Add files using upload-large-folder tool

.gitattributes

@@ -66,3 +66,10 @@ Qwen-0.5B-GRPO-Code-vllm-modified-v1/checkpoint-100/tokenizer.json filter=lfs di
 Qwen-0.5B-GRPO-Code-vllm-modified-v1/checkpoint-1600/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 llama3b-GRPO-Code-vllm/checkpoint-200/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-700/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-400/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-300/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-800/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-200/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-600/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-100/tokenizer.json filter=lfs diff=lfs merge=lfs -text

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-100/added_tokens.json

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

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-100/config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "Qwen/Qwen2.5-0.5B-Instruct",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 896,
+  "initializer_range": 0.02,
+  "intermediate_size": 4864,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 21,
+  "model_type": "qwen2",
+  "num_attention_heads": 14,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.48.2",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-100/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.1,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.48.2"
+}

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-100/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76a9c7f09fa1d2fb5198badfb65e763ac66f6972eedcca0cecc5feab433b7069
+size 14448

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-100/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f68a37892a1b445d21bb35cc10bf7a058a6f9ec8c363f5ed156ff4f49d90fb6
+size 14512

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-100/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:71d2161c9233860a0c55de4fa68dfcf16cdf4af9234c3d44fc611fcd896d1189
+size 1064

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-100/special_tokens_map.json

@@ -0,0 +1,25 @@
+{
+  "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": "<|im_end|>"
+}

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-100/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63a2951d5edfa5cc0a2346ef872f8c77a2920274cfc3b503b04e3799104dee80
+size 11422060

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-100/tokenizer_config.json

@@ -0,0 +1,208 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|im_end|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-100/trainer_state.json

