Model save

README.md

@@ -1,11 +1,9 @@
 ---
 base_model: Qwen/Qwen2.5-7B-Instruct
-datasets: open-r1/OpenR1-Math-220k
 library_name: transformers
 model_name: Qwen2.5-7B-Open-R1-GRPO
 tags:
 - generated_from_trainer
-- open-r1
 - trl
 - grpo
 licence: license
@@ -13,7 +11,7 @@ licence: license
 
 # Model Card for Qwen2.5-7B-Open-R1-GRPO
 
-This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the [open-r1/OpenR1-Math-220k](https://huggingface.co/datasets/open-r1/OpenR1-Math-220k) dataset.
+This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-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/chenyukang2020-nvidia/huggingface/runs/7lknxmxa) 
+[<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/chenyukang2020-nvidia/huggingface/runs/n1iuzt67) 
 
 
 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": 3.769922398664146e-06,
-    "train_runtime": 502.782,
+    "train_runtime": 505.2607,
     "train_samples": 93733,
-    "train_samples_per_second": 186.429,
-    "train_steps_per_second": 11.653
+    "train_samples_per_second": 185.514,
+    "train_steps_per_second": 11.596
 }

config.json

@@ -22,7 +22,7 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.52.3",
-  "use_cache": true,
+  "use_cache": false,
   "use_sliding_window": false,
   "vocab_size": 152064
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
     "train_loss": 3.769922398664146e-06,
-    "train_runtime": 502.782,
+    "train_runtime": 505.2607,
     "train_samples": 93733,
-    "train_samples_per_second": 186.429,
-    "train_steps_per_second": 11.653
+    "train_samples_per_second": 185.514,
+    "train_steps_per_second": 11.596
 }

trainer_state.json

@@ -175524,7 +175524,7 @@
       "completions/min_terminated_length": 0.0,
       "epoch": 0.9987198088247845,
       "frac_reward_zero_std": 1.0,
-      "grad_norm": 6.168677994468242e-10,
+      "grad_norm": 6.168677684106501e-10,
       "kl": 0.071533203125,
       "learning_rate": 1.4376004014216017e-10,
       "loss": 0.0029,
@@ -175554,7 +175554,7 @@
       "completions/min_terminated_length": 0.0,
       "epoch": 0.9988905009814799,
       "frac_reward_zero_std": 1.0,
-      "grad_norm": 5.005282155356922e-10,
+      "grad_norm": 5.00527050853811e-10,
       "kl": 0.06787109375,
       "learning_rate": 1.1358823698404131e-10,
       "loss": 0.0027,
@@ -175584,7 +175584,7 @@
       "completions/min_terminated_length": 0.0,
       "epoch": 0.9990611931381753,
       "frac_reward_zero_std": 1.0,
-      "grad_norm": 4.5292041093372395e-10,
+      "grad_norm": 4.529207300217758e-10,
       "kl": 0.0650634765625,
       "learning_rate": 8.696603252866808e-11,
       "loss": 0.0026,
@@ -175614,7 +175614,7 @@
       "completions/min_terminated_length": 0.0,
       "epoch": 0.9992318852948707,
       "frac_reward_zero_std": 1.0,
-      "grad_norm": 6.505299056700279e-10,
+      "grad_norm": 6.505254899019415e-10,
       "kl": 0.072265625,
       "learning_rate": 6.389343622403844e-11,
       "loss": 0.0029,
@@ -175644,7 +175644,7 @@
       "completions/min_terminated_length": 0.0,
       "epoch": 0.9994025774515661,
       "frac_reward_zero_std": 1.0,
-      "grad_norm": 7.184587540539436e-10,
+      "grad_norm": 7.184588305303324e-10,
       "kl": 0.06884765625,
       "learning_rate": 4.437045625915737e-11,
       "loss": 0.0028,
@@ -175674,7 +175674,7 @@
       "completions/min_terminated_length": 0.0,
       "epoch": 0.9995732696082615,
       "frac_reward_zero_std": 1.0,
-      "grad_norm": 5.548114154084975e-10,
+      "grad_norm": 5.548119591819174e-10,
       "kl": 0.0714111328125,
       "learning_rate": 2.839709956625747e-11,
       "loss": 0.0029,
@@ -175704,7 +175704,7 @@
       "completions/min_terminated_length": 0.0,
       "epoch": 0.999743961764957,
       "frac_reward_zero_std": 1.0,
-      "grad_norm": 4.3193985862669656e-10,
+      "grad_norm": 4.3193996079624006e-10,
       "kl": 0.0662841796875,
       "learning_rate": 1.5973371813027273e-11,
       "loss": 0.0027,
@@ -175734,7 +175734,7 @@
       "completions/min_terminated_length": 0.0,
       "epoch": 0.9999146539216524,
       "frac_reward_zero_std": 1.0,
-      "grad_norm": 4.106144063560871e-10,
+      "grad_norm": 4.10613893427095e-10,
       "kl": 0.0687255859375,
       "learning_rate": 7.099277411493077e-12,
       "loss": 0.0028,
@@ -175754,9 +175754,9 @@
       "step": 5858,
       "total_flos": 0.0,
       "train_loss": 3.769922398664146e-06,
-      "train_runtime": 502.782,
-      "train_samples_per_second": 186.429,
-      "train_steps_per_second": 11.653
+      "train_runtime": 505.2607,
+      "train_samples_per_second": 185.514,
+      "train_steps_per_second": 11.596
     }
   ],
   "logging_steps": 1,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:63904ccc51b3f0c21106f66d629c7a09825a8a38ec9cd75c9fbc015dc398b196
+oid sha256:20297eedde9646fca6464e949b57008f04406ed98f5c4b86718e3f405dd728c2
 size 8440