初始化项目，由ModelHub XC社区提供模型

Model: harsha070/exp2-qwen-island-s42-lambda-0p35 Source: Original Platform
2026-06-16 06:33:17 +08:00
commit cfde30fbb5
33 changed files with 983 additions and 0 deletions
--- a/.gitattributes
+++ b/.gitattributes
@@ -0,0 +1,36 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz 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
+tokenizer.json filter=lfs diff=lfs merge=lfs -text
--- a/README.md
+++ b/README.md
@@ -0,0 +1,67 @@
+---
+base_model: Qwen/Qwen2.5-3B-Instruct
+library_name: transformers
+model_name: exp2-qwen-island-s42-lambda-0p35
+tags:
+- generated_from_trainer
+- trl
+- grpo
+licence: license
+---
+
+# Model Card for exp2-qwen-island-s42-lambda-0p35
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-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="harsha070/exp2-qwen-island-s42-lambda-0p35", 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/models-self5933/obfuscation-early-warning/runs/573g97b3) 
+
+
+
+This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://huggingface.co/papers/2402.03300).
+
+### Framework versions
+
+- TRL: 1.3.0
+- Transformers: 5.7.0
+- Pytorch: 2.11.0
+- Datasets: 4.8.5
+- Tokenizers: 0.22.2
+
+## Citations
+
+Cite GRPO as:
+
+```bibtex
+@article{shao2024deepseekmath,
+    title        = {{DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models}},
+    author       = {Zhihong Shao and Peiyi Wang and Qihao Zhu and Runxin Xu and Junxiao Song and Mingchuan Zhang and Y. K. Li and Y. Wu and Daya Guo},
+    year         = 2024,
+    eprint       = {arXiv:2402.03300},
+}
+```
+
+Cite TRL as:
+    
+```bibtex
+@software{vonwerra2020trl,
+  title   = {{TRL: Transformers Reinforcement Learning}},
+  author  = {von Werra, Leandro and Belkada, Younes and Tunstall, Lewis and Beeching, Edward and Thrush, Tristan and Lambert, Nathan and Huang, Shengyi and Rasul, Kashif and Gallouédec, Quentin},
+  license = {Apache-2.0},
+  url     = {https://github.com/huggingface/trl},
+  year    = {2020}
+}
+```
--- a/chat_template.jinja
+++ b/chat_template.jinja
@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- 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' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\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 %}
--- a/completions/completions_00010.parquet
+++ b/completions/completions_00010.parquet
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2d0ec2a61212762f74fe599abe0b5d24da2f33ed3586186fde17944268fa727b
+size 18183
--- a/completions/completions_00020.parquet
+++ b/completions/completions_00020.parquet
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4fe7b67227f75e1af2f085503ecfca657f880e74f1144f97ca02b56bcf6e7e7b
+size 17875
--- a/completions/completions_00030.parquet
+++ b/completions/completions_00030.parquet
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:92add775d7b2486a7b23bc4561c6ad3c8d8d6fc55276136d95aa6568fbad6813
+size 18451
--- a/completions/completions_00040.parquet
+++ b/completions/completions_00040.parquet
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87513bab6fcd724b56906a8ed3d105f84b59839ba963ee62cb8f24f081a052e2
+size 19537
--- a/completions/completions_00050.parquet
+++ b/completions/completions_00050.parquet
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5c8381b7ca8b51ed59defdd1d551a9ef791dbd015a96940a5344b56c30a108f
+size 22693
--- a/completions/completions_00060.parquet
+++ b/completions/completions_00060.parquet
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aca128dfc302866aa3860d4dca0acd9bc0108ebdda12cde82325429472e4a2e9
+size 17265
--- a/completions/completions_00070.parquet
+++ b/completions/completions_00070.parquet
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d033ea56500a6d821797219fe4beba372b7c5ab2ba11ec7416ddf07633bc0a8
+size 21510
--- a/completions/completions_00080.parquet
+++ b/completions/completions_00080.parquet
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8012cb52e3773d8ad09da8af73014330980cf9190397fa6f7e6c6fc78e8655f2
+size 19301
--- a/completions/completions_00090.parquet
+++ b/completions/completions_00090.parquet
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:82dfe0dd207de7e940dba95c2b6c415a11fcbd88474195d440f118569b4a8f24
+size 19704
--- a/completions/completions_00100.parquet
+++ b/completions/completions_00100.parquet
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2fb05e4e1cecbd2f197269dfd4f2cb2eb0b63cb9166c2cc1ef72d82536b34100
+size 21852
--- a/completions/completions_00110.parquet
+++ b/completions/completions_00110.parquet
