Upload trained Qwen3-VL model

.gitattributes

@@ -33,3 +33,13 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+checkpoint-1000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-1500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-2000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-2500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-3000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-3500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-4000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-4320/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+final_model/tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,58 @@
+---
+base_model: Qwen/Qwen3-VL-2B-Instruct
+library_name: transformers
+model_name: Qwen3_checkpoint
+tags:
+- generated_from_trainer
+- sft
+- trl
+licence: license
+---
+
+# Model Card for Qwen3_checkpoint
+
+This model is a fine-tuned version of [Qwen/Qwen3-VL-2B-Instruct](https://huggingface.co/Qwen/Qwen3-VL-2B-Instruct).
+It has been trained using [TRL](https://github.com/huggingface/trl).
+
+## Quick start
+
+```python
+from transformers import pipeline
+
+question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
+generator = pipeline("text-generation", model="None", device="cuda")
+output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
+print(output["generated_text"])
+```
+
+## Training procedure
+
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/configint/qwen3-vl-sft/runs/uuci5pet) 
+
+
+This model was trained with SFT.
+
+### Framework versions
+
+- TRL: 0.23.1
+- Transformers: 4.57.0
+- Pytorch: 2.8.0
+- Datasets: 4.2.0
+- Tokenizers: 0.22.1
+
+## Citations
+
+
+
+Cite TRL as:
+    
+```bibtex
+@misc{vonwerra2022trl,
+	title        = {{TRL: Transformer Reinforcement Learning}},
+	author       = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallou{\'e}dec},
+	year         = 2020,
+	journal      = {GitHub repository},
+	publisher    = {GitHub},
+	howpublished = {\url{https://github.com/huggingface/trl}}
+}
+```

accelerate_config.yaml

@@ -0,0 +1,13 @@
+compute_environment: LOCAL_MACHINE
+debug: false
+distributed_type: MULTI_GPU
+downcast_bf16: 'no'
+gpu_ids: all
+main_training_function: main
+mixed_precision: bf16
+num_machines: 1
+num_processes: 8
+tpu_env: []
+tpu_use_cluster: false
+tpu_use_sudo: false
+use_cpu: false

checkpoint-1000/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
+}

checkpoint-1000/chat_template.jinja

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

checkpoint-1000/config.json

@@ -0,0 +1,67 @@
+{
+  "architectures": [
+    "Qwen3VLForConditionalGeneration"
+  ],
+  "dtype": "bfloat16",
+  "eos_token_id": 151645,
+  "image_token_id": 151655,
+  "model_type": "qwen3_vl",
+  "pad_token_id": 151643,
+  "text_config": {
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "bos_token_id": 151643,
+    "dtype": "bfloat16",
+    "eos_token_id": 151645,
+    "head_dim": 128,
+    "hidden_act": "silu",
+    "hidden_size": 2048,
+    "initializer_range": 0.02,
+    "intermediate_size": 6144,
+    "max_position_embeddings": 262144,
+    "model_type": "qwen3_vl_text",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 28,
+    "num_key_value_heads": 8,
+    "rms_norm_eps": 1e-06,
+    "rope_scaling": {
+      "mrope_interleaved": true,
+      "mrope_section": [
+        24,
+        20,
+        20
+      ],
+      "rope_type": "default"
+    },
+    "rope_theta": 5000000,
+    "tie_word_embeddings": true,
+    "use_cache": true,
+    "vocab_size": 151936
+  },
+  "tie_word_embeddings": true,
+  "transformers_version": "4.57.0",
+  "video_token_id": 151656,
+  "vision_config": {
+    "deepstack_visual_indexes": [
+      5,
+      11,
+      17
+    ],
+    "depth": 24,
+    "dtype": "bfloat16",
+    "hidden_act": "gelu_pytorch_tanh",
+    "hidden_size": 1024,
+    "in_channels": 3,
+    "initializer_range": 0.02,
+    "intermediate_size": 4096,
+    "model_type": "qwen3_vl",
+    "num_heads": 16,
+    "num_position_embeddings": 2304,
+    "out_hidden_size": 2048,
+    "patch_size": 16,
+    "spatial_merge_size": 2,
+    "temporal_patch_size": 2
+  },
+  "vision_end_token_id": 151653,
+  "vision_start_token_id": 151652
+}

checkpoint-1000/generation_config.json

@@ -0,0 +1,13 @@
+{
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.57.0"
+}

checkpoint-1000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60bee21b510c724c2ccb9b218eba003ddf592ac7a8c0b4c38caa23f65a14cedc
+size 4255140312

checkpoint-1000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e706a37013c99bdf0096b6a9fb1994b9400923833ecd156e8d9a4d4cf98adc50
+size 8510679398

checkpoint-1000/preprocessor_config.json

@@ -0,0 +1,39 @@
+{
+  "crop_size": null,
+  "data_format": "channels_first",
+  "default_to_square": true,
+  "device": null,
+  "disable_grouping": null,
+  "do_center_crop": null,
+  "do_convert_rgb": true,
+  "do_normalize": true,
+  "do_pad": null,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_processor_type": "Qwen2VLImageProcessorFast",
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "input_data_format": null,
+  "max_pixels": null,
+  "merge_size": 2,
+  "min_pixels": null,
+  "pad_size": null,
+  "patch_size": 16,
+  "processor_class": "Qwen3VLProcessor",
+  "resample": 3,
+  "rescale_factor": 0.00392156862745098,
+  "return_tensors": null,
+  "size": {
+    "longest_edge": 16777216,
+    "shortest_edge": 65536
+  },
+  "temporal_patch_size": 2
+}

checkpoint-1000/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c8e5f48b40f283c2be57ffeca20c84e74d5bad51d76da17d127991b78da5289d
+size 16389

checkpoint-1000/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c3696205cbf4679a7f2faf48351e924c79ada4cfd81a956e03e4319cf3d05b5
+size 16389

checkpoint-1000/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:403766b1961af9b11405f33af2b272a2d6287cefdee9eb7e26e1dc71e3ed3707
+size 16389

checkpoint-1000/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54f6e3e08fea79c105dbd92657538a3cb15e89872360a3c3072b758835ca5ebc
+size 16389

checkpoint-1000/rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0203808a2df05d235c57b83a0a087bae98ef6e47f06d685a0bddaa460a16c910
+size 16389

checkpoint-1000/rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6b492b10406611c53f05fd065d6968c7e5e3797fd0008fca46cad009936c67e
+size 16389

checkpoint-1000/rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d44646ad1e32ad260756d9c2e6927a7a101af3a5c9265d8db8a1d4def7d446cb
+size 16389

checkpoint-1000/rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b83b78288d5cea0b10985a2d48c3c0b12d2ff3f99104720c1b3281409b3b752
+size 16389

checkpoint-1000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4cc81b6515b3846e01beabc95700d2ae76e21c13cebaaa8b0b7741d224738863
+size 1465

checkpoint-1000/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
+  }
+}

checkpoint-1000/tokenizer.json

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

checkpoint-1000/tokenizer_config.json

@@ -0,0 +1,240 @@
+{
+  "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,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 262144,
+  "pad_token": "<|endoftext|>",
+  "processor_class": "Qwen3VLProcessor",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