@@ -0,0 +1,1633 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.02675943270002676,
+  "eval_steps": 500,
+  "global_step": 100,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "completion_length": 98.5,
+      "epoch": 0.0002675943270002676,
+      "grad_norm": 30.0,
+      "kl": 0.0,
+      "learning_rate": 5.3475935828877005e-08,
+      "loss": 0.0,
+      "reward": 1.0875000953674316,
+      "reward_std": 0.4607583284378052,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.6000000238418579,
+      "rewards/code_quality_reward_func": 0.17500001192092896,
+      "rewards/reasoning_quality_reward_func": 0.21250000596046448,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 1
+    },
+    {
+      "completion_length": 168.75,
+      "epoch": 0.0005351886540005352,
+      "grad_norm": 15.0625,
+      "kl": 0.0,
+      "learning_rate": 1.0695187165775401e-07,
+      "loss": -0.0,
+      "reward": 1.3125,
+      "reward_std": 0.6351761817932129,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.7250000238418579,
+      "rewards/code_quality_reward_func": 0.1875,
+      "rewards/reasoning_quality_reward_func": 0.2875000238418579,
+      "rewards/xml_structure_reward_func": 0.11250000447034836,
+      "step": 2
+    },
+    {
+      "completion_length": 138.75,
+      "epoch": 0.0008027829810008028,
+      "grad_norm": 7.84375,
+      "kl": 0.0004256499232724309,
+      "learning_rate": 1.6042780748663104e-07,
+      "loss": 0.0,
+      "reward": 0.9000000953674316,
+      "reward_std": 0.5,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.42500004172325134,
+      "rewards/code_quality_reward_func": 0.08750000596046448,
+      "rewards/reasoning_quality_reward_func": 0.2875000238418579,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 3
+    },
+    {
+      "completion_length": 161.125,
+      "epoch": 0.0010703773080010704,
+      "grad_norm": 6.96875,
+      "kl": 0.0005086653982289135,
+      "learning_rate": 2.1390374331550802e-07,
+      "loss": 0.0,
+      "reward": 1.25,
+      "reward_std": 0.7432811856269836,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.7250000238418579,
+      "rewards/code_quality_reward_func": 0.15000000596046448,
+      "rewards/reasoning_quality_reward_func": 0.25,
+      "rewards/xml_structure_reward_func": 0.125,
+      "step": 4
+    },
+    {
+      "completion_length": 108.125,
+      "epoch": 0.001337971635001338,
+      "grad_norm": 12.625,
+      "kl": 0.0007201232365332544,
+      "learning_rate": 2.6737967914438503e-07,
+      "loss": 0.0,
+      "reward": 0.8500000238418579,
+      "reward_std": 0.30000001192092896,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.30000001192092896,
+      "rewards/code_quality_reward_func": 0.125,
+      "rewards/reasoning_quality_reward_func": 0.32500001788139343,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 5
+    },
+    {
+      "completion_length": 158.125,
+      "epoch": 0.0016055659620016055,
+      "grad_norm": 9.875,
+      "kl": 0.0006567862583324313,
+      "learning_rate": 3.208556149732621e-07,
+      "loss": 0.0,
+      "reward": 0.8625000715255737,
+      "reward_std": 0.3379955291748047,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.40000003576278687,
+      "rewards/code_quality_reward_func": 0.16250000894069672,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 6
+    },
+    {
+      "completion_length": 103.875,
+      "epoch": 0.0018731602890018732,
+      "grad_norm": 11.75,
+      "kl": 0.0005918288370594382,
+      "learning_rate": 3.7433155080213904e-07,
+      "loss": 0.0,
+      "reward": 1.524999976158142,
+      "reward_std": 0.7623251676559448,
+      "rewards/basic_format_reward_func": 0.0625,
+      "rewards/code_execution_reward": 0.925000011920929,
+      "rewards/code_quality_reward_func": 0.21250000596046448,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.125,
+      "step": 7
+    },
+    {
+      "completion_length": 102.0,
+      "epoch": 0.0021407546160021407,
+      "grad_norm": 16.0,
+      "kl": 0.0005299834883771837,
+      "learning_rate": 4.2780748663101604e-07,
+      "loss": 0.0,
+      "reward": 0.9750000238418579,
+      "reward_std": 0.43620064854621887,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.5,
+      "rewards/code_quality_reward_func": 0.16249999403953552,
+      "rewards/reasoning_quality_reward_func": 0.21249999105930328,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 8
+    },
+    {
+      "completion_length": 168.25,
+      "epoch": 0.002408348943002408,
+      "grad_norm": 16.375,
+      "kl": 0.0011368428822606802,
+      "learning_rate": 4.812834224598931e-07,
+      "loss": 0.0,
+      "reward": 0.75,
+      "reward_std": 0.2748348116874695,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.30000001192092896,
+      "rewards/code_quality_reward_func": 0.08750000596046448,
+      "rewards/reasoning_quality_reward_func": 0.26249998807907104,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 9
+    },
+    {
+      "completion_length": 113.875,
+      "epoch": 0.002675943270002676,
+      "grad_norm": 10.875,
+      "kl": 0.0007178444066084921,
+      "learning_rate": 5.347593582887701e-07,
+      "loss": 0.0,
+      "reward": 0.9500000476837158,
+      "reward_std": 0.47081488370895386,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.42500004172325134,
+      "rewards/code_quality_reward_func": 0.16250000894069672,
+      "rewards/reasoning_quality_reward_func": 0.26249998807907104,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 10
+    },
+    {
+      "completion_length": 66.625,
+      "epoch": 0.0029435375970029436,
+      "grad_norm": 116.0,
+      "kl": 0.0015349595341831446,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 0.0,
+      "reward": 1.2625000476837158,
+      "reward_std": 0.6949611902236938,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.824999988079071,
+      "rewards/code_quality_reward_func": 0.13750000298023224,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 11
+    },
+    {
+      "completion_length": 92.5,
+      "epoch": 0.003211131924003211,
+      "grad_norm": 10.6875,
+      "kl": 0.0008166864863596857,
+      "learning_rate": 6.417112299465242e-07,
+      "loss": 0.0,
+      "reward": 1.8875000476837158,
+      "reward_std": 0.42763128876686096,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.274999976158142,
+      "rewards/code_quality_reward_func": 0.26249998807907104,
+      "rewards/reasoning_quality_reward_func": 0.25,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 12
+    },
+    {
+      "completion_length": 97.5,
+      "epoch": 0.0034787262510034785,
+      "grad_norm": 47.25,
+      "kl": 0.0013502974761649966,
+      "learning_rate": 6.951871657754011e-07,
+      "loss": 0.0,
+      "reward": 1.2000000476837158,
+      "reward_std": 0.7858108282089233,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.7000000476837158,
+      "rewards/code_quality_reward_func": 0.15000000596046448,
+      "rewards/reasoning_quality_reward_func": 0.2500000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 13
+    },
+    {
+      "completion_length": 115.75,
+      "epoch": 0.0037463205780037465,
+      "grad_norm": 10.8125,
+      "kl": 0.0006711420137435198,
+      "learning_rate": 7.486631016042781e-07,
+      "loss": 0.0,
+      "reward": 2.0500001907348633,
+      "reward_std": 1.0827668905258179,
+      "rewards/basic_format_reward_func": 0.125,
+      "rewards/code_execution_reward": 1.2999999523162842,
+      "rewards/code_quality_reward_func": 0.25,
+      "rewards/reasoning_quality_reward_func": 0.25,
+      "rewards/xml_structure_reward_func": 0.125,
+      "step": 14
+    },
+    {
+      "completion_length": 193.25,
+      "epoch": 0.004013914905004014,
+      "grad_norm": 8.875,
+      "kl": 0.0012524605263024569,
+      "learning_rate": 8.021390374331551e-07,
+      "loss": 0.0,
+      "reward": 0.7124999761581421,
+      "reward_std": 0.05386751890182495,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.20000000298023224,
+      "rewards/code_quality_reward_func": 0.0625,
+      "rewards/reasoning_quality_reward_func": 0.3499999940395355,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 15
+    },
+    {
+      "completion_length": 154.375,
+      "epoch": 0.004281509232004281,
+      "grad_norm": 9.0625,
+      "kl": 0.0007793009281158447,
+      "learning_rate": 8.556149732620321e-07,
+      "loss": 0.0,
+      "reward": 0.8250000476837158,
+      "reward_std": 0.2288675308227539,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.30000001192092896,
