 0.7582,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41197420348445474,
+      "grad_norm": 0.21019771695137024,
+      "learning_rate": 6.643194658312987e-06,
+      "loss": 0.7482,
+      "step": 1070
+    },
+    {
+      "epoch": 0.415824429685244,
+      "grad_norm": 0.2048809677362442,
+      "learning_rate": 6.584174093857676e-06,
+      "loss": 0.772,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4196746558860333,
+      "grad_norm": 0.22884605824947357,
+      "learning_rate": 6.5249071294110735e-06,
+      "loss": 0.757,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4235248820868226,
+      "grad_norm": 0.21377931535243988,
+      "learning_rate": 6.465402983265274e-06,
+      "loss": 0.7707,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4273751082876119,
+      "grad_norm": 0.2175796926021576,
+      "learning_rate": 6.405670910603243e-06,
+      "loss": 0.7671,
+      "step": 1110
+    },
+    {
+      "epoch": 0.43122533448840117,
+      "grad_norm": 0.22576279938220978,
+      "learning_rate": 6.345720202059291e-06,
+      "loss": 0.7623,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4350755606891905,
+      "grad_norm": 0.21094109117984772,
+      "learning_rate": 6.285560182274008e-06,
+      "loss": 0.7397,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4389257868899798,
+      "grad_norm": 0.20463645458221436,
+      "learning_rate": 6.225200208443936e-06,
+      "loss": 0.7619,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4427760130907691,
+      "grad_norm": 0.20472411811351776,
+      "learning_rate": 6.16464966886616e-06,
+      "loss": 0.7523,
+      "step": 1150
+    },
+    {
+      "epoch": 0.44662623929155837,
+      "grad_norm": 0.24761930108070374,
+      "learning_rate": 6.103917981478065e-06,
+      "loss": 0.7773,
+      "step": 1160
+    },
+    {
+      "epoch": 0.45047646549234766,
+      "grad_norm": 0.21116669476032257,
+      "learning_rate": 6.043014592392489e-06,
+      "loss": 0.7879,
+      "step": 1170
+    },
+    {
+      "epoch": 0.454326691693137,
+      "grad_norm": 0.19568389654159546,
+      "learning_rate": 5.981948974428485e-06,
+      "loss": 0.7603,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4581769178939263,
+      "grad_norm": 0.2199745625257492,
+      "learning_rate": 5.920730625637934e-06,
+      "loss": 0.7395,
+      "step": 1190
+    },
+    {
+      "epoch": 0.46202714409471557,
+      "grad_norm": 0.215942844748497,
+      "learning_rate": 5.8593690678282355e-06,
+      "loss": 0.7535,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46587737029550486,
+      "grad_norm": 0.21790876984596252,
+      "learning_rate": 5.797873845081301e-06,
+      "loss": 0.7554,
+      "step": 1210
+    },
+    {
+      "epoch": 0.46972759649629414,
+      "grad_norm": 0.2154826521873474,
+      "learning_rate": 5.736254522269075e-06,
+      "loss": 0.772,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47357782269708343,
+      "grad_norm": 0.2088659554719925,
+      "learning_rate": 5.6745206835658386e-06,
+      "loss": 0.7767,
+      "step": 1230
+    },
+    {
+      "epoch": 0.47742804889787277,
+      "grad_norm": 0.20226368308067322,
+      "learning_rate": 5.612681930957495e-06,
+      "loss": 0.7446,
+      "step": 1240
+    },
+    {
+      "epoch": 0.48127827509866206,
+      "grad_norm": 0.1928059309720993,
+      "learning_rate": 5.550747882748095e-06,
+      "loss": 0.7422,
+      "step": 1250
+    },
+    {
+      "epoch": 0.48512850129945134,
+      "grad_norm": 0.22634586691856384,
+      "learning_rate": 5.4887281720638195e-06,
+      "loss": 0.7527,
+      "step": 1260
+    },
+    {
+      "epoch": 0.48897872750024063,
+      "grad_norm": 0.23566408455371857,
+      "learning_rate": 5.42663244535466e-06,
+      "loss": 0.7341,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4928289537010299,
+      "grad_norm": 0.20938095450401306,
+      "learning_rate": 5.3644703608940205e-06,
+      "loss": 0.7435,
+      "step": 1280
+    },
+    {
+      "epoch": 0.49667917990181926,
+      "grad_norm": 0.21409724652767181,
+      "learning_rate": 5.302251587276491e-06,
+      "loss": 0.7536,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5005294061026085,
+      "grad_norm": 0.2033991664648056,
+      "learning_rate": 5.2399858019140005e-06,
+      "loss": 0.7559,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5043796323033978,
+      "grad_norm": 0.20759224891662598,
+      "learning_rate": 5.177682689530615e-06,
+      "loss": 0.7678,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5082298585041871,
+      "grad_norm": 0.21629536151885986,
+      "learning_rate": 5.11535194065619e-06,
+      "loss": 0.7387,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5120800847049765,
+      "grad_norm": 0.2309747189283371,
+      "learning_rate": 5.053003250119113e-06,
+      "loss": 0.7466,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5159303109057657,
+      "grad_norm": 0.2208855003118515,
+      "learning_rate": 4.9906463155383925e-06,
+      "loss": 0.754,
+      "step": 1340
+    },
+    {
+      "epoch": 0.519780537106555,
+      "grad_norm": 0.21024666726589203,
+      "learning_rate": 4.9282908358153046e-06,
+      "loss": 0.7489,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5236307633073443,
+      "grad_norm": 0.23770283162593842,
+      "learning_rate": 4.8659465096248345e-06,
+      "loss": 0.7649,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5274809895081336,
+      "grad_norm": 0.21441885828971863,
+      "learning_rate": 4.803623033907169e-06,
+      "loss": 0.7611,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5313312157089229,
+      "grad_norm": 0.2043793797492981,
+      "learning_rate": 4.741330102359441e-06,
+      "loss": 0.7695,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5351814419097122,
+      "grad_norm": 0.22266699373722076,
+      "learning_rate": 4.679077403927996e-06,
+      "loss": 0.7483,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5390316681105015,
+      "grad_norm": 0.20998956263065338,
+      "learning_rate": 4.616874621301373e-06,
+      "loss": 0.7494,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5428818943112907,
+      "grad_norm": 0.21371926367282867,
+      "learning_rate": 4.554731429404293e-06,
+      "loss": 0.763,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5467321205120801,
+      "grad_norm": 0.22156395018100739,
+      "learning_rate": 4.492657493892822e-06,
+      "loss": 0.7465,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5505823467128694,
+      "grad_norm": 0.22652852535247803,
+      "learning_rate": 4.430662469650999e-06,
+      "loss": 0.7632,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5544325729136587,
+      "grad_norm": 0.22102586925029755,
+      "learning_rate": 4.368755999289118e-06,
+      "loss": 0.7676,
+      "step": 1440
+    },
+    {
+      "epoch": 0.558282799114448,
+      "grad_norm": 0.2220848649740219,
+      "learning_rate": 4.306947711643949e-06,
+      "loss": 0.7474,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5621330253152372,
+      "grad_norm": 0.2216523140668869,
+      "learning_rate": 4.245247220281068e-06,
+      "loss": 0.754,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5659832515160266,
+      "grad_norm": 0.2109246551990509,
+      "learning_rate": 4.183664121999587e-06,
+      "loss": 0.7498,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5698334777168159,
+      "grad_norm": 0.2210337072610855,
+      "learning_rate": 4.122207995339474e-06,
+      "loss": 0.747,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5736837039176051,
+      "grad_norm": 0.20343822240829468,
+      "learning_rate": 4.060888399091734e-06,
+      "loss": 0.7523,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5775339301183945,
+      "grad_norm": 0.20101144909858704,
+      "learning_rate": 3.999714870811639e-06,
+      "loss": 0.763,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5813841563191837,
+      "grad_norm": 0.21249139308929443,
+      "learning_rate": 3.938696925335275e-06,
+      "loss": 0.7558,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5852343825199731,
+      "grad_norm": 0.2062484622001648,
+      "learning_rate": 3.87784405329962e-06,
+      "loss": 0.7634,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5890846087207623,
+      "grad_norm": 0.21326501667499542,
+      "learning_rate": 3.817165719666382e-06,
+      "loss": 0.7642,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5929348349215516,
+      "grad_norm": 0.20311339199543,
+      "learning_rate": 3.756671362249841e-06,
+      "loss": 0.7702,
+      "step": 1540
+    },
+    {
+      "epoch": 0.596785061122341,
+      "grad_norm": 0.21383582055568695,
+      "learning_rate": 3.6963703902488907e-06,
+      "loss": 0.7561,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6006352873231302,
+      "grad_norm": 0.20350584387779236,
+      "learning_rate": 3.6362721827835655e-06,
+      "loss": 0.7703,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6044855135239195,
+      "grad_norm": 0.22511819005012512,
+      "learning_rate": 3.576386087436212e-06,
+      "loss": 0.7679,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6083357397247088,
+      "grad_norm": 0.22315967082977295,
+      "learning_rate": 3.5167214187975913e-06,
+      "loss": 0.7573,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6121859659254981,
+      "grad_norm": 0.18645760416984558,
+      "learning_rate": 3.457287457018086e-06,
+      "loss": 0.7593,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6160361921262875,
+      "grad_norm": 0.20026874542236328,
+      "learning_rate": 3.3980934463643003e-06,
+      "loss": 0.7481,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6198864183270767,
+      "grad_norm": 0.20088103413581848,
+      "learning_rate": 3.3391485937812053e-06,
+      "loss": 0.73,
+      "step": 1610
+    },
+    {
+      "epoch": 0.623736644527866,
+      "grad_norm": 0.21048805117607117,
+      "learning_rate": 3.280462067460115e-06,
+      "loss": 0.7285,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6275868707286553,
+      "grad_norm": 0.20529092848300934,
+      "learning_rate": 3.222042995412669e-06,
+      "loss": 0.7611,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6314370969294446,
+      "grad_norm": 0.18651160597801208,
+      "learning_rate": 3.163900464051089e-06,
+      "loss": 0.7494,
+      "step": 1640
+    },
+    {
+      "epoch": 0.635287323130234,
+      "grad_norm": 0.20862846076488495,
+      "learning_rate": 3.1060435167748826e-06,
+      "loss": 0.7614,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6391375493310232,
+      "grad_norm": 0.21366997063159943,
+      "learning_rate": 3.048481152564253e-06,
+      "loss": 0.7408,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6429877755318125,
+      "grad_norm": 0.2224278599023819,
+      "learning_rate": 2.9912223245804017e-06,
+      "loss": 0.7614,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6468380017326018,
+      "grad_norm": 0.22007468342781067,
+      "learning_rate": 2.934275938772981e-06,
+      "loss": 0.7557,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6506882279333911,
+      "grad_norm": 0.2168564647436142,
+      "learning_rate": 2.8776508524948676e-06,
+      "loss": 0.7465,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6545384541341804,
+      "grad_norm": 0.18673951923847198,
+      "learning_rate": 2.821355873124503e-06,
+      "loss": 0.7455,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6583886803349697,
+      "grad_norm": 0.21472465991973877,
+      "learning_rate": 2.7653997566960144e-06,
+      "loss": 0.759,
+      "step": 1710
+    },
+    {
+      "epoch": 0.662238906535759,
+      "grad_norm": 0.2146589457988739,
+      "learning_rate": 2.709791206537309e-06,
+      "loss": 0.7476,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6660891327365482,
+      "grad_norm": 0.21385711431503296,
+      "learning_rate": 2.6545388719163747e-06,
+      "loss": 0.7514,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6699393589373376,
+      "grad_norm": 0.21231603622436523,
+      "learning_rate": 2.599651346695979e-06,
+      "loss": 0.7531,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6737895851381268,
+      "grad_norm": 0.20154190063476562,
+      "learning_rate": 2.545137167997007e-06,
+      "loss": 0.7513,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6776398113389162,
+      "grad_norm": 0.19922277331352234,
+      "learning_rate": 2.4910048148705972e-06,
+      "loss": 0.7397,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6814900375397055,
+      "grad_norm": 0.2174701988697052,
+      "learning_rate": 2.4372627069793327e-06,
+      "loss": 0.7836,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6853402637404947,
+      "grad_norm": 0.22188161313533783,
+      "learning_rate": 2.3839192032876486e-06,
+      "loss": 0.7674,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6891904899412841,
+      "grad_norm": 0.20601807534694672,
+      "learning_rate": 2.3309826007617047e-06,
+      "loss": 0.7789,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6930407161420733,
+      "grad_norm": 0.20942874252796173,
+      "learning_rate": 2.278461133078881e-06,
+      "loss": 0.7414,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6968909423428626,
+      "grad_norm": 0.22094810009002686,
+      "learning_rate": 2.2263629693471258e-06,
+      "loss": 0.7347,
+      "step": 1810
+    },
+    {
+      "epoch": 0.700741168543652,
+      "grad_norm": 0.19887880980968475,
+      "learning_rate": 2.1746962128343414e-06,
+      "loss": 0.7553,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7045913947444412,
+      "grad_norm": 0.2053019255399704,
+      "learning_rate": 2.1234688997080183e-06,
+      "loss": 0.7471,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7084416209452306,
+      "grad_norm": 0.221902996301651,
+      "learning_rate": 2.0726889977853053e-06,
+      "loss": 0.752,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7122918471460198,