checkpoint-1000/trainer_state.json

@@ -0,0 +1,1078 @@
+{
+  "best_global_step": 1000,
+  "best_metric": 0.0312611423432827,
+  "best_model_checkpoint": "/workspace/Qwen3_checkpoint/checkpoint-1000",
+  "epoch": 0.6945047313134821,
+  "eval_steps": 250,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "entropy": 1.1519597232341767,
+      "epoch": 0.00694504731313482,
+      "grad_norm": 540.0,
+      "learning_rate": 6.249999999999999e-07,
+      "loss": 22.6265,
+      "mean_token_accuracy": 0.07503694812767207,
+      "num_tokens": 1640499.0,
+      "step": 10
+    },
+    {
+      "entropy": 1.1664627104997636,
+      "epoch": 0.01389009462626964,
+      "grad_norm": 466.0,
+      "learning_rate": 1.3194444444444444e-06,
+      "loss": 22.1144,
+      "mean_token_accuracy": 0.08692597970366478,
+      "num_tokens": 3288480.0,
+      "step": 20
+    },
+    {
+      "entropy": 1.239394597709179,
+      "epoch": 0.020835141939404462,
+      "grad_norm": 462.0,
+      "learning_rate": 2.0138888888888893e-06,
+      "loss": 21.7485,
+      "mean_token_accuracy": 0.07649312568828463,
+      "num_tokens": 4958023.0,
+      "step": 30
+    },
+    {
+      "entropy": 1.4183061853051186,
+      "epoch": 0.02778018925253928,
+      "grad_norm": 436.0,
+      "learning_rate": 2.7083333333333334e-06,
+      "loss": 20.1258,
+      "mean_token_accuracy": 0.07965942500159144,
+      "num_tokens": 6627704.0,
+      "step": 40
+    },
+    {
+      "entropy": 1.9329519897699357,
+      "epoch": 0.034725236565674104,
+      "grad_norm": 378.0,
+      "learning_rate": 3.402777777777778e-06,
+      "loss": 16.576,
+      "mean_token_accuracy": 0.07870438289828599,
+      "num_tokens": 8340232.0,
+      "step": 50
+    },
+    {
+      "entropy": 4.602115747332573,
+      "epoch": 0.041670283878808924,
+      "grad_norm": 184.0,
+      "learning_rate": 4.097222222222222e-06,
+      "loss": 12.1758,
+      "mean_token_accuracy": 0.07569821006618441,
+      "num_tokens": 10025658.0,
+      "step": 60
+    },
+    {
+      "entropy": 7.328701210021973,
+      "epoch": 0.04861533119194374,
+      "grad_norm": 114.5,
+      "learning_rate": 4.791666666666667e-06,
+      "loss": 9.8369,
+      "mean_token_accuracy": 0.08327204268425703,
+      "num_tokens": 11741238.0,
+      "step": 70
+    },
+    {
+      "entropy": 7.635936880111695,
+      "epoch": 0.05556037850507856,
+      "grad_norm": 113.5,
+      "learning_rate": 5.486111111111111e-06,
+      "loss": 8.3162,
+      "mean_token_accuracy": 0.0876343248412013,
+      "num_tokens": 13404022.0,
+      "step": 80
+    },
+    {
+      "entropy": 7.46761246919632,
+      "epoch": 0.06250542581821339,
+      "grad_norm": 123.5,
+      "learning_rate": 6.180555555555556e-06,
+      "loss": 7.63,
+      "mean_token_accuracy": 0.08659355682320893,
+      "num_tokens": 15112991.0,
+      "step": 90
+    },
+    {
+      "entropy": 7.1816710293293,
+      "epoch": 0.06945047313134821,
+      "grad_norm": 148.0,
+      "learning_rate": 6.875e-06,
+      "loss": 7.0576,
+      "mean_token_accuracy": 0.08753767567686736,
+      "num_tokens": 16756210.0,
+      "step": 100
+    },
+    {
+      "entropy": 7.2137157917022705,
+      "epoch": 0.07639552044448303,
+      "grad_norm": 169.0,
+      "learning_rate": 7.569444444444445e-06,
+      "loss": 6.6052,
+      "mean_token_accuracy": 0.08361622425727547,
+      "num_tokens": 18470889.0,
+      "step": 110
+    },
+    {
+      "entropy": 7.177587866783142,
+      "epoch": 0.08334056775761785,
+      "grad_norm": 156.0,
+      "learning_rate": 8.263888888888888e-06,
+      "loss": 6.0531,
+      "mean_token_accuracy": 0.08918410916812718,
+      "num_tokens": 20159140.0,
+      "step": 120
+    },
+    {
+      "entropy": 7.2117412209510805,
+      "epoch": 0.09028561507075267,
+      "grad_norm": 160.0,
+      "learning_rate": 8.958333333333334e-06,
+      "loss": 5.4473,
+      "mean_token_accuracy": 0.08560893354006112,
+      "num_tokens": 21857491.0,
+      "step": 130
+    },
+    {
+      "entropy": 7.211903780698776,
+      "epoch": 0.09723066238388749,
+      "grad_norm": 278.0,
+      "learning_rate": 9.652777777777777e-06,
+      "loss": 4.788,
+      "mean_token_accuracy": 0.127999250870198,
+      "num_tokens": 23545940.0,
+      "step": 140
+    },
+    {
+      "entropy": 7.21578094959259,
+      "epoch": 0.1041757096970223,
+      "grad_norm": 163.0,
+      "learning_rate": 1.0347222222222223e-05,
+      "loss": 4.1808,
+      "mean_token_accuracy": 0.5043030813336372,
+      "num_tokens": 25200650.0,
+      "step": 150
+    },
+    {
+      "entropy": 7.225958931446075,
+      "epoch": 0.11112075701015713,
+      "grad_norm": 146.0,
+      "learning_rate": 1.1041666666666668e-05,
+      "loss": 3.4839,
+      "mean_token_accuracy": 0.8470314003527164,
+      "num_tokens": 26879379.0,
+      "step": 160
+    },
+    {
+      "entropy": 7.03481342792511,
+      "epoch": 0.11806580432329196,
+      "grad_norm": 160.0,
+      "learning_rate": 1.1736111111111112e-05,
+      "loss": 2.6644,
+      "mean_token_accuracy": 0.9146032109856606,
+      "num_tokens": 28559842.0,
+      "step": 170
+    },
+    {
+      "entropy": 6.437038451433182,
+      "epoch": 0.12501085163642678,
+      "grad_norm": 135.0,
+      "learning_rate": 1.2430555555555557e-05,
+      "loss": 1.8016,
+      "mean_token_accuracy": 0.9310876823961735,
+      "num_tokens": 30223208.0,
+      "step": 180
+    },
+    {
+      "entropy": 4.774793094396591,
+      "epoch": 0.13195589894956158,
+      "grad_norm": 72.0,
+      "learning_rate": 1.3125e-05,
+      "loss": 1.0024,
+      "mean_token_accuracy": 0.9337419532239437,
+      "num_tokens": 31915078.0,
+      "step": 190
+    },
+    {
+      "entropy": 2.502578377723694,
+      "epoch": 0.13890094626269642,
+      "grad_norm": 18.5,
+      "learning_rate": 1.3819444444444444e-05,
+      "loss": 0.5206,
+      "mean_token_accuracy": 0.9370624087750912,
+      "num_tokens": 33627632.0,
+      "step": 200
+    },
+    {
+      "entropy": 1.1439476042985917,
+      "epoch": 0.14584599357583122,
+      "grad_norm": 6.5,
+      "learning_rate": 1.451388888888889e-05,
+      "loss": 0.314,
+      "mean_token_accuracy": 0.9395098507404327,
+      "num_tokens": 35335117.0,
+      "step": 210
+    },
+    {
+      "entropy": 0.625144999101758,
+      "epoch": 0.15279104088896606,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.5208333333333333e-05,
+      "loss": 0.2057,
+      "mean_token_accuracy": 0.9517911829054355,
+      "num_tokens": 37011186.0,
+      "step": 220
+    },
+    {
+      "entropy": 0.34807259757071735,
+      "epoch": 0.1597360882021009,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.590277777777778e-05,
+      "loss": 0.1098,
+      "mean_token_accuracy": 0.9706214666366577,
+      "num_tokens": 38673894.0,
+      "step": 230
+    },
+    {
+      "entropy": 0.22573072295635938,
+      "epoch": 0.1666811355152357,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.6597222222222222e-05,
+      "loss": 0.0648,
+      "mean_token_accuracy": 0.9812032662332058,
+      "num_tokens": 40360862.0,
+      "step": 240
+    },
+    {
+      "entropy": 0.19475232288241387,
+      "epoch": 0.17362618282837053,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7291666666666666e-05,
+      "loss": 0.0472,
+      "mean_token_accuracy": 0.9857565425336361,
+      "num_tokens": 42047917.0,
+      "step": 250
+    },
+    {
+      "epoch": 0.17362618282837053,
+      "eval_entropy": 0.22951670004171507,
+      "eval_loss": 0.05014730989933014,
+      "eval_mean_token_accuracy": 0.985958794625192,
+      "eval_num_tokens": 42047917.0,
+      "eval_runtime": 81.9972,
+      "eval_samples_per_second": 121.468,
+      "eval_steps_per_second": 7.598,
+      "step": 250
+    },
+    {
+      "entropy": 0.1923621967434883,
+      "epoch": 0.18057123014150533,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7986111111111113e-05,
+      "loss": 0.0397,
+      "mean_token_accuracy": 0.9877771027386189,
+      "num_tokens": 43714135.0,
+      "step": 260
+    },
+    {
+      "entropy": 0.18295048763975502,
+      "epoch": 0.18751627745464017,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8680555555555556e-05,
+      "loss": 0.0395,
+      "mean_token_accuracy": 0.9877975657582283,
+      "num_tokens": 45420354.0,
+      "step": 270
+    },
+    {
+      "entropy": 0.18508334569633006,
+      "epoch": 0.19446132476777497,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9375e-05,
+      "loss": 0.0365,
+      "mean_token_accuracy": 0.9889688648283481,
+      "num_tokens": 47094373.0,
+      "step": 280
+    },
+    {
+      "entropy": 0.17800553049892187,
+      "epoch": 0.2014063720809098,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 2.0069444444444447e-05,
+      "loss": 0.0334,
+      "mean_token_accuracy": 0.9898182846605778,
+      "num_tokens": 48737014.0,
+      "step": 290
+    },
+    {
+      "entropy": 0.17542087566107512,
+      "epoch": 0.2083514193940446,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.076388888888889e-05,
+      "loss": 0.0332,
+      "mean_token_accuracy": 0.9898306861519813,
+      "num_tokens": 50411139.0,
+      "step": 300
+    },
+    {
+      "entropy": 0.17356558851897716,
+      "epoch": 0.21529646670717945,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.1458333333333334e-05,
+      "loss": 0.0329,
+      "mean_token_accuracy": 0.9897275142371654,
+      "num_tokens": 52101776.0,
+      "step": 310
+    },
+    {
+      "entropy": 0.18331474922597407,
+      "epoch": 0.22224151402031425,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.2152777777777778e-05,
+      "loss": 0.0323,
+      "mean_token_accuracy": 0.9898556746542454,
+      "num_tokens": 53761706.0,
+      "step": 320
+    },
+    {
+      "entropy": 0.16903471825644373,
+      "epoch": 0.22918656133344908,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.284722222222222e-05,
+      "loss": 0.0303,
+      "mean_token_accuracy": 0.9905840046703815,
+      "num_tokens": 55510137.0,
+      "step": 330
+    },
+    {
+      "entropy": 0.1726135764271021,
+      "epoch": 0.23613160864658392,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 2.354166666666667e-05,
+      "loss": 0.0308,
+      "mean_token_accuracy": 0.9905642926692962,
+      "num_tokens": 57209080.0,
+      "step": 340
+    },
+    {
+      "entropy": 0.17743745557963847,
+      "epoch": 0.24307665595971872,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.4236111111111112e-05,
+      "loss": 0.0304,
+      "mean_token_accuracy": 0.9905851900577545,
+      "num_tokens": 58867303.0,
+      "step": 350
+    },
+    {
+      "entropy": 0.18280480420216919,
+      "epoch": 0.25002170327285356,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.4930555555555556e-05,
+      "loss": 0.0315,
+      "mean_token_accuracy": 0.9901946425437927,
+      "num_tokens": 60510859.0,
+      "step": 360
+    },
+    {
+      "entropy": 0.1761154976673424,
+      "epoch": 0.2569667505859884,
+      "grad_norm": 0.5,
+      "learning_rate": 2.5625e-05,
+      "loss": 0.0298,
+      "mean_token_accuracy": 0.9909366421401501,
+      "num_tokens": 62171999.0,
+      "step": 370
+    },
+    {
+      "entropy": 0.1671930664218962,
+      "epoch": 0.26391179789912317,
+      "grad_norm": 0.31640625,
+      "learning_rate": 2.6319444444444443e-05,
+      "loss": 0.0284,
+      "mean_token_accuracy": 0.9912083707749844,
+      "num_tokens": 63906245.0,
+      "step": 380
+    },
+    {
+      "entropy": 0.1754794458858669,
+      "epoch": 0.270856845212258,
+      "grad_norm": 0.240234375,
+      "learning_rate": 2.7013888888888887e-05,
+      "loss": 0.0303,
+      "mean_token_accuracy": 0.9904192194342614,
+      "num_tokens": 65589663.0,
+      "step": 390
+    },
+    {
+      "entropy": 0.1771222472190857,
+      "epoch": 0.27780189252539284,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.7708333333333337e-05,
+      "loss": 0.0299,
+      "mean_token_accuracy": 0.9907479107379913,
+      "num_tokens": 67241611.0,
+      "step": 400
+    },
+    {
+      "entropy": 0.17829248458147048,
+      "epoch": 0.28474693983852767,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.840277777777778e-05,
+      "loss": 0.0297,
+      "mean_token_accuracy": 0.9906024843454361,
+      "num_tokens": 68889437.0,
+      "step": 410
+    },
+    {
+      "entropy": 0.17638491541147233,
+      "epoch": 0.29169198715166245,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.9097222222222224e-05,
+      "loss": 0.0284,
+      "mean_token_accuracy": 0.9913305290043354,
+      "num_tokens": 70618362.0,
+      "step": 420
+    },
+    {
+      "entropy": 0.18182099591940643,
+      "epoch": 0.2986370344647973,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.9791666666666668e-05,
+      "loss": 0.0285,
+      "mean_token_accuracy": 0.9912928201258182,
+      "num_tokens": 72313577.0,
+      "step": 430
+    },
+    {
+      "entropy": 0.18310831850394607,
+      "epoch": 0.3055820817779321,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.9999760059577934e-05,
+      "loss": 0.029,
+      "mean_token_accuracy": 0.9910079762339592,
+      "num_tokens": 73988194.0,
+      "step": 440
+    },
+    {
+      "entropy": 0.17772579193115234,
+      "epoch": 0.31252712909106695,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.9998584859663212e-05,
+      "loss": 0.0284,
+      "mean_token_accuracy": 0.9910575695335865,
+      "num_tokens": 75663076.0,
+      "step": 450
+    },
+    {
+      "entropy": 0.17988401055335998,
+      "epoch": 0.3194721764042018,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.9996430406198636e-05,
+      "loss": 0.028,
+      "mean_token_accuracy": 0.9914611585438251,
+      "num_tokens": 77335287.0,
+      "step": 460
+    },
+    {
+      "entropy": 0.173457539267838,
+      "epoch": 0.32641722371733656,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.9993296839847897e-05,
+      "loss": 0.0287,
+      "mean_token_accuracy": 0.9910615682601929,
+      "num_tokens": 79052668.0,
+      "step": 470
+    },
+    {
+      "entropy": 0.16980679212138056,
+      "epoch": 0.3333622710304714,
+      "grad_norm": 0.234375,
+      "learning_rate": 2.9989184365200695e-05,
+      "loss": 0.0266,
+      "mean_token_accuracy": 0.9917186461389065,
+      "num_tokens": 80733082.0,
+      "step": 480
+    },
+    {
+      "entropy": 0.16687583839520811,
+      "epoch": 0.3403073183436062,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.998409325075938e-05,
+      "loss": 0.0274,
+      "mean_token_accuracy": 0.9914394572377205,
+      "num_tokens": 82456201.0,
+      "step": 490
+    },
+    {
+      "entropy": 0.16694002971053123,
+      "epoch": 0.34725236565674106,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.997802382892144e-05,
+      "loss": 0.0271,
+      "mean_token_accuracy": 0.9916702680289745,
+      "num_tokens": 84174465.0,
+      "step": 500
+    },
+    {
+      "epoch": 0.34725236565674106,
+      "eval_entropy": 0.2104673289826938,
+      "eval_loss": 0.03412040323019028,
+      "eval_mean_token_accuracy": 0.9906957189688522,
+      "eval_num_tokens": 84174465.0,
+      "eval_runtime": 81.2363,
+      "eval_samples_per_second": 122.605,
+      "eval_steps_per_second": 7.669,
+      "step": 500
+    },
+    {
+      "entropy": 0.16945981308817865,