+      "rewards/code_quality_reward_func": 0.15000000596046448,
+      "rewards/reasoning_quality_reward_func": 0.2750000059604645,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 16
+    },
+    {
+      "completion_length": 119.5,
+      "epoch": 0.004549103559004549,
+      "grad_norm": 16.5,
+      "kl": 0.0010260213166475296,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 0.0,
+      "reward": 0.7625000476837158,
+      "reward_std": 0.23393532633781433,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.30000001192092896,
+      "rewards/code_quality_reward_func": 0.11250000447034836,
+      "rewards/reasoning_quality_reward_func": 0.25,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 17
+    },
+    {
+      "completion_length": 138.625,
+      "epoch": 0.004816697886004816,
+      "grad_norm": 10.75,
+      "kl": 0.0008449122542515397,
+      "learning_rate": 9.625668449197862e-07,
+      "loss": 0.0,
+      "reward": 0.9500000476837158,
+      "reward_std": 0.4422738254070282,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.4000000059604645,
+      "rewards/code_quality_reward_func": 0.16250000894069672,
+      "rewards/reasoning_quality_reward_func": 0.2875000238418579,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 18
+    },
+    {
+      "completion_length": 113.625,
+      "epoch": 0.005084292213005084,
+      "grad_norm": 12.0625,
+      "kl": 0.0007207915186882019,
+      "learning_rate": 1.0160427807486633e-06,
+      "loss": 0.0,
+      "reward": 1.0125000476837158,
+      "reward_std": 0.4223129451274872,
+      "rewards/basic_format_reward_func": 0.0625,
+      "rewards/code_execution_reward": 0.4000000059604645,
+      "rewards/code_quality_reward_func": 0.1875,
+      "rewards/reasoning_quality_reward_func": 0.25,
+      "rewards/xml_structure_reward_func": 0.11250000447034836,
+      "step": 19
+    },
+    {
+      "completion_length": 109.75,
+      "epoch": 0.005351886540005352,
+      "grad_norm": 14.1875,
+      "kl": 0.0009593804134055972,
+      "learning_rate": 1.0695187165775401e-06,
+      "loss": 0.0,
+      "reward": 1.1124999523162842,
+      "reward_std": 0.4327978491783142,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.6000000238418579,
+      "rewards/code_quality_reward_func": 0.1875000149011612,
+      "rewards/reasoning_quality_reward_func": 0.22500000894069672,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 20
+    },
+    {
+      "completion_length": 167.75,
+      "epoch": 0.00561948086700562,
+      "grad_norm": 9.4375,
+      "kl": 0.0019687237218022346,
+      "learning_rate": 1.1229946524064172e-06,
+      "loss": 0.0,
+      "reward": 0.7125000357627869,
+      "reward_std": 0.09787134826183319,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.20000000298023224,
+      "rewards/code_quality_reward_func": 0.11249999701976776,
+      "rewards/reasoning_quality_reward_func": 0.30000001192092896,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 21
+    },
+    {
+      "completion_length": 159.25,
+      "epoch": 0.005887075194005887,
+      "grad_norm": 8.4375,
+      "kl": 0.0009809941984713078,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 0.0,
+      "reward": 0.8125000596046448,
+      "reward_std": 0.3002312183380127,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.30000001192092896,
+      "rewards/code_quality_reward_func": 0.15000000596046448,
+      "rewards/reasoning_quality_reward_func": 0.25,
+      "rewards/xml_structure_reward_func": 0.11250000447034836,
+      "step": 22
+    },
+    {
+      "completion_length": 200.625,
+      "epoch": 0.006154669521006155,
+      "grad_norm": 10.375,
+      "kl": 0.0009175295126624405,
+      "learning_rate": 1.2299465240641713e-06,
+      "loss": 0.0,
+      "reward": 1.0,
+      "reward_std": 0.4392969608306885,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.40000003576278687,
+      "rewards/code_quality_reward_func": 0.26249998807907104,
+      "rewards/reasoning_quality_reward_func": 0.23749999701976776,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 23
+    },
+    {
+      "completion_length": 139.375,
+      "epoch": 0.006422263848006422,
+      "grad_norm": 22.5,
+      "kl": 0.0015095111448317766,
+      "learning_rate": 1.2834224598930483e-06,
+      "loss": 0.0,
+      "reward": 0.800000011920929,
+      "reward_std": 0.4590751528739929,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.375,
+      "rewards/code_quality_reward_func": 0.11250000447034836,
+      "rewards/reasoning_quality_reward_func": 0.21250000596046448,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 24
+    },
+    {
+      "completion_length": 50.25,
+      "epoch": 0.00668985817500669,
+      "grad_norm": 0.00201416015625,
+      "kl": 0.000902252912055701,
+      "learning_rate": 1.3368983957219254e-06,
+      "loss": 0.0,
+      "reward": 2.0,
+      "reward_std": 0.0,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.5,
+      "rewards/code_quality_reward_func": 0.25,
+      "rewards/reasoning_quality_reward_func": 0.15000000596046448,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 25
+    },
+    {
+      "completion_length": 88.875,
+      "epoch": 0.006957452502006957,
+      "grad_norm": 15.6875,
+      "kl": 0.0012402876745909452,
+      "learning_rate": 1.3903743315508022e-06,
+      "loss": 0.0,
+      "reward": 1.4124999046325684,
+      "reward_std": 0.9211608171463013,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.9750000238418579,
+      "rewards/code_quality_reward_func": 0.15000000596046448,
+      "rewards/reasoning_quality_reward_func": 0.1875,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 26
+    },
+    {
+      "completion_length": 202.5,
+      "epoch": 0.0072250468290072254,
+      "grad_norm": 9.9375,
+      "kl": 0.001193880569189787,
+      "learning_rate": 1.4438502673796793e-06,
+      "loss": 0.0,
+      "reward": 0.8125000596046448,
+      "reward_std": 0.2467355877161026,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.30000001192092896,
+      "rewards/code_quality_reward_func": 0.11250000447034836,
+      "rewards/reasoning_quality_reward_func": 0.30000001192092896,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 27
+    },
+    {
+      "completion_length": 217.5,
+      "epoch": 0.007492641156007493,
+      "grad_norm": 6.65625,
+      "kl": 0.002084344858303666,
+      "learning_rate": 1.4973262032085562e-06,
+      "loss": 0.0,
+      "reward": 0.987500011920929,
+      "reward_std": 0.2808602452278137,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.5,
+      "rewards/code_quality_reward_func": 0.13750000298023224,
+      "rewards/reasoning_quality_reward_func": 0.25,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 28
+    },
+    {
+      "completion_length": 144.375,
+      "epoch": 0.00776023548300776,
+      "grad_norm": 11.375,
+      "kl": 0.0022054852452129126,
+      "learning_rate": 1.5508021390374334e-06,
+      "loss": 0.0,
+      "reward": 0.9500000476837158,
+      "reward_std": 0.2717355787754059,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.5,
+      "rewards/code_quality_reward_func": 0.11249999701976776,
+      "rewards/reasoning_quality_reward_func": 0.23749999701976776,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 29
+    },
+    {
+      "completion_length": 90.0,
+      "epoch": 0.008027829810008028,
+      "grad_norm": 142.0,
+      "kl": 0.0038667870685458183,
+      "learning_rate": 1.6042780748663103e-06,
+      "loss": 0.0,
+      "reward": 0.9500000476837158,
+      "reward_std": 0.5573897957801819,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.4750000238418579,
+      "rewards/code_quality_reward_func": 0.17500001192092896,
+      "rewards/reasoning_quality_reward_func": 0.19999998807907104,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 30
+    },
+    {
+      "completion_length": 148.5,
+      "epoch": 0.008295424137008296,
+      "grad_norm": 10.4375,
+      "kl": 0.0013046960812062025,
+      "learning_rate": 1.6577540106951873e-06,
+      "loss": 0.0,
+      "reward": 1.8499999046325684,
+      "reward_std": 0.8840129375457764,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.2999999523162842,
+      "rewards/code_quality_reward_func": 0.23750001192092896,
+      "rewards/reasoning_quality_reward_func": 0.21249999105930328,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 31
+    },
+    {
+      "completion_length": 165.875,
+      "epoch": 0.008563018464008563,
+      "grad_norm": 11.4375,
+      "kl": 0.00392666133120656,
+      "learning_rate": 1.7112299465240642e-06,
