Upload continuous evaluation checkpoints, batch 10

.gitattributes

@@ -128,3 +128,6 @@ continuous/qwen3/eupe_vit_t/finetune/tokenizer.json filter=lfs diff=lfs merge=lf
 continuous/qwen3/ijepa_vith14/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 continuous/qwen3/mc1_b16_224_2.5b/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 continuous/qwen3/mc1_b16_224_400m/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+continuous/qwen3/mc1_b32_224_2.5b/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+continuous/qwen3/mc1_b32_224_400m/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+continuous/qwen3/mc1_g14_224_2.5b/finetune/tokenizer.json filter=lfs diff=lfs merge=lfs -text

continuous/qwen3/mc1_b32_224_2.5b/finetune/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88127e5f084e225f2134a3dd65832e7dda008a9dc57e06c195eeeeec858197b5
+size 4262321400

continuous/qwen3/mc1_b32_224_2.5b/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/qwen3/mc1_b32_224_2.5b/finetune/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
+size 11422654

continuous/qwen3/mc1_b32_224_2.5b/finetune/tokenizer_config.json

@@ -0,0 +1,241 @@
+{
+  "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
+    },
+    "151665": {
+      "content": "<tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</think>",
+      "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\\n' }}\n    {%- endif %}\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    {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for message in messages[::-1] %}\n    {%- set index = (messages|length - 1) - loop.index0 %}\n    {%- if ns.multi_step_tool and message.role == \"user\" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}\n        {%- set ns.multi_step_tool = false %}\n        {%- set ns.last_query_index = index %}\n    {%- endif %}\n{%- endfor %}\n{%- for message in messages %}\n    {%- if message.content is string %}\n        {%- set content = message.content %}\n    {%- else %}\n        {%- set content = '' %}\n    {%- endif %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n        {{- '<|im_start|>' + message.role + '\\n' + content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set reasoning_content = '' %}\n        {%- if message.reasoning_content is string %}\n            {%- set reasoning_content = message.reasoning_content %}\n        {%- else %}\n            {%- if '</think>' in content %}\n                {%- set reasoning_content = content.split('</think>')[0].rstrip('\\n').split('<think>')[-1].lstrip('\\n') %}\n                {%- set content = content.split('</think>')[-1].lstrip('\\n') %}\n            {%- endif %}\n        {%- endif %}\n        {%- if loop.index0 > ns.last_query_index %}\n            {%- if loop.last or (not loop.last and reasoning_content) %}\n                {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content.strip('\\n') + '\\n</think>\\n\\n' + content.lstrip('\\n') }}\n            {%- else %}\n                {{- '<|im_start|>' + message.role + '\\n' + content }}\n            {%- endif %}\n        {%- else %}\n            {{- '<|im_start|>' + message.role + '\\n' + content }}\n        {%- endif %}\n        {%- if message.tool_calls %}\n            {%- for tool_call in message.tool_calls %}\n                {%- if (loop.first and content) or (not loop.first) %}\n                    {{- '\\n' }}\n                {%- endif %}\n                {%- if tool_call.function %}\n                    {%- set tool_call = tool_call.function %}\n                {%- endif %}\n                {{- '<tool_call>\\n{\"name\": \"' }}\n                {{- tool_call.name }}\n                {{- '\", \"arguments\": ' }}\n                {%- if tool_call.arguments is string %}\n                    {{- tool_call.arguments }}\n                {%- else %}\n                    {{- tool_call.arguments | tojson }}\n                {%- endif %}\n                {{- '}\\n</tool_call>' }}\n            {%- endfor %}\n        {%- endif %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if loop.first or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- 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    {%- if enable_thinking is defined and enable_thinking is false %}\n        {{- '<think>\\n\\n</think>\\n\\n' }}\n    {%- endif %}\n{%- endif %}",
+  "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/qwen3/mc1_b32_224_2.5b/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": 10.897425651550293,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 2.4726,
+      "step": 10
+    },
+    {
+      "epoch": 0.007700452401578592,
+      "grad_norm": 3.6208925247192383,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 1.9591,
+      "step": 20
+    },
+    {
+      "epoch": 0.01155067860236789,
+      "grad_norm": 2.3730409145355225,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.5486,
+      "step": 30
+    },
+    {
+      "epoch": 0.015400904803157185,
+      "grad_norm": 0.568038284778595,
+      "learning_rate": 5e-06,
+      "loss": 1.1835,
+      "step": 40
+    },
+    {
+      "epoch": 0.019251131003946483,
+      "grad_norm": 0.46500590443611145,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 1.0766,
+      "step": 50
+    },
+    {
+      "epoch": 0.02310135720473578,
+      "grad_norm": 0.24020300805568695,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 1.0001,
+      "step": 60
+    },
+    {
+      "epoch": 0.026951583405525073,
+      "grad_norm": 0.1972605586051941,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.9579,
+      "step": 70
+    },
+    {
+      "epoch": 0.03080180960631437,
+      "grad_norm": 0.1741226315498352,
+      "learning_rate": 9.999996111488723e-06,
+      "loss": 0.9548,
+      "step": 80
+    },
+    {
+      "epoch": 0.034652035807103666,
+      "grad_norm": 0.17555946111679077,
+      "learning_rate": 9.999529497453782e-06,
+      "loss": 0.9283,
+      "step": 90
+    },
+    {
+      "epoch": 0.038502262007892966,
+      "grad_norm": 0.14926737546920776,
+      "learning_rate": 9.998285264324193e-06,
+      "loss": 0.9028,
+      "step": 100
+    },
+    {
+      "epoch": 0.04235248820868226,
+      "grad_norm": 0.16021673381328583,
+      "learning_rate": 9.996263605626052e-06,
+      "loss": 0.8789,
+      "step": 110
+    },
+    {
+      "epoch": 0.04620271440947156,
+      "grad_norm": 0.14228788018226624,
+      "learning_rate": 9.99346483580503e-06,
+      "loss": 0.8851,
+      "step": 120
+    },
+    {
+      "epoch": 0.05005294061026085,
+      "grad_norm": 0.14278294146060944,
+      "learning_rate": 9.989889390177467e-06,
+      "loss": 0.8779,
+      "step": 130
+    },
+    {
+      "epoch": 0.053903166811050146,
+      "grad_norm": 0.16158427298069,
+      "learning_rate": 9.985537824862647e-06,
+      "loss": 0.8739,
+      "step": 140
+    },
+    {
+      "epoch": 0.057753393011839446,
+      "grad_norm": 0.1686583012342453,
+      "learning_rate": 9.980410816696317e-06,
+      "loss": 0.8693,
+      "step": 150
+    },
+    {
+      "epoch": 0.06160361921262874,
+      "grad_norm": 0.14870119094848633,
+      "learning_rate": 9.97450916312541e-06,
+      "loss": 0.8642,
+      "step": 160
+    },
+    {
+      "epoch": 0.06545384541341803,
+      "grad_norm": 0.12532837688922882,
+      "learning_rate": 9.967833782084004e-06,
+      "loss": 0.8618,
+      "step": 170
+    },
+    {
+      "epoch": 0.06930407161420733,
+      "grad_norm": 0.13481761515140533,
+      "learning_rate": 9.960385711850554e-06,
+      "loss": 0.8861,
+      "step": 180
+    },
+    {
+      "epoch": 0.07315429781499663,
+      "grad_norm": 0.14263096451759338,
+      "learning_rate": 9.952166110886394e-06,
+      "loss": 0.884,
+      "step": 190
+    },
+    {
+      "epoch": 0.07700452401578593,
+      "grad_norm": 0.151332288980484,
+      "learning_rate": 9.943176257655567e-06,
+      "loss": 0.8641,
+      "step": 200
+    },
+    {
+      "epoch": 0.08085475021657522,
+      "grad_norm": 0.14818933606147766,
+      "learning_rate": 9.933417550425952e-06,
+      "loss": 0.8686,
+      "step": 210
+    },
+    {
+      "epoch": 0.08470497641736452,
+      "grad_norm": 0.13357235491275787,
+      "learning_rate": 9.9228915070518e-06,
+      "loss": 0.8483,
+      "step": 220
+    },
+    {
+      "epoch": 0.08855520261815382,
+      "grad_norm": 0.14479365944862366,
+      "learning_rate": 9.911599764737632e-06,
+      "loss": 0.8542,
+      "step": 230
+    },
+    {
+      "epoch": 0.09240542881894312,
+      "grad_norm": 0.14021635055541992,
+      "learning_rate": 9.899544079783606e-06,
+      "loss": 0.841,
+      "step": 240
+    },
+    {
+      "epoch": 0.0962556550197324,
+      "grad_norm": 0.1429843008518219,
+      "learning_rate": 9.886726327312338e-06,
+      "loss": 0.8568,
+      "step": 250
+    },
+    {
+      "epoch": 0.1001058812205217,
+      "grad_norm": 0.13500994443893433,
+      "learning_rate": 9.87314850097724e-06,
+      "loss": 0.8607,
+      "step": 260
+    },
+    {
+      "epoch": 0.103956107421311,
+      "grad_norm": 0.12818260490894318,
+      "learning_rate": 9.858812712652452e-06,
+      "loss": 0.846,
+      "step": 270
+    },
+    {
+      "epoch": 0.10780633362210029,
+      "grad_norm": 0.1297636777162552,
+      "learning_rate": 9.843721192104341e-06,
+      "loss": 0.8489,
+      "step": 280
+    },
+    {
+      "epoch": 0.11165655982288959,
+      "grad_norm": 0.1320681869983673,
+      "learning_rate": 9.827876286644698e-06,
+      "loss": 0.8618,
+      "step": 290
+    },
+    {
+      "epoch": 0.11550678602367889,
+      "grad_norm": 0.12005192041397095,
+      "learning_rate": 9.811280460765643e-06,
+      "loss": 0.8223,
+      "step": 300
+    },
+    {
+      "epoch": 0.11935701222446819,
+      "grad_norm": 0.12621133029460907,
+      "learning_rate": 9.793936295756292e-06,
+      "loss": 0.8394,
+      "step": 310
+    },
+    {
+      "epoch": 0.12320723842525748,
+      "grad_norm": 0.13515962660312653,
+      "learning_rate": 9.775846489301279e-06,
+      "loss": 0.8345,
+      "step": 320
+    },
+    {
+      "epoch": 0.12705746462604678,
+      "grad_norm": 0.13688556849956512,
+      "learning_rate": 9.757013855061157e-06,
+      "loss": 0.8375,
+      "step": 330
+    },
+    {
+      "epoch": 0.13090769082683607,
+      "grad_norm": 0.1404041200876236,
+      "learning_rate": 9.737441322234758e-06,
+      "loss": 0.8482,
+      "step": 340
+    },
+    {
+      "epoch": 0.13475791702762538,
+      "grad_norm": 0.14022520184516907,
+      "learning_rate": 9.717131935103605e-06,
+      "loss": 0.8519,
+      "step": 350
+    },
+    {
+      "epoch": 0.13860814322841467,
+      "grad_norm": 0.12540209293365479,
+      "learning_rate": 9.696088852558398e-06,
+      "loss": 0.8356,
+      "step": 360
+    },
+    {
+      "epoch": 0.14245836942920398,
+      "grad_norm": 0.13122379779815674,
+      "learning_rate": 9.674315347607686e-06,
+      "loss": 0.8608,
+      "step": 370
+    },
+    {
+      "epoch": 0.14630859562999327,
+      "grad_norm": 0.11826972663402557,
+      "learning_rate": 9.651814806868796e-06,
+      "loss": 0.8282,
+      "step": 380
+    },
+    {
+      "epoch": 0.15015882183078255,
+      "grad_norm": 0.13075265288352966,
+      "learning_rate": 9.62859073004107e-06,
+      "loss": 0.8349,
+      "step": 390
+    },
+    {
+      "epoch": 0.15400904803157187,
+      "grad_norm": 0.12260919064283371,
+      "learning_rate": 9.604646729361547e-06,
+      "loss": 0.8365,
+      "step": 400
+    },
+    {
+      "epoch": 0.15785927423236115,
+      "grad_norm": 0.12474191933870316,
+      "learning_rate": 9.579986529043099e-06,
+      "loss": 0.8181,
+      "step": 410
+    },
+    {
+      "epoch": 0.16170950043315044,
+      "grad_norm": 0.12459273636341095,
+      "learning_rate": 9.554613964695189e-06,
+      "loss": 0.8331,
+      "step": 420
+    },
+    {
+      "epoch": 0.16555972663393975,
+      "grad_norm": 0.14229892194271088,
+      "learning_rate": 9.528532982727281e-06,
+      "loss": 0.8252,
+      "step": 430
+    },
+    {
+      "epoch": 0.16940995283472904,
+      "grad_norm": 0.13695360720157623,
+      "learning_rate": 9.501747639735018e-06,
+      "loss": 0.8253,
+      "step": 440
+    },
+    {
+      "epoch": 0.17326017903551832,
+      "grad_norm": 0.13475476205348969,
+      "learning_rate": 9.474262101869272e-06,
+      "loss": 0.8091,
+      "step": 450
+    },
+    {
+      "epoch": 0.17711040523630764,
+      "grad_norm": 0.13842882215976715,
+      "learning_rate": 9.446080644188141e-06,
+      "loss": 0.8255,
+      "step": 460
+    },
+    {
+      "epoch": 0.18096063143709692,
+      "grad_norm": 0.14203888177871704,
+      "learning_rate": 9.417207649992017e-06,
+      "loss": 0.8304,
+      "step": 470
+    },
+    {
+      "epoch": 0.18481085763788624,
+      "grad_norm": 0.1274341344833374,
+      "learning_rate": 9.387647610141816e-06,
+      "loss": 0.8235,
+      "step": 480
+    },
+    {
+      "epoch": 0.18866108383867553,
+      "grad_norm": 0.12942607700824738,
+      "learning_rate": 9.35740512236047e-06,
+      "loss": 0.814,
+      "step": 490
+    },
+    {
+      "epoch": 0.1925113100394648,
+      "grad_norm": 0.11805500090122223,
+      "learning_rate": 9.326484890517809e-06,
+      "loss": 0.8158,
+      "step": 500
+    },
+    {
+      "epoch": 0.19636153624025413,
+      "grad_norm": 0.12576976418495178,
+      "learning_rate": 9.294891723898926e-06,
+      "loss": 0.8029,
+      "step": 510
+    },
+    {
+      "epoch": 0.2002117624410434,
+      "grad_norm": 0.1276092827320099,
+      "learning_rate": 9.262630536456153e-06,
+      "loss": 0.8433,
+      "step": 520
+    },
+    {
+      "epoch": 0.2040619886418327,
+      "grad_norm": 0.13986025750637054,
+      "learning_rate": 9.229706346044749e-06,
+      "loss": 0.8099,
+      "step": 530
+    },
+    {
+      "epoch": 0.207912214842622,
+      "grad_norm": 0.14010626077651978,
+      "learning_rate": 9.196124273642422e-06,
+      "loss": 0.8299,
+      "step": 540
+    },
+    {
+      "epoch": 0.2117624410434113,
+      "grad_norm": 0.11608286201953888,
+      "learning_rate": 9.161889542552842e-06,
+      "loss": 0.833,
+      "step": 550
+    },
+    {
+      "epoch": 0.21561266724420058,
+      "grad_norm": 0.13119152188301086,
+      "learning_rate": 9.127007477593193e-06,
+      "loss": 0.8354,
+      "step": 560
+    },
+    {
+      "epoch": 0.2194628934449899,
+      "grad_norm": 0.14081628620624542,
+      "learning_rate": 9.091483504265974e-06,
+      "loss": 0.8455,
+      "step": 570
+    },
+    {
+      "epoch": 0.22331311964577918,
+      "grad_norm": 0.13097400963306427,
+      "learning_rate": 9.055323147915116e-06,
+      "loss": 0.8186,
+      "step": 580
+    },
+    {
+      "epoch": 0.2271633458465685,
+      "grad_norm": 0.1269870400428772,
+      "learning_rate": 9.018532032866585e-06,
+      "loss": 0.8377,
+      "step": 590
+    },
+    {
+      "epoch": 0.23101357204735778,
+      "grad_norm": 0.18563932180404663,
+      "learning_rate": 8.981115881553579e-06,
+      "loss": 0.8324,
+      "step": 600
+    },
+    {
+      "epoch": 0.23486379824814707,
+      "grad_norm": 0.13851842284202576,
+      "learning_rate": 8.943080513626476e-06,
+      "loss": 0.8421,
+      "step": 610
+    },
+    {
+      "epoch": 0.23871402444893638,
+      "grad_norm": 0.13284507393836975,
+      "learning_rate": 8.904431845047641e-06,
+      "loss": 0.8332,
+      "step": 620
+    },
+    {
+      "epoch": 0.24256425064972567,
+      "grad_norm": 0.1355530172586441,
+      "learning_rate": 8.86517588717128e-06,
+      "loss": 0.8392,
+      "step": 630
+    },
+    {
+      "epoch": 0.24641447685051496,
+      "grad_norm": 0.12457861006259918,
+      "learning_rate": 8.82531874580844e-06,
+      "loss": 0.8322,
+      "step": 640
+    },
+    {
+      "epoch": 0.25026470305130427,
+      "grad_norm": 0.1154821515083313,
+      "learning_rate": 8.784866620277314e-06,
+      "loss": 0.7882,
+      "step": 650
+    },
+    {
+      "epoch": 0.25411492925209356,
+      "grad_norm": 0.13651302456855774,
+      "learning_rate": 8.743825802439018e-06,
+      "loss": 0.8192,
+      "step": 660
+    },
+    {
+      "epoch": 0.25796515545288284,
+      "grad_norm": 0.13111267983913422,
+      "learning_rate": 8.702202675718956e-06,
+      "loss": 0.8328,
+      "step": 670
+    },
+    {
+      "epoch": 0.26181538165367213,
+      "grad_norm": 0.12157490104436874,
+      "learning_rate": 8.660003714113959e-06,
+      "loss": 0.8078,
+      "step": 680
+    },
+    {
+      "epoch": 0.26566560785446147,
+      "grad_norm": 0.11961985379457474,
+      "learning_rate": 8.617235481185323e-06,
+      "loss": 0.8223,
+      "step": 690
+    },
+    {
+      "epoch": 0.26951583405525076,
+      "grad_norm": 0.12829197943210602,
+      "learning_rate": 8.573904629037931e-06,
+      "loss": 0.8273,
+      "step": 700
+    },
+    {
+      "epoch": 0.27336606025604004,
+      "grad_norm": 0.11970342695713043,
+      "learning_rate": 8.530017897285585e-06,
+      "loss": 0.8114,
+      "step": 710
+    },
+    {
+      "epoch": 0.27721628645682933,
+      "grad_norm": 0.11392735689878464,
+      "learning_rate": 8.485582112002746e-06,
+      "loss": 0.8139,
+      "step": 720
+    },
+    {
+      "epoch": 0.2810665126576186,
+      "grad_norm": 0.1288517713546753,
+      "learning_rate": 8.440604184662811e-06,
+      "loss": 0.8263,
+      "step": 730
+    },
+    {
+      "epoch": 0.28491673885840796,
+      "grad_norm": 0.1366633176803589,
+      "learning_rate": 8.395091111063118e-06,
+      "loss": 0.8074,
+      "step": 740
+    },
+    {
+      "epoch": 0.28876696505919724,
+      "grad_norm": 0.11629980802536011,
+      "learning_rate": 8.349049970236822e-06,
+      "loss": 0.8068,
+      "step": 750
+    },
+    {
+      "epoch": 0.29261719125998653,
+      "grad_norm": 0.12513957917690277,
+      "learning_rate": 8.302487923351842e-06,
+      "loss": 0.7996,
+      "step": 760
+    },
+    {
+      "epoch": 0.2964674174607758,
+      "grad_norm": 0.1326008439064026,
+      "learning_rate": 8.255412212597021e-06,
+      "loss": 0.8166,
+      "step": 770
+    },
+    {
+      "epoch": 0.3003176436615651,
+      "grad_norm": 0.12974192202091217,
+      "learning_rate": 8.207830160055688e-06,
+      "loss": 0.8101,
+      "step": 780
+    },
+    {
+      "epoch": 0.3041678698623544,
+      "grad_norm": 0.13810661435127258,
+      "learning_rate": 8.159749166566786e-06,
+      "loss": 0.7992,
+      "step": 790
+    },
+    {
+      "epoch": 0.30801809606314373,
+      "grad_norm": 0.13205720484256744,
+      "learning_rate": 8.11117671057377e-06,
+      "loss": 0.8271,
+      "step": 800
+    },
+    {
+      "epoch": 0.311868322263933,
+      "grad_norm": 0.12387409806251526,
+      "learning_rate": 8.062120346961404e-06,
+      "loss": 0.8136,
+      "step": 810
+    },
+    {
+      "epoch": 0.3157185484647223,
+      "grad_norm": 0.11777382344007492,
+      "learning_rate": 8.012587705880705e-06,
+      "loss": 0.8146,
+      "step": 820
+    },
+    {
+      "epoch": 0.3195687746655116,
+      "grad_norm": 0.12526744604110718,
+      "learning_rate": 7.962586491562146e-06,
+      "loss": 0.8028,
+      "step": 830
+    },
+    {
+      "epoch": 0.3234190008663009,
+      "grad_norm": 0.13423044979572296,
+      "learning_rate": 7.912124481117357e-06,
+      "loss": 0.807,
+      "step": 840
+    },
+    {
+      "epoch": 0.3272692270670902,
+      "grad_norm": 0.12047898769378662,
+      "learning_rate": 7.861209523329494e-06,
+      "loss": 0.8066,
+      "step": 850
+    },
+    {
+      "epoch": 0.3311194532678795,
+      "grad_norm": 0.12419753521680832,
+      "learning_rate": 7.809849537432432e-06,
+      "loss": 0.8009,
+      "step": 860
+    },
+    {
+      "epoch": 0.3349696794686688,
+      "grad_norm": 0.12232045829296112,
+      "learning_rate": 7.758052511879035e-06,
+      "loss": 0.7989,
+      "step": 870
+    },
+    {
+      "epoch": 0.3388199056694581,
+      "grad_norm": 0.11979101598262787,
+      "learning_rate": 7.705826503098642e-06,
+      "loss": 0.8045,
+      "step": 880
+    },
+    {
+      "epoch": 0.34267013187024736,
+      "grad_norm": 0.12611697614192963,
+      "learning_rate": 7.653179634243979e-06,
+      "loss": 0.7951,
+      "step": 890
+    },
+    {
+      "epoch": 0.34652035807103665,
+      "grad_norm": 0.12231084704399109,
+      "learning_rate": 7.600120093927696e-06,
+      "loss": 0.801,
+      "step": 900
+    },
+    {
+      "epoch": 0.350370584271826,
+      "grad_norm": 0.13126608729362488,
+      "learning_rate": 7.546656134948729e-06,
+      "loss": 0.8021,
+      "step": 910
+    },
+    {
+      "epoch": 0.3542208104726153,
+      "grad_norm": 0.11952194571495056,
+      "learning_rate": 7.492796073008661e-06,
+      "loss": 0.788,
+      "step": 920
+    },
+    {
+      "epoch": 0.35807103667340456,
+      "grad_norm": 0.1300327181816101,
+      "learning_rate": 7.438548285418326e-06,
+      "loss": 0.7817,
+      "step": 930
+    },
+    {
+      "epoch": 0.36192126287419385,
+      "grad_norm": 0.1310899257659912,
+      "learning_rate": 7.3839212097948e-06,
+      "loss": 0.818,
+      "step": 940
+    },
+    {
+      "epoch": 0.36577148907498314,
+      "grad_norm": 0.12046364694833755,
+      "learning_rate": 7.328923342749039e-06,
+      "loss": 0.829,
+      "step": 950
+    },
+    {
+      "epoch": 0.3696217152757725,
+      "grad_norm": 0.15349863469600677,
+      "learning_rate": 7.273563238564319e-06,
+      "loss": 0.8067,
+      "step": 960
+    },
+    {
+      "epoch": 0.37347194147656176,
+      "grad_norm": 0.12952382862567902,
+      "learning_rate": 7.217849507865724e-06,
+      "loss": 0.7873,
+      "step": 970
+    },
+    {
+      "epoch": 0.37732216767735105,
+      "grad_norm": 0.11644799262285233,
+      "learning_rate": 7.161790816280858e-06,
+      "loss": 0.7965,
+      "step": 980
+    },
+    {
+      "epoch": 0.38117239387814034,
+      "grad_norm": 0.13094845414161682,
+      "learning_rate": 7.1053958830920125e-06,
+      "loss": 0.7837,
+      "step": 990
+    },
+    {
+      "epoch": 0.3850226200789296,
+      "grad_norm": 0.12655293941497803,
+      "learning_rate": 7.0486734798799715e-06,
+      "loss": 0.8067,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3888728462797189,
+      "grad_norm": 0.13061296939849854,
+      "learning_rate": 6.991632429159708e-06,
+      "loss": 0.8262,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39272307248050825,
+      "grad_norm": 0.13151878118515015,
+      "learning_rate": 6.934281603008132e-06,
+      "loss": 0.8032,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39657329868129754,
+      "grad_norm": 0.11932211369276047,
+      "learning_rate": 6.876629921684143e-06,
+      "loss": 0.8141,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4004235248820868,
+      "grad_norm": 0.12402301281690598,
+      "learning_rate": 6.818686352241199e-06,
+      "loss": 0.8123,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4042737510828761,
+      "grad_norm": 0.11400502175092697,
+      "learning_rate": 6.760459907132573e-06,
+      "loss": 0.8023,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4081239772836654,
+      "grad_norm": 0.1250140368938446,
+      "learning_rate": 6.701959642809589e-06,
+      "loss": 0.8075,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41197420348445474,
+      "grad_norm": 0.12766261398792267,
+      "learning_rate": 6.643194658312987e-06,
+      "loss": 0.7903,
+      "step": 1070
+    },
+    {
+      "epoch": 0.415824429685244,
+      "grad_norm": 0.12118445336818695,
+      "learning_rate": 6.584174093857676e-06,
+      "loss": 0.8226,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4196746558860333,
+      "grad_norm": 0.14290851354599,
+      "learning_rate": 6.5249071294110735e-06,
+      "loss": 0.7979,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4235248820868226,
+      "grad_norm": 0.14045311510562897,
+      "learning_rate": 6.465402983265274e-06,
+      "loss": 0.817,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4273751082876119,
+      "grad_norm": 0.13146770000457764,
+      "learning_rate": 6.405670910603243e-06,
+      "loss": 0.8106,
+      "step": 1110
+    },
+    {
+      "epoch": 0.43122533448840117,
+      "grad_norm": 0.14061707258224487,
+      "learning_rate": 6.345720202059291e-06,
+      "loss": 0.8053,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4350755606891905,
+      "grad_norm": 0.13088859617710114,
+      "learning_rate": 6.285560182274008e-06,
+      "loss": 0.7935,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4389257868899798,
+      "grad_norm": 0.12028706073760986,
+      "learning_rate": 6.225200208443936e-06,
+      "loss": 0.7983,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4427760130907691,
+      "grad_norm": 0.11800989508628845,
+      "learning_rate": 6.16464966886616e-06,
+      "loss": 0.7946,
+      "step": 1150
+    },
+    {
+      "epoch": 0.44662623929155837,
+      "grad_norm": 0.14426986873149872,
+      "learning_rate": 6.103917981478065e-06,
+      "loss": 0.8207,
+      "step": 1160
+    },
+    {
+      "epoch": 0.45047646549234766,
+      "grad_norm": 0.12402904778718948,
+      "learning_rate": 6.043014592392489e-06,
+      "loss": 0.8279,
+      "step": 1170
+    },
+    {
+      "epoch": 0.454326691693137,
+      "grad_norm": 0.11201053112745285,
+      "learning_rate": 5.981948974428485e-06,
+      "loss": 0.8007,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4581769178939263,
+      "grad_norm": 0.126358300447464,
+      "learning_rate": 5.920730625637934e-06,
+      "loss": 0.7821,
+      "step": 1190
+    },
+    {