+      "epoch": 0.35419741296987584,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.997097649595779e-05,
+      "loss": 0.0281,
+      "mean_token_accuracy": 0.9912286765873433,
+      "num_tokens": 85878591.0,
+      "step": 510
+    },
+    {
+      "entropy": 0.16488271821290254,
+      "epoch": 0.36114246028301067,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.9962951711986897e-05,
+      "loss": 0.0277,
+      "mean_token_accuracy": 0.9914339393377304,
+      "num_tokens": 87573271.0,
+      "step": 520
+    },
+    {
+      "entropy": 0.17190398359671236,
+      "epoch": 0.3680875075961455,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.9953950000944724e-05,
+      "loss": 0.0291,
+      "mean_token_accuracy": 0.9910320229828358,
+      "num_tokens": 89226804.0,
+      "step": 530
+    },
+    {
+      "entropy": 0.16401167614385487,
+      "epoch": 0.37503255490928034,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.994397195055056e-05,
+      "loss": 0.0268,
+      "mean_token_accuracy": 0.9915202751755714,
+      "num_tokens": 90931627.0,
+      "step": 540
+    },
+    {
+      "entropy": 0.1601521703414619,
+      "epoch": 0.3819776022224151,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.9933018212268607e-05,
+      "loss": 0.0256,
+      "mean_token_accuracy": 0.9919740296900272,
+      "num_tokens": 92669023.0,
+      "step": 550
+    },
+    {
+      "entropy": 0.1690834665670991,
+      "epoch": 0.38892264953554995,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.9921089501265493e-05,
+      "loss": 0.0272,
+      "mean_token_accuracy": 0.9918020509183407,
+      "num_tokens": 94351832.0,
+      "step": 560
+    },
+    {
+      "entropy": 0.16791587956249715,
+      "epoch": 0.3958676968486848,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.990818659636352e-05,
+      "loss": 0.0269,
+      "mean_token_accuracy": 0.9915785849094391,
+      "num_tokens": 96026408.0,
+      "step": 570
+    },
+    {
+      "entropy": 0.1673963068984449,
+      "epoch": 0.4028127441618196,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.989431033998986e-05,
+      "loss": 0.0263,
+      "mean_token_accuracy": 0.9915929511189461,
+      "num_tokens": 97667765.0,
+      "step": 580
+    },
+    {
+      "entropy": 0.1658677328377962,
+      "epoch": 0.40975779147495445,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.987946163812155e-05,
+      "loss": 0.0263,
+      "mean_token_accuracy": 0.9918196611106396,
+      "num_tokens": 99392767.0,
+      "step": 590
+    },
+    {
+      "entropy": 0.16773838149383663,
+      "epoch": 0.4167028387880892,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.9863641460226304e-05,
+      "loss": 0.0269,
+      "mean_token_accuracy": 0.9917811565101147,
+      "num_tokens": 101073637.0,
+      "step": 600
+    },
+    {
+      "entropy": 0.1649629818275571,
+      "epoch": 0.42364788610122406,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.984685083919928e-05,
+      "loss": 0.0274,
+      "mean_token_accuracy": 0.9912688471376896,
+      "num_tokens": 102740766.0,
+      "step": 610
+    },
+    {
+      "entropy": 0.15838869903236627,
+      "epoch": 0.4305929334143589,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.9829090871295564e-05,
+      "loss": 0.0251,
+      "mean_token_accuracy": 0.9922874167561531,
+      "num_tokens": 104488780.0,
+      "step": 620
+    },
+    {
+      "entropy": 0.17186203664168714,
+      "epoch": 0.4375379807274937,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.981036271605867e-05,
+      "loss": 0.0263,
+      "mean_token_accuracy": 0.9918930843472481,
+      "num_tokens": 106195449.0,
+      "step": 630
+    },
+    {
+      "entropy": 0.1648147610016167,
+      "epoch": 0.4444830280406285,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.979066759624479e-05,
+      "loss": 0.0252,
+      "mean_token_accuracy": 0.992139758169651,
+      "num_tokens": 107902411.0,
+      "step": 640
+    },
+    {
+      "entropy": 0.16121056992560626,
+      "epoch": 0.45142807535376334,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 2.9770006797742963e-05,
+      "loss": 0.025,
+      "mean_token_accuracy": 0.9919275738298893,
+      "num_tokens": 109608103.0,
+      "step": 650
+    },
+    {
+      "entropy": 0.17468737829476594,
+      "epoch": 0.45837312266689817,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.974838166949114e-05,
+      "loss": 0.0272,
+      "mean_token_accuracy": 0.9917291469871998,
+      "num_tokens": 111293326.0,
+      "step": 660
+    },
+    {
+      "entropy": 0.1678467271849513,
+      "epoch": 0.465318169980033,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.9725793623388097e-05,
+      "loss": 0.0255,
+      "mean_token_accuracy": 0.9922867320477963,
+      "num_tokens": 113010182.0,
+      "step": 670
+    },
+    {
+      "entropy": 0.16656696163117884,
+      "epoch": 0.47226321729316784,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.9702244134201263e-05,
+      "loss": 0.0258,
+      "mean_token_accuracy": 0.9920145235955715,
+      "num_tokens": 114702373.0,
+      "step": 680
+    },
+    {
+      "entropy": 0.1741402018815279,
+      "epoch": 0.4792082646063026,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.9677734739470407e-05,
+      "loss": 0.0282,
+      "mean_token_accuracy": 0.9915049582719803,
+      "num_tokens": 116371688.0,
+      "step": 690
+    },
+    {
+      "entropy": 0.17028953870758415,
+      "epoch": 0.48615331191943745,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.9652267039407297e-05,
+      "loss": 0.0273,
+      "mean_token_accuracy": 0.9915660627186298,
+      "num_tokens": 118057500.0,
+      "step": 700
+    },
+    {
+      "entropy": 0.17608886091038584,
+      "epoch": 0.4930983592325723,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.962584269679117e-05,
+      "loss": 0.0274,
+      "mean_token_accuracy": 0.9914011798799038,
+      "num_tokens": 119691073.0,
+      "step": 710
+    },
+    {
+      "entropy": 0.16367722414433955,
+      "epoch": 0.5000434065457071,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.959846343686022e-05,
+      "loss": 0.0251,
+      "mean_token_accuracy": 0.9922043934464455,
+      "num_tokens": 121405698.0,
+      "step": 720
+    },
+    {
+      "entropy": 0.16646005054935814,
+      "epoch": 0.506988453858842,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.957013104719892e-05,
+      "loss": 0.0257,
+      "mean_token_accuracy": 0.9920815974473953,
+      "num_tokens": 123140318.0,
+      "step": 730
+    },
+    {
+      "entropy": 0.16303647235035895,
+      "epoch": 0.5139335011719768,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.9540847377621337e-05,
+      "loss": 0.0264,
+      "mean_token_accuracy": 0.991966935992241,
+      "num_tokens": 124857523.0,
+      "step": 740
+    },
+    {
+      "entropy": 0.16988610094413162,
+      "epoch": 0.5208785484851115,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.9510614340050325e-05,
+      "loss": 0.0256,
+      "mean_token_accuracy": 0.9918926700949668,
+      "num_tokens": 126537184.0,
+      "step": 750
+    },
+    {
+      "epoch": 0.5208785484851115,
+      "eval_entropy": 0.2028083825546704,
+      "eval_loss": 0.032208461314439774,
+      "eval_mean_token_accuracy": 0.99125398544783,
+      "eval_num_tokens": 126537184.0,
+      "eval_runtime": 81.5469,
+      "eval_samples_per_second": 122.138,
+      "eval_steps_per_second": 7.64,
+      "step": 750
+    },
+    {
+      "entropy": 0.1591445118188858,
+      "epoch": 0.5278235957982463,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.9479433908392747e-05,
+      "loss": 0.025,
+      "mean_token_accuracy": 0.9921697616577149,
+      "num_tokens": 128251701.0,
+      "step": 760
+    },
+    {
+      "entropy": 0.16354208169505,
+      "epoch": 0.5347686431113812,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.9447308118410544e-05,
+      "loss": 0.0251,
+      "mean_token_accuracy": 0.9920223578810692,
+      "num_tokens": 129913493.0,
+      "step": 770
+    },
+    {
+      "entropy": 0.16334566865116357,
+      "epoch": 0.541713690424516,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.9414239067587866e-05,
+      "loss": 0.0256,
+      "mean_token_accuracy": 0.992166705429554,
+      "num_tokens": 131640183.0,
+      "step": 780
+    },
+    {
+      "entropy": 0.16315491450950503,
+      "epoch": 0.5486587377376508,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.938022891499409e-05,
+      "loss": 0.0257,
+      "mean_token_accuracy": 0.9920231319963932,
+      "num_tokens": 133328365.0,
+      "step": 790
+    },
+    {
+      "entropy": 0.1643814197741449,
+      "epoch": 0.5556037850507857,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.9345279881142887e-05,
+      "loss": 0.0257,
+      "mean_token_accuracy": 0.991948215663433,
+      "num_tokens": 134988156.0,
+      "step": 800
+    },
+    {
+      "entropy": 0.160524242464453,
+      "epoch": 0.5625488323639205,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 2.930939424784723e-05,
+      "loss": 0.0243,
+      "mean_token_accuracy": 0.9926339693367481,
+      "num_tokens": 136738001.0,
+      "step": 810
+    },
+    {
+      "entropy": 0.1635106342844665,
+      "epoch": 0.5694938796770553,
+      "grad_norm": 0.279296875,
+      "learning_rate": 2.9272574358070406e-05,
+      "loss": 0.0263,
+      "mean_token_accuracy": 0.991894131153822,
+      "num_tokens": 138411117.0,
+      "step": 820
+    },
+    {
+      "entropy": 0.16258806074038148,
+      "epoch": 0.5764389269901902,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.923482261577307e-05,
+      "loss": 0.0265,
+      "mean_token_accuracy": 0.9918580889701843,
+      "num_tokens": 140144400.0,
+      "step": 830
+    },
+    {
+      "entropy": 0.17904180409386755,
+      "epoch": 0.5833839743033249,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.9196141485756267e-05,
+      "loss": 0.0283,
+      "mean_token_accuracy": 0.9913281343877316,
+      "num_tokens": 141787622.0,
+      "step": 840
+    },
+    {
+      "entropy": 0.1736303588375449,
+      "epoch": 0.5903290216164597,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 2.915653349350051e-05,
+      "loss": 0.0262,
+      "mean_token_accuracy": 0.9919952727854252,
+      "num_tokens": 143416525.0,
+      "step": 850
+    },
+    {
+      "entropy": 0.17081214878708123,
+      "epoch": 0.5972740689295946,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 2.9116001225000912e-05,
+      "loss": 0.0265,
+      "mean_token_accuracy": 0.9918745383620262,
+      "num_tokens": 145100905.0,
+      "step": 860
+    },
+    {
+      "entropy": 0.17302428921684623,
+      "epoch": 0.6042191162427294,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.9074547326598316e-05,
+      "loss": 0.0259,
+      "mean_token_accuracy": 0.991954755783081,
+      "num_tokens": 146789995.0,
+      "step": 870
+    },
+    {
+      "entropy": 0.16971431467682124,
+      "epoch": 0.6111641635558642,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.9032174504806546e-05,
+      "loss": 0.0254,
+      "mean_token_accuracy": 0.991959635913372,
+      "num_tokens": 148472897.0,
+      "step": 880
+    },
+    {
+      "entropy": 0.17342975698411464,
+      "epoch": 0.6181092108689991,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.8988885526135672e-05,
+      "loss": 0.0259,
+      "mean_token_accuracy": 0.9921123184263706,
+      "num_tokens": 150147383.0,
+      "step": 890
+    },
+    {
+      "entropy": 0.16509362338110806,
+      "epoch": 0.6250542581821339,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.894468321691141e-05,
+      "loss": 0.0247,
+      "mean_token_accuracy": 0.9921708039939403,
+      "num_tokens": 151859058.0,
+      "step": 900
+    },
+    {
+      "entropy": 0.17183032808825374,
+      "epoch": 0.6319993054952687,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.889957046309058e-05,
+      "loss": 0.0259,
+      "mean_token_accuracy": 0.9919849313795567,
+      "num_tokens": 153511113.0,
+      "step": 910
+    },
+    {
+      "entropy": 0.1668629450723529,
+      "epoch": 0.6389443528084036,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.8853550210072676e-05,
+      "loss": 0.0242,
+      "mean_token_accuracy": 0.9923402860760688,
+      "num_tokens": 155209381.0,
+      "step": 920
+    },
+    {
+      "entropy": 0.165106045268476,
+      "epoch": 0.6458894001215383,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.8806625462507573e-05,
+      "loss": 0.0244,
+      "mean_token_accuracy": 0.9924936473369599,
+      "num_tokens": 156930591.0,
+      "step": 930
+    },
+    {
+      "entropy": 0.1726138608530164,
+      "epoch": 0.6528344474346731,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.8758799284099357e-05,
+      "loss": 0.0255,
+      "mean_token_accuracy": 0.9920349515974521,
+      "num_tokens": 158611580.0,
+      "step": 940
+    },
+    {
+      "entropy": 0.1682711481116712,
+      "epoch": 0.659779494747808,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.8710074797406277e-05,
+      "loss": 0.0242,
+      "mean_token_accuracy": 0.9924891531467438,
+      "num_tokens": 160303915.0,
+      "step": 950
+    },
+    {
+      "entropy": 0.1658643066883087,
+      "epoch": 0.6667245420609428,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.8660455183636893e-05,
+      "loss": 0.0255,
+      "mean_token_accuracy": 0.9920015670359135,
+      "num_tokens": 161990176.0,
+      "step": 960
+    },
+    {
+      "entropy": 0.16630794005468488,
+      "epoch": 0.6736695893740776,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.860994368244238e-05,
+      "loss": 0.0246,
+      "mean_token_accuracy": 0.9922889664769172,
+      "num_tokens": 163672500.0,
+      "step": 970
+    },
+    {
+      "entropy": 0.17136559821665287,
+      "epoch": 0.6806146366872124,
+      "grad_norm": 0.162109375,
+      "learning_rate": 2.8558543591704984e-05,
+      "loss": 0.0263,
+      "mean_token_accuracy": 0.9918027445673943,
+      "num_tokens": 165338616.0,
+      "step": 980
+    },
+    {
+      "entropy": 0.16459389431402088,
+      "epoch": 0.6875596840003473,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.8506258267322738e-05,
+      "loss": 0.0251,
+      "mean_token_accuracy": 0.9921816885471344,
+      "num_tokens": 167056620.0,
+      "step": 990
+    },
+    {
+      "entropy": 0.1651211366057396,
+      "epoch": 0.6945047313134821,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.8453091122990325e-05,
+      "loss": 0.0242,
+      "mean_token_accuracy": 0.9924872562289238,
+      "num_tokens": 168766550.0,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6945047313134821,
+      "eval_entropy": 0.20344850389857927,
+      "eval_loss": 0.0312611423432827,
+      "eval_mean_token_accuracy": 0.9914905528386945,
+      "eval_num_tokens": 168766550.0,
+      "eval_runtime": 82.313,
+      "eval_samples_per_second": 121.002,
+      "eval_steps_per_second": 7.569,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 4320,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.5961741992379023e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-1000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8363ddab34be1be669fb9c3b0e477751c99c8dd6938033052e7430da744c5331
+size 6289