+      "loss": 0.0,
+      "reward": 0.8875000476837158,
+      "reward_std": 0.4786737263202667,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.4000000059604645,
+      "rewards/code_quality_reward_func": 0.125,
+      "rewards/reasoning_quality_reward_func": 0.26249998807907104,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 32
+    },
+    {
+      "completion_length": 97.75,
+      "epoch": 0.008830612791008831,
+      "grad_norm": 14.125,
+      "kl": 0.0017194543033838272,
+      "learning_rate": 1.7647058823529414e-06,
+      "loss": 0.0,
+      "reward": 1.350000023841858,
+      "reward_std": 0.6012930870056152,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.8250000476837158,
+      "rewards/code_quality_reward_func": 0.21249999105930328,
+      "rewards/reasoning_quality_reward_func": 0.21249999105930328,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 33
+    },
+    {
+      "completion_length": 65.125,
+      "epoch": 0.009098207118009098,
+      "grad_norm": 14.375,
+      "kl": 0.006373521871864796,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.0,
+      "reward": 1.4249999523162842,
+      "reward_std": 0.30773499608039856,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.8999999761581421,
+      "rewards/code_quality_reward_func": 0.25,
+      "rewards/reasoning_quality_reward_func": 0.17499999701976776,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 34
+    },
+    {
+      "completion_length": 166.125,
+      "epoch": 0.009365801445009366,
+      "grad_norm": 11.3125,
+      "kl": 0.00691020954400301,
+      "learning_rate": 1.8716577540106954e-06,
+      "loss": 0.0,
+      "reward": 1.0750000476837158,
+      "reward_std": 0.6594308614730835,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.625,
+      "rewards/code_quality_reward_func": 0.10000000149011612,
+      "rewards/reasoning_quality_reward_func": 0.25,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 35
+    },
+    {
+      "completion_length": 93.625,
+      "epoch": 0.009633395772009633,
+      "grad_norm": 7.90625,
+      "kl": 0.005570035893470049,
+      "learning_rate": 1.9251336898395724e-06,
+      "loss": 0.0,
+      "reward": 1.5250000953674316,
+      "reward_std": 0.5560275316238403,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.9249999523162842,
+      "rewards/code_quality_reward_func": 0.23750001192092896,
+      "rewards/reasoning_quality_reward_func": 0.25,
+      "rewards/xml_structure_reward_func": 0.11249999701976776,
+      "step": 36
+    },
+    {
+      "completion_length": 228.25,
+      "epoch": 0.009900990099009901,
+      "grad_norm": 10.875,
+      "kl": 0.006528891623020172,
+      "learning_rate": 1.9786096256684497e-06,
+      "loss": 0.0,
+      "reward": 1.6499998569488525,
+      "reward_std": 0.4349328875541687,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.0750000476837158,
+      "rewards/code_quality_reward_func": 0.22499999403953552,
+      "rewards/reasoning_quality_reward_func": 0.25,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 37
+    },
+    {
+      "completion_length": 92.75,
+      "epoch": 0.010168584426010168,
+      "grad_norm": 10.125,
+      "kl": 0.004976172000169754,
+      "learning_rate": 2.0320855614973265e-06,
+      "loss": 0.0,
+      "reward": 1.2625000476837158,
+      "reward_std": 0.7337645292282104,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.7250000238418579,
+      "rewards/code_quality_reward_func": 0.22500000894069672,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.11250000447034836,
+      "step": 38
+    },
+    {
+      "completion_length": 113.25,
+      "epoch": 0.010436178753010436,
+      "grad_norm": 13.6875,
+      "kl": 0.006215811241418123,
+      "learning_rate": 2.0855614973262034e-06,
+      "loss": 0.0,
+      "reward": 1.0125000476837158,
+      "reward_std": 0.4158177375793457,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.5,
+      "rewards/code_quality_reward_func": 0.1875,
+      "rewards/reasoning_quality_reward_func": 0.22499999403953552,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 39
+    },
+    {
+      "completion_length": 134.25,
+      "epoch": 0.010703773080010704,
+      "grad_norm": 10.1875,
+      "kl": 0.002335732337087393,
+      "learning_rate": 2.1390374331550802e-06,
+      "loss": 0.0,
+      "reward": 1.600000023841858,
+      "reward_std": 0.05000003054738045,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.0,
+      "rewards/code_quality_reward_func": 0.2875000238418579,
+      "rewards/reasoning_quality_reward_func": 0.21249999105930328,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 40
+    },
+    {
+      "completion_length": 64.0,
+      "epoch": 0.010971367407010971,
+      "grad_norm": 24.875,
+      "kl": 0.004402505233883858,
+      "learning_rate": 2.1925133689839575e-06,
+      "loss": 0.0,
+      "reward": 1.9249999523162842,
+      "reward_std": 0.6726604700088501,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.399999976158142,
+      "rewards/code_quality_reward_func": 0.23749999701976776,
+      "rewards/reasoning_quality_reward_func": 0.1875,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 41
+    },
+    {
+      "completion_length": 102.75,
+      "epoch": 0.01123896173401124,
+      "grad_norm": 12.6875,
+      "kl": 0.008235585875809193,
+      "learning_rate": 2.2459893048128343e-06,
+      "loss": 0.0,
+      "reward": 1.1124999523162842,
+      "reward_std": 0.4286736845970154,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.6000000238418579,
+      "rewards/code_quality_reward_func": 0.20000000298023224,
+      "rewards/reasoning_quality_reward_func": 0.21250000596046448,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 42
+    },
+    {
+      "completion_length": 98.25,
+      "epoch": 0.011506556061011506,
+      "grad_norm": 38.0,
+      "kl": 0.006902622990310192,
+      "learning_rate": 2.2994652406417116e-06,
+      "loss": 0.0,
+      "reward": 1.0750000476837158,
+      "reward_std": 0.520801305770874,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.5750000476837158,
+      "rewards/code_quality_reward_func": 0.17500001192092896,
+      "rewards/reasoning_quality_reward_func": 0.22500000894069672,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 43
+    },
+    {
+      "completion_length": 158.75,
+      "epoch": 0.011774150388011774,
+      "grad_norm": 9.25,
+      "kl": 0.012133742682635784,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 0.0,
+      "reward": 1.225000023841858,
+      "reward_std": 0.4723786413669586,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.699999988079071,
+      "rewards/code_quality_reward_func": 0.17499999701976776,
+      "rewards/reasoning_quality_reward_func": 0.25,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 44
+    },
+    {
+      "completion_length": 119.25,
+      "epoch": 0.012041744715012041,
+      "grad_norm": 19.375,
+      "kl": 0.007604178972542286,
+      "learning_rate": 2.4064171122994653e-06,
+      "loss": 0.0,
+      "reward": 1.0125000476837158,
+      "reward_std": 0.5095207691192627,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.5,
+      "rewards/code_quality_reward_func": 0.1875,
+      "rewards/reasoning_quality_reward_func": 0.22500000894069672,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 45
+    },
+    {
+      "completion_length": 128.625,
+      "epoch": 0.01230933904201231,
+      "grad_norm": 10.375,
+      "kl": 0.007734506856650114,
+      "learning_rate": 2.4598930481283426e-06,
+      "loss": 0.0,
+      "reward": 1.0125000476837158,
+      "reward_std": 0.4967355728149414,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.5,
+      "rewards/code_quality_reward_func": 0.22500000894069672,
+      "rewards/reasoning_quality_reward_func": 0.1875,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 46
+    },
+    {
+      "completion_length": 100.25,
+      "epoch": 0.012576933369012578,
+      "grad_norm": 6.5,
+      "kl": 0.005467721726745367,
+      "learning_rate": 2.5133689839572194e-06,
+      "loss": 0.0,
+      "reward": 1.7374999523162842,
+      "reward_std": 0.2428133189678192,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.125,
+      "rewards/code_quality_reward_func": 0.2875000238418579,
+      "rewards/reasoning_quality_reward_func": 0.22499999403953552,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 47
+    },
+    {
+      "completion_length": 122.875,
+      "epoch": 0.012844527696012844,
+      "grad_norm": 8.875,
+      "kl": 0.010364975780248642,