+      "epoch": 0.46202714409471557,
+      "grad_norm": 0.13829633593559265,
+      "learning_rate": 5.8593690678282355e-06,
+      "loss": 0.7893,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46587737029550486,
+      "grad_norm": 0.12676101922988892,
+      "learning_rate": 5.797873845081301e-06,
+      "loss": 0.7944,
+      "step": 1210
+    },
+    {
+      "epoch": 0.46972759649629414,
+      "grad_norm": 0.12875781953334808,
+      "learning_rate": 5.736254522269075e-06,
+      "loss": 0.8096,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47357782269708343,
+      "grad_norm": 0.12596870958805084,
+      "learning_rate": 5.6745206835658386e-06,
+      "loss": 0.8105,
+      "step": 1230
+    },
+    {
+      "epoch": 0.47742804889787277,
+      "grad_norm": 0.11838774383068085,
+      "learning_rate": 5.612681930957495e-06,
+      "loss": 0.7897,
+      "step": 1240
+    },
+    {
+      "epoch": 0.48127827509866206,
+      "grad_norm": 0.12332314997911453,
+      "learning_rate": 5.550747882748095e-06,
+      "loss": 0.7875,
+      "step": 1250
+    },
+    {
+      "epoch": 0.48512850129945134,
+      "grad_norm": 0.13126707077026367,
+      "learning_rate": 5.4887281720638195e-06,
+      "loss": 0.7874,
+      "step": 1260
+    },
+    {
+      "epoch": 0.48897872750024063,
+      "grad_norm": 0.1406238079071045,
+      "learning_rate": 5.42663244535466e-06,
+      "loss": 0.7774,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4928289537010299,
+      "grad_norm": 0.13077838718891144,
+      "learning_rate": 5.3644703608940205e-06,
+      "loss": 0.7823,
+      "step": 1280
+    },
+    {
+      "epoch": 0.49667917990181926,
+      "grad_norm": 0.12134639173746109,
+      "learning_rate": 5.302251587276491e-06,
+      "loss": 0.8048,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5005294061026085,
+      "grad_norm": 0.12754687666893005,
+      "learning_rate": 5.2399858019140005e-06,
+      "loss": 0.7971,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5043796323033978,
+      "grad_norm": 0.14385583996772766,
+      "learning_rate": 5.177682689530615e-06,
+      "loss": 0.8125,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5082298585041871,
+      "grad_norm": 0.1386464536190033,
+      "learning_rate": 5.11535194065619e-06,
+      "loss": 0.7891,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5120800847049765,
+      "grad_norm": 0.14445945620536804,
+      "learning_rate": 5.053003250119113e-06,
+      "loss": 0.7918,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5159303109057657,
+      "grad_norm": 0.13509704172611237,
+      "learning_rate": 4.9906463155383925e-06,
+      "loss": 0.7969,
+      "step": 1340
+    },
+    {
+      "epoch": 0.519780537106555,
+      "grad_norm": 0.1331833302974701,
+      "learning_rate": 4.9282908358153046e-06,
+      "loss": 0.7868,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5236307633073443,
+      "grad_norm": 0.1402117908000946,
+      "learning_rate": 4.8659465096248345e-06,
+      "loss": 0.7996,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5274809895081336,
+      "grad_norm": 0.13380764424800873,
+      "learning_rate": 4.803623033907169e-06,
+      "loss": 0.8064,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5313312157089229,
+      "grad_norm": 0.12465987354516983,
+      "learning_rate": 4.741330102359441e-06,
+      "loss": 0.8095,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5351814419097122,
+      "grad_norm": 0.13688521087169647,
+      "learning_rate": 4.679077403927996e-06,
+      "loss": 0.7864,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5390316681105015,
+      "grad_norm": 0.12673544883728027,
+      "learning_rate": 4.616874621301373e-06,
+      "loss": 0.7909,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5428818943112907,
+      "grad_norm": 0.12841714918613434,
+      "learning_rate": 4.554731429404293e-06,
+      "loss": 0.8078,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5467321205120801,
+      "grad_norm": 0.13569600880146027,
+      "learning_rate": 4.492657493892822e-06,
+      "loss": 0.792,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5505823467128694,
+      "grad_norm": 0.13387185335159302,
+      "learning_rate": 4.430662469650999e-06,
+      "loss": 0.8017,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5544325729136587,
+      "grad_norm": 0.13110393285751343,
+      "learning_rate": 4.368755999289118e-06,
+      "loss": 0.8048,
+      "step": 1440
+    },
+    {
+      "epoch": 0.558282799114448,
+      "grad_norm": 0.142400860786438,
+      "learning_rate": 4.306947711643949e-06,
+      "loss": 0.7886,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5621330253152372,
+      "grad_norm": 0.12259452790021896,
+      "learning_rate": 4.245247220281068e-06,
+      "loss": 0.7927,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5659832515160266,
+      "grad_norm": 0.1265445500612259,
+      "learning_rate": 4.183664121999587e-06,
+      "loss": 0.7899,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5698334777168159,
+      "grad_norm": 0.14089535176753998,
+      "learning_rate": 4.122207995339474e-06,
+      "loss": 0.7872,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5736837039176051,
+      "grad_norm": 0.13186399638652802,
+      "learning_rate": 4.060888399091734e-06,
+      "loss": 0.7935,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5775339301183945,
+      "grad_norm": 0.13513027131557465,
+      "learning_rate": 3.999714870811639e-06,
+      "loss": 0.8013,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5813841563191837,
+      "grad_norm": 0.12127958983182907,
+      "learning_rate": 3.938696925335275e-06,
+      "loss": 0.7917,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5852343825199731,
+      "grad_norm": 0.137859508395195,
+      "learning_rate": 3.87784405329962e-06,
+      "loss": 0.8001,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5890846087207623,
+      "grad_norm": 0.1322646588087082,
+      "learning_rate": 3.817165719666382e-06,
+      "loss": 0.8004,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5929348349215516,
+      "grad_norm": 0.1266612559556961,
+      "learning_rate": 3.756671362249841e-06,
+      "loss": 0.8066,
+      "step": 1540
+    },
+    {
+      "epoch": 0.596785061122341,
+      "grad_norm": 0.12909898161888123,
+      "learning_rate": 3.6963703902488907e-06,
+      "loss": 0.7987,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6006352873231302,
+      "grad_norm": 0.13215824961662292,
+      "learning_rate": 3.6362721827835655e-06,
+      "loss": 0.8142,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6044855135239195,
+      "grad_norm": 0.20093874633312225,
+      "learning_rate": 3.576386087436212e-06,
+      "loss": 0.8,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6083357397247088,
+      "grad_norm": 0.1288207322359085,
+      "learning_rate": 3.5167214187975913e-06,
+      "loss": 0.8015,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6121859659254981,
+      "grad_norm": 0.11233828961849213,
+      "learning_rate": 3.457287457018086e-06,
+      "loss": 0.7999,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6160361921262875,
+      "grad_norm": 0.12344548106193542,
+      "learning_rate": 3.3980934463643003e-06,
+      "loss": 0.7819,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6198864183270767,
+      "grad_norm": 0.11386287957429886,
+      "learning_rate": 3.3391485937812053e-06,
+      "loss": 0.7654,
+      "step": 1610
+    },
+    {
+      "epoch": 0.623736644527866,
+      "grad_norm": 0.12353255599737167,
+      "learning_rate": 3.280462067460115e-06,
+      "loss": 0.7687,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6275868707286553,
+      "grad_norm": 0.1225055530667305,
+      "learning_rate": 3.222042995412669e-06,
+      "loss": 0.8041,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6314370969294446,
+      "grad_norm": 0.11444497853517532,
+      "learning_rate": 3.163900464051089e-06,
+      "loss": 0.7869,
+      "step": 1640
+    },
+    {
+      "epoch": 0.635287323130234,
+      "grad_norm": 0.1250477135181427,
+      "learning_rate": 3.1060435167748826e-06,
+      "loss": 0.797,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6391375493310232,
+      "grad_norm": 0.12704256176948547,
+      "learning_rate": 3.048481152564253e-06,
+      "loss": 0.7816,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6429877755318125,
+      "grad_norm": 0.1315152496099472,
+      "learning_rate": 2.9912223245804017e-06,
+      "loss": 0.8051,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6468380017326018,
+      "grad_norm": 0.12690001726150513,
+      "learning_rate": 2.934275938772981e-06,
+      "loss": 0.7917,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6506882279333911,
+      "grad_norm": 0.13660277426242828,
+      "learning_rate": 2.8776508524948676e-06,
+      "loss": 0.78,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6545384541341804,
+      "grad_norm": 0.11632051318883896,
+      "learning_rate": 2.821355873124503e-06,
+      "loss": 0.7766,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6583886803349697,
+      "grad_norm": 0.1328650712966919,
+      "learning_rate": 2.7653997566960144e-06,
+      "loss": 0.7974,
+      "step": 1710
+    },
+    {
+      "epoch": 0.662238906535759,
+      "grad_norm": 0.1264372617006302,
+      "learning_rate": 2.709791206537309e-06,
+      "loss": 0.7861,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6660891327365482,
+      "grad_norm": 0.11998788267374039,
+      "learning_rate": 2.6545388719163747e-06,
+      "loss": 0.7945,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6699393589373376,
+      "grad_norm": 0.12283533066511154,
+      "learning_rate": 2.599651346695979e-06,
+      "loss": 0.7966,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6737895851381268,
+      "grad_norm": 0.1326475590467453,
+      "learning_rate": 2.545137167997007e-06,
+      "loss": 0.7889,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6776398113389162,
+      "grad_norm": 0.1134677603840828,
+      "learning_rate": 2.4910048148705972e-06,
+      "loss": 0.7784,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6814900375397055,
+      "grad_norm": 0.13002271950244904,
+      "learning_rate": 2.4372627069793327e-06,
+      "loss": 0.8159,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6853402637404947,
+      "grad_norm": 0.12709783017635345,
+      "learning_rate": 2.3839192032876486e-06,
+      "loss": 0.8008,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6891904899412841,
+      "grad_norm": 0.11819744110107422,
+      "learning_rate": 2.3309826007617047e-06,
+      "loss": 0.8134,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6930407161420733,
+      "grad_norm": 0.11854253709316254,
+      "learning_rate": 2.278461133078881e-06,
+      "loss": 0.7783,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6968909423428626,
+      "grad_norm": 0.121813103556633,
+      "learning_rate": 2.2263629693471258e-06,
+      "loss": 0.7738,
+      "step": 1810
+    },
+    {
+      "epoch": 0.700741168543652,
+      "grad_norm": 0.11929237842559814,
+      "learning_rate": 2.1746962128343414e-06,
+      "loss": 0.7986,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7045913947444412,
+      "grad_norm": 0.11955148726701736,
+      "learning_rate": 2.1234688997080183e-06,
+      "loss": 0.7934,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7084416209452306,
+      "grad_norm": 0.1406019628047943,
+      "learning_rate": 2.0726889977853053e-06,
+      "loss": 0.7917,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7122918471460198,
+      "grad_norm": 0.13085028529167175,
+      "learning_rate": 2.022364405293703e-06,
+      "loss": 0.7997,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7161420733468091,
+      "grad_norm": 0.1215239018201828,
+      "learning_rate": 1.9725029496425814e-06,
+      "loss": 0.7968,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7199922995475985,
+      "grad_norm": 0.12898334860801697,
+      "learning_rate": 1.923112386205727e-06,
+      "loss": 0.7927,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7238425257483877,
+      "grad_norm": 0.12893567979335785,
+      "learning_rate": 1.874200397115079e-06,
+      "loss": 0.7978,
+      "step": 1880
+    },
+    {
+      "epoch": 0.727692751949177,
+      "grad_norm": 0.12435682862997055,
+      "learning_rate": 1.825774590065857e-06,
+      "loss": 0.7854,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7315429781499663,
+      "grad_norm": 0.13746371865272522,
+      "learning_rate": 1.7778424971332813e-06,
+      "loss": 0.7929,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7353932043507556,
+      "grad_norm": 0.11700835824012756,
+      "learning_rate": 1.7304115736010407e-06,
+      "loss": 0.7947,
+      "step": 1910
+    },
+    {
+      "epoch": 0.739243430551545,
+      "grad_norm": 0.12289503216743469,
+      "learning_rate": 1.6834891968017108e-06,
+      "loss": 0.7845,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7430936567523342,
+      "grad_norm": 0.1301659494638443,
+      "learning_rate": 1.6370826649692884e-06,
+      "loss": 0.8004,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7469438829531235,
+      "grad_norm": 0.12746968865394592,
+      "learning_rate": 1.5911991961040463e-06,
+      "loss": 0.8088,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7507941091539128,
+      "grad_norm": 0.12817108631134033,
+      "learning_rate": 1.545845926849851e-06,
+      "loss": 0.7952,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7546443353547021,
+      "grad_norm": 0.1319892704486847,
+      "learning_rate": 1.5010299113841397e-06,
+      "loss": 0.8117,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7584945615554913,
+      "grad_norm": 0.12761011719703674,
+      "learning_rate": 1.4567581203207238e-06,
+      "loss": 0.8066,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7623447877562807,
+      "grad_norm": 0.12489989399909973,
+      "learning_rate": 1.413037439625593e-06,
+      "loss": 0.7918,
+      "step": 1980
+    },
+    {
+      "epoch": 0.76619501395707,
+      "grad_norm": 0.14400576055049896,
+      "learning_rate": 1.3698746695458847e-06,
+      "loss": 0.794,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7700452401578592,
+      "grad_norm": 0.12268252670764923,
+      "learning_rate": 1.3272765235521745e-06,
+      "loss": 0.7939,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7738954663586486,
+      "grad_norm": 0.12257197499275208,
+      "learning_rate": 1.2852496272942877e-06,
+      "loss": 0.7907,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7777456925594378,
+      "grad_norm": 0.11626384407281876,
+      "learning_rate": 1.2438005175707396e-06,
+      "loss": 0.7812,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7815959187602272,
+      "grad_norm": 0.13488733768463135,
+      "learning_rate": 1.2029356413120297e-06,
+      "loss": 0.7684,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7854461449610165,
+      "grad_norm": 0.13001152873039246,
+      "learning_rate": 1.1626613545778786e-06,
+      "loss": 0.7854,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7892963711618057,
+      "grad_norm": 0.11777208745479584,
+      "learning_rate": 1.1229839215686294e-06,
+      "loss": 0.791,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7931465973625951,
+      "grad_norm": 0.12041101604700089,
+      "learning_rate": 1.0839095136509188e-06,
+      "loss": 0.7746,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7969968235633843,
+      "grad_norm": 0.1391756534576416,
+      "learning_rate": 1.045444208397791e-06,
+      "loss": 0.7879,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8008470497641736,
+      "grad_norm": 0.13181354105472565,
+      "learning_rate": 1.007593988643402e-06,
+      "loss": 0.7718,
+      "step": 2080
+    },
+    {
+      "epoch": 0.804697275964963,
+      "grad_norm": 0.12342358380556107,
+      "learning_rate": 9.703647415524613e-07,
+      "loss": 0.7961,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8085475021657522,
+      "grad_norm": 0.12620916962623596,
+      "learning_rate": 9.337622577045485e-07,
+      "loss": 0.7894,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8123977283665416,
+      "grad_norm": 0.12857596576213837,
+      "learning_rate": 8.977922301934588e-07,
+      "loss": 0.8002,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8162479545673308,
+      "grad_norm": 0.12429419904947281,
+      "learning_rate": 8.624602537417009e-07,
+      "loss": 0.7986,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8200981807681201,
+      "grad_norm": 0.12372923642396927,
+      "learning_rate": 8.277718238303123e-07,
+      "loss": 0.7845,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8239484069689095,
+      "grad_norm": 0.12361016124486923,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.7961,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8277986331696987,
+      "grad_norm": 0.13457311689853668,
+      "learning_rate": 7.603470842324223e-07,
+      "loss": 0.7816,
+      "step": 2150
+    },
+    {
+      "epoch": 0.831648859370488,
+      "grad_norm": 0.12227658927440643,
+      "learning_rate": 7.276212616857575e-07,
+      "loss": 0.7965,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8354990855712773,
+      "grad_norm": 0.11822185665369034,
+      "learning_rate": 6.955599583279837e-07,
+      "loss": 0.8001,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8393493117720666,
+      "grad_norm": 0.12294301390647888,
+      "learning_rate": 6.641681609246981e-07,
+      "loss": 0.7877,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8431995379728559,
+      "grad_norm": 0.13270793855190277,
+      "learning_rate": 6.334507521075711e-07,
+      "loss": 0.7997,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8470497641736452,
+      "grad_norm": 0.13098740577697754,
+      "learning_rate": 6.034125096149179e-07,
+      "loss": 0.7819,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8508999903744345,
+      "grad_norm": 0.13126066327095032,
+      "learning_rate": 5.740581055485722e-07,
+      "loss": 0.7706,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8547502165752238,
+      "grad_norm": 0.11594405025243759,
+      "learning_rate": 5.453921056471961e-07,
+      "loss": 0.7949,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8586004427760131,
+      "grad_norm": 0.13695979118347168,
+      "learning_rate": 5.174189685761288e-07,
+      "loss": 0.7842,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8624506689768023,
+      "grad_norm": 0.1290534883737564,
+      "learning_rate": 4.901430452338979e-07,
+      "loss": 0.8034,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8663008951775917,
+      "grad_norm": 0.12216270714998245,
+      "learning_rate": 4.6356857807548286e-07,
+      "loss": 0.7863,
+      "step": 2250
+    },
+    {
+      "epoch": 0.870151121378381,
+      "grad_norm": 0.13465340435504913,
+      "learning_rate": 4.376997004524541e-07,
+      "loss": 0.7745,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8740013475791703,
+      "grad_norm": 0.12048375606536865,
+      "learning_rate": 4.12540435970073e-07,
+      "loss": 0.781,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8778515737799596,
+      "grad_norm": 0.12305070459842682,
+      "learning_rate": 3.880946978614703e-07,
+      "loss": 0.795,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8817017999807488,
+      "grad_norm": 0.1224573627114296,
+      "learning_rate": 3.643662883789878e-07,
+      "loss": 0.8007,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8855520261815382,
+      "grad_norm": 0.11956695467233658,
+      "learning_rate": 3.41358898202781e-07,
+      "loss": 0.7802,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8894022523823275,
+      "grad_norm": 0.11497985571622849,
+      "learning_rate": 3.190761058667741e-07,
+      "loss": 0.782,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8932524785831167,
+      "grad_norm": 0.11563810706138611,
+      "learning_rate": 2.9752137720206463e-07,
+      "loss": 0.7923,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8971027047839061,
+      "grad_norm": 0.12866128981113434,
+      "learning_rate": 2.7669806479785067e-07,
+      "loss": 0.8033,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9009529309846953,
+      "grad_norm": 0.1292843073606491,
+      "learning_rate": 2.566094074799763e-07,
+      "loss": 0.7939,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9048031571854847,
+      "grad_norm": 0.12106771767139435,
+      "learning_rate": 2.3725852980716557e-07,
+      "loss": 0.7743,
+      "step": 2350
+    },
+    {
+      "epoch": 0.908653383386274,
+      "grad_norm": 0.12277862429618835,
+      "learning_rate": 2.1864844158503983e-07,
+      "loss": 0.7939,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9125036095870632,
+      "grad_norm": 0.13536040484905243,
+      "learning_rate": 2.007820373979702e-07,
+      "loss": 0.7789,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9163538357878526,
+      "grad_norm": 0.1295584738254547,
+      "learning_rate": 1.8366209615886532e-07,
+      "loss": 0.7829,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9202040619886418,
+      "grad_norm": 0.12073605507612228,
+      "learning_rate": 1.6729128067693613e-07,
+      "loss": 0.8011,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9240542881894311,
+      "grad_norm": 0.13767144083976746,
+      "learning_rate": 1.5167213724353426e-07,
+      "loss": 0.7915,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9279045143902205,
+      "grad_norm": 0.12394948303699493,
+      "learning_rate": 1.3680709523610058e-07,
+      "loss": 0.7815,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9317547405910097,
+      "grad_norm": 0.11637039482593536,
+      "learning_rate": 1.2269846674030418e-07,
+      "loss": 0.8056,
+      "step": 2420
+    },
+    {
+      "epoch": 0.935604966791799,
+      "grad_norm": 0.1259966939687729,
+      "learning_rate": 1.0934844619042595e-07,
+      "loss": 0.8014,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9394551929925883,
+      "grad_norm": 0.13369697332382202,
+      "learning_rate": 9.675911002803961e-08,
+      "loss": 0.789,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9433054191933776,
+      "grad_norm": 0.1341429501771927,
+      "learning_rate": 8.493241637904381e-08,
+      "loss": 0.788,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9471556453941669,
+      "grad_norm": 0.1291925311088562,
+      "learning_rate": 7.387020474909689e-08,
+      "loss": 0.8073,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9510058715949562,
+      "grad_norm": 0.12333804368972778,
+      "learning_rate": 6.357419573750856e-08,
+      "loss": 0.7915,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9548560977957455,
+      "grad_norm": 0.12546703219413757,
+      "learning_rate": 5.4045990769613365e-08,
+      "loss": 0.7875,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9587063239965348,
+      "grad_norm": 0.14507997035980225,
+      "learning_rate": 4.5287071847694434e-08,
+      "loss": 0.7888,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9625565501973241,
+      "grad_norm": 0.12240733206272125,
+      "learning_rate": 3.7298801320467817e-08,
+      "loss": 0.7818,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9664067763981133,
+      "grad_norm": 0.1249658614397049,
+      "learning_rate": 3.0082421671192576e-08,
+      "loss": 0.7789,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9702570025989027,
+      "grad_norm": 0.11926795542240143,
+      "learning_rate": 2.3639055324410355e-08,
+      "loss": 0.7942,
+      "step": 2520
+    },
+    {
+      "epoch": 0.974107228799692,
+      "grad_norm": 0.12819339334964752,
+      "learning_rate": 1.7969704471369477e-08,
+      "loss": 0.7828,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9779574550004813,
+      "grad_norm": 0.11736205220222473,
+      "learning_rate": 1.3075250914142413e-08,
+      "loss": 0.7888,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9818076812012706,
+      "grad_norm": 0.11923353374004364,
+      "learning_rate": 8.956455928476049e-09,
+      "loss": 0.7976,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9856579074020598,
+      "grad_norm": 0.11851949989795685,
+      "learning_rate": 5.6139601453786275e-09,
+      "loss": 0.7779,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9895081336028492,
+      "grad_norm": 0.12628617882728577,
+      "learning_rate": 3.0482834514805602e-09,
+      "loss": 0.7794,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9933583598036385,
+      "grad_norm": 0.11309806257486343,
+      "learning_rate": 1.2598249081713366e-09,
+      "loss": 0.7704,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9972085860044277,
+      "grad_norm": 0.1322404444217682,
+      "learning_rate": 2.488626895302826e-10,
+      "loss": 0.7949,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9999037443449803,
+      "step": 2597,
+      "total_flos": 2.995607458728116e+18,
+      "train_loss": 0.8253552240732316,
+      "train_runtime": 9332.6825,
+      "train_samples_per_second": 71.245,
+      "train_steps_per_second": 0.278
+    }
+  ],
+  "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": 2.995607458728116e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3/mc1_b32_224_2.5b/finetune/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8af37103c1f7fa3b5b171b42c5e71f2d475306c2885f25fb567db4165af8d58a
+size 6136