checkpoint-1000/video_preprocessor_config.json

@@ -0,0 +1,41 @@
+{
+  "crop_size": null,
+  "data_format": "channels_first",
+  "default_to_square": true,
+  "device": null,
+  "do_center_crop": null,
+  "do_convert_rgb": true,
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "do_sample_frames": true,
+  "fps": 2,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "input_data_format": null,
+  "max_frames": 768,
+  "merge_size": 2,
+  "min_frames": 4,
+  "num_frames": null,
+  "pad_size": null,
+  "patch_size": 16,
+  "processor_class": "Qwen3VLProcessor",
+  "resample": 3,
+  "rescale_factor": 0.00392156862745098,
+  "return_metadata": false,
+  "size": {
+    "longest_edge": 25165824,
+    "shortest_edge": 4096
+  },
+  "temporal_patch_size": 2,
+  "video_metadata": null,
+  "video_processor_type": "Qwen3VLVideoProcessor"
+}

checkpoint-1500/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
+}

checkpoint-1500/chat_template.jinja

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

checkpoint-1500/config.json

@@ -0,0 +1,67 @@
+{
+  "architectures": [
+    "Qwen3VLForConditionalGeneration"
+  ],
+  "dtype": "bfloat16",
+  "eos_token_id": 151645,
+  "image_token_id": 151655,
+  "model_type": "qwen3_vl",
+  "pad_token_id": 151643,
+  "text_config": {
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "bos_token_id": 151643,
+    "dtype": "bfloat16",
+    "eos_token_id": 151645,
+    "head_dim": 128,
+    "hidden_act": "silu",
+    "hidden_size": 2048,
+    "initializer_range": 0.02,
+    "intermediate_size": 6144,
+    "max_position_embeddings": 262144,
+    "model_type": "qwen3_vl_text",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 28,
+    "num_key_value_heads": 8,
+    "rms_norm_eps": 1e-06,
+    "rope_scaling": {
+      "mrope_interleaved": true,
+      "mrope_section": [
+        24,
+        20,
+        20
+      ],
+      "rope_type": "default"
+    },
+    "rope_theta": 5000000,
+    "tie_word_embeddings": true,
+    "use_cache": true,
+    "vocab_size": 151936
+  },
+  "tie_word_embeddings": true,
+  "transformers_version": "4.57.0",
+  "video_token_id": 151656,
+  "vision_config": {
+    "deepstack_visual_indexes": [
+      5,
+      11,
+      17
+    ],
+    "depth": 24,
+    "dtype": "bfloat16",
+    "hidden_act": "gelu_pytorch_tanh",
+    "hidden_size": 1024,
+    "in_channels": 3,
+    "initializer_range": 0.02,
+    "intermediate_size": 4096,
+    "model_type": "qwen3_vl",
+    "num_heads": 16,
+    "num_position_embeddings": 2304,
+    "out_hidden_size": 2048,
+    "patch_size": 16,
+    "spatial_merge_size": 2,
+    "temporal_patch_size": 2
+  },
+  "vision_end_token_id": 151653,
+  "vision_start_token_id": 151652
+}

checkpoint-1500/generation_config.json

@@ -0,0 +1,13 @@
+{
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.57.0"
+}

checkpoint-1500/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5530396e0c21e6d9f6af1e8472f7cb21fe08ae240781e73d1775d8edc9cb5200
+size 4255140312

checkpoint-1500/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:951c6c3b2d54e59bc114cb3f21b654e3286993754c6ca960677f78ba4682edd2
+size 8510679398

checkpoint-1500/preprocessor_config.json

@@ -0,0 +1,39 @@
+{
+  "crop_size": null,
+  "data_format": "channels_first",
+  "default_to_square": true,
+  "device": null,
+  "disable_grouping": null,
+  "do_center_crop": null,
+  "do_convert_rgb": true,
+  "do_normalize": true,
+  "do_pad": null,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_processor_type": "Qwen2VLImageProcessorFast",
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "input_data_format": null,
+  "max_pixels": null,
+  "merge_size": 2,
+  "min_pixels": null,
+  "pad_size": null,
+  "patch_size": 16,
+  "processor_class": "Qwen3VLProcessor",
+  "resample": 3,
+  "rescale_factor": 0.00392156862745098,
+  "return_tensors": null,
+  "size": {
+    "longest_edge": 16777216,
+    "shortest_edge": 65536
+  },
+  "temporal_patch_size": 2
+}

checkpoint-1500/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:706722fc41436bdb24c4016ab83fbbb62526893d41d4cf49a009fdb7bb4abb5e
+size 16389

checkpoint-1500/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9cd3c5bfb3f44d6324570cdf01e259d25d57145e7c5bcf8a356f6f7b56afdf24
+size 16389

checkpoint-1500/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ada3d67296c7e68b4d424298393788e9698d123c0f49e21ffd505f02a1def057
+size 16389

checkpoint-1500/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6abe268928ee44107d52539186bb325f6915b9fe7dafd490b17c58d29bd50783
+size 16389

checkpoint-1500/rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bdabaa1468bbe8afa177e64e4dd55a0b2c5ac41e68d229b3d790a6cdd8fbf6fb
+size 16389

checkpoint-1500/rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20c4cea43dfd62d3a3364a4603fb39a847fae2cd95a7b7d8f43fb21cef815cde
+size 16389

checkpoint-1500/rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c22599770b4a3567bdcbd40583c1c29ca81866b4f2fb6dbcf94bafecdef10d2
+size 16389

checkpoint-1500/rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:12f20736861ad603467227a1c08b4d09fd872a184e0c54c650a3ac14d130dfcd
+size 16389

checkpoint-1500/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2090e53456e02d38e003eadec700e1e47336ea9d8442b5e9e6d344c82e3d5dfb
+size 1465

checkpoint-1500/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
+  }
+}

checkpoint-1500/tokenizer.json

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

checkpoint-1500/tokenizer_config.json

@@ -0,0 +1,240 @@
+{
+  "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,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 262144,
+  "pad_token": "<|endoftext|>",
+  "processor_class": "Qwen3VLProcessor",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

