Model save

README.md

@@ -1,11 +1,9 @@
 ---
 base_model: Qwen/Qwen2.5-1.5B-Instruct
-datasets: Muennighoff/openaimath-100
 library_name: transformers
 model_name: Qwen2.5-1.5B-hl-false-v16
 tags:
 - generated_from_trainer
-- open-r1
 - trl
 - grpo
 licence: license
@@ -13,7 +11,7 @@ licence: license
 
 # Model Card for Qwen2.5-1.5B-hl-false-v16
 
-This model is a fine-tuned version of [Qwen/Qwen2.5-1.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-1.5B-Instruct) on the [Muennighoff/openaimath-100](https://huggingface.co/datasets/Muennighoff/openaimath-100) dataset.
+This model is a fine-tuned version of [Qwen/Qwen2.5-1.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-1.5B-Instruct).
 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/halos/runs/ns6mzzo7) 
+[<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/halos/runs/sp4dty7e) 
 
 
 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": 0.010697796469139575,
-    "train_runtime": 57363.3108,
+    "train_loss": 1.3027437901511778e-05,
+    "train_runtime": 176.4542,
     "train_samples": 1200000,
-    "train_samples_per_second": 24.992,
-    "train_steps_per_second": 0.028
+    "train_samples_per_second": 8124.487,
+    "train_steps_per_second": 9.068
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 0.010697796469139575,
-    "train_runtime": 57363.3108,
+    "train_loss": 1.3027437901511778e-05,
+    "train_runtime": 176.4542,
     "train_samples": 1200000,
-    "train_samples_per_second": 24.992,
-    "train_steps_per_second": 0.028
+    "train_samples_per_second": 8124.487,
+    "train_steps_per_second": 9.068
 }

trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.14933296000093332,
+  "epoch": 0.14942629310093392,
   "eval_steps": 500,
-  "global_step": 1600,
+  "global_step": 1601,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -38410,18 +38410,43 @@
       "step": 1600
     },
     {
-      "epoch": 0.14933296000093332,
-      "step": 1600,
+      "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": 0.013392857142857095,
+      "completions/max_length": 4096.0,
+      "completions/max_terminated_length": 2653.0,
+      "completions/mean_length": 653.8092041015625,
+      "completions/mean_terminated_length": 607.0825805664062,
+      "completions/min_length": 135.0,
+      "completions/min_terminated_length": 135.0,
+      "epoch": 0.14942629310093392,
+      "frac_reward_zero_std": 0.6071428656578064,
+      "grad_norm": 0.12178564816713333,
+      "learning_rate": 2e-07,
+      "loss": 0.0209,
+      "num_tokens": 1034400635.0,
+      "reward": 0.5680803656578064,
+      "reward_std": 0.16954627633094788,
+      "rewards/simpleverify_reward/mean": 0.5680803656578064,
+      "rewards/simpleverify_reward/std": 0.4956200420856476,
+      "step": 1601
+    },
+    {
+      "epoch": 0.14942629310093392,
+      "step": 1601,
       "total_flos": 0.0,
-      "train_loss": 0.010697796469139575,
-      "train_runtime": 57363.3108,
-      "train_samples_per_second": 24.992,
-      "train_steps_per_second": 0.028
+      "train_loss": 1.3027437901511778e-05,
+      "train_runtime": 176.4542,
+      "train_samples_per_second": 8124.487,
+      "train_steps_per_second": 9.068
     }
   ],
   "logging_steps": 1,
   "max_steps": 1600,
-  "num_input_tokens_seen": 1033719638,
+  "num_input_tokens_seen": 1034400635,
   "num_train_epochs": 1,
   "save_steps": 160,
   "stateful_callbacks": {