continuous/qwen3/mc1_b32_224_2.5b/pretrain/config.json

@@ -0,0 +1,60 @@
+{
+  "_attn_implementation_autoset": true,
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 6144,
+  "max_position_embeddings": 40960,
+  "max_window_layers": 28,
+  "mm_hidden_size": 768,
+  "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-B-32",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "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/mc1/mc1_b32_224_2.5b.pt",
+  "vocab_size": 151936
+}

continuous/qwen3/mc1_b32_224_2.5b/pretrain/mm_projector.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:500bd16450aef8134206e210bd91f94ba042aa22f647279d86fe243e55636eb1
+size 23087160

continuous/qwen3/mc1_b32_224_2.5b/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": 5.1200408935546875,
+      "learning_rate": 0.00032142857142857147,
+      "loss": 9.5168,
+      "step": 10
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 1.2918411493301392,
+      "learning_rate": 0.0006785714285714287,
+      "loss": 6.8369,
+      "step": 20
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 1.7169702053070068,
+      "learning_rate": 0.0009999907688176173,
+      "loss": 5.7871,
+      "step": 30
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 0.26706209778785706,
+      "learning_rate": 0.0009988834393115766,
+      "loss": 5.0836,
+      "step": 40
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 0.14363327622413635,
+      "learning_rate": 0.0009959345572940772,
+      "loss": 4.8486,
+      "step": 50
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 0.0969676598906517,
+      "learning_rate": 0.0009911550081155983,
+      "loss": 4.7717,
+      "step": 60
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.07326562702655792,
+      "learning_rate": 0.0009845624347566433,
+      "loss": 4.7132,
+      "step": 70
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 0.17319589853286743,
+      "learning_rate": 0.0009761811727013547,
+      "loss": 4.6836,
+      "step": 80
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 0.1214609369635582,
+      "learning_rate": 0.0009660421601067666,
+      "loss": 4.6484,
+      "step": 90
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 0.12182394415140152,
+      "learning_rate": 0.0009541828235992926,
+      "loss": 4.6265,
+      "step": 100
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 0.13561159372329712,
+      "learning_rate": 0.0009406469401200151,
+      "loss": 4.592,
+      "step": 110
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 1.075958013534546,
+      "learning_rate": 0.0009254844753287492,
+      "loss": 4.5791,
+      "step": 120
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 0.19902704656124115,
+      "learning_rate": 0.0009087513991633924,
+      "loss": 4.7492,
+      "step": 130
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 0.10625278949737549,
+      "learning_rate": 0.0008905094792353917,
+      "loss": 4.6275,
+      "step": 140
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 0.09212242066860199,
+      "learning_rate": 0.0008708260528239789,
+      "loss": 4.5873,
+      "step": 150
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.07144082337617874,
+      "learning_rate": 0.0008497737783108237,
+      "loss": 4.5503,
+      "step": 160
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 0.07914894074201584,
+      "learning_rate": 0.0008274303669726426,
+      "loss": 4.5228,
+      "step": 170
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 0.07171554863452911,
+      "learning_rate": 0.0008038782961218137,
+      "loss": 4.4996,
+      "step": 180
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.06347852200269699,
+      "learning_rate": 0.000779204504653894,
+      "loss": 4.4874,
+      "step": 190
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 0.08318465948104858,
+      "learning_rate": 0.0007535000721258791,
+      "loss": 4.4587,
+      "step": 200
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 0.08610600978136063,
+      "learning_rate": 0.00072685988254984,
+      "loss": 4.4456,
+      "step": 210
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 0.08051843196153641,
+      "learning_rate": 0.0006993822741429906,
+      "loss": 4.4512,
+      "step": 220
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.10511162877082825,
+      "learning_rate": 0.0006711686763270818,
+      "loss": 4.4418,
+      "step": 230
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 0.10291241854429245,
+      "learning_rate": 0.0006423232353170831,
+      "loss": 4.421,
+      "step": 240
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 0.09839367121458054,
+      "learning_rate": 0.0006129524296812335,
+      "loss": 4.3999,
+      "step": 250
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 0.08939998596906662,
+      "learning_rate": 0.0005831646772915651,
+      "loss": 4.4045,
+      "step": 260
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.11251821368932724,
+      "learning_rate": 0.0005530699351157809,
+      "loss": 4.3819,
+      "step": 270
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 0.09093751013278961,
+      "learning_rate": 0.0005227792933277943,
+      "loss": 4.3766,
+      "step": 280
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 0.11293774098157883,
+      "learning_rate": 0.0004924045652352092,
+      "loss": 4.3566,
+      "step": 290
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 0.11815927922725677,
+      "learning_rate": 0.0004620578745374633,
+      "loss": 4.352,
+      "step": 300
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 0.10566351562738419,
+      "learning_rate": 0.0004318512414382059,
+      "loss": 4.3667,
+      "step": 310
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 0.09344740211963654,
+      "learning_rate": 0.0004018961691397148,
+      "loss": 4.3545,
+      "step": 320
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 0.10354839265346527,
+      "learning_rate": 0.0003723032322457458,
+      "loss": 4.3335,
+      "step": 330
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.10503047704696655,
+      "learning_rate": 0.0003431816685921625,
+      "loss": 4.335,
+      "step": 340
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 0.09549292176961899,
+      "learning_rate": 0.0003146389760120416,
+      "loss": 4.3376,
+      "step": 350
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 0.12899629771709442,
+      "learning_rate": 0.00028678051552373485,
+      "loss": 4.3436,
+      "step": 360
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 0.09264609217643738,
+      "learning_rate": 0.0002597091224066581,
+      "loss": 4.3233,
+      "step": 370
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 0.10274000465869904,
+      "learning_rate": 0.0002335247266004629,
+      "loss": 4.3381,
+      "step": 380
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 0.09702663123607635,
+      "learning_rate": 0.000208323983828828,
+      "loss": 4.3188,
+      "step": 390
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.10859082639217377,
+      "learning_rate": 0.0001841999188095224,
+      "loss": 4.3201,
+      "step": 400
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 0.08885742723941803,
+      "learning_rate": 0.00016124158186777676,
+      "loss": 4.3189,
+      "step": 410
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 0.10839273780584335,
+      "learning_rate": 0.0001395337202205211,
+      "loss": 4.311,
+      "step": 420
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 0.10854825377464294,
+      "learning_rate": 0.00011915646514489292,
+      "loss": 4.3166,
+      "step": 430
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 0.11812586337327957,
+      "learning_rate": 0.00010018503618578818,
+      "loss": 4.3219,
+      "step": 440
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.10054424405097961,
+      "learning_rate": 8.268946349432583e-05,
+      "loss": 4.3119,
+      "step": 450
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 0.09567540884017944,
+      "learning_rate": 6.673432932217117e-05,
+      "loss": 4.3048,
+      "step": 460
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.08656623214483261,
+      "learning_rate": 5.2378529625954684e-05,
+      "loss": 4.2913,
+      "step": 470
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.10255353152751923,
+      "learning_rate": 3.9675056661785556e-05,
+      "loss": 4.3022,
+      "step": 480
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.09151938557624817,
+      "learning_rate": 2.8670803372379538e-05,
+      "loss": 4.3026,
+      "step": 490
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 0.08885455876588821,
+      "learning_rate": 1.9406390288876586e-05,
+      "loss": 4.315,
+      "step": 500
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 0.08728527277708054,
+      "learning_rate": 1.191601558631461e-05,
+      "loss": 4.3032,
+      "step": 510
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 0.08805518597364426,
+      "learning_rate": 6.2273288462559305e-06,
+      "loss": 4.2959,
+      "step": 520
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 0.0912131741642952,
+      "learning_rate": 2.361328992552314e-06,
+      "loss": 4.2979,
+      "step": 530
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 0.08712775260210037,
+      "learning_rate": 3.322867770044202e-07,
+      "loss": 4.3045,
+      "step": 540
+    },
+    {
+      "epoch": 1.0,
+      "step": 545,
+      "total_flos": 2.988888507453276e+17,
+      "train_loss": 4.6064112672018345,
+      "train_runtime": 5622.3353,
+      "train_samples_per_second": 99.27,
+      "train_steps_per_second": 0.097
+    }
+  ],
+  "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": 2.988888507453276e+17,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3/mc1_b32_224_400m/finetune/added_tokens.json

@@ -0,0 +1,28 @@
+{
+  "</think>": 151668,
+  "</tool_call>": 151658,
+  "</tool_response>": 151666,
+  "<think>": 151667,
+  "<tool_call>": 151657,
+  "<tool_response>": 151665,
+  "<|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/qwen3/mc1_b32_224_400m/finetune/config.json

@@ -0,0 +1,59 @@
+{
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "LlavaQwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 6144,
+  "max_position_embeddings": 40960,
+  "max_window_layers": 28,
+  "mm_hidden_size": 768,
+  "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-B-32",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "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/mc1/mc1_b32_224_400m.pt",
+  "vocab_size": 151936
+}

continuous/qwen3/mc1_b32_224_400m/finetune/generation_config.json

@@ -0,0 +1,13 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "temperature": 0.6,
+  "top_k": 20,
+  "top_p": 0.95,
+  "transformers_version": "4.51.3"
+}

continuous/qwen3/mc1_b32_224_400m/finetune/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:84cd26e9dd21e17a2ea8eb080d6a6350e56bccada45e86fdaa9c1cda88352f9c
+size 4262321400

continuous/qwen3/mc1_b32_224_400m/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/qwen3/mc1_b32_224_400m/finetune/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
+size 11422654