+      "grad_norm": 0.22843322157859802,
+      "learning_rate": 2.022364405293703e-06,
+      "loss": 0.7632,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7161420733468091,
+      "grad_norm": 0.2099401205778122,
+      "learning_rate": 1.9725029496425814e-06,
+      "loss": 0.7577,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7199922995475985,
+      "grad_norm": 0.1918313205242157,
+      "learning_rate": 1.923112386205727e-06,
+      "loss": 0.753,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7238425257483877,
+      "grad_norm": 0.2015622854232788,
+      "learning_rate": 1.874200397115079e-06,
+      "loss": 0.7612,
+      "step": 1880
+    },
+    {
+      "epoch": 0.727692751949177,
+      "grad_norm": 0.22247175872325897,
+      "learning_rate": 1.825774590065857e-06,
+      "loss": 0.744,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7315429781499663,
+      "grad_norm": 0.2192992866039276,
+      "learning_rate": 1.7778424971332813e-06,
+      "loss": 0.7552,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7353932043507556,
+      "grad_norm": 0.20429269969463348,
+      "learning_rate": 1.7304115736010407e-06,
+      "loss": 0.7509,
+      "step": 1910
+    },
+    {
+      "epoch": 0.739243430551545,
+      "grad_norm": 0.20754073560237885,
+      "learning_rate": 1.6834891968017108e-06,
+      "loss": 0.7466,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7430936567523342,
+      "grad_norm": 0.212954580783844,
+      "learning_rate": 1.6370826649692884e-06,
+      "loss": 0.7612,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7469438829531235,
+      "grad_norm": 0.22112978994846344,
+      "learning_rate": 1.5911991961040463e-06,
+      "loss": 0.7674,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7507941091539128,
+      "grad_norm": 0.21691519021987915,
+      "learning_rate": 1.545845926849851e-06,
+      "loss": 0.7562,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7546443353547021,
+      "grad_norm": 0.22664524614810944,
+      "learning_rate": 1.5010299113841397e-06,
+      "loss": 0.7735,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7584945615554913,
+      "grad_norm": 0.20755329728126526,
+      "learning_rate": 1.4567581203207238e-06,
+      "loss": 0.7632,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7623447877562807,
+      "grad_norm": 0.21877416968345642,
+      "learning_rate": 1.413037439625593e-06,
+      "loss": 0.7622,
+      "step": 1980
+    },
+    {
+      "epoch": 0.76619501395707,
+      "grad_norm": 0.22153635323047638,
+      "learning_rate": 1.3698746695458847e-06,
+      "loss": 0.7556,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7700452401578592,
+      "grad_norm": 0.2072059065103531,
+      "learning_rate": 1.3272765235521745e-06,
+      "loss": 0.7478,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7738954663586486,
+      "grad_norm": 0.2077191174030304,
+      "learning_rate": 1.2852496272942877e-06,
+      "loss": 0.7576,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7777456925594378,
+      "grad_norm": 0.19917038083076477,
+      "learning_rate": 1.2438005175707396e-06,
+      "loss": 0.7435,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7815959187602272,
+      "grad_norm": 0.20422153174877167,
+      "learning_rate": 1.2029356413120297e-06,
+      "loss": 0.728,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7854461449610165,
+      "grad_norm": 0.20686648786067963,
+      "learning_rate": 1.1626613545778786e-06,
+      "loss": 0.7455,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7892963711618057,
+      "grad_norm": 0.20089182257652283,
+      "learning_rate": 1.1229839215686294e-06,
+      "loss": 0.7504,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7931465973625951,
+      "grad_norm": 0.19403326511383057,
+      "learning_rate": 1.0839095136509188e-06,
+      "loss": 0.7325,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7969968235633843,
+      "grad_norm": 0.22272174060344696,
+      "learning_rate": 1.045444208397791e-06,
+      "loss": 0.7552,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8008470497641736,
+      "grad_norm": 0.1956993043422699,
+      "learning_rate": 1.007593988643402e-06,
+      "loss": 0.7403,
+      "step": 2080
+    },
+    {
+      "epoch": 0.804697275964963,
+      "grad_norm": 0.2041662633419037,
+      "learning_rate": 9.703647415524613e-07,
+      "loss": 0.7607,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8085475021657522,
+      "grad_norm": 0.21103177964687347,
+      "learning_rate": 9.337622577045485e-07,
+      "loss": 0.7491,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8123977283665416,
+      "grad_norm": 0.20936903357505798,
+      "learning_rate": 8.977922301934588e-07,
+      "loss": 0.7591,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8162479545673308,
+      "grad_norm": 0.21097984910011292,
+      "learning_rate": 8.624602537417009e-07,
+      "loss": 0.7598,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8200981807681201,
+      "grad_norm": 0.2232218086719513,
+      "learning_rate": 8.277718238303123e-07,
+      "loss": 0.7455,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8239484069689095,
+      "grad_norm": 0.20916861295700073,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.7507,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8277986331696987,
+      "grad_norm": 0.22733628749847412,
+      "learning_rate": 7.603470842324223e-07,
+      "loss": 0.7403,
+      "step": 2150
+    },
+    {
+      "epoch": 0.831648859370488,
+      "grad_norm": 0.20365650951862335,
+      "learning_rate": 7.276212616857575e-07,
+      "loss": 0.7597,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8354990855712773,
+      "grad_norm": 0.2051434963941574,
+      "learning_rate": 6.955599583279837e-07,
+      "loss": 0.759,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8393493117720666,
+      "grad_norm": 0.2052781581878662,
+      "learning_rate": 6.641681609246981e-07,
+      "loss": 0.7509,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8431995379728559,
+      "grad_norm": 0.2215193510055542,
+      "learning_rate": 6.334507521075711e-07,
+      "loss": 0.7578,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8470497641736452,
+      "grad_norm": 0.22075481712818146,
+      "learning_rate": 6.034125096149179e-07,
+      "loss": 0.7539,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8508999903744345,
+      "grad_norm": 0.20797300338745117,
+      "learning_rate": 5.740581055485722e-07,
+      "loss": 0.7234,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8547502165752238,
+      "grad_norm": 0.197932168841362,
+      "learning_rate": 5.453921056471961e-07,
+      "loss": 0.7563,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8586004427760131,
+      "grad_norm": 0.20865581929683685,
+      "learning_rate": 5.174189685761288e-07,
+      "loss": 0.7532,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8624506689768023,
+      "grad_norm": 0.19626477360725403,
+      "learning_rate": 4.901430452338979e-07,
+      "loss": 0.7633,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8663008951775917,
+      "grad_norm": 0.20305222272872925,
+      "learning_rate": 4.6356857807548286e-07,
+      "loss": 0.748,
+      "step": 2250
+    },
+    {
+      "epoch": 0.870151121378381,
+      "grad_norm": 0.20897530019283295,
+      "learning_rate": 4.376997004524541e-07,
+      "loss": 0.7372,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8740013475791703,
+      "grad_norm": 0.2076602727174759,
+      "learning_rate": 4.12540435970073e-07,
+      "loss": 0.7449,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8778515737799596,
+      "grad_norm": 0.21175378561019897,
+      "learning_rate": 3.880946978614703e-07,
+      "loss": 0.7567,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8817017999807488,
+      "grad_norm": 0.2099006325006485,
+      "learning_rate": 3.643662883789878e-07,
+      "loss": 0.7596,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8855520261815382,
+      "grad_norm": 0.19856323301792145,
+      "learning_rate": 3.41358898202781e-07,
+      "loss": 0.7388,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8894022523823275,
+      "grad_norm": 0.19641050696372986,
+      "learning_rate": 3.190761058667741e-07,
+      "loss": 0.7507,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8932524785831167,
+      "grad_norm": 0.20263437926769257,
+      "learning_rate": 2.9752137720206463e-07,
+      "loss": 0.754,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8971027047839061,
+      "grad_norm": 0.2089318484067917,
+      "learning_rate": 2.7669806479785067e-07,
+      "loss": 0.7592,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9009529309846953,
+      "grad_norm": 0.20733408629894257,
+      "learning_rate": 2.566094074799763e-07,
+      "loss": 0.7525,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9048031571854847,
+      "grad_norm": 0.20699486136436462,
+      "learning_rate": 2.3725852980716557e-07,
+      "loss": 0.7379,
+      "step": 2350
+    },
+    {
+      "epoch": 0.908653383386274,
+      "grad_norm": 0.20353306829929352,
+      "learning_rate": 2.1864844158503983e-07,
+      "loss": 0.7484,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9125036095870632,
+      "grad_norm": 0.22034503519535065,
+      "learning_rate": 2.007820373979702e-07,
+      "loss": 0.7434,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9163538357878526,
+      "grad_norm": 0.21327997744083405,
+      "learning_rate": 1.8366209615886532e-07,
+      "loss": 0.7514,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9202040619886418,
+      "grad_norm": 0.2092486470937729,
+      "learning_rate": 1.6729128067693613e-07,
+      "loss": 0.763,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9240542881894311,
+      "grad_norm": 0.2224603146314621,
+      "learning_rate": 1.5167213724353426e-07,
+      "loss": 0.7624,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9279045143902205,
+      "grad_norm": 0.20579323172569275,
+      "learning_rate": 1.3680709523610058e-07,
+      "loss": 0.7424,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9317547405910097,
+      "grad_norm": 0.20996473729610443,
+      "learning_rate": 1.2269846674030418e-07,
+      "loss": 0.7674,
+      "step": 2420
+    },
+    {
+      "epoch": 0.935604966791799,
+      "grad_norm": 0.21724142134189606,
+      "learning_rate": 1.0934844619042595e-07,
+      "loss": 0.7626,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9394551929925883,
+      "grad_norm": 0.22044630348682404,
+      "learning_rate": 9.675911002803961e-08,
+      "loss": 0.751,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9433054191933776,
+      "grad_norm": 0.2259121686220169,
+      "learning_rate": 8.493241637904381e-08,
+      "loss": 0.7483,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9471556453941669,
+      "grad_norm": 0.21290327608585358,
+      "learning_rate": 7.387020474909689e-08,
+      "loss": 0.7681,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9510058715949562,
+      "grad_norm": 0.22176212072372437,
+      "learning_rate": 6.357419573750856e-08,
+      "loss": 0.7629,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9548560977957455,
+      "grad_norm": 0.20919044315814972,
+      "learning_rate": 5.4045990769613365e-08,
+      "loss": 0.7441,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9587063239965348,
+      "grad_norm": 0.21189755201339722,
+      "learning_rate": 4.5287071847694434e-08,
+      "loss": 0.7506,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9625565501973241,
+      "grad_norm": 0.20441526174545288,
+      "learning_rate": 3.7298801320467817e-08,
+      "loss": 0.7482,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9664067763981133,
+      "grad_norm": 0.20946770906448364,
+      "learning_rate": 3.0082421671192576e-08,
+      "loss": 0.7414,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9702570025989027,
+      "grad_norm": 0.20916658639907837,
+      "learning_rate": 2.3639055324410355e-08,
+      "loss": 0.7616,
+      "step": 2520
+    },
+    {
+      "epoch": 0.974107228799692,
+      "grad_norm": 0.20734639465808868,
+      "learning_rate": 1.7969704471369477e-08,
+      "loss": 0.7468,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9779574550004813,
+      "grad_norm": 0.20028896629810333,
+      "learning_rate": 1.3075250914142413e-08,
+      "loss": 0.7478,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9818076812012706,
+      "grad_norm": 0.20245103538036346,
+      "learning_rate": 8.956455928476049e-09,
+      "loss": 0.757,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9856579074020598,
+      "grad_norm": 0.20345163345336914,
+      "learning_rate": 5.6139601453786275e-09,
+      "loss": 0.7493,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9895081336028492,
+      "grad_norm": 0.2047918289899826,
+      "learning_rate": 3.0482834514805602e-09,
+      "loss": 0.7391,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9933583598036385,
+      "grad_norm": 0.20237474143505096,
+      "learning_rate": 1.2598249081713366e-09,
+      "loss": 0.7306,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9972085860044277,
+      "grad_norm": 0.21342609822750092,
+      "learning_rate": 2.488626895302826e-10,
+      "loss": 0.7641,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9999037443449803,
+      "step": 2597,
+      "total_flos": 3.3979476276590674e+18,
+      "train_loss": 0.7725522557634273,
+      "train_runtime": 8184.592,
+      "train_samples_per_second": 81.238,
+      "train_steps_per_second": 0.317
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2597,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.3979476276590674e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen25/mc2_m16_224_mt5/finetune/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3f7115b70074cf3a3bc9e8e913f3c3e9db7a8964a706cad816133915d099419
+size 6136