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5501d7ffea37a608c2ca3a8242af637664683dd9b36aefe96f8104b21bdb984
+size 20281
--- a/completions/completions_00120.parquet
+++ b/completions/completions_00120.parquet
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7db9eaeb135fe13ba55177c5087299fea8bd855aabc9352088474636bd0af817
+size 19643
--- a/completions/completions_00130.parquet
+++ b/completions/completions_00130.parquet
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69e29e068e1b2d96432d13edb1dc3cf6544b6a7487c1b08da6d2d3f3d0da6a09
+size 19171
--- a/completions/completions_00140.parquet
+++ b/completions/completions_00140.parquet
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce69bf3cfa6e466f94f7129d68e4c3da76e14883d601fba41d8700d20d3d39c6
+size 21633
--- a/completions/completions_00150.parquet
+++ b/completions/completions_00150.parquet
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23716785f881bc42d405ca9bf3e32b4f54e5f0ba3f46e26d42e90fd1633c85d1
+size 18123
--- a/completions/completions_00160.parquet
+++ b/completions/completions_00160.parquet
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b4ec9dec91007f2f32caf8def9a25bdcb008eb6b56bd8507d485a6771e3a7784
+size 18597
--- a/config.json
+++ b/config.json
@@ -0,0 +1,69 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": null,
+  "dtype": "bfloat16",
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "intermediate_size": 11008,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 70,
+  "model_type": "qwen2",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 2,
+  "pad_token_id": 151643,
+  "rms_norm_eps": 1e-06,
+  "rope_parameters": {
+    "rope_theta": 1000000.0,
+    "rope_type": "default"
+  },
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.7.0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}
--- a/generation_config.json
+++ b/generation_config.json
@@ -0,0 +1,13 @@
+{
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "5.7.0"
+}
--- a/last-checkpoint/chat_template.jinja
+++ b/last-checkpoint/chat_template.jinja
@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- 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' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\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 %}
--- a/last-checkpoint/config.json
+++ b/last-checkpoint/config.json
@@ -0,0 +1,69 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": null,
+  "dtype": "bfloat16",
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "intermediate_size": 11008,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 70,
+  "model_type": "qwen2",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 2,
+  "pad_token_id": 151643,
+  "rms_norm_eps": 1e-06,
+  "rope_parameters": {
+    "rope_theta": 1000000.0,
+    "rope_type": "default"
+  },
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.7.0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}
--- a/last-checkpoint/generation_config.json
+++ b/last-checkpoint/generation_config.json
@@ -0,0 +1,13 @@
+{
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "5.7.0"
+}
--- a/last-checkpoint/model.safetensors
+++ b/last-checkpoint/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0feb0dd2ec4125bfaf367198c52c9eb8d2fd0461e96e1797c629ba61162aeb71
+size 6171927112
--- a/last-checkpoint/tokenizer.json
+++ b/last-checkpoint/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
--- a/last-checkpoint/tokenizer_config.json
+++ b/last-checkpoint/tokenizer_config.json
@@ -0,0 +1,30 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": false,
+  "local_files_only": false,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}
--- a/last-checkpoint/trainer_state.json
+++ b/last-checkpoint/trainer_state.json
@@ -0,0 +1,482 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.28,
+  "eval_steps": 500,
+  "global_step": 160,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 697.1,
+      "completions/max_terminated_length": 697.1,
+      "completions/mean_length": 444.4875,
+      "completions/mean_terminated_length": 444.4875,
+      "completions/min_length": 229.5,
+      "completions/min_terminated_length": 229.5,
+      "entropy": 0.7007812947034836,
+      "epoch": 0.08,
+      "frac_reward_zero_std": 0.275,
+      "grad_norm": 2.53125,
+      "kl": 0.07940242243275861,
+      "learning_rate": 9.4375e-06,
+      "loss": 0.008688435703516007,
+      "num_tokens": 56203.0,
+      "reward": 0.6825366348028183,
+      "reward_std": 0.42148933857679366,