+      "learning_rate": 2.5668449197860967e-06,
+      "loss": 0.0,
+      "reward": 1.2000000476837158,
+      "reward_std": 0.4000000059604645,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.6000000238418579,
+      "rewards/code_quality_reward_func": 0.2750000059604645,
+      "rewards/reasoning_quality_reward_func": 0.22499999403953552,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 48
+    },
+    {
+      "completion_length": 91.625,
+      "epoch": 0.013112122023013113,
+      "grad_norm": 11.1875,
+      "kl": 0.007891715504229069,
+      "learning_rate": 2.6203208556149735e-06,
+      "loss": 0.0,
+      "reward": 1.3624999523162842,
+      "reward_std": 0.5999225974082947,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.8250000476837158,
+      "rewards/code_quality_reward_func": 0.25,
+      "rewards/reasoning_quality_reward_func": 0.1875,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 49
+    },
+    {
+      "completion_length": 98.875,
+      "epoch": 0.01337971635001338,
+      "grad_norm": 9.8125,
+      "kl": 0.011884085834026337,
+      "learning_rate": 2.673796791443851e-06,
+      "loss": 0.0,
+      "reward": 1.649999976158142,
+      "reward_std": 0.5500816106796265,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.0499999523162842,
+      "rewards/code_quality_reward_func": 0.2750000059604645,
+      "rewards/reasoning_quality_reward_func": 0.22499999403953552,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 50
+    },
+    {
+      "completion_length": 167.875,
+      "epoch": 0.013647310677013648,
+      "grad_norm": 7.78125,
+      "kl": 0.024216562509536743,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 0.0,
+      "reward": 1.5499999523162842,
+      "reward_std": 0.6633045077323914,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.9500000476837158,
+      "rewards/code_quality_reward_func": 0.23749999701976776,
+      "rewards/reasoning_quality_reward_func": 0.26249998807907104,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 51
+    },
+    {
+      "completion_length": 103.75,
+      "epoch": 0.013914905004013914,
+      "grad_norm": 13.6875,
+      "kl": 0.008454602211713791,
+      "learning_rate": 2.7807486631016045e-06,
+      "loss": 0.0,
+      "reward": 1.4874999523162842,
+      "reward_std": 0.6402456760406494,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.925000011920929,
+      "rewards/code_quality_reward_func": 0.23750001192092896,
+      "rewards/reasoning_quality_reward_func": 0.22499999403953552,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 52
+    },
+    {
+      "completion_length": 56.125,
+      "epoch": 0.014182499331014183,
+      "grad_norm": 16.75,
+      "kl": 0.01777804270386696,
+      "learning_rate": 2.8342245989304818e-06,
+      "loss": 0.0,
+      "reward": 1.7374999523162842,
+      "reward_std": 0.5997022390365601,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.149999976158142,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.1875,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 53
+    },
+    {
+      "completion_length": 112.25,
+      "epoch": 0.014450093658014451,
+      "grad_norm": 11.3125,
+      "kl": 0.011312819086015224,
+      "learning_rate": 2.8877005347593586e-06,
+      "loss": 0.0,
+      "reward": 1.0625,
+      "reward_std": 0.4679570198059082,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.5,
+      "rewards/code_quality_reward_func": 0.23749999701976776,
+      "rewards/reasoning_quality_reward_func": 0.22499999403953552,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 54
+    },
+    {
+      "completion_length": 92.375,
+      "epoch": 0.014717687985014717,
+      "grad_norm": 5.71875,
+      "kl": 0.021408643573522568,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 0.0,
+      "reward": 1.4874999523162842,
+      "reward_std": 0.22499997913837433,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.8999999761581421,
+      "rewards/code_quality_reward_func": 0.2750000059604645,
+      "rewards/reasoning_quality_reward_func": 0.21250000596046448,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 55
+    },
+    {
+      "completion_length": 107.625,
+      "epoch": 0.014985282312014986,
+      "grad_norm": 7.1875,
+      "kl": 0.015812650322914124,
+      "learning_rate": 2.9946524064171123e-06,
+      "loss": 0.0,
+      "reward": 1.7374999523162842,
+      "reward_std": 0.2428133189678192,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.125,
+      "rewards/code_quality_reward_func": 0.3125,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 56
+    },
+    {
+      "completion_length": 83.125,
+      "epoch": 0.015252876639015252,
+      "grad_norm": 6.71875,
+      "kl": 0.009970920160412788,
+      "learning_rate": 3.0481283422459896e-06,
+      "loss": 0.0,
+      "reward": 1.1875,
+      "reward_std": 0.5055937170982361,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.6499999761581421,
+      "rewards/code_quality_reward_func": 0.22499999403953552,
+      "rewards/reasoning_quality_reward_func": 0.21250000596046448,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 57
+    },
+    {
+      "completion_length": 124.375,
+      "epoch": 0.01552047096601552,
+      "grad_norm": 7.15625,
+      "kl": 0.015437191352248192,
+      "learning_rate": 3.101604278074867e-06,
+      "loss": 0.0,
+      "reward": 1.5250000953674316,
+      "reward_std": 0.5685557126998901,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.925000011920929,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 58
+    },
+    {
+      "completion_length": 77.125,
+      "epoch": 0.01578806529301579,
+      "grad_norm": 10.25,
+      "kl": 0.027593251317739487,
+      "learning_rate": 3.1550802139037433e-06,
+      "loss": 0.0,
+      "reward": 2.0874998569488525,
+      "reward_std": 0.5924705266952515,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.5,
+      "rewards/code_quality_reward_func": 0.2875000238418579,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 59
+    },
+    {
+      "completion_length": 126.125,
+      "epoch": 0.016055659620016056,
+      "grad_norm": 8.5625,
+      "kl": 0.01432301290333271,
+      "learning_rate": 3.2085561497326205e-06,
+      "loss": 0.0,
+      "reward": 1.0625,
+      "reward_std": 0.24787135422229767,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.5,
+      "rewards/code_quality_reward_func": 0.26250001788139343,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 60
+    },
+    {
+      "completion_length": 119.875,
+      "epoch": 0.016323253947016322,
+      "grad_norm": 8.9375,
+      "kl": 0.024657145142555237,
+      "learning_rate": 3.262032085561498e-06,
+      "loss": 0.0,
+      "reward": 1.6375000476837158,
+      "reward_std": 0.6750324368476868,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.0749999284744263,
+      "rewards/code_quality_reward_func": 0.23749999701976776,
+      "rewards/reasoning_quality_reward_func": 0.22500000894069672,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 61
+    },
+    {
+      "completion_length": 128.5,
+      "epoch": 0.016590848274016592,
+      "grad_norm": 6.28125,
+      "kl": 0.011644430458545685,
+      "learning_rate": 3.3155080213903747e-06,
+      "loss": 0.0,
+      "reward": 1.625,
+      "reward_std": 0.3685557246208191,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.024999976158142,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 62
+    },
+    {
+      "completion_length": 103.125,
+      "epoch": 0.01685844260101686,
+      "grad_norm": 7.34375,
+      "kl": 0.013543471693992615,
+      "learning_rate": 3.368983957219252e-06,
+      "loss": 0.0,
+      "reward": 1.4875000715255737,
+      "reward_std": 0.2250000238418579,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.8999999761581421,
+      "rewards/code_quality_reward_func": 0.2875000238418579,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 63
+    },
+    {
+      "completion_length": 313.125,
+      "epoch": 0.017126036928017126,
+      "grad_norm": 10.1875,
+      "kl": 0.03165314346551895,
+      "learning_rate": 3.4224598930481284e-06,
+      "loss": 0.0,
+      "reward": 1.6124999523162842,
+      "reward_std": 0.6787031888961792,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.0499999523162842,
+      "rewards/code_quality_reward_func": 0.2750000059604645,
+      "rewards/reasoning_quality_reward_func": 0.1875,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 64
+    },
+    {
+      "completion_length": 115.25,
+      "epoch": 0.017393631255017392,
+      "grad_norm": 9.5625,