checkpoint-1500/trainer_state.json

@@ -0,0 +1,1600 @@
+{
+  "best_global_step": 1500,
+  "best_metric": 0.030296573415398598,
+  "best_model_checkpoint": "/workspace/Qwen3_checkpoint/checkpoint-1500",
+  "epoch": 1.041670283878809,
+  "eval_steps": 250,
+  "global_step": 1500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "entropy": 1.1519597232341767,
+      "epoch": 0.00694504731313482,
+      "grad_norm": 540.0,
+      "learning_rate": 6.249999999999999e-07,
+      "loss": 22.6265,
+      "mean_token_accuracy": 0.07503694812767207,
+      "num_tokens": 1640499.0,
+      "step": 10
+    },
+    {
+      "entropy": 1.1664627104997636,
+      "epoch": 0.01389009462626964,
+      "grad_norm": 466.0,
+      "learning_rate": 1.3194444444444444e-06,
+      "loss": 22.1144,
+      "mean_token_accuracy": 0.08692597970366478,
+      "num_tokens": 3288480.0,
+      "step": 20
+    },
+    {
+      "entropy": 1.239394597709179,
+      "epoch": 0.020835141939404462,
+      "grad_norm": 462.0,
+      "learning_rate": 2.0138888888888893e-06,
+      "loss": 21.7485,
+      "mean_token_accuracy": 0.07649312568828463,
+      "num_tokens": 4958023.0,
+      "step": 30
+    },
+    {
+      "entropy": 1.4183061853051186,
+      "epoch": 0.02778018925253928,
+      "grad_norm": 436.0,
+      "learning_rate": 2.7083333333333334e-06,
+      "loss": 20.1258,
+      "mean_token_accuracy": 0.07965942500159144,
+      "num_tokens": 6627704.0,
+      "step": 40
+    },
+    {
+      "entropy": 1.9329519897699357,
+      "epoch": 0.034725236565674104,
+      "grad_norm": 378.0,
+      "learning_rate": 3.402777777777778e-06,
+      "loss": 16.576,
+      "mean_token_accuracy": 0.07870438289828599,
+      "num_tokens": 8340232.0,
+      "step": 50
+    },
+    {
+      "entropy": 4.602115747332573,
+      "epoch": 0.041670283878808924,
+      "grad_norm": 184.0,
+      "learning_rate": 4.097222222222222e-06,
+      "loss": 12.1758,
+      "mean_token_accuracy": 0.07569821006618441,
+      "num_tokens": 10025658.0,
+      "step": 60
+    },
+    {
+      "entropy": 7.328701210021973,
+      "epoch": 0.04861533119194374,
+      "grad_norm": 114.5,
+      "learning_rate": 4.791666666666667e-06,
+      "loss": 9.8369,
+      "mean_token_accuracy": 0.08327204268425703,
+      "num_tokens": 11741238.0,
+      "step": 70
+    },
+    {
+      "entropy": 7.635936880111695,
+      "epoch": 0.05556037850507856,
+      "grad_norm": 113.5,
+      "learning_rate": 5.486111111111111e-06,
+      "loss": 8.3162,
+      "mean_token_accuracy": 0.0876343248412013,
+      "num_tokens": 13404022.0,
+      "step": 80
+    },
+    {
+      "entropy": 7.46761246919632,
+      "epoch": 0.06250542581821339,
+      "grad_norm": 123.5,
+      "learning_rate": 6.180555555555556e-06,
+      "loss": 7.63,
+      "mean_token_accuracy": 0.08659355682320893,
+      "num_tokens": 15112991.0,
+      "step": 90
+    },
+    {
+      "entropy": 7.1816710293293,
+      "epoch": 0.06945047313134821,
+      "grad_norm": 148.0,
+      "learning_rate": 6.875e-06,
+      "loss": 7.0576,
+      "mean_token_accuracy": 0.08753767567686736,
+      "num_tokens": 16756210.0,
+      "step": 100
+    },
+    {
+      "entropy": 7.2137157917022705,
+      "epoch": 0.07639552044448303,
+      "grad_norm": 169.0,
+      "learning_rate": 7.569444444444445e-06,
+      "loss": 6.6052,
+      "mean_token_accuracy": 0.08361622425727547,
+      "num_tokens": 18470889.0,
+      "step": 110
+    },
+    {
+      "entropy": 7.177587866783142,
+      "epoch": 0.08334056775761785,
+      "grad_norm": 156.0,
+      "learning_rate": 8.263888888888888e-06,
+      "loss": 6.0531,
+      "mean_token_accuracy": 0.08918410916812718,
+      "num_tokens": 20159140.0,
+      "step": 120
+    },
+    {
+      "entropy": 7.2117412209510805,
+      "epoch": 0.09028561507075267,
+      "grad_norm": 160.0,
+      "learning_rate": 8.958333333333334e-06,
+      "loss": 5.4473,
+      "mean_token_accuracy": 0.08560893354006112,
+      "num_tokens": 21857491.0,
+      "step": 130
+    },
+    {
+      "entropy": 7.211903780698776,
+      "epoch": 0.09723066238388749,
+      "grad_norm": 278.0,
+      "learning_rate": 9.652777777777777e-06,
+      "loss": 4.788,
+      "mean_token_accuracy": 0.127999250870198,
+      "num_tokens": 23545940.0,
+      "step": 140
+    },
+    {
+      "entropy": 7.21578094959259,
+      "epoch": 0.1041757096970223,
+      "grad_norm": 163.0,
+      "learning_rate": 1.0347222222222223e-05,
+      "loss": 4.1808,
+      "mean_token_accuracy": 0.5043030813336372,
+      "num_tokens": 25200650.0,
+      "step": 150
+    },
+    {
+      "entropy": 7.225958931446075,
+      "epoch": 0.11112075701015713,
+      "grad_norm": 146.0,
+      "learning_rate": 1.1041666666666668e-05,
+      "loss": 3.4839,
+      "mean_token_accuracy": 0.8470314003527164,
+      "num_tokens": 26879379.0,
+      "step": 160
+    },
+    {
+      "entropy": 7.03481342792511,
+      "epoch": 0.11806580432329196,
+      "grad_norm": 160.0,
+      "learning_rate": 1.1736111111111112e-05,
+      "loss": 2.6644,
+      "mean_token_accuracy": 0.9146032109856606,
+      "num_tokens": 28559842.0,
+      "step": 170
+    },
+    {
+      "entropy": 6.437038451433182,
+      "epoch": 0.12501085163642678,
+      "grad_norm": 135.0,
+      "learning_rate": 1.2430555555555557e-05,
+      "loss": 1.8016,
+      "mean_token_accuracy": 0.9310876823961735,
+      "num_tokens": 30223208.0,
+      "step": 180
+    },
+    {
+      "entropy": 4.774793094396591,
+      "epoch": 0.13195589894956158,
+      "grad_norm": 72.0,
+      "learning_rate": 1.3125e-05,
+      "loss": 1.0024,
+      "mean_token_accuracy": 0.9337419532239437,
+      "num_tokens": 31915078.0,
+      "step": 190
+    },
+    {
+      "entropy": 2.502578377723694,
+      "epoch": 0.13890094626269642,
+      "grad_norm": 18.5,
+      "learning_rate": 1.3819444444444444e-05,
+      "loss": 0.5206,
+      "mean_token_accuracy": 0.9370624087750912,
+      "num_tokens": 33627632.0,
+      "step": 200
+    },
+    {
+      "entropy": 1.1439476042985917,
+      "epoch": 0.14584599357583122,
+      "grad_norm": 6.5,
+      "learning_rate": 1.451388888888889e-05,
+      "loss": 0.314,
+      "mean_token_accuracy": 0.9395098507404327,
+      "num_tokens": 35335117.0,
+      "step": 210
+    },
+    {
+      "entropy": 0.625144999101758,
+      "epoch": 0.15279104088896606,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.5208333333333333e-05,
+      "loss": 0.2057,
+      "mean_token_accuracy": 0.9517911829054355,
+      "num_tokens": 37011186.0,
+      "step": 220
+    },
+    {
+      "entropy": 0.34807259757071735,
+      "epoch": 0.1597360882021009,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.590277777777778e-05,
+      "loss": 0.1098,
+      "mean_token_accuracy": 0.9706214666366577,
+      "num_tokens": 38673894.0,
+      "step": 230
+    },
+    {
+      "entropy": 0.22573072295635938,
+      "epoch": 0.1666811355152357,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.6597222222222222e-05,
+      "loss": 0.0648,
+      "mean_token_accuracy": 0.9812032662332058,
+      "num_tokens": 40360862.0,
+      "step": 240
+    },
+    {
+      "entropy": 0.19475232288241387,
+      "epoch": 0.17362618282837053,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7291666666666666e-05,
+      "loss": 0.0472,
+      "mean_token_accuracy": 0.9857565425336361,
+      "num_tokens": 42047917.0,
+      "step": 250
+    },
+    {
+      "epoch": 0.17362618282837053,
+      "eval_entropy": 0.22951670004171507,
+      "eval_loss": 0.05014730989933014,
+      "eval_mean_token_accuracy": 0.985958794625192,
+      "eval_num_tokens": 42047917.0,
+      "eval_runtime": 81.9972,
+      "eval_samples_per_second": 121.468,
+      "eval_steps_per_second": 7.598,
+      "step": 250
+    },
+    {
+      "entropy": 0.1923621967434883,
+      "epoch": 0.18057123014150533,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7986111111111113e-05,
+      "loss": 0.0397,
+      "mean_token_accuracy": 0.9877771027386189,
+      "num_tokens": 43714135.0,
+      "step": 260
+    },
+    {
+      "entropy": 0.18295048763975502,
+      "epoch": 0.18751627745464017,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8680555555555556e-05,
+      "loss": 0.0395,
+      "mean_token_accuracy": 0.9877975657582283,
+      "num_tokens": 45420354.0,
+      "step": 270
+    },
+    {
+      "entropy": 0.18508334569633006,
+      "epoch": 0.19446132476777497,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9375e-05,
+      "loss": 0.0365,
+      "mean_token_accuracy": 0.9889688648283481,
+      "num_tokens": 47094373.0,
+      "step": 280
+    },
+    {
+      "entropy": 0.17800553049892187,
+      "epoch": 0.2014063720809098,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 2.0069444444444447e-05,
+      "loss": 0.0334,
+      "mean_token_accuracy": 0.9898182846605778,
+      "num_tokens": 48737014.0,
+      "step": 290
+    },
+    {
+      "entropy": 0.17542087566107512,
+      "epoch": 0.2083514193940446,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.076388888888889e-05,
+      "loss": 0.0332,
+      "mean_token_accuracy": 0.9898306861519813,
+      "num_tokens": 50411139.0,
+      "step": 300
+    },
+    {
+      "entropy": 0.17356558851897716,
+      "epoch": 0.21529646670717945,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.1458333333333334e-05,
+      "loss": 0.0329,
+      "mean_token_accuracy": 0.9897275142371654,
+      "num_tokens": 52101776.0,
+      "step": 310
+    },
+    {
+      "entropy": 0.18331474922597407,
+      "epoch": 0.22224151402031425,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.2152777777777778e-05,
+      "loss": 0.0323,
+      "mean_token_accuracy": 0.9898556746542454,
+      "num_tokens": 53761706.0,
+      "step": 320
+    },
+    {
+      "entropy": 0.16903471825644373,
+      "epoch": 0.22918656133344908,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.284722222222222e-05,
+      "loss": 0.0303,
+      "mean_token_accuracy": 0.9905840046703815,
+      "num_tokens": 55510137.0,
+      "step": 330
+    },
+    {
+      "entropy": 0.1726135764271021,
+      "epoch": 0.23613160864658392,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 2.354166666666667e-05,
+      "loss": 0.0308,
+      "mean_token_accuracy": 0.9905642926692962,
+      "num_tokens": 57209080.0,
+      "step": 340
+    },
+    {
+      "entropy": 0.17743745557963847,
+      "epoch": 0.24307665595971872,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.4236111111111112e-05,
+      "loss": 0.0304,
+      "mean_token_accuracy": 0.9905851900577545,
+      "num_tokens": 58867303.0,
+      "step": 350
+    },
+    {
+      "entropy": 0.18280480420216919,
+      "epoch": 0.25002170327285356,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.4930555555555556e-05,
+      "loss": 0.0315,
+      "mean_token_accuracy": 0.9901946425437927,
+      "num_tokens": 60510859.0,
+      "step": 360
+    },
+    {
+      "entropy": 0.1761154976673424,
+      "epoch": 0.2569667505859884,
+      "grad_norm": 0.5,
+      "learning_rate": 2.5625e-05,
+      "loss": 0.0298,
+      "mean_token_accuracy": 0.9909366421401501,
+      "num_tokens": 62171999.0,
+      "step": 370
+    },
+    {
+      "entropy": 0.1671930664218962,
+      "epoch": 0.26391179789912317,
+      "grad_norm": 0.31640625,
+      "learning_rate": 2.6319444444444443e-05,
+      "loss": 0.0284,
+      "mean_token_accuracy": 0.9912083707749844,
+      "num_tokens": 63906245.0,
+      "step": 380
+    },
+    {
+      "entropy": 0.1754794458858669,
+      "epoch": 0.270856845212258,
+      "grad_norm": 0.240234375,
+      "learning_rate": 2.7013888888888887e-05,
+      "loss": 0.0303,