+      "rewards/JointRewardFunction/mean": 0.6825366348028183,
+      "rewards/JointRewardFunction/std": 0.4214893504977226,
+      "step": 10,
+      "step_time": 29.87573713920101
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 655.1,
+      "completions/max_terminated_length": 655.1,
+      "completions/mean_length": 424.4125,
+      "completions/mean_terminated_length": 424.4125,
+      "completions/min_length": 240.3,
+      "completions/min_terminated_length": 240.3,
+      "entropy": 0.7277415141463279,
+      "epoch": 0.16,
+      "frac_reward_zero_std": 0.225,
+      "grad_norm": 1.765625,
+      "kl": 0.09351013670675457,
+      "learning_rate": 8.8125e-06,
+      "loss": 0.0028222408145666122,
+      "num_tokens": 110800.0,
+      "reward": 0.7511120676994324,
+      "reward_std": 0.4809179216623306,
+      "rewards/JointRewardFunction/mean": 0.7511120676994324,
+      "rewards/JointRewardFunction/std": 0.4809179395437241,
+      "step": 20,
+      "step_time": 29.239146613800404
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 681.9,
+      "completions/max_terminated_length": 681.9,
+      "completions/mean_length": 482.1,
+      "completions/mean_terminated_length": 482.1,
+      "completions/min_length": 293.0,
+      "completions/min_terminated_length": 293.0,
+      "entropy": 0.8137854047119617,
+      "epoch": 0.24,
+      "frac_reward_zero_std": 0.275,
+      "grad_norm": 2.34375,
+      "kl": 0.08033584761433303,
+      "learning_rate": 8.1875e-06,
+      "loss": 0.019247914850711822,
+      "num_tokens": 169812.0,
+      "reward": 0.8227136462926865,
+      "reward_std": 0.3767334818840027,
+      "rewards/JointRewardFunction/mean": 0.8227136462926865,
+      "rewards/JointRewardFunction/std": 0.37673348784446714,
+      "step": 30,
+      "step_time": 30.690014854901893
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 699.3,
+      "completions/max_terminated_length": 699.3,
+      "completions/mean_length": 469.3375,
+      "completions/mean_terminated_length": 469.3375,
+      "completions/min_length": 286.5,
+      "completions/min_terminated_length": 286.5,
+      "entropy": 0.8683047238737345,
+      "epoch": 0.32,
+      "frac_reward_zero_std": 0.2,
+      "grad_norm": 2.296875,
+      "kl": 0.08547290097922086,
+      "learning_rate": 7.5625e-06,
+      "loss": 0.005851123481988907,
+      "num_tokens": 227933.0,
+      "reward": 0.8848474144935607,
+      "reward_std": 0.4506288319826126,
+      "rewards/JointRewardFunction/mean": 0.8848474144935607,
+      "rewards/JointRewardFunction/std": 0.4506288468837738,
+      "step": 40,
+      "step_time": 31.253191193498786
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 713.1,
+      "completions/max_terminated_length": 713.1,
+      "completions/mean_length": 510.825,
+      "completions/mean_terminated_length": 510.825,
+      "completions/min_length": 317.9,
+      "completions/min_terminated_length": 317.9,
+      "entropy": 0.8964162725955248,
+      "epoch": 0.4,
+      "frac_reward_zero_std": 0.2,
+      "grad_norm": 1.71875,
+      "kl": 0.07821776089258492,
+      "learning_rate": 6.9375e-06,
+      "loss": 0.013301727175712586,
+      "num_tokens": 289629.0,
+      "reward": 1.038847678899765,
+      "reward_std": 0.3691652432084084,
+      "rewards/JointRewardFunction/mean": 1.038847678899765,
+      "rewards/JointRewardFunction/std": 0.36916525810956957,
+      "step": 50,
+      "step_time": 31.799948546098314
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 746.0,
+      "completions/max_terminated_length": 746.0,
+      "completions/mean_length": 523.425,
+      "completions/mean_terminated_length": 523.425,
+      "completions/min_length": 309.1,
+      "completions/min_terminated_length": 309.1,
+      "entropy": 0.8802642215043306,
+      "epoch": 0.48,
+      "frac_reward_zero_std": 0.175,
+      "grad_norm": 2.0,
+      "kl": 0.0804219183512032,
+      "learning_rate": 6.3125e-06,
+      "loss": 0.03207117319107056,
+      "num_tokens": 351843.0,
+      "reward": 0.9487841844558715,
+      "reward_std": 0.42972691655158995,
+      "rewards/JointRewardFunction/mean": 0.9487841844558715,
+      "rewards/JointRewardFunction/std": 0.4297269403934479,
+      "step": 60,
+      "step_time": 32.83071275450202
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0125,
+      "completions/max_length": 715.3,
+      "completions/max_terminated_length": 695.1,
+      "completions/mean_length": 510.3,
+      "completions/mean_terminated_length": 503.83750305175784,