continuous/qwen25/mc2_m16_224_mt5/pretrain/config.json

@@ -0,0 +1,58 @@
+{
+  "_attn_implementation_autoset": true,
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 21,
+  "mm_hidden_size": 512,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "flat",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_tunable_parts": "mm_mlp_adapter",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "open_clip_hub:ViT-M-16",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen2",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 32768,
+  "tie_word_embeddings": true,
+  "tokenizer_model_max_length": 8192,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.3",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_tower_pretrained": "/cache/ckpt/download/tokenizer/continuous/mc2/mc2_m16_224_mt5.pt",
+  "vocab_size": 151936
+}

continuous/qwen25/mc2_m16_224_mt5/pretrain/mm_projector.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2bcb11ca67ab82f37344b3b603e6c1099f13c4b98b791d74a2eaf4298c8b7799
+size 12597304

continuous/qwen25/mc2_m16_224_mt5/pretrain/trainer_state.json

@@ -0,0 +1,421 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 545,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01834862385321101,
+      "grad_norm": 2.5407330989837646,
+      "learning_rate": 0.00032142857142857147,
+      "loss": 4.9861,
+      "step": 10
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 2.388478994369507,
+      "learning_rate": 0.0006785714285714287,
+      "loss": 3.7066,
+      "step": 20
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 0.23131787776947021,
+      "learning_rate": 0.0009999907688176173,
+      "loss": 3.3009,
+      "step": 30
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 0.19612620770931244,
+      "learning_rate": 0.0009988834393115766,
+      "loss": 3.2115,
+      "step": 40
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 0.14647743105888367,
+      "learning_rate": 0.0009959345572940772,
+      "loss": 3.1723,
+      "step": 50
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 0.1307649314403534,
+      "learning_rate": 0.0009911550081155983,
+      "loss": 3.122,
+      "step": 60
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.17433635890483856,
+      "learning_rate": 0.0009845624347566433,
+      "loss": 3.0736,
+      "step": 70
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 0.19048425555229187,
+      "learning_rate": 0.0009761811727013547,
+      "loss": 3.0331,
+      "step": 80
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 0.19643640518188477,
+      "learning_rate": 0.0009660421601067666,
+      "loss": 2.9892,
+      "step": 90
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 0.1609342396259308,
+      "learning_rate": 0.0009541828235992926,
+      "loss": 2.9628,
+      "step": 100
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 0.22121073305606842,
+      "learning_rate": 0.0009406469401200151,
+      "loss": 2.9307,
+      "step": 110
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 0.2380785495042801,
+      "learning_rate": 0.0009254844753287492,
+      "loss": 2.9183,
+      "step": 120
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 0.24339230358600616,
+      "learning_rate": 0.0009087513991633924,
+      "loss": 2.8844,
+      "step": 130
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 0.16946633160114288,
+      "learning_rate": 0.0008905094792353917,
+      "loss": 2.866,
+      "step": 140
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 0.1883004754781723,
+      "learning_rate": 0.0008708260528239789,
+      "loss": 2.8571,
+      "step": 150
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.2265540510416031,
+      "learning_rate": 0.0008497737783108237,
+      "loss": 2.8272,
+      "step": 160
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 0.2263176590204239,
+      "learning_rate": 0.0008274303669726426,
+      "loss": 2.8033,
+      "step": 170
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 0.20723441243171692,
+      "learning_rate": 0.0008038782961218137,
+      "loss": 2.7739,
+      "step": 180
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.20400987565517426,
+      "learning_rate": 0.000779204504653894,
+      "loss": 2.753,
+      "step": 190
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 0.19698961079120636,
+      "learning_rate": 0.0007535000721258791,
+      "loss": 2.724,
+      "step": 200
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 0.212701216340065,
+      "learning_rate": 0.00072685988254984,
+      "loss": 2.7125,
+      "step": 210
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 0.24121539294719696,
+      "learning_rate": 0.0006993822741429906,
+      "loss": 2.7173,
+      "step": 220
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.20718953013420105,
+      "learning_rate": 0.0006711686763270818,
+      "loss": 2.7081,
+      "step": 230
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 0.19447745382785797,
+      "learning_rate": 0.0006423232353170831,
+      "loss": 2.6926,
+      "step": 240
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 0.1900073140859604,
+      "learning_rate": 0.0006129524296812335,
+      "loss": 2.6872,
+      "step": 250
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 0.17140918970108032,
+      "learning_rate": 0.0005831646772915651,
+      "loss": 2.6935,
+      "step": 260
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.19301597774028778,
+      "learning_rate": 0.0005530699351157809,
+      "loss": 2.6727,
+      "step": 270
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 0.20544083416461945,
+      "learning_rate": 0.0005227792933277943,
+      "loss": 2.6745,
+      "step": 280
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 0.17813275754451752,
+      "learning_rate": 0.0004924045652352092,
+      "loss": 2.6609,
+      "step": 290
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 0.1750490665435791,
+      "learning_rate": 0.0004620578745374633,
+      "loss": 2.6585,
+      "step": 300
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 0.2616715729236603,
+      "learning_rate": 0.0004318512414382059,
+      "loss": 2.6658,
+      "step": 310
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 0.22556813061237335,
+      "learning_rate": 0.0004018961691397148,
+      "loss": 2.6623,
+      "step": 320
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 0.16471098363399506,
+      "learning_rate": 0.0003723032322457458,
+      "loss": 2.6481,
+      "step": 330
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.2748624086380005,
+      "learning_rate": 0.0003431816685921625,
+      "loss": 2.6505,
+      "step": 340
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 0.1866309493780136,
+      "learning_rate": 0.0003146389760120416,
+      "loss": 2.6489,
+      "step": 350
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 0.2029782086610794,
+      "learning_rate": 0.00028678051552373485,
+      "loss": 2.6532,
+      "step": 360
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 0.16878151893615723,
+      "learning_rate": 0.0002597091224066581,
+      "loss": 2.6434,
+      "step": 370
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 0.1693291813135147,
+      "learning_rate": 0.0002335247266004629,
+      "loss": 2.6537,
+      "step": 380
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 0.2060406506061554,
+      "learning_rate": 0.000208323983828828,
+      "loss": 2.642,
+      "step": 390
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.16578562557697296,
+      "learning_rate": 0.0001841999188095224,
+      "loss": 2.6343,
+      "step": 400
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 0.171912282705307,
+      "learning_rate": 0.00016124158186777676,
+      "loss": 2.6368,
+      "step": 410
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 0.17923152446746826,
+      "learning_rate": 0.0001395337202205211,
+      "loss": 2.6358,
+      "step": 420
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 0.1589379459619522,
+      "learning_rate": 0.00011915646514489292,
+      "loss": 2.6325,
+      "step": 430
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 0.17125694453716278,
+      "learning_rate": 0.00010018503618578818,
+      "loss": 2.6429,
+      "step": 440
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.15847378969192505,
+      "learning_rate": 8.268946349432583e-05,
+      "loss": 2.6343,
+      "step": 450
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 0.16015057265758514,
+      "learning_rate": 6.673432932217117e-05,
+      "loss": 2.6302,
+      "step": 460
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.1783066987991333,
+      "learning_rate": 5.2378529625954684e-05,
+      "loss": 2.6195,
+      "step": 470
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.1806861013174057,
+      "learning_rate": 3.9675056661785556e-05,
+      "loss": 2.6241,
+      "step": 480
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.17161089181900024,
+      "learning_rate": 2.8670803372379538e-05,
+      "loss": 2.6316,
+      "step": 490
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 0.1572864055633545,
+      "learning_rate": 1.9406390288876586e-05,
+      "loss": 2.6405,
+      "step": 500
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 0.17861419916152954,
+      "learning_rate": 1.191601558631461e-05,
+      "loss": 2.6279,
+      "step": 510
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 0.1613374650478363,
+      "learning_rate": 6.2273288462559305e-06,
+      "loss": 2.6241,
+      "step": 520
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 0.14807172119617462,
+      "learning_rate": 2.361328992552314e-06,
+      "loss": 2.6293,
+      "step": 530
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 0.15403467416763306,
+      "learning_rate": 3.322867770044202e-07,
+      "loss": 2.6231,
+      "step": 540
+    },
+    {
+      "epoch": 1.0,
+      "step": 545,
+      "total_flos": 2.789803883013079e+17,
+      "train_loss": 2.8152197706590005,
+      "train_runtime": 5321.336,
+      "train_samples_per_second": 104.885,
+      "train_steps_per_second": 0.102
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 545,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.789803883013079e+17,
+  "train_batch_size": 128,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen25/mc2_m16_384/finetune/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