+      "kl": 0.017762087285518646,
+      "learning_rate": 3.4759358288770056e-06,
+      "loss": 0.0,
+      "reward": 1.774999976158142,
+      "reward_std": 0.4809401035308838,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.1749999523162842,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 65
+    },
+    {
+      "completion_length": 53.125,
+      "epoch": 0.017661225582017662,
+      "grad_norm": 25.125,
+      "kl": 0.025264494121074677,
+      "learning_rate": 3.529411764705883e-06,
+      "loss": 0.0,
+      "reward": 1.5499999523162842,
+      "reward_std": 0.09999998658895493,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.0,
+      "rewards/code_quality_reward_func": 0.2750000059604645,
+      "rewards/reasoning_quality_reward_func": 0.17499999701976776,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 66
+    },
+    {
+      "completion_length": 85.0,
+      "epoch": 0.01792881990901793,
+      "grad_norm": 0.00439453125,
+      "kl": 0.0155054721981287,
+      "learning_rate": 3.5828877005347597e-06,
+      "loss": 0.0,
+      "reward": 2.049999952316284,
+      "reward_std": 0.0,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.5,
+      "rewards/code_quality_reward_func": 0.25,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 67
+    },
+    {
+      "completion_length": 97.625,
+      "epoch": 0.018196414236018196,
+      "grad_norm": 9.875,
+      "kl": 0.02109808288514614,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.0,
+      "reward": 1.625,
+      "reward_std": 0.44999995827674866,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.024999976158142,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 68
+    },
+    {
+      "completion_length": 96.75,
+      "epoch": 0.018464008563018466,
+      "grad_norm": 10.0625,
+      "kl": 0.014365495182573795,
+      "learning_rate": 3.6898395721925134e-06,
+      "loss": 0.0,
+      "reward": 1.8874999284744263,
+      "reward_std": 0.45966240763664246,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.274999976158142,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.21249999105930328,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 69
+    },
+    {
+      "completion_length": 63.5,
+      "epoch": 0.018731602890018732,
+      "grad_norm": 8.8125,
+      "kl": 0.015685059130191803,
+      "learning_rate": 3.7433155080213907e-06,
+      "loss": 0.0,
+      "reward": 2.012500047683716,
+      "reward_std": 0.22499997913837433,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.399999976158142,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.21249999105930328,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 70
+    },
+    {
+      "completion_length": 79.5,
+      "epoch": 0.018999197217019,
+      "grad_norm": 6.5625,
+      "kl": 0.025880061089992523,
+      "learning_rate": 3.796791443850268e-06,
+      "loss": 0.0,
+      "reward": 2.2249999046325684,
+      "reward_std": 0.24999995529651642,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.625,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 71
+    },
+    {
+      "completion_length": 71.75,
+      "epoch": 0.019266791544019266,
+      "grad_norm": 8.75,
+      "kl": 0.026266396045684814,
+      "learning_rate": 3.850267379679145e-06,
+      "loss": 0.0,
+      "reward": 1.5,
+      "reward_std": 0.20000000298023224,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.8999999761581421,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 72
+    },
+    {
+      "completion_length": 123.5,
+      "epoch": 0.019534385871019536,
+      "grad_norm": 7.65625,
+      "kl": 0.040041256695985794,
+      "learning_rate": 3.903743315508022e-06,
+      "loss": 0.0,
+      "reward": 1.1749999523162842,
+      "reward_std": 0.6809400916099548,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.625,
+      "rewards/code_quality_reward_func": 0.25,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 73
+    },
+    {
+      "completion_length": 66.125,
+      "epoch": 0.019801980198019802,
+      "grad_norm": 14.1875,
+      "kl": 0.015473383478820324,
+      "learning_rate": 3.957219251336899e-06,
+      "loss": 0.0,
+      "reward": 1.9499999284744263,
+      "reward_std": 0.2788674831390381,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.375,
+      "rewards/code_quality_reward_func": 0.2750000059604645,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 74
+    },
+    {
+      "completion_length": 133.625,
+      "epoch": 0.02006957452502007,
+      "grad_norm": 7.9375,
+      "kl": 0.01520618237555027,
+      "learning_rate": 4.010695187165775e-06,
+      "loss": 0.0,
+      "reward": 1.4625000953674316,
+      "reward_std": 0.21746647357940674,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.800000011920929,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.26249998807907104,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 75
+    },
+    {
+      "completion_length": 99.125,
+      "epoch": 0.020337168852020335,
+      "grad_norm": 7.0,
+      "kl": 0.021426944062113762,
+      "learning_rate": 4.064171122994653e-06,
+      "loss": 0.0,
+      "reward": 1.625,
+      "reward_std": 0.49074774980545044,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.0499999523162842,
+      "rewards/code_quality_reward_func": 0.2875000238418579,
+      "rewards/reasoning_quality_reward_func": 0.1875,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 76
+    },
+    {
+      "completion_length": 121.375,
+      "epoch": 0.020604763179020606,
+      "grad_norm": 3.828125,
+      "kl": 0.01213662326335907,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 0.0,
+      "reward": 1.75,
+      "reward_std": 0.4358898997306824,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.149999976158142,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 77
+    },
+    {
+      "completion_length": 55.75,
+      "epoch": 0.020872357506020872,
+      "grad_norm": 11.25,
+      "kl": 0.042267389595508575,
+      "learning_rate": 4.171122994652407e-06,
+      "loss": 0.0,
+      "reward": 1.5374999046325684,
+      "reward_std": 0.05386751890182495,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.0,
+      "rewards/code_quality_reward_func": 0.2875000238418579,
+      "rewards/reasoning_quality_reward_func": 0.15000000596046448,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 78
+    },
+    {
+      "completion_length": 75.5,
+      "epoch": 0.02113995183302114,
+      "grad_norm": 13.5625,
+      "kl": 0.04684869945049286,
+      "learning_rate": 4.224598930481284e-06,
+      "loss": 0.0,
+      "reward": 1.5250000953674316,
+      "reward_std": 0.6907477378845215,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.9500000476837158,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.17499999701976776,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 79
+    },
+    {
+      "completion_length": 68.0,
+      "epoch": 0.02140754616002141,
+      "grad_norm": 11.0,
+      "kl": 0.05405609309673309,
+      "learning_rate": 4.2780748663101604e-06,
+      "loss": 0.0001,
+      "reward": 1.600000023841858,
+      "reward_std": 0.07886753231287003,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.0,
+      "rewards/code_quality_reward_func": 0.3125,
+      "rewards/reasoning_quality_reward_func": 0.1875,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 80
+    },
+    {
+      "completion_length": 76.875,
+      "epoch": 0.021675140487021675,
+      "grad_norm": 6.0625,
+      "kl": 0.020545657724142075,
+      "learning_rate": 4.331550802139038e-06,
+      "loss": 0.0,
+      "reward": 1.6624999046325684,
+      "reward_std": 0.25940635800361633,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.125,
+      "rewards/code_quality_reward_func": 0.25,
+      "rewards/reasoning_quality_reward_func": 0.1875,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 81
+    },
+    {
+      "completion_length": 52.0,
+      "epoch": 0.021942734814021942,
+      "grad_norm": 13.3125,
+      "kl": 0.04687309265136719,
+      "learning_rate": 4.385026737967915e-06,
+      "loss": 0.0,
+      "reward": 1.9499999284744263,
+      "reward_std": 0.5563814043998718,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.375,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.17500001192092896,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 82
+    },
+    {
+      "completion_length": 83.5,
+      "epoch": 0.02221032914102221,
+      "grad_norm": 8.5,
+      "kl": 0.020440079271793365,
+      "learning_rate": 4.438502673796792e-06,
+      "loss": 0.0,
+      "reward": 1.625,