+      "mean_token_accuracy": 0.9904192194342614,
+      "num_tokens": 65589663.0,
+      "step": 390
+    },
+    {
+      "entropy": 0.1771222472190857,
+      "epoch": 0.27780189252539284,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.7708333333333337e-05,
+      "loss": 0.0299,
+      "mean_token_accuracy": 0.9907479107379913,
+      "num_tokens": 67241611.0,
+      "step": 400
+    },
+    {
+      "entropy": 0.17829248458147048,
+      "epoch": 0.28474693983852767,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.840277777777778e-05,
+      "loss": 0.0297,
+      "mean_token_accuracy": 0.9906024843454361,
+      "num_tokens": 68889437.0,
+      "step": 410
+    },
+    {
+      "entropy": 0.17638491541147233,
+      "epoch": 0.29169198715166245,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.9097222222222224e-05,
+      "loss": 0.0284,
+      "mean_token_accuracy": 0.9913305290043354,
+      "num_tokens": 70618362.0,
+      "step": 420
+    },
+    {
+      "entropy": 0.18182099591940643,
+      "epoch": 0.2986370344647973,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.9791666666666668e-05,
+      "loss": 0.0285,
+      "mean_token_accuracy": 0.9912928201258182,
+      "num_tokens": 72313577.0,
+      "step": 430
+    },
+    {
+      "entropy": 0.18310831850394607,
+      "epoch": 0.3055820817779321,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.9999760059577934e-05,
+      "loss": 0.029,
+      "mean_token_accuracy": 0.9910079762339592,
+      "num_tokens": 73988194.0,
+      "step": 440
+    },
+    {
+      "entropy": 0.17772579193115234,
+      "epoch": 0.31252712909106695,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.9998584859663212e-05,
+      "loss": 0.0284,
+      "mean_token_accuracy": 0.9910575695335865,
+      "num_tokens": 75663076.0,
+      "step": 450
+    },
+    {
+      "entropy": 0.17988401055335998,
+      "epoch": 0.3194721764042018,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.9996430406198636e-05,
+      "loss": 0.028,
+      "mean_token_accuracy": 0.9914611585438251,
+      "num_tokens": 77335287.0,
+      "step": 460
+    },
+    {
+      "entropy": 0.173457539267838,
+      "epoch": 0.32641722371733656,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.9993296839847897e-05,
+      "loss": 0.0287,
+      "mean_token_accuracy": 0.9910615682601929,
+      "num_tokens": 79052668.0,
+      "step": 470
+    },
+    {
+      "entropy": 0.16980679212138056,
+      "epoch": 0.3333622710304714,
+      "grad_norm": 0.234375,
+      "learning_rate": 2.9989184365200695e-05,
+      "loss": 0.0266,
+      "mean_token_accuracy": 0.9917186461389065,
+      "num_tokens": 80733082.0,
+      "step": 480
+    },
+    {
+      "entropy": 0.16687583839520811,
+      "epoch": 0.3403073183436062,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.998409325075938e-05,
+      "loss": 0.0274,
+      "mean_token_accuracy": 0.9914394572377205,
+      "num_tokens": 82456201.0,
+      "step": 490
+    },
+    {
+      "entropy": 0.16694002971053123,
+      "epoch": 0.34725236565674106,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.997802382892144e-05,
+      "loss": 0.0271,
+      "mean_token_accuracy": 0.9916702680289745,
+      "num_tokens": 84174465.0,
+      "step": 500
+    },
+    {
+      "epoch": 0.34725236565674106,
+      "eval_entropy": 0.2104673289826938,
+      "eval_loss": 0.03412040323019028,
+      "eval_mean_token_accuracy": 0.9906957189688522,
+      "eval_num_tokens": 84174465.0,
+      "eval_runtime": 81.2363,
+      "eval_samples_per_second": 122.605,
+      "eval_steps_per_second": 7.669,
+      "step": 500
+    },
+    {
+      "entropy": 0.16945981308817865,
+      "epoch": 0.35419741296987584,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.997097649595779e-05,
+      "loss": 0.0281,
+      "mean_token_accuracy": 0.9912286765873433,
+      "num_tokens": 85878591.0,
+      "step": 510
+    },
+    {
+      "entropy": 0.16488271821290254,
+      "epoch": 0.36114246028301067,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.9962951711986897e-05,
+      "loss": 0.0277,
+      "mean_token_accuracy": 0.9914339393377304,
+      "num_tokens": 87573271.0,
+      "step": 520
+    },
+    {
+      "entropy": 0.17190398359671236,
+      "epoch": 0.3680875075961455,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.9953950000944724e-05,
+      "loss": 0.0291,
+      "mean_token_accuracy": 0.9910320229828358,
+      "num_tokens": 89226804.0,
+      "step": 530
+    },
+    {
+      "entropy": 0.16401167614385487,
+      "epoch": 0.37503255490928034,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.994397195055056e-05,
+      "loss": 0.0268,
+      "mean_token_accuracy": 0.9915202751755714,
+      "num_tokens": 90931627.0,
+      "step": 540
+    },
+    {
+      "entropy": 0.1601521703414619,
+      "epoch": 0.3819776022224151,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.9933018212268607e-05,
+      "loss": 0.0256,
+      "mean_token_accuracy": 0.9919740296900272,
+      "num_tokens": 92669023.0,
+      "step": 550
+    },
+    {
+      "entropy": 0.1690834665670991,
+      "epoch": 0.38892264953554995,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.9921089501265493e-05,
+      "loss": 0.0272,
+      "mean_token_accuracy": 0.9918020509183407,
+      "num_tokens": 94351832.0,
+      "step": 560
+    },
+    {
+      "entropy": 0.16791587956249715,
+      "epoch": 0.3958676968486848,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.990818659636352e-05,
+      "loss": 0.0269,
+      "mean_token_accuracy": 0.9915785849094391,
+      "num_tokens": 96026408.0,
+      "step": 570
+    },
+    {
+      "entropy": 0.1673963068984449,
+      "epoch": 0.4028127441618196,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.989431033998986e-05,
+      "loss": 0.0263,
+      "mean_token_accuracy": 0.9915929511189461,
+      "num_tokens": 97667765.0,
+      "step": 580
+    },
+    {
+      "entropy": 0.1658677328377962,
+      "epoch": 0.40975779147495445,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.987946163812155e-05,
+      "loss": 0.0263,
+      "mean_token_accuracy": 0.9918196611106396,
+      "num_tokens": 99392767.0,
+      "step": 590
+    },
+    {
+      "entropy": 0.16773838149383663,
+      "epoch": 0.4167028387880892,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.9863641460226304e-05,
+      "loss": 0.0269,
+      "mean_token_accuracy": 0.9917811565101147,
+      "num_tokens": 101073637.0,
+      "step": 600
+    },
+    {
+      "entropy": 0.1649629818275571,
+      "epoch": 0.42364788610122406,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.984685083919928e-05,
+      "loss": 0.0274,
+      "mean_token_accuracy": 0.9912688471376896,
+      "num_tokens": 102740766.0,
+      "step": 610
+    },
+    {
+      "entropy": 0.15838869903236627,
+      "epoch": 0.4305929334143589,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.9829090871295564e-05,
+      "loss": 0.0251,
+      "mean_token_accuracy": 0.9922874167561531,
+      "num_tokens": 104488780.0,
+      "step": 620
+    },
+    {
+      "entropy": 0.17186203664168714,
+      "epoch": 0.4375379807274937,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.981036271605867e-05,
+      "loss": 0.0263,
+      "mean_token_accuracy": 0.9918930843472481,
+      "num_tokens": 106195449.0,
+      "step": 630
+    },
+    {
+      "entropy": 0.1648147610016167,
+      "epoch": 0.4444830280406285,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.979066759624479e-05,
+      "loss": 0.0252,
+      "mean_token_accuracy": 0.992139758169651,
+      "num_tokens": 107902411.0,
+      "step": 640
+    },
+    {
+      "entropy": 0.16121056992560626,
+      "epoch": 0.45142807535376334,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 2.9770006797742963e-05,
+      "loss": 0.025,
+      "mean_token_accuracy": 0.9919275738298893,
+      "num_tokens": 109608103.0,
+      "step": 650
+    },
+    {
+      "entropy": 0.17468737829476594,
+      "epoch": 0.45837312266689817,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.974838166949114e-05,
+      "loss": 0.0272,
+      "mean_token_accuracy": 0.9917291469871998,
+      "num_tokens": 111293326.0,
+      "step": 660
+    },
+    {
+      "entropy": 0.1678467271849513,
+      "epoch": 0.465318169980033,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.9725793623388097e-05,
+      "loss": 0.0255,
+      "mean_token_accuracy": 0.9922867320477963,
+      "num_tokens": 113010182.0,
+      "step": 670
+    },
+    {
+      "entropy": 0.16656696163117884,
+      "epoch": 0.47226321729316784,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.9702244134201263e-05,
+      "loss": 0.0258,
+      "mean_token_accuracy": 0.9920145235955715,
+      "num_tokens": 114702373.0,
+      "step": 680
+    },
+    {
+      "entropy": 0.1741402018815279,
+      "epoch": 0.4792082646063026,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.9677734739470407e-05,
+      "loss": 0.0282,
+      "mean_token_accuracy": 0.9915049582719803,
+      "num_tokens": 116371688.0,
+      "step": 690
+    },
+    {
+      "entropy": 0.17028953870758415,
+      "epoch": 0.48615331191943745,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.9652267039407297e-05,
+      "loss": 0.0273,
+      "mean_token_accuracy": 0.9915660627186298,
+      "num_tokens": 118057500.0,
+      "step": 700
+    },
+    {
+      "entropy": 0.17608886091038584,
+      "epoch": 0.4930983592325723,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.962584269679117e-05,
+      "loss": 0.0274,
+      "mean_token_accuracy": 0.9914011798799038,
+      "num_tokens": 119691073.0,
+      "step": 710
+    },
+    {
+      "entropy": 0.16367722414433955,
+      "epoch": 0.5000434065457071,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.959846343686022e-05,
+      "loss": 0.0251,
+      "mean_token_accuracy": 0.9922043934464455,
+      "num_tokens": 121405698.0,
+      "step": 720
+    },
+    {
+      "entropy": 0.16646005054935814,
+      "epoch": 0.506988453858842,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.957013104719892e-05,
+      "loss": 0.0257,
+      "mean_token_accuracy": 0.9920815974473953,
+      "num_tokens": 123140318.0,
+      "step": 730
+    },
+    {
+      "entropy": 0.16303647235035895,
+      "epoch": 0.5139335011719768,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.9540847377621337e-05,
+      "loss": 0.0264,
+      "mean_token_accuracy": 0.991966935992241,
+      "num_tokens": 124857523.0,
+      "step": 740
+    },
+    {
+      "entropy": 0.16988610094413162,
+      "epoch": 0.5208785484851115,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.9510614340050325e-05,
+      "loss": 0.0256,
+      "mean_token_accuracy": 0.9918926700949668,
+      "num_tokens": 126537184.0,
+      "step": 750
+    },
+    {
+      "epoch": 0.5208785484851115,
+      "eval_entropy": 0.2028083825546704,
+      "eval_loss": 0.032208461314439774,
+      "eval_mean_token_accuracy": 0.99125398544783,
+      "eval_num_tokens": 126537184.0,
+      "eval_runtime": 81.5469,
+      "eval_samples_per_second": 122.138,
+      "eval_steps_per_second": 7.64,
+      "step": 750
+    },
+    {
+      "entropy": 0.1591445118188858,
+      "epoch": 0.5278235957982463,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.9479433908392747e-05,
+      "loss": 0.025,
+      "mean_token_accuracy": 0.9921697616577149,
+      "num_tokens": 128251701.0,
+      "step": 760
+    },
+    {
+      "entropy": 0.16354208169505,
+      "epoch": 0.5347686431113812,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.9447308118410544e-05,
+      "loss": 0.0251,
+      "mean_token_accuracy": 0.9920223578810692,