continuous/qwen3/mc1_b32_224_400m/finetune/tokenizer_config.json

@@ -0,0 +1,241 @@
+{
+  "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
+    },
+    "151665": {
+      "content": "<tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</think>",
+      "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\\n' }}\n    {%- endif %}\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    {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for message in messages[::-1] %}\n    {%- set index = (messages|length - 1) - loop.index0 %}\n    {%- if ns.multi_step_tool and message.role == \"user\" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}\n        {%- set ns.multi_step_tool = false %}\n        {%- set ns.last_query_index = index %}\n    {%- endif %}\n{%- endfor %}\n{%- for message in messages %}\n    {%- if message.content is string %}\n        {%- set content = message.content %}\n    {%- else %}\n        {%- set content = '' %}\n    {%- endif %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n        {{- '<|im_start|>' + message.role + '\\n' + content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set reasoning_content = '' %}\n        {%- if message.reasoning_content is string %}\n            {%- set reasoning_content = message.reasoning_content %}\n        {%- else %}\n            {%- if '</think>' in content %}\n                {%- set reasoning_content = content.split('</think>')[0].rstrip('\\n').split('<think>')[-1].lstrip('\\n') %}\n                {%- set content = content.split('</think>')[-1].lstrip('\\n') %}\n            {%- endif %}\n        {%- endif %}\n        {%- if loop.index0 > ns.last_query_index %}\n            {%- if loop.last or (not loop.last and reasoning_content) %}\n                {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content.strip('\\n') + '\\n</think>\\n\\n' + content.lstrip('\\n') }}\n            {%- else %}\n                {{- '<|im_start|>' + message.role + '\\n' + content }}\n            {%- endif %}\n        {%- else %}\n            {{- '<|im_start|>' + message.role + '\\n' + content }}\n        {%- endif %}\n        {%- if message.tool_calls %}\n            {%- for tool_call in message.tool_calls %}\n                {%- if (loop.first and content) or (not loop.first) %}\n                    {{- '\\n' }}\n                {%- endif %}\n                {%- if tool_call.function %}\n                    {%- set tool_call = tool_call.function %}\n                {%- endif %}\n                {{- '<tool_call>\\n{\"name\": \"' }}\n                {{- tool_call.name }}\n                {{- '\", \"arguments\": ' }}\n                {%- if tool_call.arguments is string %}\n                    {{- tool_call.arguments }}\n                {%- else %}\n                    {{- tool_call.arguments | tojson }}\n                {%- endif %}\n                {{- '}\\n</tool_call>' }}\n            {%- endfor %}\n        {%- endif %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if loop.first or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- 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    {%- if enable_thinking is defined and enable_thinking is false %}\n        {{- '<think>\\n\\n</think>\\n\\n' }}\n    {%- endif %}\n{%- endif %}",
+  "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/qwen3/mc1_b32_224_400m/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": 10.257267951965332,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 2.6494,
+      "step": 10
+    },
+    {
+      "epoch": 0.007700452401578592,
+      "grad_norm": 4.281728744506836,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 2.1298,
+      "step": 20
+    },
+    {
+      "epoch": 0.01155067860236789,
+      "grad_norm": 2.4569058418273926,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.5923,
+      "step": 30
+    },
+    {
+      "epoch": 0.015400904803157185,
+      "grad_norm": 0.7289968132972717,
+      "learning_rate": 5e-06,
+      "loss": 1.2174,
+      "step": 40
+    },
+    {
+      "epoch": 0.019251131003946483,
+      "grad_norm": 0.4764060080051422,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 1.1002,
+      "step": 50
+    },
+    {
+      "epoch": 0.02310135720473578,
+      "grad_norm": 0.2536356449127197,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 1.0199,
+      "step": 60
+    },
+    {
+      "epoch": 0.026951583405525073,
+      "grad_norm": 0.20425951480865479,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.9744,
+      "step": 70
+    },
+    {
+      "epoch": 0.03080180960631437,
+      "grad_norm": 0.1767134815454483,
+      "learning_rate": 9.999996111488723e-06,
+      "loss": 0.9685,
+      "step": 80
+    },
+    {
+      "epoch": 0.034652035807103666,
+      "grad_norm": 0.1766105741262436,
+      "learning_rate": 9.999529497453782e-06,
+      "loss": 0.9408,
+      "step": 90
+    },
+    {
+      "epoch": 0.038502262007892966,
+      "grad_norm": 0.14789658784866333,
+      "learning_rate": 9.998285264324193e-06,
+      "loss": 0.9141,
+      "step": 100
+    },
+    {
+      "epoch": 0.04235248820868226,
+      "grad_norm": 0.154489666223526,
+      "learning_rate": 9.996263605626052e-06,
+      "loss": 0.8903,
+      "step": 110
+    },
+    {
+      "epoch": 0.04620271440947156,
+      "grad_norm": 0.14545302093029022,
+      "learning_rate": 9.99346483580503e-06,
+      "loss": 0.8978,
+      "step": 120
+    },
+    {
+      "epoch": 0.05005294061026085,
+      "grad_norm": 0.13559427857398987,
+      "learning_rate": 9.989889390177467e-06,
+      "loss": 0.8893,
+      "step": 130
+    },
+    {
+      "epoch": 0.053903166811050146,
+      "grad_norm": 0.16056084632873535,
+      "learning_rate": 9.985537824862647e-06,
+      "loss": 0.8846,
+      "step": 140
+    },
+    {
+      "epoch": 0.057753393011839446,
+      "grad_norm": 0.16464664041996002,
+      "learning_rate": 9.980410816696317e-06,
+      "loss": 0.8801,
+      "step": 150
+    },
+    {
+      "epoch": 0.06160361921262874,
+      "grad_norm": 0.14926864206790924,
+      "learning_rate": 9.97450916312541e-06,
+      "loss": 0.8745,
+      "step": 160
+    },
+    {
+      "epoch": 0.06545384541341803,
+      "grad_norm": 0.12472842633724213,
+      "learning_rate": 9.967833782084004e-06,
+      "loss": 0.8718,
+      "step": 170
+    },
+    {
+      "epoch": 0.06930407161420733,
+      "grad_norm": 0.13940542936325073,
+      "learning_rate": 9.960385711850554e-06,
+      "loss": 0.896,
+      "step": 180
+    },
+    {
+      "epoch": 0.07315429781499663,
+      "grad_norm": 0.1432742178440094,
+      "learning_rate": 9.952166110886394e-06,
+      "loss": 0.8952,
+      "step": 190
+    },
+    {
+      "epoch": 0.07700452401578593,
+      "grad_norm": 0.15245279669761658,
+      "learning_rate": 9.943176257655567e-06,
+      "loss": 0.8739,
+      "step": 200
+    },
+    {
+      "epoch": 0.08085475021657522,
+      "grad_norm": 0.14237819612026215,
+      "learning_rate": 9.933417550425952e-06,
+      "loss": 0.8787,
+      "step": 210
+    },
+    {
+      "epoch": 0.08470497641736452,
+      "grad_norm": 0.13912205398082733,
+      "learning_rate": 9.9228915070518e-06,
+      "loss": 0.858,
+      "step": 220
+    },
+    {
+      "epoch": 0.08855520261815382,
+      "grad_norm": 0.14294469356536865,
+      "learning_rate": 9.911599764737632e-06,
+      "loss": 0.8637,
+      "step": 230
+    },
+    {
+      "epoch": 0.09240542881894312,
+      "grad_norm": 0.14080236852169037,
+      "learning_rate": 9.899544079783606e-06,
+      "loss": 0.8496,
+      "step": 240
+    },
+    {
+      "epoch": 0.0962556550197324,
+      "grad_norm": 0.1424482762813568,
+      "learning_rate": 9.886726327312338e-06,
+      "loss": 0.8659,
+      "step": 250
+    },
+    {
+      "epoch": 0.1001058812205217,
+      "grad_norm": 0.13282063603401184,
+      "learning_rate": 9.87314850097724e-06,
+      "loss": 0.8702,
+      "step": 260
+    },
+    {
+      "epoch": 0.103956107421311,
+      "grad_norm": 0.13034920394420624,
+      "learning_rate": 9.858812712652452e-06,
+      "loss": 0.8541,
+      "step": 270
+    },
+    {
+      "epoch": 0.10780633362210029,
+      "grad_norm": 0.12787355482578278,
+      "learning_rate": 9.843721192104341e-06,
+      "loss": 0.8588,
+      "step": 280
+    },
+    {
+      "epoch": 0.11165655982288959,
+      "grad_norm": 0.13563811779022217,
+      "learning_rate": 9.827876286644698e-06,
+      "loss": 0.8698,
+      "step": 290
+    },
+    {
+      "epoch": 0.11550678602367889,
+      "grad_norm": 0.12509486079216003,
+      "learning_rate": 9.811280460765643e-06,
+      "loss": 0.8308,
+      "step": 300
+    },
+    {
+      "epoch": 0.11935701222446819,
+      "grad_norm": 0.13254617154598236,
+      "learning_rate": 9.793936295756292e-06,
+      "loss": 0.848,
+      "step": 310
+    },
+    {
+      "epoch": 0.12320723842525748,
+      "grad_norm": 0.13493067026138306,
+      "learning_rate": 9.775846489301279e-06,
+      "loss": 0.8425,
+      "step": 320
+    },
+    {
+      "epoch": 0.12705746462604678,
+      "grad_norm": 0.13145172595977783,
+      "learning_rate": 9.757013855061157e-06,
+      "loss": 0.8462,
+      "step": 330
+    },
+    {
+      "epoch": 0.13090769082683607,
+      "grad_norm": 0.13543322682380676,
+      "learning_rate": 9.737441322234758e-06,
+      "loss": 0.8551,
+      "step": 340
+    },
+    {
+      "epoch": 0.13475791702762538,
+      "grad_norm": 0.1493712067604065,
+      "learning_rate": 9.717131935103605e-06,
+      "loss": 0.8608,
+      "step": 350
+    },
+    {
+      "epoch": 0.13860814322841467,
+      "grad_norm": 0.12839123606681824,
+      "learning_rate": 9.696088852558398e-06,
+      "loss": 0.8438,
+      "step": 360
+    },
+    {
+      "epoch": 0.14245836942920398,
+      "grad_norm": 0.13356612622737885,
+      "learning_rate": 9.674315347607686e-06,
+      "loss": 0.8679,
+      "step": 370
+    },
+    {
+      "epoch": 0.14630859562999327,
+      "grad_norm": 0.1263704001903534,
+      "learning_rate": 9.651814806868796e-06,
+      "loss": 0.8367,
+      "step": 380
+    },
+    {
+      "epoch": 0.15015882183078255,
+      "grad_norm": 0.15660174190998077,
+      "learning_rate": 9.62859073004107e-06,
+      "loss": 0.8434,
+      "step": 390
+    },
+    {
+      "epoch": 0.15400904803157187,
+      "grad_norm": 0.1289655566215515,
+      "learning_rate": 9.604646729361547e-06,
+      "loss": 0.8448,
+      "step": 400
+    },
+    {
+      "epoch": 0.15785927423236115,
+      "grad_norm": 0.12890788912773132,
+      "learning_rate": 9.579986529043099e-06,
+      "loss": 0.8253,
+      "step": 410
+    },
+    {
+      "epoch": 0.16170950043315044,
+      "grad_norm": 0.12331058084964752,
+      "learning_rate": 9.554613964695189e-06,
+      "loss": 0.8413,
+      "step": 420
+    },
+    {
+      "epoch": 0.16555972663393975,
+      "grad_norm": 0.1267324686050415,
+      "learning_rate": 9.528532982727281e-06,
+      "loss": 0.8334,
+      "step": 430
+    },
+    {
+      "epoch": 0.16940995283472904,
+      "grad_norm": 0.14451685547828674,
+      "learning_rate": 9.501747639735018e-06,
+      "loss": 0.8348,
+      "step": 440
+    },
+    {
+      "epoch": 0.17326017903551832,
+      "grad_norm": 0.13507826626300812,
+      "learning_rate": 9.474262101869272e-06,
+      "loss": 0.8172,
+      "step": 450
+    },
+    {
+      "epoch": 0.17711040523630764,
+      "grad_norm": 0.12986357510089874,
+      "learning_rate": 9.446080644188141e-06,
+      "loss": 0.8334,
+      "step": 460
+    },
+    {
+      "epoch": 0.18096063143709692,
+      "grad_norm": 0.15500818192958832,
+      "learning_rate": 9.417207649992017e-06,
+      "loss": 0.839,
+      "step": 470
+    },
+    {
+      "epoch": 0.18481085763788624,
+      "grad_norm": 0.12644755840301514,
+      "learning_rate": 9.387647610141816e-06,
+      "loss": 0.8324,
+      "step": 480
+    },
+    {
+      "epoch": 0.18866108383867553,
+      "grad_norm": 0.12099631130695343,
+      "learning_rate": 9.35740512236047e-06,
+      "loss": 0.8219,
+      "step": 490
+    },
+    {
+      "epoch": 0.1925113100394648,
+      "grad_norm": 0.11787678301334381,
+      "learning_rate": 9.326484890517809e-06,
+      "loss": 0.8241,
+      "step": 500
+    },
+    {
+      "epoch": 0.19636153624025413,
+      "grad_norm": 0.12572579085826874,
+      "learning_rate": 9.294891723898926e-06,
+      "loss": 0.8113,
+      "step": 510
+    },
+    {
+      "epoch": 0.2002117624410434,
+      "grad_norm": 0.1257006973028183,
+      "learning_rate": 9.262630536456153e-06,
+      "loss": 0.8513,
+      "step": 520
+    },
+    {
+      "epoch": 0.2040619886418327,
+      "grad_norm": 0.138712078332901,
+      "learning_rate": 9.229706346044749e-06,
+      "loss": 0.8179,
+      "step": 530
+    },
+    {
+      "epoch": 0.207912214842622,
+      "grad_norm": 0.14471082389354706,
+      "learning_rate": 9.196124273642422e-06,
+      "loss": 0.8374,
+      "step": 540
+    },
+    {
+      "epoch": 0.2117624410434113,
+      "grad_norm": 0.12267274409532547,
+      "learning_rate": 9.161889542552842e-06,
+      "loss": 0.8413,
+      "step": 550
+    },
+    {
+      "epoch": 0.21561266724420058,
+      "grad_norm": 0.12733341753482819,
+      "learning_rate": 9.127007477593193e-06,
+      "loss": 0.8433,
+      "step": 560
+    },
+    {
+      "epoch": 0.2194628934449899,
+      "grad_norm": 0.13992628455162048,
+      "learning_rate": 9.091483504265974e-06,
+      "loss": 0.8533,
+      "step": 570
+    },
+    {
+      "epoch": 0.22331311964577918,
+      "grad_norm": 0.13929937779903412,
+      "learning_rate": 9.055323147915116e-06,
+      "loss": 0.8268,
+      "step": 580
+    },
+    {
+      "epoch": 0.2271633458465685,
+      "grad_norm": 0.1369718313217163,
+      "learning_rate": 9.018532032866585e-06,
+      "loss": 0.8462,
+      "step": 590
+    },
+    {
+      "epoch": 0.23101357204735778,
+      "grad_norm": 0.13639406859874725,
+      "learning_rate": 8.981115881553579e-06,
+      "loss": 0.8408,
+      "step": 600
+    },
+    {
+      "epoch": 0.23486379824814707,
+      "grad_norm": 0.13501191139221191,
+      "learning_rate": 8.943080513626476e-06,
+      "loss": 0.8504,
+      "step": 610
+    },
+    {
+      "epoch": 0.23871402444893638,
+      "grad_norm": 0.13820721209049225,
+      "learning_rate": 8.904431845047641e-06,
+      "loss": 0.8419,
+      "step": 620
+    },
+    {
+      "epoch": 0.24256425064972567,
+      "grad_norm": 0.14026568830013275,
+      "learning_rate": 8.86517588717128e-06,
+      "loss": 0.8477,
+      "step": 630
+    },
+    {
+      "epoch": 0.24641447685051496,
+      "grad_norm": 0.1245928406715393,
+      "learning_rate": 8.82531874580844e-06,
+      "loss": 0.8411,
+      "step": 640
+    },
+    {
+      "epoch": 0.25026470305130427,
+      "grad_norm": 0.1194000393152237,
+      "learning_rate": 8.784866620277314e-06,
+      "loss": 0.7964,
+      "step": 650
+    },
+    {
+      "epoch": 0.25411492925209356,
+      "grad_norm": 0.14160138368606567,
+      "learning_rate": 8.743825802439018e-06,
+      "loss": 0.8277,
+      "step": 660
+    },
+    {
+      "epoch": 0.25796515545288284,
+      "grad_norm": 0.13412989675998688,
+      "learning_rate": 8.702202675718956e-06,
+      "loss": 0.8407,
+      "step": 670
+    },
+    {
+      "epoch": 0.26181538165367213,
+      "grad_norm": 0.12993736565113068,
+      "learning_rate": 8.660003714113959e-06,
+      "loss": 0.8166,
+      "step": 680
+    },
+    {
+      "epoch": 0.26566560785446147,
+      "grad_norm": 0.1305348128080368,
+      "learning_rate": 8.617235481185323e-06,
+      "loss": 0.8308,
+      "step": 690
+    },
+    {
+      "epoch": 0.26951583405525076,
+      "grad_norm": 0.1301368623971939,
+      "learning_rate": 8.573904629037931e-06,
+      "loss": 0.8353,
+      "step": 700
+    },
+    {
+      "epoch": 0.27336606025604004,
+      "grad_norm": 0.12388906627893448,
+      "learning_rate": 8.530017897285585e-06,
+      "loss": 0.8205,
+      "step": 710
+    },
+    {
+      "epoch": 0.27721628645682933,
+      "grad_norm": 0.11769033223390579,
+      "learning_rate": 8.485582112002746e-06,
+      "loss": 0.8229,
+      "step": 720
+    },
+    {
+      "epoch": 0.2810665126576186,
+      "grad_norm": 0.13332229852676392,
+      "learning_rate": 8.440604184662811e-06,
+      "loss": 0.8355,
+      "step": 730
+    },
+    {
+      "epoch": 0.28491673885840796,
+      "grad_norm": 0.13542205095291138,
+      "learning_rate": 8.395091111063118e-06,
+      "loss": 0.8163,
+      "step": 740
+    },
+    {
+      "epoch": 0.28876696505919724,
+      "grad_norm": 0.12036008387804031,
+      "learning_rate": 8.349049970236822e-06,
+      "loss": 0.8143,
+      "step": 750
+    },
+    {
+      "epoch": 0.29261719125998653,
+      "grad_norm": 0.12465699017047882,
+      "learning_rate": 8.302487923351842e-06,
+      "loss": 0.808,
+      "step": 760
+    },
+    {
+      "epoch": 0.2964674174607758,
+      "grad_norm": 0.14378078281879425,
+      "learning_rate": 8.255412212597021e-06,
+      "loss": 0.8261,
+      "step": 770
+    },
+    {
+      "epoch": 0.3003176436615651,
+      "grad_norm": 0.13303640484809875,
+      "learning_rate": 8.207830160055688e-06,
+      "loss": 0.8192,
+      "step": 780
+    },
+    {
+      "epoch": 0.3041678698623544,
+      "grad_norm": 0.1425410360097885,
+      "learning_rate": 8.159749166566786e-06,
+      "loss": 0.8083,
+      "step": 790
+    },
+    {
+      "epoch": 0.30801809606314373,
+      "grad_norm": 0.13025549054145813,
+      "learning_rate": 8.11117671057377e-06,
+      "loss": 0.8348,
+      "step": 800
+    },
+    {
+      "epoch": 0.311868322263933,
+      "grad_norm": 0.12669609487056732,
+      "learning_rate": 8.062120346961404e-06,
+      "loss": 0.8219,
+      "step": 810
+    },
+    {
+      "epoch": 0.3157185484647223,
+      "grad_norm": 0.12240136414766312,
+      "learning_rate": 8.012587705880705e-06,
+      "loss": 0.8235,
+      "step": 820
+    },
+    {
+      "epoch": 0.3195687746655116,
+      "grad_norm": 0.12372560799121857,
+      "learning_rate": 7.962586491562146e-06,
+      "loss": 0.8114,
+      "step": 830
+    },
+    {
+      "epoch": 0.3234190008663009,
+      "grad_norm": 0.13104791939258575,
+      "learning_rate": 7.912124481117357e-06,
+      "loss": 0.8153,
+      "step": 840
+    },
+    {
+      "epoch": 0.3272692270670902,
+      "grad_norm": 0.12423375993967056,
+      "learning_rate": 7.861209523329494e-06,
+      "loss": 0.8145,
+      "step": 850
+    },
+    {
+      "epoch": 0.3311194532678795,
+      "grad_norm": 0.12054426968097687,
+      "learning_rate": 7.809849537432432e-06,
+      "loss": 0.809,
+      "step": 860
+    },
+    {
+      "epoch": 0.3349696794686688,
+      "grad_norm": 0.12201160937547684,
+      "learning_rate": 7.758052511879035e-06,
+      "loss": 0.8071,
+      "step": 870
+    },
+    {
+      "epoch": 0.3388199056694581,
+      "grad_norm": 0.12048336863517761,
+      "learning_rate": 7.705826503098642e-06,
+      "loss": 0.813,
+      "step": 880
+    },
+    {
+      "epoch": 0.34267013187024736,
+      "grad_norm": 0.12296102195978165,
+      "learning_rate": 7.653179634243979e-06,
+      "loss": 0.8034,
+      "step": 890
+    },
+    {
+      "epoch": 0.34652035807103665,
+      "grad_norm": 0.12015175074338913,
+      "learning_rate": 7.600120093927696e-06,
+      "loss": 0.8101,
+      "step": 900
+    },
+    {
+      "epoch": 0.350370584271826,
+      "grad_norm": 0.1260998249053955,
+      "learning_rate": 7.546656134948729e-06,
+      "loss": 0.8109,
+      "step": 910
+    },
+    {
+      "epoch": 0.3542208104726153,
+      "grad_norm": 0.12671120464801788,
+      "learning_rate": 7.492796073008661e-06,
+      "loss": 0.7965,
+      "step": 920
+    },
+    {
+      "epoch": 0.35807103667340456,
+      "grad_norm": 0.13276000320911407,
+      "learning_rate": 7.438548285418326e-06,
+      "loss": 0.7907,
+      "step": 930
+    },
+    {
+      "epoch": 0.36192126287419385,
+      "grad_norm": 0.13963459432125092,
+      "learning_rate": 7.3839212097948e-06,
+      "loss": 0.8275,
+      "step": 940
+    },
+    {
+      "epoch": 0.36577148907498314,
+      "grad_norm": 0.12105058878660202,
+      "learning_rate": 7.328923342749039e-06,
+      "loss": 0.8386,
+      "step": 950
+    },
+    {
+      "epoch": 0.3696217152757725,
+      "grad_norm": 0.158193439245224,
+      "learning_rate": 7.273563238564319e-06,
+      "loss": 0.8163,
+      "step": 960
+    },
+    {
+      "epoch": 0.37347194147656176,
+      "grad_norm": 0.12830014526844025,
+      "learning_rate": 7.217849507865724e-06,
+      "loss": 0.7972,
+      "step": 970
+    },
+    {
+      "epoch": 0.37732216767735105,
+      "grad_norm": 0.12109928578138351,
+      "learning_rate": 7.161790816280858e-06,
+      "loss": 0.8055,
+      "step": 980
+    },
+    {
+      "epoch": 0.38117239387814034,
+      "grad_norm": 0.1319950520992279,
+      "learning_rate": 7.1053958830920125e-06,
+      "loss": 0.7951,
+      "step": 990
+    },
+    {
+      "epoch": 0.3850226200789296,
+      "grad_norm": 0.12419692426919937,
+      "learning_rate": 7.0486734798799715e-06,
+      "loss": 0.8164,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3888728462797189,
+      "grad_norm": 0.12821152806282043,
+      "learning_rate": 6.991632429159708e-06,
+      "loss": 0.8358,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39272307248050825,
+      "grad_norm": 0.13065829873085022,
+      "learning_rate": 6.934281603008132e-06,
+      "loss": 0.8139,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39657329868129754,
+      "grad_norm": 0.12534597516059875,
+      "learning_rate": 6.876629921684143e-06,
+      "loss": 0.8253,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4004235248820868,
+      "grad_norm": 0.12778089940547943,
+      "learning_rate": 6.818686352241199e-06,
+      "loss": 0.8234,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4042737510828761,
+      "grad_norm": 0.12349394708871841,
+      "learning_rate": 6.760459907132573e-06,
+      "loss": 0.8149,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4081239772836654,
+      "grad_norm": 0.13253311812877655,
+      "learning_rate": 6.701959642809589e-06,
+      "loss": 0.8198,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41197420348445474,
+      "grad_norm": 0.12581288814544678,
+      "learning_rate": 6.643194658312987e-06,
+      "loss": 0.8015,
+      "step": 1070
+    },
+    {
+      "epoch": 0.415824429685244,
+      "grad_norm": 0.12181761860847473,
+      "learning_rate": 6.584174093857676e-06,
+      "loss": 0.834,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4196746558860333,
+      "grad_norm": 0.14505836367607117,
+      "learning_rate": 6.5249071294110735e-06,
+      "loss": 0.8088,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4235248820868226,
+      "grad_norm": 0.13787059485912323,
+      "learning_rate": 6.465402983265274e-06,
+      "loss": 0.8299,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4273751082876119,
+      "grad_norm": 0.1388739049434662,
+      "learning_rate": 6.405670910603243e-06,
+      "loss": 0.8229,
+      "step": 1110
+    },
+    {
+      "epoch": 0.43122533448840117,
+      "grad_norm": 0.13729701936244965,
+      "learning_rate": 6.345720202059291e-06,
+      "loss": 0.8177,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4350755606891905,
+      "grad_norm": 0.12985707819461823,
+      "learning_rate": 6.285560182274008e-06,
+      "loss": 0.8073,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4389257868899798,
+      "grad_norm": 0.12027091532945633,
+      "learning_rate": 6.225200208443936e-06,
+      "loss": 0.8113,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4427760130907691,
+      "grad_norm": 0.11886423081159592,
+      "learning_rate": 6.16464966886616e-06,
+      "loss": 0.8078,
+      "step": 1150
+    },
+    {
+      "epoch": 0.44662623929155837,
+      "grad_norm": 0.14282074570655823,
+      "learning_rate": 6.103917981478065e-06,
+      "loss": 0.8331,
+      "step": 1160
+    },
+    {
+      "epoch": 0.45047646549234766,
+      "grad_norm": 0.1263916939496994,
+      "learning_rate": 6.043014592392489e-06,
+      "loss": 0.8418,
+      "step": 1170
+    },
+    {
+      "epoch": 0.454326691693137,
+      "grad_norm": 0.11453361064195633,
+      "learning_rate": 5.981948974428485e-06,
+      "loss": 0.8133,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4581769178939263,
+      "grad_norm": 0.12965433299541473,
+      "learning_rate": 5.920730625637934e-06,
+      "loss": 0.7963,
+      "step": 1190
+    },
+    {
+      "epoch": 0.46202714409471557,
+      "grad_norm": 0.13959713280200958,
+      "learning_rate": 5.8593690678282355e-06,
+      "loss": 0.8029,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46587737029550486,
+      "grad_norm": 0.13146570324897766,
+      "learning_rate": 5.797873845081301e-06,
+      "loss": 0.8078,
+      "step": 1210
+    },
+    {
+      "epoch": 0.46972759649629414,
+      "grad_norm": 0.13008081912994385,
+      "learning_rate": 5.736254522269075e-06,