continuous/qwen25/mc2_m16_384/finetune/config.json

@@ -0,0 +1,57 @@
+{
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 21,
+  "mm_hidden_size": 512,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "flat",
+  "mm_projector_lr": 0.0001,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": null,
+  "mm_tunable_parts": "mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "open_clip_hub:ViT-M-16",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen2",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 32768,
+  "tie_word_embeddings": true,
+  "tokenizer_model_max_length": 8192,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.3",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_tower_pretrained": "/cache/ckpt/download/tokenizer/continuous/mc2/mc2_m16_384.pt",
+  "vocab_size": 151936
+}

continuous/qwen25/mc2_m16_384/finetune/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.1,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.51.3"
+}

continuous/qwen25/mc2_m16_384/finetune/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3f3dbeec65a0133b96c57abc7b1d487f0ce74ada5ee85daef207927b9a99c4b
+size 3644396136

continuous/qwen25/mc2_m16_384/finetune/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_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|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

continuous/qwen25/mc2_m16_384/finetune/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

continuous/qwen25/mc2_m16_384/finetune/tokenizer_config.json

@@ -0,0 +1,209 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_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|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# 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>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\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\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 8192,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

continuous/qwen25/mc2_m16_384/finetune/trainer_state.json

@@ -0,0 +1,1856 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999037443449803,
+  "eval_steps": 500,
+  "global_step": 2597,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003850226200789296,
+      "grad_norm": 7.502678871154785,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 1.9098,
+      "step": 10
+    },
+    {
+      "epoch": 0.007700452401578592,
+      "grad_norm": 2.763319969177246,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 1.519,
+      "step": 20
+    },
+    {
+      "epoch": 0.01155067860236789,
+      "grad_norm": 1.8780262470245361,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.1006,
+      "step": 30
+    },
+    {
+      "epoch": 0.015400904803157185,
+      "grad_norm": 0.3534908890724182,
+      "learning_rate": 5e-06,
+      "loss": 0.9235,
+      "step": 40
+    },
+    {
+      "epoch": 0.019251131003946483,
+      "grad_norm": 0.20123635232448578,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 0.899,
+      "step": 50
+    },
+    {
+      "epoch": 0.02310135720473578,
+      "grad_norm": 0.16160890460014343,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 0.8555,
+      "step": 60
+    },
+    {
+      "epoch": 0.026951583405525073,
+      "grad_norm": 0.12631648778915405,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.8309,
+      "step": 70
+    },
+    {
+      "epoch": 0.03080180960631437,
+      "grad_norm": 0.12579382956027985,
+      "learning_rate": 9.999996111488723e-06,
+      "loss": 0.8384,
+      "step": 80
+    },
+    {
+      "epoch": 0.034652035807103666,
+      "grad_norm": 0.12207123637199402,
+      "learning_rate": 9.999529497453782e-06,
+      "loss": 0.8276,
+      "step": 90
+    },
+    {
+      "epoch": 0.038502262007892966,
+      "grad_norm": 0.12030481547117233,
+      "learning_rate": 9.998285264324193e-06,
+      "loss": 0.8032,
+      "step": 100
+    },
+    {
+      "epoch": 0.04235248820868226,
+      "grad_norm": 0.11470679938793182,
+      "learning_rate": 9.996263605626052e-06,
+      "loss": 0.7906,
+      "step": 110
+    },
+    {
+      "epoch": 0.04620271440947156,
+      "grad_norm": 0.12326677143573761,
+      "learning_rate": 9.99346483580503e-06,
+      "loss": 0.802,
+      "step": 120
+    },
+    {
+      "epoch": 0.05005294061026085,
+      "grad_norm": 0.1096850037574768,
+      "learning_rate": 9.989889390177467e-06,
+      "loss": 0.7853,
+      "step": 130
+    },
+    {
+      "epoch": 0.053903166811050146,
+      "grad_norm": 0.12255839258432388,
+      "learning_rate": 9.985537824862647e-06,
+      "loss": 0.7913,
+      "step": 140
+    },
+    {
+      "epoch": 0.057753393011839446,
+      "grad_norm": 0.13147203624248505,
+      "learning_rate": 9.980410816696317e-06,
+      "loss": 0.7834,
+      "step": 150
+    },
+    {
+      "epoch": 0.06160361921262874,
+      "grad_norm": 0.11926069855690002,
+      "learning_rate": 9.97450916312541e-06,
+      "loss": 0.7827,
+      "step": 160
+    },
+    {
+      "epoch": 0.06545384541341803,
+      "grad_norm": 0.11015338450670242,
+      "learning_rate": 9.967833782084004e-06,
+      "loss": 0.7808,
+      "step": 170
+    },
+    {
+      "epoch": 0.06930407161420733,
+      "grad_norm": 0.12126290798187256,
+      "learning_rate": 9.960385711850554e-06,
+      "loss": 0.8024,
+      "step": 180
+    },
+    {
+      "epoch": 0.07315429781499663,
+      "grad_norm": 0.12068787962198257,
+      "learning_rate": 9.952166110886394e-06,
+      "loss": 0.8068,
+      "step": 190
+    },
+    {
+      "epoch": 0.07700452401578593,
+      "grad_norm": 0.1247032880783081,
+      "learning_rate": 9.943176257655567e-06,
+      "loss": 0.7845,
+      "step": 200
+    },
+    {
+      "epoch": 0.08085475021657522,
+      "grad_norm": 0.11361949890851974,
+      "learning_rate": 9.933417550425952e-06,
+      "loss": 0.7913,
+      "step": 210
+    },
+    {
+      "epoch": 0.08470497641736452,
+      "grad_norm": 0.11030127853155136,
+      "learning_rate": 9.9228915070518e-06,
+      "loss": 0.7674,
+      "step": 220
+    },
+    {
+      "epoch": 0.08855520261815382,
+      "grad_norm": 0.12313994765281677,
+      "learning_rate": 9.911599764737632e-06,
+      "loss": 0.7743,
+      "step": 230
+    },
+    {
+      "epoch": 0.09240542881894312,
+      "grad_norm": 0.12048649042844772,
+      "learning_rate": 9.899544079783606e-06,
+      "loss": 0.7602,
+      "step": 240
+    },
+    {
+      "epoch": 0.0962556550197324,
+      "grad_norm": 0.11456608027219772,
+      "learning_rate": 9.886726327312338e-06,
+      "loss": 0.778,
+      "step": 250
+    },
+    {
+      "epoch": 0.1001058812205217,
+      "grad_norm": 0.12082139402627945,
+      "learning_rate": 9.87314850097724e-06,
+      "loss": 0.7819,
+      "step": 260
+    },
+    {
+      "epoch": 0.103956107421311,
+      "grad_norm": 0.10880251228809357,
+      "learning_rate": 9.858812712652452e-06,
+      "loss": 0.7693,
+      "step": 270
+    },
+    {
+      "epoch": 0.10780633362210029,
+      "grad_norm": 0.10744450241327286,
+      "learning_rate": 9.843721192104341e-06,
+      "loss": 0.7718,
+      "step": 280
+    },
+    {
+      "epoch": 0.11165655982288959,
+      "grad_norm": 0.1138581931591034,
+      "learning_rate": 9.827876286644698e-06,
+      "loss": 0.7874,
+      "step": 290
+    },
+    {
+      "epoch": 0.11550678602367889,
+      "grad_norm": 0.10702113062143326,
+      "learning_rate": 9.811280460765643e-06,
+      "loss": 0.75,
+      "step": 300
+    },
+    {
+      "epoch": 0.11935701222446819,
+      "grad_norm": 0.11405590921640396,
+      "learning_rate": 9.793936295756292e-06,
+      "loss": 0.7628,
+      "step": 310
+    },
+    {
+      "epoch": 0.12320723842525748,
+      "grad_norm": 0.11720357835292816,
+      "learning_rate": 9.775846489301279e-06,
+      "loss": 0.7625,
+      "step": 320
+    },
+    {
+      "epoch": 0.12705746462604678,
+      "grad_norm": 0.11461129784584045,
+      "learning_rate": 9.757013855061157e-06,
+      "loss": 0.763,
+      "step": 330
+    },
+    {
+      "epoch": 0.13090769082683607,
+      "grad_norm": 0.11235353350639343,
+      "learning_rate": 9.737441322234758e-06,
+      "loss": 0.773,
+      "step": 340
+    },
+    {
+      "epoch": 0.13475791702762538,
+      "grad_norm": 0.12045136839151382,
+      "learning_rate": 9.717131935103605e-06,
+      "loss": 0.7785,
+      "step": 350
+    },
+    {
+      "epoch": 0.13860814322841467,
+      "grad_norm": 0.11057382076978683,
+      "learning_rate": 9.696088852558398e-06,
+      "loss": 0.7623,
+      "step": 360
+    },
+    {
+      "epoch": 0.14245836942920398,
+      "grad_norm": 0.10918495059013367,
+      "learning_rate": 9.674315347607686e-06,
+      "loss": 0.7859,
+      "step": 370
+    },
+    {
+      "epoch": 0.14630859562999327,
+      "grad_norm": 0.11445240676403046,
+      "learning_rate": 9.651814806868796e-06,
+      "loss": 0.76,
+      "step": 380
+    },
+    {
+      "epoch": 0.15015882183078255,
+      "grad_norm": 0.1191505715250969,
+      "learning_rate": 9.62859073004107e-06,
+      "loss": 0.7607,
+      "step": 390
+    },
+    {
+      "epoch": 0.15400904803157187,
+      "grad_norm": 0.10534489154815674,
+      "learning_rate": 9.604646729361547e-06,
+      "loss": 0.7636,
+      "step": 400
+    },
+    {
+      "epoch": 0.15785927423236115,
+      "grad_norm": 0.1101505309343338,
+      "learning_rate": 9.579986529043099e-06,
+      "loss": 0.7465,
+      "step": 410
+    },
+    {
+      "epoch": 0.16170950043315044,
+      "grad_norm": 0.11563799530267715,
+      "learning_rate": 9.554613964695189e-06,
+      "loss": 0.7622,
+      "step": 420
+    },
+    {
+      "epoch": 0.16555972663393975,
+      "grad_norm": 0.11573365330696106,
+      "learning_rate": 9.528532982727281e-06,
+      "loss": 0.7572,
+      "step": 430
+    },
+    {
