Model save

README.md

@@ -1,11 +1,9 @@
 ---
 base_model: Qwen/Qwen3-8B
-datasets: ryzax/train_v6_filtered_math
 library_name: transformers
 model_name: 8B-v2
 tags:
 - generated_from_trainer
-- open-r1
 - trl
 - grpo
 licence: license
@@ -13,7 +11,7 @@ licence: license
 
 # Model Card for 8B-v2
 
-This model is a fine-tuned version of [Qwen/Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B) on the [ryzax/train_v6_filtered_math](https://huggingface.co/datasets/ryzax/train_v6_filtered_math) dataset.
+This model is a fine-tuned version of [Qwen/Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B).
 It has been trained using [TRL](https://github.com/huggingface/trl).
 
 ## Quick start
@@ -29,7 +27,7 @@ print(output["generated_text"])
 
 ## Training procedure
 
-[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/muennighoff/s2/runs/svl9q6ig) 
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/muennighoff/s2/runs/v6dzxsu9) 
 
 
 This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://huggingface.co/papers/2402.03300).

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 0.0017749488347908481,
-    "train_runtime": 14122.464,
+    "train_loss": 9.714211751292874e-06,
+    "train_runtime": 700.4321,
     "train_samples": 93894,
-    "train_samples_per_second": 18.127,
-    "train_steps_per_second": 0.071
+    "train_samples_per_second": 365.489,
+    "train_steps_per_second": 1.428
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 0.0017749488347908481,
-    "train_runtime": 14122.464,
+    "train_loss": 9.714211751292874e-06,
+    "train_runtime": 700.4321,
     "train_samples": 93894,
-    "train_samples_per_second": 18.127,
-    "train_steps_per_second": 0.071
+    "train_samples_per_second": 365.489,
+    "train_steps_per_second": 1.428
 }

trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.1704049247023239,
+  "epoch": 0.17057532962702623,
   "eval_steps": 500,
-  "global_step": 1000,
+  "global_step": 1001,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -36010,18 +36010,54 @@
       "tools/num_tool_detect_error": 0.0
     },
     {
-      "epoch": 0.1704049247023239,
-      "step": 1000,
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.1171875,
+      "completions/max_length": 2048.0,
+      "completions/max_terminated_length": 2033.0,
+      "completions/mean_length": 1211.97265625,
+      "completions/mean_terminated_length": 1101.0,
+      "completions/min_length": 20.0,
+      "completions/min_terminated_length": 20.0,
+      "entropy": 0.16800379753112793,
+      "epoch": 0.17057532962702623,
+      "frac_reward_zero_std": 0.4375,
+      "grad_norm": 0.2011258453130722,
+      "learning_rate": 1e-06,
+      "loss": 0.0097,
+      "num_tokens": 425104847.0,
+      "reward": 0.6484375,
+      "reward_std": 0.22965487837791443,
+      "rewards/simpleverify_reward/mean": 0.6484375,
+      "rewards/simpleverify_reward/std": 0.47839346528053284,
+      "step": 1001,
+      "tools/generated_tokens": 4155.984375,
+      "tools/num_python": 0.0,
+      "tools/num_python_exec_error": 0.0,
+      "tools/num_retrieval": 0.0,
+      "tools/num_retriever_exec_error": 0.0,
+      "tools/num_saving": 1.4375,
+      "tools/num_saving_exec_error": 0.0,
+      "tools/num_saving_forced": 0.0,
+      "tools/num_saving_invalid_use": 0.0,
+      "tools/num_tool_detect_error": 0.0
+    },
+    {
+      "epoch": 0.17057532962702623,
+      "step": 1001,
       "total_flos": 0.0,
-      "train_loss": 0.0017749488347908481,
-      "train_runtime": 14122.464,
-      "train_samples_per_second": 18.127,
-      "train_steps_per_second": 0.071
+      "train_loss": 9.714211751292874e-06,
+      "train_runtime": 700.4321,
+      "train_samples_per_second": 365.489,
+      "train_steps_per_second": 1.428
     }
   ],
   "logging_steps": 1,
   "max_steps": 1000,
-  "num_input_tokens_seen": 424717334,
+  "num_input_tokens_seen": 425104847,
   "num_train_epochs": 1,
   "save_steps": 100,
   "stateful_callbacks": {