+      "loss": 0.8226,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47357782269708343,
+      "grad_norm": 0.11892782151699066,
+      "learning_rate": 5.6745206835658386e-06,
+      "loss": 0.8243,
+      "step": 1230
+    },
+    {
+      "epoch": 0.47742804889787277,
+      "grad_norm": 0.12362486124038696,
+      "learning_rate": 5.612681930957495e-06,
+      "loss": 0.8021,
+      "step": 1240
+    },
+    {
+      "epoch": 0.48127827509866206,
+      "grad_norm": 0.12233374267816544,
+      "learning_rate": 5.550747882748095e-06,
+      "loss": 0.8023,
+      "step": 1250
+    },
+    {
+      "epoch": 0.48512850129945134,
+      "grad_norm": 0.13705424964427948,
+      "learning_rate": 5.4887281720638195e-06,
+      "loss": 0.8005,
+      "step": 1260
+    },
+    {
+      "epoch": 0.48897872750024063,
+      "grad_norm": 0.14320459961891174,
+      "learning_rate": 5.42663244535466e-06,
+      "loss": 0.7928,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4928289537010299,
+      "grad_norm": 0.13606920838356018,
+      "learning_rate": 5.3644703608940205e-06,
+      "loss": 0.7969,
+      "step": 1280
+    },
+    {
+      "epoch": 0.49667917990181926,
+      "grad_norm": 0.12745173275470734,
+      "learning_rate": 5.302251587276491e-06,
+      "loss": 0.8193,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5005294061026085,
+      "grad_norm": 0.1359611451625824,
+      "learning_rate": 5.2399858019140005e-06,
+      "loss": 0.8106,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5043796323033978,
+      "grad_norm": 0.1348227560520172,
+      "learning_rate": 5.177682689530615e-06,
+      "loss": 0.8256,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5082298585041871,
+      "grad_norm": 0.1442672461271286,
+      "learning_rate": 5.11535194065619e-06,
+      "loss": 0.8031,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5120800847049765,
+      "grad_norm": 0.13860802352428436,
+      "learning_rate": 5.053003250119113e-06,
+      "loss": 0.8074,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5159303109057657,
+      "grad_norm": 0.1330154538154602,
+      "learning_rate": 4.9906463155383925e-06,
+      "loss": 0.8098,
+      "step": 1340
+    },
+    {
+      "epoch": 0.519780537106555,
+      "grad_norm": 0.1281103640794754,
+      "learning_rate": 4.9282908358153046e-06,
+      "loss": 0.7993,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5236307633073443,
+      "grad_norm": 0.13910508155822754,
+      "learning_rate": 4.8659465096248345e-06,
+      "loss": 0.8143,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5274809895081336,
+      "grad_norm": 0.13757608830928802,
+      "learning_rate": 4.803623033907169e-06,
+      "loss": 0.8221,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5313312157089229,
+      "grad_norm": 0.12154904752969742,
+      "learning_rate": 4.741330102359441e-06,
+      "loss": 0.824,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5351814419097122,
+      "grad_norm": 0.13247612118721008,
+      "learning_rate": 4.679077403927996e-06,
+      "loss": 0.801,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5390316681105015,
+      "grad_norm": 0.1277068555355072,
+      "learning_rate": 4.616874621301373e-06,
+      "loss": 0.805,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5428818943112907,
+      "grad_norm": 0.13025794923305511,
+      "learning_rate": 4.554731429404293e-06,
+      "loss": 0.8226,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5467321205120801,
+      "grad_norm": 0.1338622272014618,
+      "learning_rate": 4.492657493892822e-06,
+      "loss": 0.8082,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5505823467128694,
+      "grad_norm": 0.14205579459667206,
+      "learning_rate": 4.430662469650999e-06,
+      "loss": 0.8179,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5544325729136587,
+      "grad_norm": 0.13429605960845947,
+      "learning_rate": 4.368755999289118e-06,
+      "loss": 0.82,
+      "step": 1440
+    },
+    {
+      "epoch": 0.558282799114448,
+      "grad_norm": 0.14617061614990234,
+      "learning_rate": 4.306947711643949e-06,
+      "loss": 0.8044,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5621330253152372,
+      "grad_norm": 0.12459474056959152,
+      "learning_rate": 4.245247220281068e-06,
+      "loss": 0.8075,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5659832515160266,
+      "grad_norm": 0.12673023343086243,
+      "learning_rate": 4.183664121999587e-06,
+      "loss": 0.8058,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5698334777168159,
+      "grad_norm": 0.1362849473953247,
+      "learning_rate": 4.122207995339474e-06,
+      "loss": 0.8036,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5736837039176051,
+      "grad_norm": 0.1305910348892212,
+      "learning_rate": 4.060888399091734e-06,
+      "loss": 0.8105,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5775339301183945,
+      "grad_norm": 0.12834496796131134,
+      "learning_rate": 3.999714870811639e-06,
+      "loss": 0.8159,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5813841563191837,
+      "grad_norm": 0.12067916989326477,
+      "learning_rate": 3.938696925335275e-06,
+      "loss": 0.8075,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5852343825199731,
+      "grad_norm": 0.1384790986776352,
+      "learning_rate": 3.87784405329962e-06,
+      "loss": 0.816,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5890846087207623,
+      "grad_norm": 0.13708056509494781,
+      "learning_rate": 3.817165719666382e-06,
+      "loss": 0.8159,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5929348349215516,
+      "grad_norm": 0.12651465833187103,
+      "learning_rate": 3.756671362249841e-06,
+      "loss": 0.8223,
+      "step": 1540
+    },
+    {
+      "epoch": 0.596785061122341,
+      "grad_norm": 0.13078460097312927,
+      "learning_rate": 3.6963703902488907e-06,
+      "loss": 0.8158,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6006352873231302,
+      "grad_norm": 0.13632111251354218,
+      "learning_rate": 3.6362721827835655e-06,
+      "loss": 0.8315,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6044855135239195,
+      "grad_norm": 0.1410801112651825,
+      "learning_rate": 3.576386087436212e-06,
+      "loss": 0.8157,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6083357397247088,
+      "grad_norm": 0.1325787454843521,
+      "learning_rate": 3.5167214187975913e-06,
+      "loss": 0.816,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6121859659254981,
+      "grad_norm": 0.10961833596229553,
+      "learning_rate": 3.457287457018086e-06,
+      "loss": 0.8158,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6160361921262875,
+      "grad_norm": 0.12775027751922607,
+      "learning_rate": 3.3980934463643003e-06,
+      "loss": 0.7981,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6198864183270767,
+      "grad_norm": 0.1143123209476471,
+      "learning_rate": 3.3391485937812053e-06,
+      "loss": 0.7813,
+      "step": 1610
+    },
+    {
+      "epoch": 0.623736644527866,
+      "grad_norm": 0.13100531697273254,
+      "learning_rate": 3.280462067460115e-06,
+      "loss": 0.7847,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6275868707286553,
+      "grad_norm": 0.1264168918132782,
+      "learning_rate": 3.222042995412669e-06,
+      "loss": 0.8189,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6314370969294446,
+      "grad_norm": 0.11751738935709,
+      "learning_rate": 3.163900464051089e-06,
+      "loss": 0.8019,
+      "step": 1640
+    },
+    {
+      "epoch": 0.635287323130234,
+      "grad_norm": 0.12937228381633759,
+      "learning_rate": 3.1060435167748826e-06,
+      "loss": 0.8116,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6391375493310232,
+      "grad_norm": 0.1291312575340271,
+      "learning_rate": 3.048481152564253e-06,
+      "loss": 0.796,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6429877755318125,
+      "grad_norm": 0.13809122145175934,
+      "learning_rate": 2.9912223245804017e-06,
+      "loss": 0.8189,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6468380017326018,
+      "grad_norm": 0.128720223903656,
+      "learning_rate": 2.934275938772981e-06,
+      "loss": 0.8085,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6506882279333911,
+      "grad_norm": 0.13646145164966583,
+      "learning_rate": 2.8776508524948676e-06,
+      "loss": 0.7963,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6545384541341804,
+      "grad_norm": 0.11988291889429092,
+      "learning_rate": 2.821355873124503e-06,
+      "loss": 0.7926,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6583886803349697,
+      "grad_norm": 0.14029701054096222,
+      "learning_rate": 2.7653997566960144e-06,
+      "loss": 0.8125,
+      "step": 1710
+    },
+    {
+      "epoch": 0.662238906535759,
+      "grad_norm": 0.1317368894815445,
+      "learning_rate": 2.709791206537309e-06,
+      "loss": 0.8026,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6660891327365482,
+      "grad_norm": 0.119501993060112,
+      "learning_rate": 2.6545388719163747e-06,
+      "loss": 0.8101,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6699393589373376,
+      "grad_norm": 0.12434754520654678,
+      "learning_rate": 2.599651346695979e-06,
+      "loss": 0.8126,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6737895851381268,
+      "grad_norm": 0.14223319292068481,
+      "learning_rate": 2.545137167997007e-06,
+      "loss": 0.8061,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6776398113389162,
+      "grad_norm": 0.11763649433851242,
+      "learning_rate": 2.4910048148705972e-06,
+      "loss": 0.7941,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6814900375397055,
+      "grad_norm": 0.13143852353096008,
+      "learning_rate": 2.4372627069793327e-06,
+      "loss": 0.8319,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6853402637404947,
+      "grad_norm": 0.12858116626739502,
+      "learning_rate": 2.3839192032876486e-06,
+      "loss": 0.8178,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6891904899412841,
+      "grad_norm": 0.11904086917638779,
+      "learning_rate": 2.3309826007617047e-06,
+      "loss": 0.8297,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6930407161420733,
+      "grad_norm": 0.11965224891901016,
+      "learning_rate": 2.278461133078881e-06,
+      "loss": 0.7952,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6968909423428626,
+      "grad_norm": 0.1255815327167511,
+      "learning_rate": 2.2263629693471258e-06,
+      "loss": 0.7899,
+      "step": 1810
+    },
+    {
+      "epoch": 0.700741168543652,
+      "grad_norm": 0.12755729258060455,
+      "learning_rate": 2.1746962128343414e-06,
+      "loss": 0.8157,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7045913947444412,
+      "grad_norm": 0.12438524514436722,
+      "learning_rate": 2.1234688997080183e-06,
+      "loss": 0.8095,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7084416209452306,
+      "grad_norm": 0.13609172403812408,
+      "learning_rate": 2.0726889977853053e-06,
+      "loss": 0.8079,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7122918471460198,
+      "grad_norm": 0.12762044370174408,
+      "learning_rate": 2.022364405293703e-06,
+      "loss": 0.8164,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7161420733468091,
+      "grad_norm": 0.12305556237697601,
+      "learning_rate": 1.9725029496425814e-06,
+      "loss": 0.8129,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7199922995475985,
+      "grad_norm": 0.13082174956798553,
+      "learning_rate": 1.923112386205727e-06,
+      "loss": 0.8099,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7238425257483877,
+      "grad_norm": 0.14348790049552917,
+      "learning_rate": 1.874200397115079e-06,
+      "loss": 0.8128,
+      "step": 1880
+    },
+    {
+      "epoch": 0.727692751949177,
+      "grad_norm": 0.13298940658569336,
+      "learning_rate": 1.825774590065857e-06,
+      "loss": 0.801,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7315429781499663,
+      "grad_norm": 0.13528043031692505,
+      "learning_rate": 1.7778424971332813e-06,
+      "loss": 0.8089,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7353932043507556,
+      "grad_norm": 0.12048614770174026,
+      "learning_rate": 1.7304115736010407e-06,
+      "loss": 0.8103,
+      "step": 1910
+    },
+    {
+      "epoch": 0.739243430551545,
+      "grad_norm": 0.12090927362442017,
+      "learning_rate": 1.6834891968017108e-06,
+      "loss": 0.8,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7430936567523342,
+      "grad_norm": 0.13037662208080292,
+      "learning_rate": 1.6370826649692884e-06,
+      "loss": 0.819,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7469438829531235,
+      "grad_norm": 0.12699422240257263,
+      "learning_rate": 1.5911991961040463e-06,
+      "loss": 0.8264,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7507941091539128,
+      "grad_norm": 0.13237451016902924,
+      "learning_rate": 1.545845926849851e-06,
+      "loss": 0.8123,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7546443353547021,
+      "grad_norm": 0.13068468868732452,
+      "learning_rate": 1.5010299113841397e-06,
+      "loss": 0.8286,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7584945615554913,
+      "grad_norm": 0.12613104283809662,
+      "learning_rate": 1.4567581203207238e-06,
+      "loss": 0.821,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7623447877562807,
+      "grad_norm": 0.13299359381198883,
+      "learning_rate": 1.413037439625593e-06,
+      "loss": 0.8086,
+      "step": 1980
+    },
+    {
+      "epoch": 0.76619501395707,
+      "grad_norm": 0.14432688057422638,
+      "learning_rate": 1.3698746695458847e-06,
+      "loss": 0.8091,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7700452401578592,
+      "grad_norm": 0.12141236662864685,
+      "learning_rate": 1.3272765235521745e-06,
+      "loss": 0.8087,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7738954663586486,
+      "grad_norm": 0.1206887736916542,
+      "learning_rate": 1.2852496272942877e-06,
+      "loss": 0.8062,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7777456925594378,
+      "grad_norm": 0.12112224102020264,
+      "learning_rate": 1.2438005175707396e-06,
+      "loss": 0.7951,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7815959187602272,
+      "grad_norm": 0.1315736174583435,
+      "learning_rate": 1.2029356413120297e-06,
+      "loss": 0.7846,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7854461449610165,
+      "grad_norm": 0.13082782924175262,
+      "learning_rate": 1.1626613545778786e-06,
+      "loss": 0.8032,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7892963711618057,
+      "grad_norm": 0.1201329380273819,
+      "learning_rate": 1.1229839215686294e-06,
+      "loss": 0.8061,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7931465973625951,
+      "grad_norm": 0.11544676870107651,
+      "learning_rate": 1.0839095136509188e-06,
+      "loss": 0.7926,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7969968235633843,
+      "grad_norm": 0.13575564324855804,
+      "learning_rate": 1.045444208397791e-06,
+      "loss": 0.806,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8008470497641736,
+      "grad_norm": 0.12412498891353607,
+      "learning_rate": 1.007593988643402e-06,
+      "loss": 0.7872,
+      "step": 2080
+    },
+    {
+      "epoch": 0.804697275964963,
+      "grad_norm": 0.12508469820022583,
+      "learning_rate": 9.703647415524613e-07,
+      "loss": 0.8118,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8085475021657522,
+      "grad_norm": 0.1277608573436737,
+      "learning_rate": 9.337622577045485e-07,
+      "loss": 0.8059,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8123977283665416,
+      "grad_norm": 0.1231621727347374,
+      "learning_rate": 8.977922301934588e-07,
+      "loss": 0.8143,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8162479545673308,
+      "grad_norm": 0.1304914653301239,
+      "learning_rate": 8.624602537417009e-07,
+      "loss": 0.8137,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8200981807681201,
+      "grad_norm": 0.12645164132118225,
+      "learning_rate": 8.277718238303123e-07,
+      "loss": 0.8012,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8239484069689095,
+      "grad_norm": 0.12678617238998413,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.8124,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8277986331696987,
+      "grad_norm": 0.14158856868743896,
+      "learning_rate": 7.603470842324223e-07,
+      "loss": 0.7985,
+      "step": 2150
+    },
+    {
+      "epoch": 0.831648859370488,
+      "grad_norm": 0.12068299204111099,
+      "learning_rate": 7.276212616857575e-07,
+      "loss": 0.8127,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8354990855712773,
+      "grad_norm": 0.12050720304250717,
+      "learning_rate": 6.955599583279837e-07,
+      "loss": 0.8177,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8393493117720666,
+      "grad_norm": 0.1280621439218521,
+      "learning_rate": 6.641681609246981e-07,
+      "loss": 0.8039,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8431995379728559,
+      "grad_norm": 0.13108575344085693,
+      "learning_rate": 6.334507521075711e-07,
+      "loss": 0.8174,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8470497641736452,
+      "grad_norm": 0.13912294805049896,
+      "learning_rate": 6.034125096149179e-07,
+      "loss": 0.7981,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8508999903744345,
+      "grad_norm": 0.13245178759098053,
+      "learning_rate": 5.740581055485722e-07,
+      "loss": 0.7865,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8547502165752238,
+      "grad_norm": 0.11891192942857742,
+      "learning_rate": 5.453921056471961e-07,
+      "loss": 0.8108,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8586004427760131,
+      "grad_norm": 0.13235950469970703,
+      "learning_rate": 5.174189685761288e-07,
+      "loss": 0.8002,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8624506689768023,
+      "grad_norm": 0.1325034350156784,
+      "learning_rate": 4.901430452338979e-07,
+      "loss": 0.8186,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8663008951775917,
+      "grad_norm": 0.12349297106266022,
+      "learning_rate": 4.6356857807548286e-07,
+      "loss": 0.8031,
+      "step": 2250
+    },
+    {
+      "epoch": 0.870151121378381,
+      "grad_norm": 0.1357106864452362,
+      "learning_rate": 4.376997004524541e-07,
+      "loss": 0.7927,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8740013475791703,
+      "grad_norm": 0.11891859769821167,
+      "learning_rate": 4.12540435970073e-07,
+      "loss": 0.7993,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8778515737799596,
+      "grad_norm": 0.11726263165473938,
+      "learning_rate": 3.880946978614703e-07,
+      "loss": 0.81,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8817017999807488,
+      "grad_norm": 0.12104993313550949,
+      "learning_rate": 3.643662883789878e-07,
+      "loss": 0.8186,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8855520261815382,
+      "grad_norm": 0.12389664351940155,
+      "learning_rate": 3.41358898202781e-07,
+      "loss": 0.7976,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8894022523823275,
+      "grad_norm": 0.11760278046131134,
+      "learning_rate": 3.190761058667741e-07,
+      "loss": 0.7994,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8932524785831167,
+      "grad_norm": 0.1141587346792221,
+      "learning_rate": 2.9752137720206463e-07,
+      "loss": 0.8098,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8971027047839061,
+      "grad_norm": 0.12915396690368652,
+      "learning_rate": 2.7669806479785067e-07,
+      "loss": 0.8194,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9009529309846953,
+      "grad_norm": 0.12809284031391144,
+      "learning_rate": 2.566094074799763e-07,
+      "loss": 0.8115,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9048031571854847,
+      "grad_norm": 0.12270138412714005,
+      "learning_rate": 2.3725852980716557e-07,
+      "loss": 0.7892,
+      "step": 2350
+    },
+    {
+      "epoch": 0.908653383386274,
+      "grad_norm": 0.12263435125350952,
+      "learning_rate": 2.1864844158503983e-07,
+      "loss": 0.8102,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9125036095870632,
+      "grad_norm": 0.13655927777290344,
+      "learning_rate": 2.007820373979702e-07,
+      "loss": 0.7934,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9163538357878526,
+      "grad_norm": 0.12616360187530518,
+      "learning_rate": 1.8366209615886532e-07,
+      "loss": 0.7996,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9202040619886418,
+      "grad_norm": 0.12057797610759735,
+      "learning_rate": 1.6729128067693613e-07,
+      "loss": 0.8162,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9240542881894311,
+      "grad_norm": 0.1290263533592224,
+      "learning_rate": 1.5167213724353426e-07,
+      "loss": 0.8067,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9279045143902205,
+      "grad_norm": 0.12634064257144928,
+      "learning_rate": 1.3680709523610058e-07,
+      "loss": 0.7969,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9317547405910097,
+      "grad_norm": 0.11923575401306152,
+      "learning_rate": 1.2269846674030418e-07,
+      "loss": 0.8211,
+      "step": 2420
+    },
+    {
+      "epoch": 0.935604966791799,
+      "grad_norm": 0.1229824349284172,
+      "learning_rate": 1.0934844619042595e-07,
+      "loss": 0.8176,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9394551929925883,
+      "grad_norm": 0.14441080391407013,
+      "learning_rate": 9.675911002803961e-08,
+      "loss": 0.8055,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9433054191933776,
+      "grad_norm": 0.1331772804260254,
+      "learning_rate": 8.493241637904381e-08,
+      "loss": 0.8053,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9471556453941669,
+      "grad_norm": 0.1331002414226532,
+      "learning_rate": 7.387020474909689e-08,
+      "loss": 0.8224,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9510058715949562,
+      "grad_norm": 0.12446079403162003,
+      "learning_rate": 6.357419573750856e-08,
+      "loss": 0.8085,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9548560977957455,
+      "grad_norm": 0.12798036634922028,
+      "learning_rate": 5.4045990769613365e-08,
+      "loss": 0.804,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9587063239965348,
+      "grad_norm": 0.14547736942768097,
+      "learning_rate": 4.5287071847694434e-08,
+      "loss": 0.8067,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9625565501973241,
+      "grad_norm": 0.12950223684310913,
+      "learning_rate": 3.7298801320467817e-08,
+      "loss": 0.7975,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9664067763981133,
+      "grad_norm": 0.12480832636356354,
+      "learning_rate": 3.0082421671192576e-08,
+      "loss": 0.7956,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9702570025989027,
+      "grad_norm": 0.1211831271648407,
+      "learning_rate": 2.3639055324410355e-08,
+      "loss": 0.809,
+      "step": 2520
+    },
+    {
+      "epoch": 0.974107228799692,
+      "grad_norm": 0.12821049988269806,
+      "learning_rate": 1.7969704471369477e-08,
+      "loss": 0.7993,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9779574550004813,
+      "grad_norm": 0.117767833173275,
+      "learning_rate": 1.3075250914142413e-08,
+      "loss": 0.8041,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9818076812012706,
+      "grad_norm": 0.11851906031370163,
+      "learning_rate": 8.956455928476049e-09,
+      "loss": 0.8141,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9856579074020598,
+      "grad_norm": 0.11818092316389084,
+      "learning_rate": 5.6139601453786275e-09,
+      "loss": 0.7944,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9895081336028492,
+      "grad_norm": 0.12632419168949127,
+      "learning_rate": 3.0482834514805602e-09,
+      "loss": 0.7973,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9933583598036385,
+      "grad_norm": 0.11071363836526871,
+      "learning_rate": 1.2598249081713366e-09,
+      "loss": 0.7873,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9972085860044277,
+      "grad_norm": 0.13298992812633514,
+      "learning_rate": 2.488626895302826e-10,
+      "loss": 0.8111,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9999037443449803,
+      "step": 2597,
+      "total_flos": 2.995607458728116e+18,
+      "train_loss": 0.8398951575257569,
+      "train_runtime": 9394.6135,
+      "train_samples_per_second": 70.775,
+      "train_steps_per_second": 0.276
+    }
+  ],
+  "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": 2.995607458728116e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3/mc1_b32_224_400m/finetune/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b437f1bf8e03fe1192fe4c9b83c2530826c74101ac5d06bb6429e3f255c86cd
+size 6136