+      "epoch": 0.16940995283472904,
+      "grad_norm": 0.11263683438301086,
+      "learning_rate": 9.501747639735018e-06,
+      "loss": 0.7559,
+      "step": 440
+    },
+    {
+      "epoch": 0.17326017903551832,
+      "grad_norm": 0.11174523830413818,
+      "learning_rate": 9.474262101869272e-06,
+      "loss": 0.7403,
+      "step": 450
+    },
+    {
+      "epoch": 0.17711040523630764,
+      "grad_norm": 0.11205888539552689,
+      "learning_rate": 9.446080644188141e-06,
+      "loss": 0.7569,
+      "step": 460
+    },
+    {
+      "epoch": 0.18096063143709692,
+      "grad_norm": 0.11216938495635986,
+      "learning_rate": 9.417207649992017e-06,
+      "loss": 0.7577,
+      "step": 470
+    },
+    {
+      "epoch": 0.18481085763788624,
+      "grad_norm": 0.10909686237573624,
+      "learning_rate": 9.387647610141816e-06,
+      "loss": 0.7513,
+      "step": 480
+    },
+    {
+      "epoch": 0.18866108383867553,
+      "grad_norm": 0.10078215599060059,
+      "learning_rate": 9.35740512236047e-06,
+      "loss": 0.7437,
+      "step": 490
+    },
+    {
+      "epoch": 0.1925113100394648,
+      "grad_norm": 0.10402926057577133,
+      "learning_rate": 9.326484890517809e-06,
+      "loss": 0.7532,
+      "step": 500
+    },
+    {
+      "epoch": 0.19636153624025413,
+      "grad_norm": 0.12178223580121994,
+      "learning_rate": 9.294891723898926e-06,
+      "loss": 0.7339,
+      "step": 510
+    },
+    {
+      "epoch": 0.2002117624410434,
+      "grad_norm": 0.11464907228946686,
+      "learning_rate": 9.262630536456153e-06,
+      "loss": 0.7724,
+      "step": 520
+    },
+    {
+      "epoch": 0.2040619886418327,
+      "grad_norm": 0.119774729013443,
+      "learning_rate": 9.229706346044749e-06,
+      "loss": 0.7428,
+      "step": 530
+    },
+    {
+      "epoch": 0.207912214842622,
+      "grad_norm": 0.12146639078855515,
+      "learning_rate": 9.196124273642422e-06,
+      "loss": 0.7607,
+      "step": 540
+    },
+    {
+      "epoch": 0.2117624410434113,
+      "grad_norm": 0.10846410691738129,
+      "learning_rate": 9.161889542552842e-06,
+      "loss": 0.7671,
+      "step": 550
+    },
+    {
+      "epoch": 0.21561266724420058,
+      "grad_norm": 0.10587923973798752,
+      "learning_rate": 9.127007477593193e-06,
+      "loss": 0.7644,
+      "step": 560
+    },
+    {
+      "epoch": 0.2194628934449899,
+      "grad_norm": 0.11945808678865433,
+      "learning_rate": 9.091483504265974e-06,
+      "loss": 0.7759,
+      "step": 570
+    },
+    {
+      "epoch": 0.22331311964577918,
+      "grad_norm": 0.11162535101175308,
+      "learning_rate": 9.055323147915116e-06,
+      "loss": 0.7488,
+      "step": 580
+    },
+    {
+      "epoch": 0.2271633458465685,
+      "grad_norm": 0.11383433640003204,
+      "learning_rate": 9.018532032866585e-06,
+      "loss": 0.7672,
+      "step": 590
+    },
+    {
+      "epoch": 0.23101357204735778,
+      "grad_norm": 0.11380232125520706,
+      "learning_rate": 8.981115881553579e-06,
+      "loss": 0.7637,
+      "step": 600
+    },
+    {
+      "epoch": 0.23486379824814707,
+      "grad_norm": 0.11867867410182953,
+      "learning_rate": 8.943080513626476e-06,
+      "loss": 0.7752,
+      "step": 610
+    },
+    {
+      "epoch": 0.23871402444893638,
+      "grad_norm": 0.11545798927545547,
+      "learning_rate": 8.904431845047641e-06,
+      "loss": 0.7653,
+      "step": 620
+    },
+    {
+      "epoch": 0.24256425064972567,
+      "grad_norm": 0.1185082197189331,
+      "learning_rate": 8.86517588717128e-06,
+      "loss": 0.7702,
+      "step": 630
+    },
+    {
+      "epoch": 0.24641447685051496,
+      "grad_norm": 0.10728500783443451,
+      "learning_rate": 8.82531874580844e-06,
+      "loss": 0.769,
+      "step": 640
+    },
+    {
+      "epoch": 0.25026470305130427,
+      "grad_norm": 0.10467024892568588,
+      "learning_rate": 8.784866620277314e-06,
+      "loss": 0.7268,
+      "step": 650
+    },
+    {
+      "epoch": 0.25411492925209356,
+      "grad_norm": 0.1218658834695816,
+      "learning_rate": 8.743825802439018e-06,
+      "loss": 0.7464,
+      "step": 660
+    },
+    {
+      "epoch": 0.25796515545288284,
+      "grad_norm": 0.1160091683268547,
+      "learning_rate": 8.702202675718956e-06,
+      "loss": 0.7674,
+      "step": 670
+    },
+    {
+      "epoch": 0.26181538165367213,
+      "grad_norm": 0.10711445659399033,
+      "learning_rate": 8.660003714113959e-06,
+      "loss": 0.7417,
+      "step": 680
+    },
+    {
+      "epoch": 0.26566560785446147,
+      "grad_norm": 0.11315526813268661,
+      "learning_rate": 8.617235481185323e-06,
+      "loss": 0.752,
+      "step": 690
+    },
+    {
+      "epoch": 0.26951583405525076,
+      "grad_norm": 0.1117711141705513,
+      "learning_rate": 8.573904629037931e-06,
+      "loss": 0.7591,
+      "step": 700
+    },
+    {
+      "epoch": 0.27336606025604004,
+      "grad_norm": 0.1101578027009964,
+      "learning_rate": 8.530017897285585e-06,
+      "loss": 0.7453,
+      "step": 710
+    },
+    {
+      "epoch": 0.27721628645682933,
+      "grad_norm": 0.10174145549535751,
+      "learning_rate": 8.485582112002746e-06,
+      "loss": 0.7514,
+      "step": 720
+    },
+    {
+      "epoch": 0.2810665126576186,
+      "grad_norm": 0.11300273984670639,
+      "learning_rate": 8.440604184662811e-06,
+      "loss": 0.7565,
+      "step": 730
+    },
+    {
+      "epoch": 0.28491673885840796,
+      "grad_norm": 0.11383278667926788,
+      "learning_rate": 8.395091111063118e-06,
+      "loss": 0.7423,
+      "step": 740
+    },
+    {
+      "epoch": 0.28876696505919724,
+      "grad_norm": 0.1038803979754448,
+      "learning_rate": 8.349049970236822e-06,
+      "loss": 0.7432,
+      "step": 750
+    },
+    {
+      "epoch": 0.29261719125998653,
+      "grad_norm": 0.10698353499174118,
+      "learning_rate": 8.302487923351842e-06,
+      "loss": 0.7329,
+      "step": 760
+    },
+    {
+      "epoch": 0.2964674174607758,
+      "grad_norm": 0.11284562200307846,
+      "learning_rate": 8.255412212597021e-06,
+      "loss": 0.7505,
+      "step": 770
+    },
+    {
+      "epoch": 0.3003176436615651,
+      "grad_norm": 0.11700551956892014,
+      "learning_rate": 8.207830160055688e-06,
+      "loss": 0.7471,
+      "step": 780
+    },
+    {
+      "epoch": 0.3041678698623544,
+      "grad_norm": 0.1159774586558342,
+      "learning_rate": 8.159749166566786e-06,
+      "loss": 0.7365,
+      "step": 790
+    },
+    {
+      "epoch": 0.30801809606314373,
+      "grad_norm": 0.11501676589250565,
+      "learning_rate": 8.11117671057377e-06,
+      "loss": 0.7625,
+      "step": 800
+    },
+    {
+      "epoch": 0.311868322263933,
+      "grad_norm": 0.10963020473718643,
+      "learning_rate": 8.062120346961404e-06,
+      "loss": 0.7477,
+      "step": 810
+    },
+    {
+      "epoch": 0.3157185484647223,
+      "grad_norm": 0.10343271493911743,
+      "learning_rate": 8.012587705880705e-06,
+      "loss": 0.7536,
+      "step": 820
+    },
+    {
+      "epoch": 0.3195687746655116,
+      "grad_norm": 0.11455662548542023,
+      "learning_rate": 7.962586491562146e-06,
+      "loss": 0.739,
+      "step": 830
+    },
+    {
+      "epoch": 0.3234190008663009,
+      "grad_norm": 0.1201857477426529,
+      "learning_rate": 7.912124481117357e-06,
+      "loss": 0.7417,
+      "step": 840
+    },
+    {
+      "epoch": 0.3272692270670902,
+      "grad_norm": 0.10914399474859238,
+      "learning_rate": 7.861209523329494e-06,
+      "loss": 0.7461,
+      "step": 850
+    },
+    {
+      "epoch": 0.3311194532678795,
+      "grad_norm": 0.10768371820449829,
+      "learning_rate": 7.809849537432432e-06,
+      "loss": 0.7378,
+      "step": 860
+    },
+    {
+      "epoch": 0.3349696794686688,
+      "grad_norm": 0.1089480072259903,
+      "learning_rate": 7.758052511879035e-06,
+      "loss": 0.7369,
+      "step": 870
+    },
+    {
+      "epoch": 0.3388199056694581,
+      "grad_norm": 0.1081153079867363,
+      "learning_rate": 7.705826503098642e-06,
+      "loss": 0.7441,
+      "step": 880
+    },
+    {
+      "epoch": 0.34267013187024736,
+      "grad_norm": 0.10838697850704193,
+      "learning_rate": 7.653179634243979e-06,
+      "loss": 0.7346,
+      "step": 890
+    },
+    {
+      "epoch": 0.34652035807103665,
+      "grad_norm": 0.10063482820987701,
+      "learning_rate": 7.600120093927696e-06,
+      "loss": 0.7339,
+      "step": 900
+    },
+    {
+      "epoch": 0.350370584271826,
+      "grad_norm": 0.1139690950512886,
+      "learning_rate": 7.546656134948729e-06,
+      "loss": 0.7307,
+      "step": 910
+    },
+    {
+      "epoch": 0.3542208104726153,
+      "grad_norm": 0.10656514763832092,
+      "learning_rate": 7.492796073008661e-06,
+      "loss": 0.7261,
+      "step": 920
+    },
+    {
+      "epoch": 0.35807103667340456,
+      "grad_norm": 0.11891157925128937,
+      "learning_rate": 7.438548285418326e-06,
+      "loss": 0.7219,
+      "step": 930
+    },
+    {
+      "epoch": 0.36192126287419385,
+      "grad_norm": 0.12113171815872192,
+      "learning_rate": 7.3839212097948e-06,
+      "loss": 0.7577,
+      "step": 940
+    },
+    {
+      "epoch": 0.36577148907498314,
+      "grad_norm": 0.10975494235754013,
+      "learning_rate": 7.328923342749039e-06,
+      "loss": 0.7664,
+      "step": 950
+    },
+    {
+      "epoch": 0.3696217152757725,
+      "grad_norm": 0.11863745003938675,
+      "learning_rate": 7.273563238564319e-06,
+      "loss": 0.7478,
+      "step": 960
+    },
+    {
+      "epoch": 0.37347194147656176,
+      "grad_norm": 0.11509817838668823,
+      "learning_rate": 7.217849507865724e-06,
+      "loss": 0.7273,
+      "step": 970
+    },
+    {
+      "epoch": 0.37732216767735105,
+      "grad_norm": 0.10476287454366684,
+      "learning_rate": 7.161790816280858e-06,
+      "loss": 0.7386,
+      "step": 980