+      "reward_std": 0.7579997181892395,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.0499999523162842,
+      "rewards/code_quality_reward_func": 0.2750000059604645,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 83
+    },
+    {
+      "completion_length": 71.5,
+      "epoch": 0.02247792346802248,
+      "grad_norm": 9.6875,
+      "kl": 0.0274663046002388,
+      "learning_rate": 4.491978609625669e-06,
+      "loss": 0.0,
+      "reward": 1.5499999523162842,
+      "reward_std": 0.6668300032615662,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.949999988079071,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 84
+    },
+    {
+      "completion_length": 79.5,
+      "epoch": 0.022745517795022745,
+      "grad_norm": 5.3125,
+      "kl": 0.02609388716518879,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.0,
+      "reward": 1.7249999046325684,
+      "reward_std": 0.24999995529651642,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.125,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 85
+    },
+    {
+      "completion_length": 104.75,
+      "epoch": 0.023013112122023012,
+      "grad_norm": 0.0023193359375,
+      "kl": 0.018861236050724983,
+      "learning_rate": 4.598930481283423e-06,
+      "loss": 0.0,
+      "reward": 1.5499999523162842,
+      "reward_std": 0.0,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.0,
+      "rewards/code_quality_reward_func": 0.25,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 86
+    },
+    {
+      "completion_length": 53.125,
+      "epoch": 0.023280706449023282,
+      "grad_norm": 12.8125,
+      "kl": 0.04781404882669449,
+      "learning_rate": 4.6524064171123e-06,
+      "loss": 0.0,
+      "reward": 1.75,
+      "reward_std": 0.6354056596755981,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.1749999523162842,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.17499999701976776,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 87
+    },
+    {
+      "completion_length": 117.375,
+      "epoch": 0.02354830077602355,
+      "grad_norm": 6.875,
+      "kl": 0.012227240018546581,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 0.0,
+      "reward": 1.5,
+      "reward_std": 0.20000000298023224,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.8999999761581421,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 88
+    },
+    {
+      "completion_length": 54.625,
+      "epoch": 0.023815895103023815,
+      "grad_norm": 11.8125,
+      "kl": 0.05620575696229935,
+      "learning_rate": 4.759358288770054e-06,
+      "loss": 0.0001,
+      "reward": 1.475000023841858,
+      "reward_std": 0.21811051666736603,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.8999999761581421,
+      "rewards/code_quality_reward_func": 0.2875000238418579,
+      "rewards/reasoning_quality_reward_func": 0.1875,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 89
+    },
+    {
+      "completion_length": 65.0,
+      "epoch": 0.024083489430024082,
+      "grad_norm": 0.01141357421875,
+      "kl": 0.04672177881002426,
+      "learning_rate": 4.812834224598931e-06,
+      "loss": 0.0,
+      "reward": 1.600000023841858,
+      "reward_std": 0.0,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.0,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 90
+    },
+    {
+      "completion_length": 94.5,
+      "epoch": 0.024351083757024352,
+      "grad_norm": 10.4375,
+      "kl": 0.03195090591907501,
+      "learning_rate": 4.866310160427808e-06,
+      "loss": 0.0,
+      "reward": 1.2750000953674316,
+      "reward_std": 0.43933194875717163,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.7000000476837158,
+      "rewards/code_quality_reward_func": 0.2750000059604645,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 91
+    },
+    {
+      "completion_length": 108.375,
+      "epoch": 0.02461867808402462,
+      "grad_norm": 10.125,
+      "kl": 0.02047823742032051,
+      "learning_rate": 4.919786096256685e-06,
+      "loss": 0.0,
+      "reward": 1.337499976158142,
+      "reward_std": 0.4249999523162842,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.800000011920929,
+      "rewards/code_quality_reward_func": 0.23749999701976776,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 92
+    },
+    {
+      "completion_length": 59.25,
+      "epoch": 0.024886272411024885,
+      "grad_norm": 6.09375,
+      "kl": 0.033340681344270706,
+      "learning_rate": 4.973262032085562e-06,
+      "loss": 0.0,
+      "reward": 1.5125000476837158,
+      "reward_std": 0.02499997615814209,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.0,
+      "rewards/code_quality_reward_func": 0.25,
+      "rewards/reasoning_quality_reward_func": 0.16250000894069672,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 93
+    },
+    {
+      "completion_length": 72.125,
+      "epoch": 0.025153866738025155,
+      "grad_norm": 12.9375,
+      "kl": 0.0372508242726326,
+      "learning_rate": 5.026737967914439e-06,
+      "loss": 0.0,
+      "reward": 1.712499976158142,
+      "reward_std": 0.5224937200546265,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.149999976158142,
+      "rewards/code_quality_reward_func": 0.26249998807907104,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 94
+    },
+    {
+      "completion_length": 82.875,
+      "epoch": 0.025421461065025422,
+      "grad_norm": 15.75,
+      "kl": 0.04930912330746651,
+      "learning_rate": 5.0802139037433165e-06,
+      "loss": 0.0,
+      "reward": 1.475000023841858,
+      "reward_std": 0.25,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.8999999761581421,
+      "rewards/code_quality_reward_func": 0.2750000059604645,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 95
+    },
+    {
+      "completion_length": 84.25,
+      "epoch": 0.02568905539202569,
+      "grad_norm": 5.28125,
+      "kl": 0.034752584993839264,
+      "learning_rate": 5.133689839572193e-06,
+      "loss": 0.0,
+      "reward": 2.0749998092651367,
+      "reward_std": 0.028867486864328384,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.5,
+      "rewards/code_quality_reward_func": 0.25,
+      "rewards/reasoning_quality_reward_func": 0.22499999403953552,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 96
+    },
+    {
+      "completion_length": 87.5,
+      "epoch": 0.025956649719025955,
+      "grad_norm": 10.5,
+      "kl": 0.01500009186565876,
+      "learning_rate": 5.187165775401069e-06,
+      "loss": 0.0,
+      "reward": 1.5999999046325684,
+      "reward_std": 0.43484222888946533,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.024999976158142,
+      "rewards/code_quality_reward_func": 0.2750000059604645,
+      "rewards/reasoning_quality_reward_func": 0.20000000298023224,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 97
+    },
+    {
+      "completion_length": 73.625,
+      "epoch": 0.026224244046026225,
+      "grad_norm": 6.21875,
+      "kl": 0.02606140449643135,
+      "learning_rate": 5.240641711229947e-06,
+      "loss": 0.0,
+      "reward": 1.7374999523162842,
+      "reward_std": 0.2749999463558197,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 1.125,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.21249999105930328,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 98
+    },
+    {
+      "completion_length": 151.375,
+      "epoch": 0.026491838373026492,
+      "grad_norm": 6.28125,
+      "kl": 0.027492664754390717,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 0.0,
+      "reward": 1.2750000953674316,
+      "reward_std": 0.2179449498653412,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.7000000476837158,
+      "rewards/code_quality_reward_func": 0.26249998807907104,
+      "rewards/reasoning_quality_reward_func": 0.21249999105930328,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 99
+    },
+    {
+      "completion_length": 113.875,
+      "epoch": 0.02675943270002676,
+      "grad_norm": 7.625,
+      "kl": 0.016676055267453194,
+      "learning_rate": 5.347593582887702e-06,
+      "loss": 0.0,
+      "reward": 1.5625,
+      "reward_std": 0.49034643173217773,
+      "rewards/basic_format_reward_func": 0.0,
+      "rewards/code_execution_reward": 0.925000011920929,
+      "rewards/code_quality_reward_func": 0.30000001192092896,
+      "rewards/reasoning_quality_reward_func": 0.23749999701976776,
+      "rewards/xml_structure_reward_func": 0.10000000149011612,
+      "step": 100
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3737,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-200/added_tokens.json