+      "completions/min_length": 295.9,
+      "completions/min_terminated_length": 295.9,
+      "entropy": 0.9274463646113873,
+      "epoch": 0.56,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 2.265625,
+      "kl": 0.08897386915050447,
+      "learning_rate": 5.6875e-06,
+      "loss": 0.07122264504432678,
+      "num_tokens": 413123.0,
+      "reward": 1.0958777070045471,
+      "reward_std": 0.3831350475549698,
+      "rewards/JointRewardFunction/mean": 1.0958777070045471,
+      "rewards/JointRewardFunction/std": 0.3831350475549698,
+      "step": 70,
+      "step_time": 31.748691623102424
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 608.4,
+      "completions/max_terminated_length": 608.4,
+      "completions/mean_length": 440.2875,
+      "completions/mean_terminated_length": 440.2875,
+      "completions/min_length": 326.3,
+      "completions/min_terminated_length": 326.3,
+      "entropy": 0.8611678160727024,
+      "epoch": 0.64,
+      "frac_reward_zero_std": 0.275,
+      "grad_norm": 2.8125,
+      "kl": 0.0876699925865978,
+      "learning_rate": 5.0625e-06,
+      "loss": 0.019712889194488527,
+      "num_tokens": 468960.0,
+      "reward": 1.0908459782600404,
+      "reward_std": 0.35941318422555923,
+      "rewards/JointRewardFunction/mean": 1.0908459782600404,
+      "rewards/JointRewardFunction/std": 0.3594131987541914,
+      "step": 80,
+      "step_time": 27.2742907967011
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 626.7,
+      "completions/max_terminated_length": 626.7,
+      "completions/mean_length": 434.05,
+      "completions/mean_terminated_length": 434.05,
+      "completions/min_length": 263.7,
+      "completions/min_terminated_length": 263.7,
+      "entropy": 0.8135976739227772,
+      "epoch": 0.72,
+      "frac_reward_zero_std": 0.375,
+      "grad_norm": 1.6953125,
+      "kl": 0.09378819759003817,
+      "learning_rate": 4.4375e-06,
+      "loss": 0.06376264691352844,
+      "num_tokens": 524304.0,
+      "reward": 1.109246850013733,
+      "reward_std": 0.35286828482057897,
+      "rewards/JointRewardFunction/mean": 1.109246850013733,
+      "rewards/JointRewardFunction/std": 0.3528682907810435,
+      "step": 90,
+      "step_time": 28.181775099898367
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 627.0,
+      "completions/max_terminated_length": 627.0,
+      "completions/mean_length": 444.325,
+      "completions/mean_terminated_length": 444.325,
+      "completions/min_length": 307.4,
+      "completions/min_terminated_length": 307.4,
+      "entropy": 0.9360679395496845,
+      "epoch": 0.8,
+      "frac_reward_zero_std": 0.275,
+      "grad_norm": 2.3125,
+      "kl": 0.08875131588429212,
+      "learning_rate": 3.8125e-06,
+      "loss": 0.03808712363243103,
+      "num_tokens": 580516.0,
+      "reward": 1.0944311678409577,
+      "reward_std": 0.3305914536118507,
+      "rewards/JointRewardFunction/mean": 1.0944311678409577,
+      "rewards/JointRewardFunction/std": 0.3305914685130119,
+      "step": 100,
+      "step_time": 28.45329053500027
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 650.9,
+      "completions/max_terminated_length": 650.9,
+      "completions/mean_length": 442.6875,
+      "completions/mean_terminated_length": 442.6875,
+      "completions/min_length": 262.7,
+      "completions/min_terminated_length": 262.7,
+      "entropy": 0.9691729046404362,
+      "epoch": 0.88,
+      "frac_reward_zero_std": 0.275,
+      "grad_norm": 2.21875,
+      "kl": 0.08224498964846134,
+      "learning_rate": 3.1875e-06,
+      "loss": 0.027349627017974852,
+      "num_tokens": 636475.0,
+      "reward": 1.031976318359375,
+      "reward_std": 0.43738164007663727,
+      "rewards/JointRewardFunction/mean": 1.031976318359375,
+      "rewards/JointRewardFunction/std": 0.4373816430568695,
+      "step": 110,
+      "step_time": 29.40713436929982
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 587.8,
+      "completions/max_terminated_length": 587.8,
+      "completions/mean_length": 423.65,
+      "completions/mean_terminated_length": 423.65,
+      "completions/min_length": 273.3,
+      "completions/min_terminated_length": 273.3,
+      "entropy": 0.886437576636672,
+      "epoch": 0.96,
+      "frac_reward_zero_std": 0.425,
+      "grad_norm": 1.96875,
+      "kl": 0.08013711464591325,
+      "learning_rate": 2.5625e-06,
+      "loss": 0.029777994751930235,
+      "num_tokens": 690739.0,
+      "reward": 1.1266614079475403,
+      "reward_std": 0.2993911794852465,