+    },
+    {
+      "epoch": 0.38117239387814034,
+      "grad_norm": 0.11558035761117935,
+      "learning_rate": 7.1053958830920125e-06,
+      "loss": 0.7189,
+      "step": 990
+    },
+    {
+      "epoch": 0.3850226200789296,
+      "grad_norm": 0.10650355368852615,
+      "learning_rate": 7.0486734798799715e-06,
+      "loss": 0.7454,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3888728462797189,
+      "grad_norm": 0.10894986242055893,
+      "learning_rate": 6.991632429159708e-06,
+      "loss": 0.7649,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39272307248050825,
+      "grad_norm": 0.11336119472980499,
+      "learning_rate": 6.934281603008132e-06,
+      "loss": 0.7441,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39657329868129754,
+      "grad_norm": 0.10470680892467499,
+      "learning_rate": 6.876629921684143e-06,
+      "loss": 0.7544,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4004235248820868,
+      "grad_norm": 0.10975487530231476,
+      "learning_rate": 6.818686352241199e-06,
+      "loss": 0.7521,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4042737510828761,
+      "grad_norm": 0.10767613351345062,
+      "learning_rate": 6.760459907132573e-06,
+      "loss": 0.7418,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4081239772836654,
+      "grad_norm": 0.10674374550580978,
+      "learning_rate": 6.701959642809589e-06,
+      "loss": 0.7466,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41197420348445474,
+      "grad_norm": 0.11246778070926666,
+      "learning_rate": 6.643194658312987e-06,
+      "loss": 0.7272,
+      "step": 1070
+    },
+    {
+      "epoch": 0.415824429685244,
+      "grad_norm": 0.10568033903837204,
+      "learning_rate": 6.584174093857676e-06,
+      "loss": 0.7611,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4196746558860333,
+      "grad_norm": 0.11874755471944809,
+      "learning_rate": 6.5249071294110735e-06,
+      "loss": 0.7383,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4235248820868226,
+      "grad_norm": 0.11687806248664856,
+      "learning_rate": 6.465402983265274e-06,
+      "loss": 0.7588,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4273751082876119,
+      "grad_norm": 0.1129055842757225,
+      "learning_rate": 6.405670910603243e-06,
+      "loss": 0.7538,
+      "step": 1110
+    },
+    {
+      "epoch": 0.43122533448840117,
+      "grad_norm": 0.11784109473228455,
+      "learning_rate": 6.345720202059291e-06,
+      "loss": 0.7455,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4350755606891905,
+      "grad_norm": 0.10899516195058823,
+      "learning_rate": 6.285560182274008e-06,
+      "loss": 0.7333,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4389257868899798,
+      "grad_norm": 0.10470816493034363,
+      "learning_rate": 6.225200208443936e-06,
+      "loss": 0.7383,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4427760130907691,
+      "grad_norm": 0.10665585845708847,
+      "learning_rate": 6.16464966886616e-06,
+      "loss": 0.7358,
+      "step": 1150
+    },
+    {
+      "epoch": 0.44662623929155837,
+      "grad_norm": 0.12578950822353363,
+      "learning_rate": 6.103917981478065e-06,
+      "loss": 0.7633,
+      "step": 1160
+    },
+    {
+      "epoch": 0.45047646549234766,
+      "grad_norm": 0.10707557201385498,
+      "learning_rate": 6.043014592392489e-06,
+      "loss": 0.7699,
+      "step": 1170
+    },
+    {
+      "epoch": 0.454326691693137,
+      "grad_norm": 0.09822295606136322,
+      "learning_rate": 5.981948974428485e-06,
+      "loss": 0.7421,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4581769178939263,
+      "grad_norm": 0.10965897142887115,
+      "learning_rate": 5.920730625637934e-06,
+      "loss": 0.7238,
+      "step": 1190
+    },
+    {
+      "epoch": 0.46202714409471557,
+      "grad_norm": 0.11821416765451431,
+      "learning_rate": 5.8593690678282355e-06,
+      "loss": 0.7344,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46587737029550486,
+      "grad_norm": 0.11401423066854477,
+      "learning_rate": 5.797873845081301e-06,
+      "loss": 0.7374,
+      "step": 1210
+    },
+    {
+      "epoch": 0.46972759649629414,
+      "grad_norm": 0.11630196869373322,
+      "learning_rate": 5.736254522269075e-06,
+      "loss": 0.7504,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47357782269708343,
+      "grad_norm": 0.10775155574083328,
+      "learning_rate": 5.6745206835658386e-06,
+      "loss": 0.7561,
+      "step": 1230
+    },
+    {
+      "epoch": 0.47742804889787277,
+      "grad_norm": 0.10453297942876816,
+      "learning_rate": 5.612681930957495e-06,
+      "loss": 0.7324,
+      "step": 1240
+    },
+    {
+      "epoch": 0.48127827509866206,
+      "grad_norm": 0.10358171910047531,
+      "learning_rate": 5.550747882748095e-06,
+      "loss": 0.7299,
+      "step": 1250
+    },
+    {
+      "epoch": 0.48512850129945134,
+      "grad_norm": 0.11438196152448654,
+      "learning_rate": 5.4887281720638195e-06,
+      "loss": 0.7345,
+      "step": 1260
+    },
+    {
+      "epoch": 0.48897872750024063,
+      "grad_norm": 0.12303616851568222,
+      "learning_rate": 5.42663244535466e-06,
+      "loss": 0.7186,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4928289537010299,
+      "grad_norm": 0.10501397401094437,
+      "learning_rate": 5.3644703608940205e-06,
+      "loss": 0.7273,
+      "step": 1280
+    },
+    {
+      "epoch": 0.49667917990181926,
+      "grad_norm": 0.10975082963705063,
+      "learning_rate": 5.302251587276491e-06,
+      "loss": 0.7466,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5005294061026085,
+      "grad_norm": 0.10879012197256088,
+      "learning_rate": 5.2399858019140005e-06,
+      "loss": 0.7428,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5043796323033978,
+      "grad_norm": 0.11213205009698868,
+      "learning_rate": 5.177682689530615e-06,
+      "loss": 0.7547,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5082298585041871,
+      "grad_norm": 0.12394260615110397,
+      "learning_rate": 5.11535194065619e-06,
+      "loss": 0.7328,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5120800847049765,
+      "grad_norm": 0.12340950965881348,
+      "learning_rate": 5.053003250119113e-06,
+      "loss": 0.7332,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5159303109057657,
+      "grad_norm": 0.11524638533592224,
+      "learning_rate": 4.9906463155383925e-06,
+      "loss": 0.7408,
+      "step": 1340
+    },
+    {
+      "epoch": 0.519780537106555,
+      "grad_norm": 0.11275030672550201,
+      "learning_rate": 4.9282908358153046e-06,
+      "loss": 0.7346,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5236307633073443,
+      "grad_norm": 0.12697428464889526,
+      "learning_rate": 4.8659465096248345e-06,
+      "loss": 0.743,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5274809895081336,
+      "grad_norm": 0.11002713441848755,
+      "learning_rate": 4.803623033907169e-06,
+      "loss": 0.7471,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5313312157089229,
+      "grad_norm": 0.1019381731748581,
+      "learning_rate": 4.741330102359441e-06,
+      "loss": 0.7548,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5351814419097122,
+      "grad_norm": 0.11847570538520813,
+      "learning_rate": 4.679077403927996e-06,
+      "loss": 0.7288,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5390316681105015,
+      "grad_norm": 0.11314871907234192,
+      "learning_rate": 4.616874621301373e-06,
+      "loss": 0.7356,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5428818943112907,
+      "grad_norm": 0.11008746922016144,
+      "learning_rate": 4.554731429404293e-06,
+      "loss": 0.7517,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5467321205120801,
+      "grad_norm": 0.11648157238960266,
+      "learning_rate": 4.492657493892822e-06,
+      "loss": 0.7331,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5505823467128694,
+      "grad_norm": 0.11342335492372513,
+      "learning_rate": 4.430662469650999e-06,
+      "loss": 0.7478,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5544325729136587,
+      "grad_norm": 0.10998478531837463,
+      "learning_rate": 4.368755999289118e-06,
+      "loss": 0.7503,
+      "step": 1440
+    },
+    {
+      "epoch": 0.558282799114448,
+      "grad_norm": 0.12068407982587814,
+      "learning_rate": 4.306947711643949e-06,
+      "loss": 0.7288,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5621330253152372,
+      "grad_norm": 0.1107109859585762,
+      "learning_rate": 4.245247220281068e-06,
+      "loss": 0.7381,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5659832515160266,
+      "grad_norm": 0.11038530617952347,
+      "learning_rate": 4.183664121999587e-06,
+      "loss": 0.7322,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5698334777168159,
+      "grad_norm": 0.11666359007358551,
+      "learning_rate": 4.122207995339474e-06,
+      "loss": 0.7311,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5736837039176051,
+      "grad_norm": 0.11379261314868927,
+      "learning_rate": 4.060888399091734e-06,
+      "loss": 0.7385,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5775339301183945,
+      "grad_norm": 0.11306154727935791,
+      "learning_rate": 3.999714870811639e-06,
+      "loss": 0.7504,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5813841563191837,
+      "grad_norm": 0.1089501902461052,
+      "learning_rate": 3.938696925335275e-06,
+      "loss": 0.74,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5852343825199731,
+      "grad_norm": 0.11330631375312805,
+      "learning_rate": 3.87784405329962e-06,
+      "loss": 0.7474,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5890846087207623,
+      "grad_norm": 0.11857966333627701,
+      "learning_rate": 3.817165719666382e-06,