+      "num_tokens": 129913493.0,
+      "step": 770
+    },
+    {
+      "entropy": 0.16334566865116357,
+      "epoch": 0.541713690424516,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.9414239067587866e-05,
+      "loss": 0.0256,
+      "mean_token_accuracy": 0.992166705429554,
+      "num_tokens": 131640183.0,
+      "step": 780
+    },
+    {
+      "entropy": 0.16315491450950503,
+      "epoch": 0.5486587377376508,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.938022891499409e-05,
+      "loss": 0.0257,
+      "mean_token_accuracy": 0.9920231319963932,
+      "num_tokens": 133328365.0,
+      "step": 790
+    },
+    {
+      "entropy": 0.1643814197741449,
+      "epoch": 0.5556037850507857,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.9345279881142887e-05,
+      "loss": 0.0257,
+      "mean_token_accuracy": 0.991948215663433,
+      "num_tokens": 134988156.0,
+      "step": 800
+    },
+    {
+      "entropy": 0.160524242464453,
+      "epoch": 0.5625488323639205,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 2.930939424784723e-05,
+      "loss": 0.0243,
+      "mean_token_accuracy": 0.9926339693367481,
+      "num_tokens": 136738001.0,
+      "step": 810
+    },
+    {
+      "entropy": 0.1635106342844665,
+      "epoch": 0.5694938796770553,
+      "grad_norm": 0.279296875,
+      "learning_rate": 2.9272574358070406e-05,
+      "loss": 0.0263,
+      "mean_token_accuracy": 0.991894131153822,
+      "num_tokens": 138411117.0,
+      "step": 820
+    },
+    {
+      "entropy": 0.16258806074038148,
+      "epoch": 0.5764389269901902,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.923482261577307e-05,
+      "loss": 0.0265,
+      "mean_token_accuracy": 0.9918580889701843,
+      "num_tokens": 140144400.0,
+      "step": 830
+    },
+    {
+      "entropy": 0.17904180409386755,
+      "epoch": 0.5833839743033249,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.9196141485756267e-05,
+      "loss": 0.0283,
+      "mean_token_accuracy": 0.9913281343877316,
+      "num_tokens": 141787622.0,
+      "step": 840
+    },
+    {
+      "entropy": 0.1736303588375449,
+      "epoch": 0.5903290216164597,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 2.915653349350051e-05,
+      "loss": 0.0262,
+      "mean_token_accuracy": 0.9919952727854252,
+      "num_tokens": 143416525.0,
+      "step": 850
+    },
+    {
+      "entropy": 0.17081214878708123,
+      "epoch": 0.5972740689295946,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 2.9116001225000912e-05,
+      "loss": 0.0265,
+      "mean_token_accuracy": 0.9918745383620262,
+      "num_tokens": 145100905.0,
+      "step": 860
+    },
+    {
+      "entropy": 0.17302428921684623,
+      "epoch": 0.6042191162427294,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.9074547326598316e-05,
+      "loss": 0.0259,
+      "mean_token_accuracy": 0.991954755783081,
+      "num_tokens": 146789995.0,
+      "step": 870
+    },
+    {
+      "entropy": 0.16971431467682124,
+      "epoch": 0.6111641635558642,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.9032174504806546e-05,
+      "loss": 0.0254,
+      "mean_token_accuracy": 0.991959635913372,
+      "num_tokens": 148472897.0,
+      "step": 880
+    },
+    {
+      "entropy": 0.17342975698411464,
+      "epoch": 0.6181092108689991,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.8988885526135672e-05,
+      "loss": 0.0259,
+      "mean_token_accuracy": 0.9921123184263706,
+      "num_tokens": 150147383.0,
+      "step": 890
+    },
+    {
+      "entropy": 0.16509362338110806,
+      "epoch": 0.6250542581821339,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.894468321691141e-05,
+      "loss": 0.0247,
+      "mean_token_accuracy": 0.9921708039939403,
+      "num_tokens": 151859058.0,
+      "step": 900
+    },
+    {
+      "entropy": 0.17183032808825374,
+      "epoch": 0.6319993054952687,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.889957046309058e-05,
+      "loss": 0.0259,
+      "mean_token_accuracy": 0.9919849313795567,
+      "num_tokens": 153511113.0,
+      "step": 910
+    },
+    {
+      "entropy": 0.1668629450723529,
+      "epoch": 0.6389443528084036,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.8853550210072676e-05,
+      "loss": 0.0242,
+      "mean_token_accuracy": 0.9923402860760688,
+      "num_tokens": 155209381.0,
+      "step": 920
+    },
+    {
+      "entropy": 0.165106045268476,
+      "epoch": 0.6458894001215383,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.8806625462507573e-05,
+      "loss": 0.0244,
+      "mean_token_accuracy": 0.9924936473369599,
+      "num_tokens": 156930591.0,
+      "step": 930
+    },
+    {
+      "entropy": 0.1726138608530164,
+      "epoch": 0.6528344474346731,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.8758799284099357e-05,
+      "loss": 0.0255,
+      "mean_token_accuracy": 0.9920349515974521,
+      "num_tokens": 158611580.0,
+      "step": 940
+    },
+    {
+      "entropy": 0.1682711481116712,
+      "epoch": 0.659779494747808,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.8710074797406277e-05,
+      "loss": 0.0242,
+      "mean_token_accuracy": 0.9924891531467438,
+      "num_tokens": 160303915.0,
+      "step": 950
+    },
+    {
+      "entropy": 0.1658643066883087,
+      "epoch": 0.6667245420609428,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.8660455183636893e-05,
+      "loss": 0.0255,
+      "mean_token_accuracy": 0.9920015670359135,
+      "num_tokens": 161990176.0,
+      "step": 960
+    },
+    {
+      "entropy": 0.16630794005468488,
+      "epoch": 0.6736695893740776,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.860994368244238e-05,
+      "loss": 0.0246,
+      "mean_token_accuracy": 0.9922889664769172,
+      "num_tokens": 163672500.0,
+      "step": 970
+    },
+    {
+      "entropy": 0.17136559821665287,
+      "epoch": 0.6806146366872124,
+      "grad_norm": 0.162109375,
+      "learning_rate": 2.8558543591704984e-05,
+      "loss": 0.0263,
+      "mean_token_accuracy": 0.9918027445673943,
+      "num_tokens": 165338616.0,
+      "step": 980
+    },
+    {
+      "entropy": 0.16459389431402088,
+      "epoch": 0.6875596840003473,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.8506258267322738e-05,
+      "loss": 0.0251,
+      "mean_token_accuracy": 0.9921816885471344,
+      "num_tokens": 167056620.0,
+      "step": 990
+    },
+    {
+      "entropy": 0.1651211366057396,
+      "epoch": 0.6945047313134821,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.8453091122990325e-05,
+      "loss": 0.0242,
+      "mean_token_accuracy": 0.9924872562289238,
+      "num_tokens": 168766550.0,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6945047313134821,
+      "eval_entropy": 0.20344850389857927,
+      "eval_loss": 0.0312611423432827,
+      "eval_mean_token_accuracy": 0.9914905528386945,
+      "eval_num_tokens": 168766550.0,
+      "eval_runtime": 82.313,
+      "eval_samples_per_second": 121.002,
+      "eval_steps_per_second": 7.569,
+      "step": 1000
+    },
+    {
+      "entropy": 0.16957395104691386,
+      "epoch": 0.701449778626617,
+      "grad_norm": 0.248046875,
+      "learning_rate": 2.8399045629976228e-05,
+      "loss": 0.0251,
+      "mean_token_accuracy": 0.9921023808419704,
+      "num_tokens": 170430747.0,
+      "step": 1010
+    },
+    {
+      "entropy": 0.16484390227124096,
+      "epoch": 0.7083948259397517,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.8344125316896066e-05,
+      "loss": 0.0249,
+      "mean_token_accuracy": 0.9923050113022327,
+      "num_tokens": 172148573.0,
+      "step": 1020
+    },
+    {
+      "entropy": 0.16710211224853994,
+      "epoch": 0.7153398732528865,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.8288333769482234e-05,
+      "loss": 0.0258,
+      "mean_token_accuracy": 0.9921270661056042,
+      "num_tokens": 173820070.0,
+      "step": 1030
+    },
+    {
+      "entropy": 0.16856574947014452,
+      "epoch": 0.7222849205660213,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.823167463034978e-05,
+      "loss": 0.0252,
+      "mean_token_accuracy": 0.9920636892318726,
+      "num_tokens": 175479758.0,
+      "step": 1040
+    },
+    {
+      "entropy": 0.16606747973710298,
+      "epoch": 0.7292299678791562,
+      "grad_norm": 0.3125,
+      "learning_rate": 2.817415159875857e-05,
+      "loss": 0.0243,
+      "mean_token_accuracy": 0.992286852747202,
+      "num_tokens": 177172453.0,
+      "step": 1050
+    },
+    {
+      "entropy": 0.1690367877483368,
+      "epoch": 0.736175015192291,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.8115768430371784e-05,
+      "loss": 0.0256,
+      "mean_token_accuracy": 0.9920091181993484,
+      "num_tokens": 178851088.0,
+      "step": 1060
+    },
+    {
+      "entropy": 0.16904147835448385,
+      "epoch": 0.7431200625054258,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.8056528937010707e-05,
+      "loss": 0.0244,
+      "mean_token_accuracy": 0.9924891479313374,
+      "num_tokens": 180522526.0,
+      "step": 1070
+    },
+    {
+      "entropy": 0.16681075962260367,
+      "epoch": 0.7500651098185607,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.7996436986405826e-05,
+      "loss": 0.0244,
+      "mean_token_accuracy": 0.9924980387091636,
+      "num_tokens": 182232157.0,
+      "step": 1080
+    },
+    {
+      "entropy": 0.1678140834905207,
+      "epoch": 0.7570101571316955,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.7935496501944358e-05,
+      "loss": 0.0246,
+      "mean_token_accuracy": 0.9924834348261357,
+      "num_tokens": 183935604.0,
+      "step": 1090
+    },
+    {
+      "entropy": 0.16671771639958025,
+      "epoch": 0.7639552044448302,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.7873711462414057e-05,
+      "loss": 0.0248,
+      "mean_token_accuracy": 0.9925374925136566,
+      "num_tokens": 185597802.0,
+      "step": 1100
+    },
+    {
+      "entropy": 0.16695020655170084,
+      "epoch": 0.7709002517579651,
+      "grad_norm": 0.30078125,
+      "learning_rate": 2.7811085901743443e-05,
+      "loss": 0.025,
+      "mean_token_accuracy": 0.9922063417732716,
+      "num_tokens": 187315982.0,
+      "step": 1110
+    },
+    {
+      "entropy": 0.16623194608837366,
+      "epoch": 0.7778452990710999,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 2.7747623908738438e-05,
+      "loss": 0.0244,
+      "mean_token_accuracy": 0.992550103366375,
+      "num_tokens": 189065194.0,
+      "step": 1120
+    },
+    {
+      "entropy": 0.17834464814513923,
+      "epoch": 0.7847903463842347,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.7683329626815405e-05,
+      "loss": 0.0255,
+      "mean_token_accuracy": 0.9921450808644294,
+      "num_tokens": 190723953.0,
+      "step": 1130
+    },
+    {
+      "entropy": 0.16976428665220739,
+      "epoch": 0.7917353936973696,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.761820725373063e-05,
+      "loss": 0.0244,
+      "mean_token_accuracy": 0.9923642635345459,
+      "num_tokens": 192392834.0,
+      "step": 1140
+    },
+    {
+      "entropy": 0.16556692561134695,
+      "epoch": 0.7986804410105044,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.7552261041306236e-05,
+      "loss": 0.0239,
+      "mean_token_accuracy": 0.9925307177007199,
+      "num_tokens": 194102328.0,
+      "step": 1150
+    },
+    {
+      "entropy": 0.1675541820935905,
+      "epoch": 0.8056254883236392,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.7485495295152604e-05,
+      "loss": 0.0245,
+      "mean_token_accuracy": 0.9923206239938736,
+      "num_tokens": 195797798.0,
+      "step": 1160
+    },
+    {