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

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-200/config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "Qwen/Qwen2.5-0.5B-Instruct",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 896,
+  "initializer_range": 0.02,
+  "intermediate_size": 4864,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 21,
+  "model_type": "qwen2",
+  "num_attention_heads": 14,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.48.2",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-200/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.1,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.48.2"
+}

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-200/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05710297877446a04f3920efba05a55a25dde11ea4b7e46a0983c84b19d90682
+size 14448

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-200/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f68a37892a1b445d21bb35cc10bf7a058a6f9ec8c363f5ed156ff4f49d90fb6
+size 14512

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-200/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11ebcc79730a243f22c7142afc98308dacd92b09dfab19d18c4de53c27124888
+size 1064

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-200/special_tokens_map.json

@@ -0,0 +1,25 @@
+{
+  "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": "<|im_end|>"
+}

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-200/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63a2951d5edfa5cc0a2346ef872f8c77a2920274cfc3b503b04e3799104dee80
+size 11422060

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-200/tokenizer_config.json

@@ -0,0 +1,208 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|im_end|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-200/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ae303199a3f5d31692ecd76e856b1fe809229c2b385ebaf722c0cfa122f0773
+size 5624

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-300/added_tokens.json

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

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-300/config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "Qwen/Qwen2.5-0.5B-Instruct",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 896,
+  "initializer_range": 0.02,
+  "intermediate_size": 4864,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 21,
+  "model_type": "qwen2",
+  "num_attention_heads": 14,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.48.2",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-300/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.1,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.48.2"
+}

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-300/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:439cdf1c33eecff95d718422547913643700b7ebeec4e21c93e23d66ed0e4ed7
+size 988097824

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-300/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c22e71e32943d11848f83241f6bee27a2cbfd89730e917622ca26176d6564040
+size 14448

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-300/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f68a37892a1b445d21bb35cc10bf7a058a6f9ec8c363f5ed156ff4f49d90fb6
+size 14512

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-300/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d02df7af88de4b6f674b0c33cc677c6a2ad4269da378066906d80d6a74ed8b1
+size 1064

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-300/special_tokens_map.json

@@ -0,0 +1,25 @@
+{
+  "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": "<|im_end|>"
+}

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-300/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63a2951d5edfa5cc0a2346ef872f8c77a2920274cfc3b503b04e3799104dee80
+size 11422060

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-300/tokenizer_config.json

@@ -0,0 +1,208 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|im_end|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-300/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ae303199a3f5d31692ecd76e856b1fe809229c2b385ebaf722c0cfa122f0773
+size 5624

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-400/added_tokens.json

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

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-400/config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "Qwen/Qwen2.5-0.5B-Instruct",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 896,
+  "initializer_range": 0.02,
+  "intermediate_size": 4864,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 21,
+  "model_type": "qwen2",
+  "num_attention_heads": 14,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.48.2",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-400/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.1,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.48.2"
+}

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-400/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0750ea889825dde49e05b3bb75078d98703e849b9102ebd2bbcb4b4ae26d905
+size 988097824

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-400/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c537187f43c0116f21beb3dcba86f99c57a2aa139b32330c80843a8007316757
+size 14448

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-400/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f68a37892a1b445d21bb35cc10bf7a058a6f9ec8c363f5ed156ff4f49d90fb6
+size 14512

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-400/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb71be99938129e57fc6761f65091470d0f07eabfd1b414a74a83ca9a2b33782
+size 1064

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-400/special_tokens_map.json

@@ -0,0 +1,25 @@
+{
+  "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": "<|im_end|>"
+}

Qwen-0.5B-GRPO-Code-vllm-base/checkpoint-400/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63a2951d5edfa5cc0a2346ef872f8c77a2920274cfc3b503b04e3799104dee80
+size 11422060