continuous/qwen3/mc1_b32_224_400m/pretrain/config.json

@@ -0,0 +1,60 @@
+{
+  "_attn_implementation_autoset": true,
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 6144,
+  "max_position_embeddings": 40960,
+  "max_window_layers": 28,
+  "mm_hidden_size": 768,
+  "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-B-32",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "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/mc1/mc1_b32_224_400m.pt",
+  "vocab_size": 151936
+}

continuous/qwen3/mc1_b32_224_400m/pretrain/mm_projector.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7e4f739992643fb6124f5d0fafa20ebbb627dfcb0e5a066e3c625a80ee8417d
+size 23087160

continuous/qwen3/mc1_b32_224_400m/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": 11.734721183776855,
+      "learning_rate": 0.00032142857142857147,
+      "loss": 10.9071,
+      "step": 10
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 6.910419464111328,
+      "learning_rate": 0.0006785714285714287,
+      "loss": 7.8095,
+      "step": 20
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 1.156296968460083,
+      "learning_rate": 0.0009999907688176173,
+      "loss": 5.9326,
+      "step": 30
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 0.5582736134529114,
+      "learning_rate": 0.0009988834393115766,
+      "loss": 5.2173,
+      "step": 40
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 0.11524531245231628,
+      "learning_rate": 0.0009959345572940772,
+      "loss": 4.9271,
+      "step": 50
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 0.07368436455726624,
+      "learning_rate": 0.0009911550081155983,
+      "loss": 4.8259,
+      "step": 60
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.0660000666975975,
+      "learning_rate": 0.0009845624347566433,
+      "loss": 4.7681,
+      "step": 70
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 0.12348468601703644,
+      "learning_rate": 0.0009761811727013547,
+      "loss": 4.7349,
+      "step": 80
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 0.0749615803360939,
+      "learning_rate": 0.0009660421601067666,
+      "loss": 4.7015,
+      "step": 90
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 0.06812044233083725,
+      "learning_rate": 0.0009541828235992926,
+      "loss": 4.6815,
+      "step": 100
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 0.13077042996883392,
+      "learning_rate": 0.0009406469401200151,
+      "loss": 4.6514,
+      "step": 110
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 0.12117184698581696,
+      "learning_rate": 0.0009254844753287492,
+      "loss": 4.6402,
+      "step": 120
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 0.15993456542491913,
+      "learning_rate": 0.0009087513991633924,
+      "loss": 4.6119,
+      "step": 130
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 0.19460618495941162,
+      "learning_rate": 0.0008905094792353917,
+      "loss": 4.5902,
+      "step": 140
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 0.1418684422969818,
+      "learning_rate": 0.0008708260528239789,
+      "loss": 4.5806,
+      "step": 150
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.15837879478931427,
+      "learning_rate": 0.0008497737783108237,
+      "loss": 4.561,
+      "step": 160
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 0.1536322385072708,
+      "learning_rate": 0.0008274303669726426,
+      "loss": 4.5431,
+      "step": 170
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 0.10780579596757889,
+      "learning_rate": 0.0008038782961218137,
+      "loss": 4.5254,
+      "step": 180
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.19797390699386597,
+      "learning_rate": 0.000779204504653894,
+      "loss": 4.5203,
+      "step": 190
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 0.11809113621711731,
+      "learning_rate": 0.0007535000721258791,
+      "loss": 4.4945,
+      "step": 200
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 0.1315360814332962,
+      "learning_rate": 0.00072685988254984,
+      "loss": 4.484,
+      "step": 210
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 0.1793542504310608,
+      "learning_rate": 0.0006993822741429906,
+      "loss": 4.4949,
+      "step": 220
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.1461142748594284,
+      "learning_rate": 0.0006711686763270818,
+      "loss": 4.4896,
+      "step": 230
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 0.19155800342559814,
+      "learning_rate": 0.0006423232353170831,
+      "loss": 4.4696,
+      "step": 240
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 0.13212192058563232,
+      "learning_rate": 0.0006129524296812335,
+      "loss": 4.4489,
+      "step": 250
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 0.14226926863193512,
+      "learning_rate": 0.0005831646772915651,
+      "loss": 4.4585,
+      "step": 260
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.15858270227909088,
+      "learning_rate": 0.0005530699351157809,
+      "loss": 4.4379,
+      "step": 270
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 0.13316291570663452,
+      "learning_rate": 0.0005227792933277943,
+      "loss": 4.4332,
+      "step": 280
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 0.11576040834188461,
+      "learning_rate": 0.0004924045652352092,
+      "loss": 4.4182,
+      "step": 290
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 0.13734586536884308,
+      "learning_rate": 0.0004620578745374633,
+      "loss": 4.4151,
+      "step": 300
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 0.15020610392093658,
+      "learning_rate": 0.0004318512414382059,
+      "loss": 4.4297,
+      "step": 310
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 0.11820720881223679,
+      "learning_rate": 0.0004018961691397148,
+      "loss": 4.4187,
+      "step": 320
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 0.105117067694664,
+      "learning_rate": 0.0003723032322457458,
+      "loss": 4.4006,
+      "step": 330
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.11886614561080933,
+      "learning_rate": 0.0003431816685921625,
+      "loss": 4.4015,
+      "step": 340
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 0.11372286826372147,
+      "learning_rate": 0.0003146389760120416,
+      "loss": 4.407,
+      "step": 350
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 0.11862097680568695,
+      "learning_rate": 0.00028678051552373485,
+      "loss": 4.4125,
+      "step": 360
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 0.11102572828531265,
+      "learning_rate": 0.0002597091224066581,
+      "loss": 4.3924,
+      "step": 370
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 0.1277744174003601,
+      "learning_rate": 0.0002335247266004629,
+      "loss": 4.4104,
+      "step": 380
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 0.13248063623905182,
+      "learning_rate": 0.000208323983828828,
+      "loss": 4.3922,
+      "step": 390
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.16562800109386444,
+      "learning_rate": 0.0001841999188095224,
+      "loss": 4.3954,
+      "step": 400
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 0.1116979718208313,
+      "learning_rate": 0.00016124158186777676,
+      "loss": 4.3946,
+      "step": 410
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 0.1154952272772789,
+      "learning_rate": 0.0001395337202205211,
+      "loss": 4.3846,
+      "step": 420
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 0.10184711217880249,
+      "learning_rate": 0.00011915646514489292,
+      "loss": 4.3886,
+      "step": 430
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 0.11531613022089005,
+      "learning_rate": 0.00010018503618578818,
+      "loss": 4.3948,
+      "step": 440
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.10523933917284012,
+      "learning_rate": 8.268946349432583e-05,
+      "loss": 4.3883,
+      "step": 450
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 0.11972163617610931,
+      "learning_rate": 6.673432932217117e-05,
+      "loss": 4.3805,
+      "step": 460
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.10772611200809479,
+      "learning_rate": 5.2378529625954684e-05,
+      "loss": 4.3667,
+      "step": 470
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.09398901462554932,
+      "learning_rate": 3.9675056661785556e-05,
+      "loss": 4.3773,
+      "step": 480
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.1000845804810524,
+      "learning_rate": 2.8670803372379538e-05,
+      "loss": 4.3807,
+      "step": 490
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 0.09975207597017288,
+      "learning_rate": 1.9406390288876586e-05,
+      "loss": 4.3912,
+      "step": 500
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 0.09108490496873856,
+      "learning_rate": 1.191601558631461e-05,
+      "loss": 4.3786,
+      "step": 510
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 0.0881456658244133,
+      "learning_rate": 6.2273288462559305e-06,
+      "loss": 4.3717,
+      "step": 520
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 0.09469301253557205,
+      "learning_rate": 2.361328992552314e-06,
+      "loss": 4.375,
+      "step": 530
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 0.09070117771625519,
+      "learning_rate": 3.322867770044202e-07,
+      "loss": 4.3824,
+      "step": 540
+    },
+    {
+      "epoch": 1.0,
+      "step": 545,
+      "total_flos": 2.988888507453276e+17,
+      "train_loss": 4.704703752710185,
+      "train_runtime": 5505.8332,
+      "train_samples_per_second": 101.37,
+      "train_steps_per_second": 0.099
+    }
+  ],
+  "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": 2.988888507453276e+17,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3/mc1_g14_224_2.5b/finetune/added_tokens.json

@@ -0,0 +1,28 @@
+{
+  "</think>": 151668,
+  "</tool_call>": 151658,
+  "</tool_response>": 151666,
+  "<think>": 151667,
+  "<tool_call>": 151657,
+  "<tool_response>": 151665,
+  "<|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/qwen3/mc1_g14_224_2.5b/finetune/config.json

@@ -0,0 +1,59 @@
+{
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "LlavaQwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 6144,
+  "max_position_embeddings": 40960,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1664,
+  "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-bigG-14-quickgelu",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "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/mc1/mc1_g14_224_2.5b.pt",
+  "vocab_size": 151936
+}

continuous/qwen3/mc1_g14_224_2.5b/finetune/generation_config.json

@@ -0,0 +1,13 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "temperature": 0.6,
+  "top_k": 20,
+  "top_p": 0.95,
+  "transformers_version": "4.51.3"
+}

continuous/qwen3/mc1_g14_224_2.5b/finetune/model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb1b54f0192cde4e5609b760088a97260f696bd083fb3d03f654df37d2f6e6b5
+size 4999403240

continuous/qwen3/mc1_g14_224_2.5b/finetune/model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7f7c5721a12394cb173ff962d368dfc64525813ae44980f469a4ef758ae75a8
+size 2784437848

continuous/qwen3/mc1_g14_224_2.5b/finetune/model.safetensors.index.json

@@ -0,0 +1,906 @@
+{
+  "metadata": {
+    "total_size": 7783719424
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00002-of-00002.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.mm_projector.0.bias": "model-00002-of-00002.safetensors",
+    "model.mm_projector.0.weight": "model-00002-of-00002.safetensors",
+    "model.mm_projector.2.bias": "model-00002-of-00002.safetensors",
+    "model.mm_projector.2.weight": "model-00002-of-00002.safetensors",
+    "model.norm.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.class_embedding": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.conv1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.ln_post.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.ln_post.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.ln_pre.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.ln_pre.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.positional_embedding": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.proj": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.0.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.1.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.10.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.11.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.12.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.13.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.14.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.15.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.16.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.17.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.18.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.19.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.2.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.20.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.21.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.22.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.23.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.24.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.25.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.26.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.27.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.28.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.29.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.3.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.30.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.31.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.32.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.32.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.32.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.32.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.32.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.32.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.32.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.32.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.32.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.32.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.32.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.32.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.33.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.33.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.33.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.33.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.33.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.33.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.33.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.33.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.33.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.33.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.33.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.33.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.34.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.34.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.34.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.34.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.34.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.34.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.34.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.34.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.34.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.34.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.34.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.34.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.35.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.35.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.35.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.35.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.35.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.35.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.35.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.35.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.35.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.35.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.35.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.35.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.36.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.36.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.36.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.36.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.36.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.36.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.36.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.36.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.36.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.36.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.36.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.36.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.37.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.37.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.37.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.37.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.37.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.37.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.37.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.37.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.37.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.37.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.37.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.37.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.38.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.38.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.38.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.38.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.38.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.38.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.38.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.38.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.38.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.38.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.38.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.38.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.39.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.39.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.39.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.39.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.39.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.39.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.39.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.39.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.39.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.39.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.39.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.39.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.4.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.40.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.40.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.40.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.40.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.40.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.40.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.40.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.40.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.40.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.40.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.40.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.40.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.41.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.41.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.41.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.41.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.41.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.41.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.41.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.41.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.41.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.41.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.41.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.41.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.42.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.42.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.42.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.42.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.42.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.42.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.42.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.42.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.42.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.42.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.42.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.42.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.43.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.43.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.43.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.43.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.43.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.43.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.43.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.43.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.43.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.43.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.43.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.43.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.44.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.44.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.44.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.44.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.44.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.44.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.44.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.44.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.44.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.44.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.44.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.44.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.45.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.45.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.45.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.45.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.45.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.45.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.45.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.45.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.45.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.45.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.45.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.45.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.46.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.46.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.46.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.46.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.46.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.46.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.46.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.46.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.46.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.46.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.46.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.46.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.47.attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.47.attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.47.attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.47.attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.47.ln_1.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.47.ln_1.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.47.ln_2.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.47.ln_2.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.47.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.47.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.47.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.47.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.5.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.6.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.7.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.8.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.attn.in_proj_bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.attn.in_proj_weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.attn.out_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.attn.out_proj.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.ln_1.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.ln_1.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.ln_2.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.ln_2.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
+    "model.vision_tower.vision_tower.transformer.resblocks.9.mlp.c_proj.weight": "model-00001-of-00002.safetensors"
+  }
+}

continuous/qwen3/mc1_g14_224_2.5b/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/qwen3/mc1_g14_224_2.5b/finetune/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
+size 11422654