+      "loss": 0.7435,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5929348349215516,
+      "grad_norm": 0.1089237704873085,
+      "learning_rate": 3.756671362249841e-06,
+      "loss": 0.7544,
+      "step": 1540
+    },
+    {
+      "epoch": 0.596785061122341,
+      "grad_norm": 0.11136338859796524,
+      "learning_rate": 3.6963703902488907e-06,
+      "loss": 0.7457,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6006352873231302,
+      "grad_norm": 0.11523008346557617,
+      "learning_rate": 3.6362721827835655e-06,
+      "loss": 0.7546,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6044855135239195,
+      "grad_norm": 0.11708564311265945,
+      "learning_rate": 3.576386087436212e-06,
+      "loss": 0.7462,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6083357397247088,
+      "grad_norm": 0.11660383641719818,
+      "learning_rate": 3.5167214187975913e-06,
+      "loss": 0.7475,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6121859659254981,
+      "grad_norm": 0.0990515872836113,
+      "learning_rate": 3.457287457018086e-06,
+      "loss": 0.7456,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6160361921262875,
+      "grad_norm": 0.10248612612485886,
+      "learning_rate": 3.3980934463643003e-06,
+      "loss": 0.7306,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6198864183270767,
+      "grad_norm": 0.10241679847240448,
+      "learning_rate": 3.3391485937812053e-06,
+      "loss": 0.7109,
+      "step": 1610
+    },
+    {
+      "epoch": 0.623736644527866,
+      "grad_norm": 0.11149538308382034,
+      "learning_rate": 3.280462067460115e-06,
+      "loss": 0.7126,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6275868707286553,
+      "grad_norm": 0.10890234261751175,
+      "learning_rate": 3.222042995412669e-06,
+      "loss": 0.7516,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6314370969294446,
+      "grad_norm": 0.09704490751028061,
+      "learning_rate": 3.163900464051089e-06,
+      "loss": 0.7335,
+      "step": 1640
+    },
+    {
+      "epoch": 0.635287323130234,
+      "grad_norm": 0.11142604053020477,
+      "learning_rate": 3.1060435167748826e-06,
+      "loss": 0.7459,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6391375493310232,
+      "grad_norm": 0.1088608130812645,
+      "learning_rate": 3.048481152564253e-06,
+      "loss": 0.7297,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6429877755318125,
+      "grad_norm": 0.11804860830307007,
+      "learning_rate": 2.9912223245804017e-06,
+      "loss": 0.7519,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6468380017326018,
+      "grad_norm": 0.11194264143705368,
+      "learning_rate": 2.934275938772981e-06,
+      "loss": 0.7342,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6506882279333911,
+      "grad_norm": 0.10983610153198242,
+      "learning_rate": 2.8776508524948676e-06,
+      "loss": 0.7246,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6545384541341804,
+      "grad_norm": 0.10468059778213501,
+      "learning_rate": 2.821355873124503e-06,
+      "loss": 0.7242,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6583886803349697,
+      "grad_norm": 0.10993515700101852,
+      "learning_rate": 2.7653997566960144e-06,
+      "loss": 0.7425,
+      "step": 1710
+    },
+    {
+      "epoch": 0.662238906535759,
+      "grad_norm": 0.1106952428817749,
+      "learning_rate": 2.709791206537309e-06,
+      "loss": 0.7302,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6660891327365482,
+      "grad_norm": 0.11061175167560577,
+      "learning_rate": 2.6545388719163747e-06,
+      "loss": 0.7316,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6699393589373376,
+      "grad_norm": 0.10918284952640533,
+      "learning_rate": 2.599651346695979e-06,
+      "loss": 0.7412,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6737895851381268,
+      "grad_norm": 0.1102231964468956,
+      "learning_rate": 2.545137167997007e-06,
+      "loss": 0.7374,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6776398113389162,
+      "grad_norm": 0.10207434743642807,
+      "learning_rate": 2.4910048148705972e-06,
+      "loss": 0.7235,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6814900375397055,
+      "grad_norm": 0.111183762550354,
+      "learning_rate": 2.4372627069793327e-06,
+      "loss": 0.7648,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6853402637404947,
+      "grad_norm": 0.1155601441860199,
+      "learning_rate": 2.3839192032876486e-06,
+      "loss": 0.7474,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6891904899412841,
+      "grad_norm": 0.10672982037067413,
+      "learning_rate": 2.3309826007617047e-06,
+      "loss": 0.7594,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6930407161420733,
+      "grad_norm": 0.1049095094203949,
+      "learning_rate": 2.278461133078881e-06,
+      "loss": 0.7228,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6968909423428626,
+      "grad_norm": 0.11320099979639053,
+      "learning_rate": 2.2263629693471258e-06,
+      "loss": 0.7208,
+      "step": 1810
+    },
+    {
+      "epoch": 0.700741168543652,
+      "grad_norm": 0.10218384861946106,
+      "learning_rate": 2.1746962128343414e-06,
+      "loss": 0.7421,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7045913947444412,
+      "grad_norm": 0.10667248070240021,
+      "learning_rate": 2.1234688997080183e-06,
+      "loss": 0.7402,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7084416209452306,
+      "grad_norm": 0.12129062414169312,
+      "learning_rate": 2.0726889977853053e-06,
+      "loss": 0.7384,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7122918471460198,
+      "grad_norm": 0.11721149832010269,
+      "learning_rate": 2.022364405293703e-06,
+      "loss": 0.7472,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7161420733468091,
+      "grad_norm": 0.109659343957901,
+      "learning_rate": 1.9725029496425814e-06,
+      "loss": 0.7455,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7199922995475985,
+      "grad_norm": 0.10070423036813736,
+      "learning_rate": 1.923112386205727e-06,
+      "loss": 0.7398,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7238425257483877,
+      "grad_norm": 0.10945705324411392,
+      "learning_rate": 1.874200397115079e-06,
+      "loss": 0.7463,
+      "step": 1880
+    },
+    {
+      "epoch": 0.727692751949177,
+      "grad_norm": 0.11820171773433685,
+      "learning_rate": 1.825774590065857e-06,
+      "loss": 0.7319,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7315429781499663,
+      "grad_norm": 0.11620453745126724,
+      "learning_rate": 1.7778424971332813e-06,
+      "loss": 0.7429,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7353932043507556,
+      "grad_norm": 0.10518138855695724,
+      "learning_rate": 1.7304115736010407e-06,
+      "loss": 0.7412,
+      "step": 1910
+    },
+    {
+      "epoch": 0.739243430551545,
+      "grad_norm": 0.10780428349971771,
+      "learning_rate": 1.6834891968017108e-06,
+      "loss": 0.7345,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7430936567523342,
+      "grad_norm": 0.11247050762176514,
+      "learning_rate": 1.6370826649692884e-06,
+      "loss": 0.7457,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7469438829531235,
+      "grad_norm": 0.11268564313650131,
+      "learning_rate": 1.5911991961040463e-06,
+      "loss": 0.7547,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7507941091539128,
+      "grad_norm": 0.1105349212884903,
+      "learning_rate": 1.545845926849851e-06,
+      "loss": 0.7407,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7546443353547021,
+      "grad_norm": 0.1236489936709404,
+      "learning_rate": 1.5010299113841397e-06,
+      "loss": 0.7589,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7584945615554913,
+      "grad_norm": 0.11024940758943558,
+      "learning_rate": 1.4567581203207238e-06,
+      "loss": 0.755,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7623447877562807,
+      "grad_norm": 0.11217351257801056,
+      "learning_rate": 1.413037439625593e-06,
+      "loss": 0.7399,
+      "step": 1980
+    },
+    {
+      "epoch": 0.76619501395707,
+      "grad_norm": 0.12133375555276871,
+      "learning_rate": 1.3698746695458847e-06,
+      "loss": 0.7405,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7700452401578592,
+      "grad_norm": 0.10815633088350296,
+      "learning_rate": 1.3272765235521745e-06,
+      "loss": 0.7421,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7738954663586486,
+      "grad_norm": 0.10940743237733841,
+      "learning_rate": 1.2852496272942877e-06,
+      "loss": 0.7414,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7777456925594378,
+      "grad_norm": 0.10445626825094223,
+      "learning_rate": 1.2438005175707396e-06,
+      "loss": 0.7281,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7815959187602272,
+      "grad_norm": 0.11116701364517212,
+      "learning_rate": 1.2029356413120297e-06,
+      "loss": 0.7133,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7854461449610165,
+      "grad_norm": 0.11188463121652603,
+      "learning_rate": 1.1626613545778786e-06,
+      "loss": 0.7323,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7892963711618057,
+      "grad_norm": 0.10531072318553925,
+      "learning_rate": 1.1229839215686294e-06,
+      "loss": 0.7382,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7931465973625951,
+      "grad_norm": 0.1039716899394989,
+      "learning_rate": 1.0839095136509188e-06,
+      "loss": 0.718,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7969968235633843,
+      "grad_norm": 0.1172463446855545,
+      "learning_rate": 1.045444208397791e-06,
+      "loss": 0.7314,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8008470497641736,
+      "grad_norm": 0.10571017116308212,