+      "entropy": 0.16941533563658595,
+      "epoch": 0.8125705356367741,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.7417914374387244e-05,
+      "loss": 0.0254,
+      "mean_token_accuracy": 0.9921845324337483,
+      "num_tokens": 197478529.0,
+      "step": 1170
+    },
+    {
+      "entropy": 0.1745386002585292,
+      "epoch": 0.8195155829499089,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.734952269135021e-05,
+      "loss": 0.0264,
+      "mean_token_accuracy": 0.9921305790543556,
+      "num_tokens": 199155644.0,
+      "step": 1180
+    },
+    {
+      "entropy": 0.16794457482174038,
+      "epoch": 0.8264606302630436,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.7280324711315975e-05,
+      "loss": 0.0252,
+      "mean_token_accuracy": 0.9920954599976539,
+      "num_tokens": 200849027.0,
+      "step": 1190
+    },
+    {
+      "entropy": 0.17055450221523644,
+      "epoch": 0.8334056775761784,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.7210324952201956e-05,
+      "loss": 0.0248,
+      "mean_token_accuracy": 0.9922527462244034,
+      "num_tokens": 202543815.0,
+      "step": 1200
+    },
+    {
+      "entropy": 0.1680849887430668,
+      "epoch": 0.8403507248893133,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.7139527984273498e-05,
+      "loss": 0.0241,
+      "mean_token_accuracy": 0.9922554560005665,
+      "num_tokens": 204205558.0,
+      "step": 1210
+    },
+    {
+      "entropy": 0.1646501302719116,
+      "epoch": 0.8472957722024481,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.706793842984549e-05,
+      "loss": 0.0235,
+      "mean_token_accuracy": 0.992755102366209,
+      "num_tokens": 205926396.0,
+      "step": 1220
+    },
+    {
+      "entropy": 0.173645833786577,
+      "epoch": 0.854240819515583,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.6995560962980587e-05,
+      "loss": 0.0248,
+      "mean_token_accuracy": 0.9920175686478615,
+      "num_tokens": 207587498.0,
+      "step": 1230
+    },
+    {
+      "entropy": 0.1690474884584546,
+      "epoch": 0.8611858668287178,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.6922400309184022e-05,
+      "loss": 0.0228,
+      "mean_token_accuracy": 0.9925218060612678,
+      "num_tokens": 209266767.0,
+      "step": 1240
+    },
+    {
+      "entropy": 0.17194054620340465,
+      "epoch": 0.8681309141418526,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.6848461245095104e-05,
+      "loss": 0.0244,
+      "mean_token_accuracy": 0.9921990059316158,
+      "num_tokens": 210919281.0,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8681309141418526,
+      "eval_entropy": 0.21217725058811243,
+      "eval_loss": 0.030440116301178932,
+      "eval_mean_token_accuracy": 0.9917104203092535,
+      "eval_num_tokens": 210919281.0,
+      "eval_runtime": 81.2048,
+      "eval_samples_per_second": 122.653,
+      "eval_steps_per_second": 7.672,
+      "step": 1250
+    },
+    {
+      "entropy": 0.17561539998278022,
+      "epoch": 0.8750759614549874,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.6773748598175334e-05,
+      "loss": 0.0238,
+      "mean_token_accuracy": 0.9925415866076946,
+      "num_tokens": 212549619.0,
+      "step": 1260
+    },
+    {
+      "entropy": 0.1754954437725246,
+      "epoch": 0.8820210087681223,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.6698267246393226e-05,
+      "loss": 0.0255,
+      "mean_token_accuracy": 0.9922065414488316,
+      "num_tokens": 214229955.0,
+      "step": 1270
+    },
+    {
+      "entropy": 0.16669268533587456,
+      "epoch": 0.888966056081257,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.6622022117905835e-05,
+      "loss": 0.0243,
+      "mean_token_accuracy": 0.9923124983906746,
+      "num_tokens": 215930293.0,
+      "step": 1280
+    },
+    {
+      "entropy": 0.16962254513055086,
+      "epoch": 0.8959111033943918,
+      "grad_norm": 0.29296875,
+      "learning_rate": 2.6545018190736966e-05,
+      "loss": 0.0245,
+      "mean_token_accuracy": 0.9922974735498429,
+      "num_tokens": 217597783.0,
+      "step": 1290
+    },
+    {
+      "entropy": 0.16414162274450064,
+      "epoch": 0.9028561507075267,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.646726049245222e-05,
+      "loss": 0.024,
+      "mean_token_accuracy": 0.9924651831388474,
+      "num_tokens": 219266157.0,
+      "step": 1300
+    },
+    {
+      "entropy": 0.16683494308963417,
+      "epoch": 0.9098011980206615,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.6388754099830678e-05,
+      "loss": 0.0246,
+      "mean_token_accuracy": 0.992430393397808,
+      "num_tokens": 220961611.0,
+      "step": 1310
+    },
+    {
+      "entropy": 0.16680709961801768,
+      "epoch": 0.9167462453337963,
+      "grad_norm": 0.283203125,
+      "learning_rate": 2.6309504138533493e-05,
+      "loss": 0.0237,
+      "mean_token_accuracy": 0.9925533398985863,
+      "num_tokens": 222675223.0,
+      "step": 1320
+    },
+    {
+      "entropy": 0.1718313039280474,
+      "epoch": 0.9236912926469312,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.6229515782769213e-05,
+      "loss": 0.0252,
+      "mean_token_accuracy": 0.9922806732356548,
+      "num_tokens": 224320984.0,
+      "step": 1330
+    },
+    {
+      "entropy": 0.165607544593513,
+      "epoch": 0.930636339960066,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.614879425495595e-05,
+      "loss": 0.0242,
+      "mean_token_accuracy": 0.9924774184823036,
+      "num_tokens": 226008160.0,
+      "step": 1340
+    },
+    {
+      "entropy": 0.16860781861469148,
+      "epoch": 0.9375813872732008,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.6067344825380428e-05,
+      "loss": 0.0251,
+      "mean_token_accuracy": 0.9922392442822456,
+      "num_tokens": 227691055.0,
+      "step": 1350
+    },
+    {
+      "entropy": 0.15853081615641712,
+      "epoch": 0.9445264345863357,
+      "grad_norm": 0.19140625,
+      "learning_rate": 2.5985172811853883e-05,
+      "loss": 0.0229,
+      "mean_token_accuracy": 0.9926266454160213,
+      "num_tokens": 229447767.0,
+      "step": 1360
+    },
+    {
+      "entropy": 0.16538406033068895,
+      "epoch": 0.9514714818994704,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.5902283579364855e-05,
+      "loss": 0.0243,
+      "mean_token_accuracy": 0.992415937781334,
+      "num_tokens": 231143110.0,
+      "step": 1370
+    },
+    {
+      "entropy": 0.1578194869682193,
+      "epoch": 0.9584165292126052,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.5818682539728924e-05,
+      "loss": 0.0228,
+      "mean_token_accuracy": 0.9929021798074246,
+      "num_tokens": 232890680.0,
+      "step": 1380
+    },
+    {
+      "entropy": 0.16025031013414265,
+      "epoch": 0.9653615765257401,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.573437515123536e-05,
+      "loss": 0.0232,
+      "mean_token_accuracy": 0.9927003815770149,
+      "num_tokens": 234606776.0,
+      "step": 1390
+    },
+    {
+      "entropy": 0.1669618772342801,
+      "epoch": 0.9723066238388749,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.5649366918290765e-05,
+      "loss": 0.0252,
+      "mean_token_accuracy": 0.992205673456192,
+      "num_tokens": 236312207.0,
+      "step": 1400
+    },
+    {
+      "entropy": 0.1684069731272757,
+      "epoch": 0.9792516711520097,
+      "grad_norm": 0.296875,
+      "learning_rate": 2.5563663391059675e-05,
+      "loss": 0.0241,
+      "mean_token_accuracy": 0.9923799067735672,
+      "num_tokens": 238000255.0,
+      "step": 1410
+    },
+    {
+      "entropy": 0.16883580991998315,
+      "epoch": 0.9861967184651446,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.5477270165102207e-05,
+      "loss": 0.0242,
+      "mean_token_accuracy": 0.992599005997181,
+      "num_tokens": 239682996.0,
+      "step": 1420
+    },
+    {
+      "entropy": 0.1691451609134674,
+      "epoch": 0.9931417657782794,
+      "grad_norm": 0.3046875,
+      "learning_rate": 2.5390192881008734e-05,
+      "loss": 0.0254,
+      "mean_token_accuracy": 0.9921128831803798,
+      "num_tokens": 241366834.0,
+      "step": 1430
+    },
+    {
+      "entropy": 0.17308599907386152,
+      "epoch": 1.0,
+      "grad_norm": 0.23046875,
+      "learning_rate": 2.530243722403158e-05,
+      "loss": 0.0251,
+      "mean_token_accuracy": 0.9923322027242636,
+      "num_tokens": 243025309.0,
+      "step": 1440
+    },
+    {
+      "entropy": 0.1778549592010677,
+      "epoch": 1.0069450473131347,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.5214008923713855e-05,
+      "loss": 0.0255,
+      "mean_token_accuracy": 0.9925065904855728,
+      "num_tokens": 244698287.0,
+      "step": 1450
+    },
+    {
+      "entropy": 0.17283849623054265,
+      "epoch": 1.0138900946262697,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 2.512491375351539e-05,
+      "loss": 0.0243,
+      "mean_token_accuracy": 0.9923485726118088,
+      "num_tokens": 246367654.0,
+      "step": 1460
+    },
+    {
+      "entropy": 0.1658339927904308,
+      "epoch": 1.0208351419394044,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.5035157530435748e-05,
+      "loss": 0.0246,
+      "mean_token_accuracy": 0.9923050880432129,
+      "num_tokens": 248073194.0,
+      "step": 1470
+    },
+    {
+      "entropy": 0.16023449478670954,
+      "epoch": 1.0277801892525393,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.4944746114634462e-05,
+      "loss": 0.0226,
+      "mean_token_accuracy": 0.9928353972733021,
+      "num_tokens": 249843884.0,
+      "step": 1480
+    },
+    {
+      "entropy": 0.17396175684407353,
+      "epoch": 1.034725236565674,
+      "grad_norm": 0.240234375,
+      "learning_rate": 2.4853685409048427e-05,
+      "loss": 0.0251,
+      "mean_token_accuracy": 0.9921702355146408,
+      "num_tokens": 251503510.0,
+      "step": 1490
+    },
+    {
+      "entropy": 0.16662784228101374,
+      "epoch": 1.041670283878809,
+      "grad_norm": 0.3203125,
+      "learning_rate": 2.476198135900648e-05,
+      "loss": 0.0238,
+      "mean_token_accuracy": 0.9925790295004845,
+      "num_tokens": 253182254.0,
+      "step": 1500
+    },
+    {
+      "epoch": 1.041670283878809,
+      "eval_entropy": 0.2101928804123574,
+      "eval_loss": 0.030296573415398598,
+      "eval_mean_token_accuracy": 0.9917910993577581,
+      "eval_num_tokens": 253182254.0,
+      "eval_runtime": 81.0064,
+      "eval_samples_per_second": 122.953,
+      "eval_steps_per_second": 7.691,
+      "step": 1500
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 4320,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.394361222344737e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-1500/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8363ddab34be1be669fb9c3b0e477751c99c8dd6938033052e7430da744c5331
+size 6289

checkpoint-1500/video_preprocessor_config.json

@@ -0,0 +1,41 @@
+{
+  "crop_size": null,
+  "data_format": "channels_first",
+  "default_to_square": true,
+  "device": null,
+  "do_center_crop": null,
+  "do_convert_rgb": true,
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "do_sample_frames": true,
+  "fps": 2,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "input_data_format": null,
+  "max_frames": 768,
+  "merge_size": 2,
+  "min_frames": 4,
+  "num_frames": null,
+  "pad_size": null,
+  "patch_size": 16,
+  "processor_class": "Qwen3VLProcessor",
+  "resample": 3,
+  "rescale_factor": 0.00392156862745098,
+  "return_metadata": false,
+  "size": {
+    "longest_edge": 25165824,
+    "shortest_edge": 4096
+  },
+  "temporal_patch_size": 2,
+  "video_metadata": null,
+  "video_processor_type": "Qwen3VLVideoProcessor"
+}