continuous/qwen3/mc1_g14_224_2.5b/finetune/tokenizer_config.json

@@ -0,0 +1,241 @@
+{
+  "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
+    },
+    "151665": {
+      "content": "<tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</think>",
+      "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\\n' }}\n    {%- endif %}\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    {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for message in messages[::-1] %}\n    {%- set index = (messages|length - 1) - loop.index0 %}\n    {%- if ns.multi_step_tool and message.role == \"user\" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}\n        {%- set ns.multi_step_tool = false %}\n        {%- set ns.last_query_index = index %}\n    {%- endif %}\n{%- endfor %}\n{%- for message in messages %}\n    {%- if message.content is string %}\n        {%- set content = message.content %}\n    {%- else %}\n        {%- set content = '' %}\n    {%- endif %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n        {{- '<|im_start|>' + message.role + '\\n' + content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set reasoning_content = '' %}\n        {%- if message.reasoning_content is string %}\n            {%- set reasoning_content = message.reasoning_content %}\n        {%- else %}\n            {%- if '</think>' in content %}\n                {%- set reasoning_content = content.split('</think>')[0].rstrip('\\n').split('<think>')[-1].lstrip('\\n') %}\n                {%- set content = content.split('</think>')[-1].lstrip('\\n') %}\n            {%- endif %}\n        {%- endif %}\n        {%- if loop.index0 > ns.last_query_index %}\n            {%- if loop.last or (not loop.last and reasoning_content) %}\n                {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content.strip('\\n') + '\\n</think>\\n\\n' + content.lstrip('\\n') }}\n            {%- else %}\n                {{- '<|im_start|>' + message.role + '\\n' + content }}\n            {%- endif %}\n        {%- else %}\n            {{- '<|im_start|>' + message.role + '\\n' + content }}\n        {%- endif %}\n        {%- if message.tool_calls %}\n            {%- for tool_call in message.tool_calls %}\n                {%- if (loop.first and content) or (not loop.first) %}\n                    {{- '\\n' }}\n                {%- endif %}\n                {%- if tool_call.function %}\n                    {%- set tool_call = tool_call.function %}\n                {%- endif %}\n                {{- '<tool_call>\\n{\"name\": \"' }}\n                {{- tool_call.name }}\n                {{- '\", \"arguments\": ' }}\n                {%- if tool_call.arguments is string %}\n                    {{- tool_call.arguments }}\n                {%- else %}\n                    {{- tool_call.arguments | tojson }}\n                {%- endif %}\n                {{- '}\\n</tool_call>' }}\n            {%- endfor %}\n        {%- endif %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if loop.first or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- 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    {%- if enable_thinking is defined and enable_thinking is false %}\n        {{- '<think>\\n\\n</think>\\n\\n' }}\n    {%- endif %}\n{%- endif %}",
+  "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/qwen3/mc1_g14_224_2.5b/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": 15.938782691955566,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 2.1359,
+      "step": 10
+    },
+    {
+      "epoch": 0.007700452401578592,
+      "grad_norm": 6.206850051879883,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 1.7433,
+      "step": 20
+    },
+    {
+      "epoch": 0.01155067860236789,
+      "grad_norm": 4.335718154907227,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.4112,
+      "step": 30
+    },
+    {
+      "epoch": 0.015400904803157185,
+      "grad_norm": 0.9348963499069214,
+      "learning_rate": 5e-06,
+      "loss": 1.0996,
+      "step": 40
+    },
+    {
+      "epoch": 0.019251131003946483,
+      "grad_norm": 0.8000525832176208,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 1.0108,
+      "step": 50
+    },
+    {
+      "epoch": 0.02310135720473578,
+      "grad_norm": 0.4075191617012024,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 0.9303,
+      "step": 60
+    },
+    {
+      "epoch": 0.026951583405525073,
+      "grad_norm": 0.3138245642185211,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.8861,
+      "step": 70
+    },
+    {
+      "epoch": 0.03080180960631437,
+      "grad_norm": 0.2958347797393799,
+      "learning_rate": 9.999996111488723e-06,
+      "loss": 0.8843,
+      "step": 80
+    },
+    {
+      "epoch": 0.034652035807103666,
+      "grad_norm": 0.28463247418403625,
+      "learning_rate": 9.999529497453782e-06,
+      "loss": 0.855,
+      "step": 90
+    },
+    {
+      "epoch": 0.038502262007892966,
+      "grad_norm": 0.24649657309055328,
+      "learning_rate": 9.998285264324193e-06,
+      "loss": 0.8338,
+      "step": 100
+    },
+    {
+      "epoch": 0.04235248820868226,
+      "grad_norm": 0.2728543281555176,
+      "learning_rate": 9.996263605626052e-06,
+      "loss": 0.8185,
+      "step": 110
+    },
+    {
+      "epoch": 0.04620271440947156,
+      "grad_norm": 0.253468781709671,
+      "learning_rate": 9.99346483580503e-06,
+      "loss": 0.828,
+      "step": 120
+    },
+    {
+      "epoch": 0.05005294061026085,
+      "grad_norm": 0.25056809186935425,
+      "learning_rate": 9.989889390177467e-06,
+      "loss": 0.821,
+      "step": 130
+    },
+    {
+      "epoch": 0.053903166811050146,
+      "grad_norm": 0.28353169560432434,
+      "learning_rate": 9.985537824862647e-06,
+      "loss": 0.8201,
+      "step": 140
+    },
+    {
+      "epoch": 0.057753393011839446,
+      "grad_norm": 0.2922304570674896,
+      "learning_rate": 9.980410816696317e-06,
+      "loss": 0.8029,
+      "step": 150
+    },
+    {
+      "epoch": 0.06160361921262874,
+      "grad_norm": 0.2498026043176651,
+      "learning_rate": 9.97450916312541e-06,
+      "loss": 0.811,
+      "step": 160
+    },
+    {
+      "epoch": 0.06545384541341803,
+      "grad_norm": 0.2206718921661377,
+      "learning_rate": 9.967833782084004e-06,
+      "loss": 0.8122,
+      "step": 170
+    },
+    {
+      "epoch": 0.06930407161420733,
+      "grad_norm": 0.2571215033531189,
+      "learning_rate": 9.960385711850554e-06,
+      "loss": 0.8365,
+      "step": 180
+    },
+    {
+      "epoch": 0.07315429781499663,
+      "grad_norm": 0.2603733539581299,
+      "learning_rate": 9.952166110886394e-06,
+      "loss": 0.8284,
+      "step": 190
+    },
+    {
+      "epoch": 0.07700452401578593,
+      "grad_norm": 0.274894118309021,
+      "learning_rate": 9.943176257655567e-06,
+      "loss": 0.8142,
+      "step": 200
+    },
+    {
+      "epoch": 0.08085475021657522,
+      "grad_norm": 0.25042036175727844,
+      "learning_rate": 9.933417550425952e-06,
+      "loss": 0.8154,
+      "step": 210
+    },
+    {
+      "epoch": 0.08470497641736452,
+      "grad_norm": 0.2538484036922455,
+      "learning_rate": 9.9228915070518e-06,
+      "loss": 0.7979,
+      "step": 220
+    },
+    {
+      "epoch": 0.08855520261815382,
+      "grad_norm": 0.24589698016643524,
+      "learning_rate": 9.911599764737632e-06,
+      "loss": 0.7989,
+      "step": 230
+    },
+    {
+      "epoch": 0.09240542881894312,
+      "grad_norm": 0.25686389207839966,
+      "learning_rate": 9.899544079783606e-06,
+      "loss": 0.7946,
+      "step": 240
+    },
+    {
+      "epoch": 0.0962556550197324,
+      "grad_norm": 0.28709274530410767,
+      "learning_rate": 9.886726327312338e-06,
+      "loss": 0.8128,
+      "step": 250
+    },
+    {
+      "epoch": 0.1001058812205217,
+      "grad_norm": 0.2472069263458252,
+      "learning_rate": 9.87314850097724e-06,
+      "loss": 0.8056,
+      "step": 260
+    },
+    {
+      "epoch": 0.103956107421311,
+      "grad_norm": 0.24343957006931305,
+      "learning_rate": 9.858812712652452e-06,
+      "loss": 0.7962,
+      "step": 270
+    },
+    {
+      "epoch": 0.10780633362210029,
+      "grad_norm": 0.23682667315006256,
+      "learning_rate": 9.843721192104341e-06,
+      "loss": 0.7929,
+      "step": 280
+    },
+    {
+      "epoch": 0.11165655982288959,
+      "grad_norm": 0.24648794531822205,
+      "learning_rate": 9.827876286644698e-06,
+      "loss": 0.8131,
+      "step": 290
+    },
+    {
+      "epoch": 0.11550678602367889,
+      "grad_norm": 0.22092968225479126,
+      "learning_rate": 9.811280460765643e-06,
+      "loss": 0.7753,
+      "step": 300
+    },
+    {
+      "epoch": 0.11935701222446819,
+      "grad_norm": 0.2331714630126953,
+      "learning_rate": 9.793936295756292e-06,
+      "loss": 0.7866,
+      "step": 310
+    },
+    {
+      "epoch": 0.12320723842525748,
+      "grad_norm": 0.241885706782341,
+      "learning_rate": 9.775846489301279e-06,
+      "loss": 0.7868,
+      "step": 320
+    },
+    {
+      "epoch": 0.12705746462604678,
+      "grad_norm": 0.2464597374200821,
+      "learning_rate": 9.757013855061157e-06,
+      "loss": 0.7864,
+      "step": 330
+    },
+    {
+      "epoch": 0.13090769082683607,
+      "grad_norm": 0.24696575105190277,
+      "learning_rate": 9.737441322234758e-06,
+      "loss": 0.7988,
+      "step": 340
+    },
+    {
+      "epoch": 0.13475791702762538,
+      "grad_norm": 0.25845202803611755,
+      "learning_rate": 9.717131935103605e-06,
+      "loss": 0.7991,
+      "step": 350
+    },
+    {
+      "epoch": 0.13860814322841467,
+      "grad_norm": 0.23703010380268097,
+      "learning_rate": 9.696088852558398e-06,
+      "loss": 0.7862,
+      "step": 360
+    },
+    {
+      "epoch": 0.14245836942920398,
+      "grad_norm": 0.2517313063144684,
+      "learning_rate": 9.674315347607686e-06,
+      "loss": 0.8133,
+      "step": 370
+    },
+    {
+      "epoch": 0.14630859562999327,
+      "grad_norm": 0.23772449791431427,
+      "learning_rate": 9.651814806868796e-06,
+      "loss": 0.7734,
+      "step": 380
+    },
+    {
+      "epoch": 0.15015882183078255,
+      "grad_norm": 0.24716877937316895,
+      "learning_rate": 9.62859073004107e-06,
+      "loss": 0.7773,
+      "step": 390
+    },
+    {
+      "epoch": 0.15400904803157187,
+      "grad_norm": 0.2327948361635208,
+      "learning_rate": 9.604646729361547e-06,
+      "loss": 0.7787,
+      "step": 400
+    },
+    {
+      "epoch": 0.15785927423236115,
+      "grad_norm": 0.23068377375602722,
+      "learning_rate": 9.579986529043099e-06,
+      "loss": 0.7669,
+      "step": 410
+    },
+    {
+      "epoch": 0.16170950043315044,
+      "grad_norm": 0.23790355026721954,
+      "learning_rate": 9.554613964695189e-06,
+      "loss": 0.7873,
+      "step": 420
+    },
+    {
+      "epoch": 0.16555972663393975,
+      "grad_norm": 0.23381349444389343,
+      "learning_rate": 9.528532982727281e-06,
+      "loss": 0.779,
+      "step": 430
+    },
+    {
+      "epoch": 0.16940995283472904,
+      "grad_norm": 0.23572732508182526,
+      "learning_rate": 9.501747639735018e-06,
+      "loss": 0.7739,
+      "step": 440
+    },
+    {
+      "epoch": 0.17326017903551832,
+      "grad_norm": 0.2444908618927002,
+      "learning_rate": 9.474262101869272e-06,
+      "loss": 0.7642,
+      "step": 450
+    },
+    {
+      "epoch": 0.17711040523630764,
+      "grad_norm": 0.23828577995300293,
+      "learning_rate": 9.446080644188141e-06,
+      "loss": 0.7717,
+      "step": 460
+    },
+    {
+      "epoch": 0.18096063143709692,
+      "grad_norm": 0.23429028689861298,
+      "learning_rate": 9.417207649992017e-06,
+      "loss": 0.7764,
+      "step": 470
+    },
+    {
+      "epoch": 0.18481085763788624,
+      "grad_norm": 0.23025472462177277,
+      "learning_rate": 9.387647610141816e-06,
+      "loss": 0.7743,
+      "step": 480
+    },
+    {
+      "epoch": 0.18866108383867553,
+      "grad_norm": 0.22012139856815338,
+      "learning_rate": 9.35740512236047e-06,
+      "loss": 0.7625,
+      "step": 490
+    },
+    {
+      "epoch": 0.1925113100394648,
+      "grad_norm": 0.23216642439365387,
+      "learning_rate": 9.326484890517809e-06,
+      "loss": 0.7716,
+      "step": 500
+    },
+    {
+      "epoch": 0.19636153624025413,
+      "grad_norm": 0.2323516309261322,
+      "learning_rate": 9.294891723898926e-06,
+      "loss": 0.7634,
+      "step": 510
+    },
+    {
+      "epoch": 0.2002117624410434,
+      "grad_norm": 0.24509230256080627,
+      "learning_rate": 9.262630536456153e-06,
+      "loss": 0.7933,
+      "step": 520
+    },
+    {
+      "epoch": 0.2040619886418327,
+      "grad_norm": 0.24861086905002594,
+      "learning_rate": 9.229706346044749e-06,
+      "loss": 0.7649,
+      "step": 530
+    },
+    {
+      "epoch": 0.207912214842622,
+      "grad_norm": 0.2478780299425125,
+      "learning_rate": 9.196124273642422e-06,
+      "loss": 0.7799,
+      "step": 540
+    },
+    {
+      "epoch": 0.2117624410434113,
+      "grad_norm": 0.2261587679386139,
+      "learning_rate": 9.161889542552842e-06,
+      "loss": 0.7845,
+      "step": 550
+    },
+    {
+      "epoch": 0.21561266724420058,
+      "grad_norm": 0.21859097480773926,
+      "learning_rate": 9.127007477593193e-06,
+      "loss": 0.7857,
+      "step": 560
+    },
+    {
+      "epoch": 0.2194628934449899,
+      "grad_norm": 0.2597237825393677,
+      "learning_rate": 9.091483504265974e-06,
+      "loss": 0.7897,
+      "step": 570
+    },
+    {
+      "epoch": 0.22331311964577918,
+      "grad_norm": 0.24913398921489716,
+      "learning_rate": 9.055323147915116e-06,
+      "loss": 0.7748,
+      "step": 580
+    },
+    {
+      "epoch": 0.2271633458465685,
+      "grad_norm": 0.2428615838289261,
+      "learning_rate": 9.018532032866585e-06,
+      "loss": 0.7808,
+      "step": 590
+    },
+    {
+      "epoch": 0.23101357204735778,
+      "grad_norm": 0.22239165008068085,
+      "learning_rate": 8.981115881553579e-06,
+      "loss": 0.7735,
+      "step": 600
+    },
+    {
+      "epoch": 0.23486379824814707,
+      "grad_norm": 0.25532230734825134,
+      "learning_rate": 8.943080513626476e-06,
+      "loss": 0.7938,
+      "step": 610
+    },
+    {
+      "epoch": 0.23871402444893638,
+      "grad_norm": 0.23452216386795044,
+      "learning_rate": 8.904431845047641e-06,
+      "loss": 0.7802,
+      "step": 620
+    },
+    {
+      "epoch": 0.24256425064972567,
+      "grad_norm": 0.24212679266929626,
+      "learning_rate": 8.86517588717128e-06,
+      "loss": 0.7948,
+      "step": 630
+    },
+    {
+      "epoch": 0.24641447685051496,
+      "grad_norm": 0.23078350722789764,
+      "learning_rate": 8.82531874580844e-06,
+      "loss": 0.789,
+      "step": 640
+    },
+    {
+      "epoch": 0.25026470305130427,
+      "grad_norm": 0.21680957078933716,
+      "learning_rate": 8.784866620277314e-06,
+      "loss": 0.7446,
+      "step": 650
+    },
+    {
+      "epoch": 0.25411492925209356,
+      "grad_norm": 0.24869920313358307,
+      "learning_rate": 8.743825802439018e-06,
+      "loss": 0.7754,
+      "step": 660
+    },
+    {
+      "epoch": 0.25796515545288284,
+      "grad_norm": 0.2733602225780487,
+      "learning_rate": 8.702202675718956e-06,
+      "loss": 0.7862,
+      "step": 670
+    },
+    {
+      "epoch": 0.26181538165367213,
+      "grad_norm": 0.2229803502559662,
+      "learning_rate": 8.660003714113959e-06,
+      "loss": 0.7537,
+      "step": 680
+    },
+    {
+      "epoch": 0.26566560785446147,
+      "grad_norm": 0.23141805827617645,
+      "learning_rate": 8.617235481185323e-06,
+      "loss": 0.7768,
+      "step": 690
+    },
+    {
+      "epoch": 0.26951583405525076,
+      "grad_norm": 0.24003753066062927,
+      "learning_rate": 8.573904629037931e-06,
+      "loss": 0.7841,
+      "step": 700
+    },
+    {
+      "epoch": 0.27336606025604004,
+      "grad_norm": 0.23308081924915314,
+      "learning_rate": 8.530017897285585e-06,
+      "loss": 0.7658,
+      "step": 710
+    },
+    {
+      "epoch": 0.27721628645682933,
+      "grad_norm": 0.21398328244686127,
+      "learning_rate": 8.485582112002746e-06,
+      "loss": 0.7691,
+      "step": 720
+    },
+    {
+      "epoch": 0.2810665126576186,
+      "grad_norm": 0.23429159820079803,
+      "learning_rate": 8.440604184662811e-06,
+      "loss": 0.7717,
+      "step": 730
+    },
+    {
+      "epoch": 0.28491673885840796,
+      "grad_norm": 0.22904668748378754,
+      "learning_rate": 8.395091111063118e-06,
+      "loss": 0.7597,
+      "step": 740
+    },
+    {
+      "epoch": 0.28876696505919724,
+      "grad_norm": 0.22168131172657013,
+      "learning_rate": 8.349049970236822e-06,
+      "loss": 0.7635,
+      "step": 750
+    },
+    {
+      "epoch": 0.29261719125998653,
+      "grad_norm": 0.23149767518043518,
+      "learning_rate": 8.302487923351842e-06,
+      "loss": 0.7541,
+      "step": 760
+    },
+    {
+      "epoch": 0.2964674174607758,
+      "grad_norm": 0.22459255158901215,
+      "learning_rate": 8.255412212597021e-06,
+      "loss": 0.7718,
+      "step": 770
+    },
+    {
+      "epoch": 0.3003176436615651,
+      "grad_norm": 0.23251083493232727,
+      "learning_rate": 8.207830160055688e-06,
+      "loss": 0.7685,
+      "step": 780
+    },
+    {
+      "epoch": 0.3041678698623544,
+      "grad_norm": 0.23903296887874603,
+      "learning_rate": 8.159749166566786e-06,
+      "loss": 0.7527,
+      "step": 790
+    },
+    {
+      "epoch": 0.30801809606314373,
+      "grad_norm": 0.21728335320949554,
+      "learning_rate": 8.11117671057377e-06,
+      "loss": 0.7789,
+      "step": 800
+    },
+    {
+      "epoch": 0.311868322263933,
+      "grad_norm": 0.2368263155221939,
+      "learning_rate": 8.062120346961404e-06,
+      "loss": 0.771,
+      "step": 810
+    },
+    {
+      "epoch": 0.3157185484647223,
+      "grad_norm": 0.22537410259246826,
+      "learning_rate": 8.012587705880705e-06,
+      "loss": 0.7639,
+      "step": 820
+    },
+    {
+      "epoch": 0.3195687746655116,
+      "grad_norm": 0.22877825796604156,
+      "learning_rate": 7.962586491562146e-06,
+      "loss": 0.7507,
+      "step": 830
+    },
+    {
+      "epoch": 0.3234190008663009,
+      "grad_norm": 0.23388135433197021,
+      "learning_rate": 7.912124481117357e-06,
+      "loss": 0.7577,
+      "step": 840
+    },
+    {
+      "epoch": 0.3272692270670902,
+      "grad_norm": 0.24391938745975494,
+      "learning_rate": 7.861209523329494e-06,
+      "loss": 0.7707,
+      "step": 850
+    },
+    {
+      "epoch": 0.3311194532678795,
+      "grad_norm": 0.230848491191864,
+      "learning_rate": 7.809849537432432e-06,
+      "loss": 0.7505,
+      "step": 860
+    },
+    {
+      "epoch": 0.3349696794686688,
+      "grad_norm": 0.226628378033638,
+      "learning_rate": 7.758052511879035e-06,
+      "loss": 0.762,
+      "step": 870
+    },
+    {
+      "epoch": 0.3388199056694581,
+      "grad_norm": 0.22778740525245667,
+      "learning_rate": 7.705826503098642e-06,
+      "loss": 0.7582,
+      "step": 880
+    },
+    {
+      "epoch": 0.34267013187024736,
+      "grad_norm": 0.22623923420906067,
+      "learning_rate": 7.653179634243979e-06,
+      "loss": 0.7588,
+      "step": 890
+    },
+    {
+      "epoch": 0.34652035807103665,
+      "grad_norm": 0.2115490436553955,
+      "learning_rate": 7.600120093927696e-06,
+      "loss": 0.761,
+      "step": 900
+    },
+    {
+      "epoch": 0.350370584271826,
+      "grad_norm": 0.2293318510055542,
+      "learning_rate": 7.546656134948729e-06,
+      "loss": 0.7492,
+      "step": 910
+    },
+    {
+      "epoch": 0.3542208104726153,
+      "grad_norm": 0.2384779453277588,
+      "learning_rate": 7.492796073008661e-06,
+      "loss": 0.7518,
+      "step": 920
+    },
+    {
+      "epoch": 0.35807103667340456,
+      "grad_norm": 0.24575965106487274,
+      "learning_rate": 7.438548285418326e-06,
+      "loss": 0.7369,
+      "step": 930
+    },
+    {
+      "epoch": 0.36192126287419385,
+      "grad_norm": 0.2531838119029999,
+      "learning_rate": 7.3839212097948e-06,
+      "loss": 0.773,
+      "step": 940
+    },
+    {
+      "epoch": 0.36577148907498314,
+      "grad_norm": 0.23271553218364716,
+      "learning_rate": 7.328923342749039e-06,
+      "loss": 0.7876,
+      "step": 950
+    },
+    {
+      "epoch": 0.3696217152757725,
+      "grad_norm": 0.24493509531021118,
+      "learning_rate": 7.273563238564319e-06,
+      "loss": 0.7706,
+      "step": 960
+    },
+    {
+      "epoch": 0.37347194147656176,
+      "grad_norm": 0.23276367783546448,
+      "learning_rate": 7.217849507865724e-06,
+      "loss": 0.7503,
+      "step": 970
+    },
+    {
+      "epoch": 0.37732216767735105,
+      "grad_norm": 0.22332541644573212,
+      "learning_rate": 7.161790816280858e-06,
+      "loss": 0.7582,
+      "step": 980
+    },
+    {
+      "epoch": 0.38117239387814034,
+      "grad_norm": 0.2369937151670456,
+      "learning_rate": 7.1053958830920125e-06,
+      "loss": 0.7389,
+      "step": 990
+    },
+    {
+      "epoch": 0.3850226200789296,
+      "grad_norm": 0.2348017394542694,
+      "learning_rate": 7.0486734798799715e-06,
+      "loss": 0.7606,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3888728462797189,
+      "grad_norm": 0.23130328953266144,
+      "learning_rate": 6.991632429159708e-06,
+      "loss": 0.783,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39272307248050825,
+      "grad_norm": 0.2459104061126709,
+      "learning_rate": 6.934281603008132e-06,
+      "loss": 0.7715,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39657329868129754,
+      "grad_norm": 0.23144766688346863,
+      "learning_rate": 6.876629921684143e-06,
+      "loss": 0.77,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4004235248820868,
+      "grad_norm": 0.2338777482509613,
+      "learning_rate": 6.818686352241199e-06,
+      "loss": 0.7605,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4042737510828761,
+      "grad_norm": 0.21468032896518707,
+      "learning_rate": 6.760459907132573e-06,
+      "loss": 0.7552,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4081239772836654,
+      "grad_norm": 0.22308507561683655,
+      "learning_rate": 6.701959642809589e-06,
+      "loss": 0.7611,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41197420348445474,
+      "grad_norm": 0.23295892775058746,
+      "learning_rate": 6.643194658312987e-06,
+      "loss": 0.7539,
+      "step": 1070
+    },
+    {
+      "epoch": 0.415824429685244,
+      "grad_norm": 0.21973314881324768,
+      "learning_rate": 6.584174093857676e-06,
+      "loss": 0.7772,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4196746558860333,
+      "grad_norm": 0.2715476155281067,
+      "learning_rate": 6.5249071294110735e-06,
+      "loss": 0.7602,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4235248820868226,
+      "grad_norm": 0.23945797979831696,
+      "learning_rate": 6.465402983265274e-06,
+      "loss": 0.7736,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4273751082876119,
+      "grad_norm": 0.23387253284454346,
+      "learning_rate": 6.405670910603243e-06,
+      "loss": 0.768,
+      "step": 1110
+    },
+    {
+      "epoch": 0.43122533448840117,
+      "grad_norm": 0.24628262221813202,
+      "learning_rate": 6.345720202059291e-06,
+      "loss": 0.7645,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4350755606891905,
+      "grad_norm": 0.2389136254787445,
+      "learning_rate": 6.285560182274008e-06,
+      "loss": 0.7449,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4389257868899798,
+      "grad_norm": 0.23808903992176056,
+      "learning_rate": 6.225200208443936e-06,
+      "loss": 0.7676,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4427760130907691,
+      "grad_norm": 0.21955297887325287,
+      "learning_rate": 6.16464966886616e-06,
+      "loss": 0.7541,
+      "step": 1150
+    },
+    {
+      "epoch": 0.44662623929155837,
+      "grad_norm": 0.2581087350845337,
+      "learning_rate": 6.103917981478065e-06,
+      "loss": 0.7792,
+      "step": 1160
+    },
+    {
+      "epoch": 0.45047646549234766,
+      "grad_norm": 0.22595995664596558,
+      "learning_rate": 6.043014592392489e-06,
+      "loss": 0.79,
+      "step": 1170
+    },
+    {
+      "epoch": 0.454326691693137,
+      "grad_norm": 0.22840270400047302,
+      "learning_rate": 5.981948974428485e-06,
+      "loss": 0.7651,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4581769178939263,
+      "grad_norm": 0.2432698756456375,
+      "learning_rate": 5.920730625637934e-06,
+      "loss": 0.7414,
+      "step": 1190
+    },
+    {
+      "epoch": 0.46202714409471557,
+      "grad_norm": 0.24247057735919952,
+      "learning_rate": 5.8593690678282355e-06,
+      "loss": 0.7553,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46587737029550486,
+      "grad_norm": 0.24246333539485931,
+      "learning_rate": 5.797873845081301e-06,
+      "loss": 0.7599,
+      "step": 1210
+    },
+    {
+      "epoch": 0.46972759649629414,
+      "grad_norm": 0.22863660752773285,
+      "learning_rate": 5.736254522269075e-06,
+      "loss": 0.7761,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47357782269708343,
+      "grad_norm": 0.22712889313697815,
+      "learning_rate": 5.6745206835658386e-06,
+      "loss": 0.7793,
+      "step": 1230
+    },
+    {
+      "epoch": 0.47742804889787277,
+      "grad_norm": 0.22796353697776794,
+      "learning_rate": 5.612681930957495e-06,
+      "loss": 0.7496,
+      "step": 1240
+    },
+    {
+      "epoch": 0.48127827509866206,
+      "grad_norm": 0.22127237915992737,
+      "learning_rate": 5.550747882748095e-06,
+      "loss": 0.7465,
+      "step": 1250
+    },
+    {
+      "epoch": 0.48512850129945134,
+      "grad_norm": 0.24317073822021484,
+      "learning_rate": 5.4887281720638195e-06,
+      "loss": 0.7534,
+      "step": 1260
+    },
+    {
+      "epoch": 0.48897872750024063,
+      "grad_norm": 0.2698855996131897,
+      "learning_rate": 5.42663244535466e-06,
+      "loss": 0.7368,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4928289537010299,
+      "grad_norm": 0.2446776181459427,
+      "learning_rate": 5.3644703608940205e-06,
+      "loss": 0.7478,
+      "step": 1280
+    },
+    {
+      "epoch": 0.49667917990181926,
+      "grad_norm": 0.22836153209209442,
+      "learning_rate": 5.302251587276491e-06,
+      "loss": 0.7574,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5005294061026085,
+      "grad_norm": 0.2217899113893509,
+      "learning_rate": 5.2399858019140005e-06,
+      "loss": 0.758,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5043796323033978,
+      "grad_norm": 0.22215962409973145,
+      "learning_rate": 5.177682689530615e-06,
+      "loss": 0.7747,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5082298585041871,
+      "grad_norm": 0.2372257113456726,
+      "learning_rate": 5.11535194065619e-06,
+      "loss": 0.7402,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5120800847049765,
+      "grad_norm": 0.2604122459888458,
+      "learning_rate": 5.053003250119113e-06,
+      "loss": 0.7497,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5159303109057657,
+      "grad_norm": 0.2306693196296692,
+      "learning_rate": 4.9906463155383925e-06,
+      "loss": 0.7593,
+      "step": 1340
+    },
+    {
+      "epoch": 0.519780537106555,
+      "grad_norm": 0.22874194383621216,
+      "learning_rate": 4.9282908358153046e-06,
+      "loss": 0.7534,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5236307633073443,
+      "grad_norm": 0.26147574186325073,
+      "learning_rate": 4.8659465096248345e-06,
+      "loss": 0.7673,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5274809895081336,
+      "grad_norm": 0.24484500288963318,
+      "learning_rate": 4.803623033907169e-06,
+      "loss": 0.7652,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5313312157089229,
+      "grad_norm": 0.2214219719171524,
+      "learning_rate": 4.741330102359441e-06,
+      "loss": 0.772,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5351814419097122,
+      "grad_norm": 0.25001591444015503,
+      "learning_rate": 4.679077403927996e-06,
+      "loss": 0.7509,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5390316681105015,
+      "grad_norm": 0.23428945243358612,
+      "learning_rate": 4.616874621301373e-06,
+      "loss": 0.7524,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5428818943112907,
+      "grad_norm": 0.23315560817718506,
+      "learning_rate": 4.554731429404293e-06,
+      "loss": 0.7662,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5467321205120801,
+      "grad_norm": 0.24245242774486542,
+      "learning_rate": 4.492657493892822e-06,
+      "loss": 0.7494,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5505823467128694,
+      "grad_norm": 0.2631940543651581,
+      "learning_rate": 4.430662469650999e-06,
+      "loss": 0.7651,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5544325729136587,
+      "grad_norm": 0.25806495547294617,
+      "learning_rate": 4.368755999289118e-06,
+      "loss": 0.7702,
+      "step": 1440
+    },
+    {
+      "epoch": 0.558282799114448,
+      "grad_norm": 0.24815551936626434,
+      "learning_rate": 4.306947711643949e-06,
+      "loss": 0.7534,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5621330253152372,
+      "grad_norm": 0.2405899316072464,
+      "learning_rate": 4.245247220281068e-06,
+      "loss": 0.7561,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5659832515160266,
+      "grad_norm": 0.22825482487678528,
+      "learning_rate": 4.183664121999587e-06,
+      "loss": 0.7525,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5698334777168159,
+      "grad_norm": 0.23008424043655396,
+      "learning_rate": 4.122207995339474e-06,
+      "loss": 0.7524,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5736837039176051,
+      "grad_norm": 0.2216154783964157,
+      "learning_rate": 4.060888399091734e-06,
+      "loss": 0.7542,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5775339301183945,
+      "grad_norm": 0.23121920228004456,
+      "learning_rate": 3.999714870811639e-06,
+      "loss": 0.7655,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5813841563191837,
+      "grad_norm": 0.2267971932888031,
+      "learning_rate": 3.938696925335275e-06,
+      "loss": 0.7573,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5852343825199731,
+      "grad_norm": 0.23328624665737152,
+      "learning_rate": 3.87784405329962e-06,
+      "loss": 0.7653,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5890846087207623,
+      "grad_norm": 0.22738058865070343,
+      "learning_rate": 3.817165719666382e-06,
+      "loss": 0.7673,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5929348349215516,
+      "grad_norm": 0.22763855755329132,
+      "learning_rate": 3.756671362249841e-06,
+      "loss": 0.7723,
+      "step": 1540
+    },
+    {
+      "epoch": 0.596785061122341,
+      "grad_norm": 0.22799792885780334,
+      "learning_rate": 3.6963703902488907e-06,
+      "loss": 0.7573,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6006352873231302,
+      "grad_norm": 0.24795284867286682,
+      "learning_rate": 3.6362721827835655e-06,
+      "loss": 0.7727,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6044855135239195,
+      "grad_norm": 0.23241494596004486,
+      "learning_rate": 3.576386087436212e-06,
+      "loss": 0.7698,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6083357397247088,
+      "grad_norm": 0.24409981071949005,
+      "learning_rate": 3.5167214187975913e-06,
+      "loss": 0.7582,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6121859659254981,
+      "grad_norm": 0.2116926908493042,
+      "learning_rate": 3.457287457018086e-06,
+      "loss": 0.7632,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6160361921262875,
+      "grad_norm": 0.21925297379493713,
+      "learning_rate": 3.3980934463643003e-06,
+      "loss": 0.7497,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6198864183270767,
+      "grad_norm": 0.21595044434070587,
+      "learning_rate": 3.3391485937812053e-06,
+      "loss": 0.7322,
+      "step": 1610
+    },
+    {
+      "epoch": 0.623736644527866,
+      "grad_norm": 0.23575180768966675,
+      "learning_rate": 3.280462067460115e-06,
+      "loss": 0.7311,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6275868707286553,
+      "grad_norm": 0.22942230105400085,
+      "learning_rate": 3.222042995412669e-06,
+      "loss": 0.7623,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6314370969294446,
+      "grad_norm": 0.21496590971946716,
+      "learning_rate": 3.163900464051089e-06,
+      "loss": 0.7526,
+      "step": 1640
+    },
+    {
+      "epoch": 0.635287323130234,
+      "grad_norm": 0.22745965421199799,
+      "learning_rate": 3.1060435167748826e-06,
+      "loss": 0.7621,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6391375493310232,
+      "grad_norm": 0.228306844830513,
+      "learning_rate": 3.048481152564253e-06,
+      "loss": 0.7431,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6429877755318125,
+      "grad_norm": 0.2448883205652237,
+      "learning_rate": 2.9912223245804017e-06,
+      "loss": 0.7665,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6468380017326018,
+      "grad_norm": 0.24057932198047638,
+      "learning_rate": 2.934275938772981e-06,
+      "loss": 0.7587,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6506882279333911,
+      "grad_norm": 0.2448817789554596,
+      "learning_rate": 2.8776508524948676e-06,
+      "loss": 0.7509,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6545384541341804,
+      "grad_norm": 0.209664985537529,
+      "learning_rate": 2.821355873124503e-06,
+      "loss": 0.7475,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6583886803349697,
+      "grad_norm": 0.24985933303833008,
+      "learning_rate": 2.7653997566960144e-06,
+      "loss": 0.7642,
+      "step": 1710
+    },
+    {
+      "epoch": 0.662238906535759,
+      "grad_norm": 0.2464679777622223,
+      "learning_rate": 2.709791206537309e-06,
+      "loss": 0.7538,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6660891327365482,
+      "grad_norm": 0.23413266241550446,
+      "learning_rate": 2.6545388719163747e-06,
+      "loss": 0.7558,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6699393589373376,
+      "grad_norm": 0.22719594836235046,
+      "learning_rate": 2.599651346695979e-06,
+      "loss": 0.7557,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6737895851381268,
+      "grad_norm": 0.23457537591457367,
+      "learning_rate": 2.545137167997007e-06,
+      "loss": 0.7534,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6776398113389162,
+      "grad_norm": 0.2279079556465149,
+      "learning_rate": 2.4910048148705972e-06,
+      "loss": 0.7393,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6814900375397055,
+      "grad_norm": 0.2485709935426712,
+      "learning_rate": 2.4372627069793327e-06,
+      "loss": 0.7858,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6853402637404947,
+      "grad_norm": 0.23209841549396515,
+      "learning_rate": 2.3839192032876486e-06,
+      "loss": 0.7695,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6891904899412841,
+      "grad_norm": 0.22678497433662415,
+      "learning_rate": 2.3309826007617047e-06,
+      "loss": 0.7799,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6930407161420733,
+      "grad_norm": 0.22950664162635803,
+      "learning_rate": 2.278461133078881e-06,
+      "loss": 0.7459,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6968909423428626,
+      "grad_norm": 0.23777832090854645,
+      "learning_rate": 2.2263629693471258e-06,
+      "loss": 0.7355,
+      "step": 1810
+    },
+    {
+      "epoch": 0.700741168543652,
+      "grad_norm": 0.22598639130592346,
+      "learning_rate": 2.1746962128343414e-06,
+      "loss": 0.7592,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7045913947444412,
+      "grad_norm": 0.22603057324886322,
+      "learning_rate": 2.1234688997080183e-06,
+      "loss": 0.7496,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7084416209452306,
+      "grad_norm": 0.24104034900665283,
+      "learning_rate": 2.0726889977853053e-06,
+      "loss": 0.7532,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7122918471460198,
+      "grad_norm": 0.24585062265396118,
+      "learning_rate": 2.022364405293703e-06,
+      "loss": 0.7677,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7161420733468091,
+      "grad_norm": 0.2339080423116684,
+      "learning_rate": 1.9725029496425814e-06,
+      "loss": 0.7586,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7199922995475985,
+      "grad_norm": 0.23538461327552795,
+      "learning_rate": 1.923112386205727e-06,
+      "loss": 0.7562,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7238425257483877,
+      "grad_norm": 0.21733485162258148,
+      "learning_rate": 1.874200397115079e-06,
+      "loss": 0.7637,
+      "step": 1880
+    },
+    {
+      "epoch": 0.727692751949177,
+      "grad_norm": 0.23714548349380493,
+      "learning_rate": 1.825774590065857e-06,
+      "loss": 0.747,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7315429781499663,
+      "grad_norm": 0.23281361162662506,
+      "learning_rate": 1.7778424971332813e-06,
+      "loss": 0.7549,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7353932043507556,
+      "grad_norm": 0.22572797536849976,
+      "learning_rate": 1.7304115736010407e-06,
+      "loss": 0.7521,
+      "step": 1910
+    },
+    {
+      "epoch": 0.739243430551545,
+      "grad_norm": 0.23069904744625092,
+      "learning_rate": 1.6834891968017108e-06,
+      "loss": 0.7459,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7430936567523342,
+      "grad_norm": 0.23634332418441772,
+      "learning_rate": 1.6370826649692884e-06,
+      "loss": 0.7619,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7469438829531235,
+      "grad_norm": 0.24165770411491394,
+      "learning_rate": 1.5911991961040463e-06,
+      "loss": 0.773,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7507941091539128,
+      "grad_norm": 0.24403640627861023,
+      "learning_rate": 1.545845926849851e-06,
+      "loss": 0.7605,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7546443353547021,
+      "grad_norm": 0.25386133790016174,
+      "learning_rate": 1.5010299113841397e-06,
+      "loss": 0.7762,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7584945615554913,
+      "grad_norm": 0.23420223593711853,
+      "learning_rate": 1.4567581203207238e-06,
+      "loss": 0.766,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7623447877562807,
+      "grad_norm": 0.24760258197784424,
+      "learning_rate": 1.413037439625593e-06,
+      "loss": 0.7632,
+      "step": 1980
+    },
+    {
+      "epoch": 0.76619501395707,
+      "grad_norm": 0.25393593311309814,
+      "learning_rate": 1.3698746695458847e-06,
+      "loss": 0.7609,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7700452401578592,
+      "grad_norm": 0.22682346403598785,
+      "learning_rate": 1.3272765235521745e-06,
+      "loss": 0.7516,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7738954663586486,
+      "grad_norm": 0.22389163076877594,
+      "learning_rate": 1.2852496272942877e-06,
+      "loss": 0.7594,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7777456925594378,
+      "grad_norm": 0.22006818652153015,
+      "learning_rate": 1.2438005175707396e-06,
+      "loss": 0.7462,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7815959187602272,
+      "grad_norm": 0.22562545537948608,
+      "learning_rate": 1.2029356413120297e-06,
+      "loss": 0.7305,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7854461449610165,
+      "grad_norm": 0.22267329692840576,
+      "learning_rate": 1.1626613545778786e-06,
+      "loss": 0.7479,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7892963711618057,
+      "grad_norm": 0.22151796519756317,
+      "learning_rate": 1.1229839215686294e-06,
+      "loss": 0.7526,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7931465973625951,
+      "grad_norm": 0.21686570346355438,
+      "learning_rate": 1.0839095136509188e-06,
+      "loss": 0.7361,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7969968235633843,
+      "grad_norm": 0.241536483168602,
+      "learning_rate": 1.045444208397791e-06,
+      "loss": 0.7582,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8008470497641736,
+      "grad_norm": 0.2291547805070877,
+      "learning_rate": 1.007593988643402e-06,
+      "loss": 0.7397,
+      "step": 2080
+    },
+    {
+      "epoch": 0.804697275964963,
+      "grad_norm": 0.22582273185253143,
+      "learning_rate": 9.703647415524613e-07,
+      "loss": 0.7624,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8085475021657522,
+      "grad_norm": 0.22347891330718994,
+      "learning_rate": 9.337622577045485e-07,
+      "loss": 0.7501,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8123977283665416,
+      "grad_norm": 0.23543784022331238,
+      "learning_rate": 8.977922301934588e-07,
+      "loss": 0.7617,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8162479545673308,
+      "grad_norm": 0.23100546002388,
+      "learning_rate": 8.624602537417009e-07,
+      "loss": 0.7631,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8200981807681201,
+      "grad_norm": 0.24178721010684967,
+      "learning_rate": 8.277718238303123e-07,
+      "loss": 0.7514,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8239484069689095,
+      "grad_norm": 0.23035883903503418,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.753,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8277986331696987,
+      "grad_norm": 0.23890909552574158,
+      "learning_rate": 7.603470842324223e-07,
+      "loss": 0.7466,
+      "step": 2150
+    },
+    {
+      "epoch": 0.831648859370488,
+      "grad_norm": 0.22017061710357666,
+      "learning_rate": 7.276212616857575e-07,
+      "loss": 0.763,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8354990855712773,
+      "grad_norm": 0.2293674647808075,
+      "learning_rate": 6.955599583279837e-07,
+      "loss": 0.7605,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8393493117720666,
+      "grad_norm": 0.23517408967018127,
+      "learning_rate": 6.641681609246981e-07,
+      "loss": 0.7545,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8431995379728559,
+      "grad_norm": 0.2190673053264618,
+      "learning_rate": 6.334507521075711e-07,
+      "loss": 0.7592,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8470497641736452,
+      "grad_norm": 0.25151410698890686,
+      "learning_rate": 6.034125096149179e-07,
+      "loss": 0.7546,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8508999903744345,
+      "grad_norm": 0.2384350746870041,
+      "learning_rate": 5.740581055485722e-07,
+      "loss": 0.7279,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8547502165752238,
+      "grad_norm": 0.22007137537002563,
+      "learning_rate": 5.453921056471961e-07,
+      "loss": 0.7602,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8586004427760131,
+      "grad_norm": 0.23960334062576294,
+      "learning_rate": 5.174189685761288e-07,
+      "loss": 0.7547,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8624506689768023,
+      "grad_norm": 0.22983159124851227,
+      "learning_rate": 4.901430452338979e-07,
+      "loss": 0.7676,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8663008951775917,
+      "grad_norm": 0.23354536294937134,
+      "learning_rate": 4.6356857807548286e-07,
+      "loss": 0.7481,
+      "step": 2250
+    },
+    {
+      "epoch": 0.870151121378381,
+      "grad_norm": 0.2309393584728241,
+      "learning_rate": 4.376997004524541e-07,
+      "loss": 0.7384,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8740013475791703,
+      "grad_norm": 0.23504047095775604,
+      "learning_rate": 4.12540435970073e-07,
+      "loss": 0.7468,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8778515737799596,
+      "grad_norm": 0.22716949880123138,
+      "learning_rate": 3.880946978614703e-07,
+      "loss": 0.7612,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8817017999807488,
+      "grad_norm": 0.2293882966041565,
+      "learning_rate": 3.643662883789878e-07,
+      "loss": 0.7616,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8855520261815382,
+      "grad_norm": 0.2187264859676361,
+      "learning_rate": 3.41358898202781e-07,
+      "loss": 0.7408,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8894022523823275,
+      "grad_norm": 0.22620826959609985,
+      "learning_rate": 3.190761058667741e-07,
+      "loss": 0.751,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8932524785831167,
+      "grad_norm": 0.2234366238117218,
+      "learning_rate": 2.9752137720206463e-07,
+      "loss": 0.7567,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8971027047839061,
+      "grad_norm": 0.22655124962329865,
+      "learning_rate": 2.7669806479785067e-07,
+      "loss": 0.7631,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9009529309846953,
+      "grad_norm": 0.23965324461460114,
+      "learning_rate": 2.566094074799763e-07,
+      "loss": 0.756,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9048031571854847,
+      "grad_norm": 0.22608576714992523,
+      "learning_rate": 2.3725852980716557e-07,
+      "loss": 0.7384,
+      "step": 2350
+    },
+    {
+      "epoch": 0.908653383386274,
+      "grad_norm": 0.2230716049671173,
+      "learning_rate": 2.1864844158503983e-07,
+      "loss": 0.7512,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9125036095870632,
+      "grad_norm": 0.25821709632873535,
+      "learning_rate": 2.007820373979702e-07,
+      "loss": 0.7434,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9163538357878526,
+      "grad_norm": 0.22779373824596405,
+      "learning_rate": 1.8366209615886532e-07,
+      "loss": 0.7509,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9202040619886418,
+      "grad_norm": 0.22679728269577026,
+      "learning_rate": 1.6729128067693613e-07,
+      "loss": 0.7654,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9240542881894311,
+      "grad_norm": 0.24930810928344727,
+      "learning_rate": 1.5167213724353426e-07,
+      "loss": 0.7656,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9279045143902205,
+      "grad_norm": 0.22208335995674133,
+      "learning_rate": 1.3680709523610058e-07,
+      "loss": 0.7488,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9317547405910097,
+      "grad_norm": 0.22280748188495636,
+      "learning_rate": 1.2269846674030418e-07,
+      "loss": 0.7681,
+      "step": 2420
+    },
+    {
+      "epoch": 0.935604966791799,
+      "grad_norm": 0.22854870557785034,
+      "learning_rate": 1.0934844619042595e-07,
+      "loss": 0.767,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9394551929925883,
+      "grad_norm": 0.23945195972919464,
+      "learning_rate": 9.675911002803961e-08,
+      "loss": 0.7526,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9433054191933776,
+      "grad_norm": 0.2414969950914383,
+      "learning_rate": 8.493241637904381e-08,
+      "loss": 0.75,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9471556453941669,
+      "grad_norm": 0.23870866000652313,
+      "learning_rate": 7.387020474909689e-08,
+      "loss": 0.7706,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9510058715949562,
+      "grad_norm": 0.24387870728969574,
+      "learning_rate": 6.357419573750856e-08,
+      "loss": 0.7653,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9548560977957455,
+      "grad_norm": 0.23101165890693665,
+      "learning_rate": 5.4045990769613365e-08,
+      "loss": 0.7489,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9587063239965348,
+      "grad_norm": 0.27624988555908203,
+      "learning_rate": 4.5287071847694434e-08,
+      "loss": 0.7535,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9625565501973241,
+      "grad_norm": 0.22432252764701843,
+      "learning_rate": 3.7298801320467817e-08,
+      "loss": 0.7485,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9664067763981133,
+      "grad_norm": 0.23001599311828613,
+      "learning_rate": 3.0082421671192576e-08,
+      "loss": 0.7433,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9702570025989027,
+      "grad_norm": 0.23459212481975555,
+      "learning_rate": 2.3639055324410355e-08,
+      "loss": 0.7656,
+      "step": 2520
+    },
+    {
+      "epoch": 0.974107228799692,
+      "grad_norm": 0.22097252309322357,
+      "learning_rate": 1.7969704471369477e-08,
+      "loss": 0.7475,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9779574550004813,
+      "grad_norm": 0.2188444584608078,
+      "learning_rate": 1.3075250914142413e-08,
+      "loss": 0.7516,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9818076812012706,
+      "grad_norm": 0.22081074118614197,
+      "learning_rate": 8.956455928476049e-09,
+      "loss": 0.7587,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9856579074020598,
+      "grad_norm": 0.23428404331207275,
+      "learning_rate": 5.6139601453786275e-09,
+      "loss": 0.7511,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9895081336028492,
+      "grad_norm": 0.22915489971637726,
+      "learning_rate": 3.0482834514805602e-09,
+      "loss": 0.7407,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9933583598036385,
+      "grad_norm": 0.20707334578037262,
+      "learning_rate": 1.2598249081713366e-09,
+      "loss": 0.734,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9972085860044277,
+      "grad_norm": 0.24879834055900574,
+      "learning_rate": 2.488626895302826e-10,
+      "loss": 0.7674,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9999037443449803,
+      "step": 2597,
+      "total_flos": 8.19764490638721e+18,
+      "train_loss": 0.7811364003865591,
+      "train_runtime": 11806.1854,
+      "train_samples_per_second": 56.318,
+      "train_steps_per_second": 0.22
+    }
+  ],
+  "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": 8.19764490638721e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3/mc1_g14_224_2.5b/finetune/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4eb8071e432f5432851ccae987d59ac1399976273f5bc388f35ed55d84152856
+size 6136