+      "learning_rate": 1.007593988643402e-06,
+      "loss": 0.7203,
+      "step": 2080
+    },
+    {
+      "epoch": 0.804697275964963,
+      "grad_norm": 0.10255500674247742,
+      "learning_rate": 9.703647415524613e-07,
+      "loss": 0.7427,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8085475021657522,
+      "grad_norm": 0.1099574863910675,
+      "learning_rate": 9.337622577045485e-07,
+      "loss": 0.7377,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8123977283665416,
+      "grad_norm": 0.10934717953205109,
+      "learning_rate": 8.977922301934588e-07,
+      "loss": 0.749,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8162479545673308,
+      "grad_norm": 0.11289958655834198,
+      "learning_rate": 8.624602537417009e-07,
+      "loss": 0.7448,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8200981807681201,
+      "grad_norm": 0.11194674670696259,
+      "learning_rate": 8.277718238303123e-07,
+      "loss": 0.7302,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8239484069689095,
+      "grad_norm": 0.10882336646318436,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.742,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8277986331696987,
+      "grad_norm": 0.12826266884803772,
+      "learning_rate": 7.603470842324223e-07,
+      "loss": 0.7263,
+      "step": 2150
+    },
+    {
+      "epoch": 0.831648859370488,
+      "grad_norm": 0.10797886550426483,
+      "learning_rate": 7.276212616857575e-07,
+      "loss": 0.7442,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8354990855712773,
+      "grad_norm": 0.1093938872218132,
+      "learning_rate": 6.955599583279837e-07,
+      "loss": 0.7461,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8393493117720666,
+      "grad_norm": 0.10914472490549088,
+      "learning_rate": 6.641681609246981e-07,
+      "loss": 0.734,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8431995379728559,
+      "grad_norm": 0.10774393379688263,
+      "learning_rate": 6.334507521075711e-07,
+      "loss": 0.7452,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8470497641736452,
+      "grad_norm": 0.1155761256814003,
+      "learning_rate": 6.034125096149179e-07,
+      "loss": 0.7305,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8508999903744345,
+      "grad_norm": 0.12554535269737244,
+      "learning_rate": 5.740581055485722e-07,
+      "loss": 0.716,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8547502165752238,
+      "grad_norm": 0.10167080163955688,
+      "learning_rate": 5.453921056471961e-07,
+      "loss": 0.7387,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8586004427760131,
+      "grad_norm": 0.11908018589019775,
+      "learning_rate": 5.174189685761288e-07,
+      "loss": 0.7334,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8624506689768023,
+      "grad_norm": 0.10595957189798355,
+      "learning_rate": 4.901430452338979e-07,
+      "loss": 0.7513,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8663008951775917,
+      "grad_norm": 0.10767778009176254,
+      "learning_rate": 4.6356857807548286e-07,
+      "loss": 0.7346,
+      "step": 2250
+    },
+    {
+      "epoch": 0.870151121378381,
+      "grad_norm": 0.11260323226451874,
+      "learning_rate": 4.376997004524541e-07,
+      "loss": 0.7222,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8740013475791703,
+      "grad_norm": 0.10652253031730652,
+      "learning_rate": 4.12540435970073e-07,
+      "loss": 0.7241,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8778515737799596,
+      "grad_norm": 0.10995321720838547,
+      "learning_rate": 3.880946978614703e-07,
+      "loss": 0.7421,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8817017999807488,
+      "grad_norm": 0.10731273144483566,
+      "learning_rate": 3.643662883789878e-07,
+      "loss": 0.7436,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8855520261815382,
+      "grad_norm": 0.1065845638513565,
+      "learning_rate": 3.41358898202781e-07,
+      "loss": 0.7282,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8894022523823275,
+      "grad_norm": 0.10125146806240082,
+      "learning_rate": 3.190761058667741e-07,
+      "loss": 0.7292,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8932524785831167,
+      "grad_norm": 0.10210753977298737,
+      "learning_rate": 2.9752137720206463e-07,
+      "loss": 0.7375,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8971027047839061,
+      "grad_norm": 0.11130058765411377,
+      "learning_rate": 2.7669806479785067e-07,
+      "loss": 0.7472,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9009529309846953,
+      "grad_norm": 0.1070486307144165,
+      "learning_rate": 2.566094074799763e-07,
+      "loss": 0.7383,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9048031571854847,
+      "grad_norm": 0.10964521765708923,
+      "learning_rate": 2.3725852980716557e-07,
+      "loss": 0.7239,
+      "step": 2350
+    },
+    {
+      "epoch": 0.908653383386274,
+      "grad_norm": 0.10990218073129654,
+      "learning_rate": 2.1864844158503983e-07,
+      "loss": 0.739,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9125036095870632,
+      "grad_norm": 0.11391717940568924,
+      "learning_rate": 2.007820373979702e-07,
+      "loss": 0.7288,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9163538357878526,
+      "grad_norm": 0.10987786948680878,
+      "learning_rate": 1.8366209615886532e-07,
+      "loss": 0.7309,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9202040619886418,
+      "grad_norm": 0.10975030064582825,
+      "learning_rate": 1.6729128067693613e-07,
+      "loss": 0.7522,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9240542881894311,
+      "grad_norm": 0.11256835609674454,
+      "learning_rate": 1.5167213724353426e-07,
+      "loss": 0.7414,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9279045143902205,
+      "grad_norm": 0.11043158918619156,
+      "learning_rate": 1.3680709523610058e-07,
+      "loss": 0.7275,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9317547405910097,
+      "grad_norm": 0.10945302248001099,
+      "learning_rate": 1.2269846674030418e-07,
+      "loss": 0.755,
+      "step": 2420
+    },
+    {
+      "epoch": 0.935604966791799,
+      "grad_norm": 0.11428770422935486,
+      "learning_rate": 1.0934844619042595e-07,
+      "loss": 0.7469,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9394551929925883,
+      "grad_norm": 0.11273124068975449,
+      "learning_rate": 9.675911002803961e-08,
+      "loss": 0.7352,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9433054191933776,
+      "grad_norm": 0.11690922826528549,
+      "learning_rate": 8.493241637904381e-08,
+      "loss": 0.7335,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9471556453941669,
+      "grad_norm": 0.11335375159978867,
+      "learning_rate": 7.387020474909689e-08,
+      "loss": 0.7519,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9510058715949562,
+      "grad_norm": 0.11252011358737946,
+      "learning_rate": 6.357419573750856e-08,
+      "loss": 0.7393,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9548560977957455,
+      "grad_norm": 0.10769885778427124,
+      "learning_rate": 5.4045990769613365e-08,
+      "loss": 0.7317,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9587063239965348,
+      "grad_norm": 0.11015380918979645,
+      "learning_rate": 4.5287071847694434e-08,
+      "loss": 0.7347,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9625565501973241,
+      "grad_norm": 0.10759741812944412,
+      "learning_rate": 3.7298801320467817e-08,
+      "loss": 0.7332,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9664067763981133,
+      "grad_norm": 0.11376994103193283,
+      "learning_rate": 3.0082421671192576e-08,
+      "loss": 0.7284,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9702570025989027,
+      "grad_norm": 0.10622478276491165,
+      "learning_rate": 2.3639055324410355e-08,
+      "loss": 0.7431,
+      "step": 2520
+    },
+    {
+      "epoch": 0.974107228799692,
+      "grad_norm": 0.11014197766780853,
+      "learning_rate": 1.7969704471369477e-08,
+      "loss": 0.7316,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9779574550004813,
+      "grad_norm": 0.1028149425983429,
+      "learning_rate": 1.3075250914142413e-08,
+      "loss": 0.7345,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9818076812012706,
+      "grad_norm": 0.1015634685754776,
+      "learning_rate": 8.956455928476049e-09,
+      "loss": 0.7442,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9856579074020598,
+      "grad_norm": 0.09613681584596634,
+      "learning_rate": 5.6139601453786275e-09,
+      "loss": 0.726,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9895081336028492,
+      "grad_norm": 0.11620547622442245,
+      "learning_rate": 3.0482834514805602e-09,
+      "loss": 0.7295,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9933583598036385,
+      "grad_norm": 0.10735750198364258,
+      "learning_rate": 1.2598249081713366e-09,
+      "loss": 0.715,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9972085860044277,
+      "grad_norm": 0.10855001956224442,
+      "learning_rate": 2.488626895302826e-10,
+      "loss": 0.7451,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9999037443449803,
+      "step": 2597,
+      "total_flos": 2.6951376519959675e+18,
+      "train_loss": 0.7579362606332078,
+      "train_runtime": 10243.5332,
+      "train_samples_per_second": 64.91,
+      "train_steps_per_second": 0.254
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2597,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.6951376519959675e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen25/mc2_m16_384/finetune/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6ebe80c882c611d9d684124757d8071444b0e549f4310987a62a094f50527529
+size 6136