+      "rewards/JointRewardFunction/mean": 1.1266614079475403,
+      "rewards/JointRewardFunction/std": 0.29939117534086107,
+      "step": 120,
+      "step_time": 26.532396802499353
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 548.8,
+      "completions/max_terminated_length": 548.8,
+      "completions/mean_length": 413.5,
+      "completions/mean_terminated_length": 413.5,
+      "completions/min_length": 296.6,
+      "completions/min_terminated_length": 296.6,
+      "entropy": 0.8794000033289194,
+      "epoch": 1.04,
+      "frac_reward_zero_std": 0.475,
+      "grad_norm": 1.1875,
+      "kl": 0.0789637949783355,
+      "learning_rate": 1.9375e-06,
+      "loss": -0.010777493566274643,
+      "num_tokens": 744437.0,
+      "reward": 1.1596606612205504,
+      "reward_std": 0.2929948531091213,
+      "rewards/JointRewardFunction/mean": 1.1596606612205504,
+      "rewards/JointRewardFunction/std": 0.29299486204981806,
+      "step": 130,
+      "step_time": 25.31505165609851
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 600.5,
+      "completions/max_terminated_length": 600.5,
+      "completions/mean_length": 419.575,
+      "completions/mean_terminated_length": 419.575,
+      "completions/min_length": 260.8,
+      "completions/min_terminated_length": 260.8,
+      "entropy": 0.9347391467541456,
+      "epoch": 1.12,
+      "frac_reward_zero_std": 0.275,
+      "grad_norm": 2.765625,
+      "kl": 0.08032974656671285,
+      "learning_rate": 1.3125000000000001e-06,
+      "loss": 0.02551303505897522,
+      "num_tokens": 798651.0,
+      "reward": 1.0246130645275116,
+      "reward_std": 0.39613366425037383,
+      "rewards/JointRewardFunction/mean": 1.0246130645275116,
+      "rewards/JointRewardFunction/std": 0.3961336761713028,
+      "step": 140,
+      "step_time": 27.637355703599678
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 558.5,
+      "completions/max_terminated_length": 558.5,
+      "completions/mean_length": 394.375,
+      "completions/mean_terminated_length": 394.375,
+      "completions/min_length": 255.9,
+      "completions/min_terminated_length": 255.9,
+      "entropy": 0.841333020478487,
+      "epoch": 1.2,
+      "frac_reward_zero_std": 0.5,
+      "grad_norm": 0.11767578125,
+      "kl": 0.08763318308629095,
+      "learning_rate": 6.875000000000001e-07,
+      "loss": 0.025534918904304503,
+      "num_tokens": 850777.0,
+      "reward": 1.1621960639953612,
+      "reward_std": 0.28392985463142395,
+      "rewards/JointRewardFunction/mean": 1.1621960639953612,
+      "rewards/JointRewardFunction/std": 0.2839298591017723,
+      "step": 150,
+      "step_time": 25.526650765698285
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 569.3,
+      "completions/max_terminated_length": 569.3,
+      "completions/mean_length": 417.65,
+      "completions/mean_terminated_length": 417.65,
+      "completions/min_length": 253.4,
+      "completions/min_terminated_length": 253.4,
+      "entropy": 0.9602293692529201,
+      "epoch": 1.28,
+      "frac_reward_zero_std": 0.35,
+      "grad_norm": 1.8671875,
+      "kl": 0.08848655968904495,
+      "learning_rate": 6.250000000000001e-08,
+      "loss": 0.002139208652079105,
+      "num_tokens": 905091.0,
+      "reward": 1.0879150748252868,
+      "reward_std": 0.3240631450258661,
+      "rewards/JointRewardFunction/mean": 1.0879150748252868,
+      "rewards/JointRewardFunction/std": 0.32406314978725276,
+      "step": 160,
+      "step_time": 26.06666781829881
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 160,
+  "num_input_tokens_seen": 905091,
+  "num_train_epochs": 2,
+  "save_steps": 10,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
--- a/last-checkpoint/training_args.bin
+++ b/last-checkpoint/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db782e4d76747adb11d7171c927a87cf0f1d59c124b6da85b01ec6e29b1259b7
+size 7249
--- a/model.safetensors
+++ b/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0feb0dd2ec4125bfaf367198c52c9eb8d2fd0461e96e1797c629ba61162aeb71
+size 6171927112
--- a/tokenizer.json
+++ b/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892
--- a/tokenizer_config.json
+++ b/tokenizer_config.json
@@ -0,0 +1,30 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": false,
+  "local_files_only": false,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}
--- a/training_args.bin
+++ b/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db782e4d76747adb11d7171c927a87cf0f1d59c124b6da85b01ec6e29b1259b7
+size 7249