continuous/qwen3/mc1_g14_224_2.5b/pretrain/config.json

@@ -0,0 +1,60 @@
+{
+  "_attn_implementation_autoset": true,
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 6144,
+  "max_position_embeddings": 40960,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1664,
+  "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-bigG-14-quickgelu",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "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/mc1/mc1_g14_224_2.5b.pt",
+  "vocab_size": 151936
+}

continuous/qwen3/mc1_g14_224_2.5b/pretrain/mm_projector.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d7c141c49e460a8fb8cb826881bcd496e1c17eadc365c772fe8d1fcf21ae0b4
+size 30427192

continuous/qwen3/mc1_g14_224_2.5b/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": 14.902135848999023,
+      "learning_rate": 0.00032142857142857147,
+      "loss": 9.511,
+      "step": 10
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 1.3558763265609741,
+      "learning_rate": 0.0006785714285714287,
+      "loss": 6.1307,
+      "step": 20
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 0.2772393524646759,
+      "learning_rate": 0.0009999907688176173,
+      "loss": 5.0257,
+      "step": 30
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 0.17750561237335205,
+      "learning_rate": 0.0009988834393115766,
+      "loss": 4.763,
+      "step": 40
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 0.14377839863300323,
+      "learning_rate": 0.0009959345572940772,
+      "loss": 4.663,
+      "step": 50
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 0.11337585747241974,
+      "learning_rate": 0.0009911550081155983,
+      "loss": 4.5948,
+      "step": 60
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.14086376130580902,
+      "learning_rate": 0.0009845624347566433,
+      "loss": 4.525,
+      "step": 70
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 0.13531823456287384,
+      "learning_rate": 0.0009761811727013547,
+      "loss": 4.4785,
+      "step": 80
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 0.14350971579551697,
+      "learning_rate": 0.0009660421601067666,
+      "loss": 4.4206,
+      "step": 90
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 0.1301591545343399,
+      "learning_rate": 0.0009541828235992926,
+      "loss": 4.3902,
+      "step": 100
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 0.1490766555070877,
+      "learning_rate": 0.0009406469401200151,
+      "loss": 4.3428,
+      "step": 110
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 0.1536417305469513,
+      "learning_rate": 0.0009254844753287492,
+      "loss": 4.3096,
+      "step": 120
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 0.189791738986969,
+      "learning_rate": 0.0009087513991633924,
+      "loss": 4.2671,
+      "step": 130
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 0.14721496403217316,
+      "learning_rate": 0.0008905094792353917,
+      "loss": 4.2352,
+      "step": 140
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 0.1607334464788437,
+      "learning_rate": 0.0008708260528239789,
+      "loss": 4.221,
+      "step": 150
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.13652758300304413,
+      "learning_rate": 0.0008497737783108237,
+      "loss": 4.1903,
+      "step": 160
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 0.1317805051803589,
+      "learning_rate": 0.0008274303669726426,
+      "loss": 4.1718,
+      "step": 170
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 0.16506560146808624,
+      "learning_rate": 0.0008038782961218137,
+      "loss": 4.1495,
+      "step": 180
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.1686951369047165,
+      "learning_rate": 0.000779204504653894,
+      "loss": 4.1526,
+      "step": 190
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 0.1232810840010643,
+      "learning_rate": 0.0007535000721258791,
+      "loss": 4.1284,
+      "step": 200
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 0.11644615232944489,
+      "learning_rate": 0.00072685988254984,
+      "loss": 4.0977,
+      "step": 210
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 0.14301501214504242,
+      "learning_rate": 0.0006993822741429906,
+      "loss": 4.0906,
+      "step": 220
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.13296952843666077,
+      "learning_rate": 0.0006711686763270818,
+      "loss": 4.0577,
+      "step": 230
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 0.10941395163536072,
+      "learning_rate": 0.0006423232353170831,
+      "loss": 4.0249,
+      "step": 240
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 0.10947754234075546,
+      "learning_rate": 0.0006129524296812335,
+      "loss": 3.9972,
+      "step": 250
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 0.10655362904071808,
+      "learning_rate": 0.0005831646772915651,
+      "loss": 3.9917,
+      "step": 260
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.14006240665912628,
+      "learning_rate": 0.0005530699351157809,
+      "loss": 3.9636,
+      "step": 270
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 0.13570858538150787,
+      "learning_rate": 0.0005227792933277943,
+      "loss": 3.9562,
+      "step": 280
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 0.12570670247077942,
+      "learning_rate": 0.0004924045652352092,
+      "loss": 3.9441,
+      "step": 290
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 0.1151735708117485,
+      "learning_rate": 0.0004620578745374633,
+      "loss": 3.9384,
+      "step": 300
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 0.1195785254240036,
+      "learning_rate": 0.0004318512414382059,
+      "loss": 3.9464,
+      "step": 310
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 0.12299112975597382,
+      "learning_rate": 0.0004018961691397148,
+      "loss": 3.9352,
+      "step": 320
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 0.13300324976444244,
+      "learning_rate": 0.0003723032322457458,
+      "loss": 3.9202,
+      "step": 330
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.13058723509311676,
+      "learning_rate": 0.0003431816685921625,
+      "loss": 3.9164,
+      "step": 340
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 0.12159353494644165,
+      "learning_rate": 0.0003146389760120416,
+      "loss": 3.9184,
+      "step": 350
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 0.11880973726511002,
+      "learning_rate": 0.00028678051552373485,
+      "loss": 3.9214,
+      "step": 360
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 0.1039758026599884,
+      "learning_rate": 0.0002597091224066581,
+      "loss": 3.906,
+      "step": 370
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 0.11654747277498245,
+      "learning_rate": 0.0002335247266004629,
+      "loss": 3.9192,
+      "step": 380
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 0.11479203402996063,
+      "learning_rate": 0.000208323983828828,
+      "loss": 3.9012,
+      "step": 390
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.10277805477380753,
+      "learning_rate": 0.0001841999188095224,
+      "loss": 3.8993,
+      "step": 400
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 0.10028940439224243,
+      "learning_rate": 0.00016124158186777676,
+      "loss": 3.9036,
+      "step": 410
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 0.10197778046131134,
+      "learning_rate": 0.0001395337202205211,
+      "loss": 3.8971,
+      "step": 420
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 0.10389617830514908,
+      "learning_rate": 0.00011915646514489292,
+      "loss": 3.8987,
+      "step": 430
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 0.1071009710431099,
+      "learning_rate": 0.00010018503618578818,
+      "loss": 3.9047,
+      "step": 440
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.09905208647251129,
+      "learning_rate": 8.268946349432583e-05,
+      "loss": 3.8934,
+      "step": 450
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 0.10149159282445908,
+      "learning_rate": 6.673432932217117e-05,
+      "loss": 3.8881,
+      "step": 460
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.0972200408577919,
+      "learning_rate": 5.2378529625954684e-05,
+      "loss": 3.8776,
+      "step": 470
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.10216016322374344,
+      "learning_rate": 3.9675056661785556e-05,
+      "loss": 3.8818,
+      "step": 480
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.0967029556632042,
+      "learning_rate": 2.8670803372379538e-05,
+      "loss": 3.8877,
+      "step": 490
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 0.10418641567230225,
+      "learning_rate": 1.9406390288876586e-05,
+      "loss": 3.8954,
+      "step": 500
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 0.10167988389730453,
+      "learning_rate": 1.191601558631461e-05,
+      "loss": 3.8906,
+      "step": 510
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 0.0986390933394432,
+      "learning_rate": 6.2273288462559305e-06,
+      "loss": 3.8842,
+      "step": 520
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 0.09852931648492813,
+      "learning_rate": 2.361328992552314e-06,
+      "loss": 3.8873,
+      "step": 530
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 0.10029914230108261,
+      "learning_rate": 3.322867770044202e-07,
+      "loss": 3.8852,
+      "step": 540
+    },
+    {
+      "epoch": 1.0,
+      "step": 545,
+      "total_flos": 6.486578758701548e+17,
+      "train_loss": 4.226680549131621,
+      "train_runtime": 5724.576,
+      "train_samples_per_second": 97.497,
+      "train_steps_per_second": 0.095
+    }
+  ],
+  "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": 6.486578758701548e+17,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}

continuous/qwen3/mc1_h14_224_2.5b/finetune/added_tokens.json

@@ -0,0 +1,28 @@
+{
+  "</think>": 151668,
+  "</tool_call>": 151658,
+  "</tool_response>": 151666,
+  "<think>": 151667,
+  "<tool_call>": 151657,
+  "<tool_response>": 151665,
+  "<|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/qwen3/mc1_h14_224_2.5b/finetune/config.json

@@ -0,0 +1,59 @@
+{
+  "add_faster_video": false,
+  "add_time_instruction": false,
+  "architectures": [
+    "LlavaQwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": false,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "image_aspect_ratio": "square",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": null,
+  "image_split_resolution": null,
+  "initializer_range": 0.02,
+  "intermediate_size": 6144,
+  "max_position_embeddings": 40960,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1280,
+  "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-H-14",
+  "mm_vision_tower_lr": null,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "pos_skipping_range": 4096,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "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/mc1/mc1_h14_224_2.5b.pt",
+  "vocab_size": 151936
+}

continuous/qwen3/mc1_h14_224_2.5b/finetune/generation_config.json

@@ -0,0 +1,13 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "temperature": 0.6,
+  "top_k": 20,
+  "top_p": 0.95,
+  "transformers_version": "4.51.3"
+}