Model save

README.md

@@ -1,10 +1,8 @@
 ---
-datasets: AI-MO/NuminaMath-TIR
 library_name: transformers
 model_name: Qwen2.5-1.5B-Open-R1-GRPO
 tags:
 - generated_from_trainer
-- open-r1
 - trl
 - grpo
 licence: license
@@ -12,7 +10,7 @@ licence: license
 
 # Model Card for Qwen2.5-1.5B-Open-R1-GRPO
 
-This model is a fine-tuned version of [None](https://huggingface.co/None) on the [AI-MO/NuminaMath-TIR](https://huggingface.co/datasets/AI-MO/NuminaMath-TIR) dataset.
+This model is a fine-tuned version of [None](https://huggingface.co/None).
 It has been trained using [TRL](https://github.com/huggingface/trl).
 
 ## Quick start
@@ -28,18 +26,18 @@ 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/cuiluyi/huggingface/runs/9ncklfe9) 
+[<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/cuiluyi/huggingface/runs/evidr78o) 
 
 
 This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://huggingface.co/papers/2402.03300).
 
 ### Framework versions
 
-- TRL: 0.15.0.dev0
-- Transformers: 4.49.0.dev0
-- Pytorch: 2.5.1
-- Datasets: 3.2.0
-- Tokenizers: 0.21.0
+- TRL: 0.17.0.dev0
+- Transformers: 4.51.2
+- Pytorch: 2.6.0
+- Datasets: 3.5.0
+- Tokenizers: 0.21.1
 
 ## Citations
 
@@ -60,7 +58,7 @@ Cite TRL as:
 ```bibtex
 @misc{vonwerra2022trl,
 	title        = {{TRL: Transformer Reinforcement Learning}},
-	author       = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallouédec},
+	author       = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallou{\'e}dec},
 	year         = 2020,
 	journal      = {GitHub repository},
 	publisher    = {GitHub},

all_results.json

@@ -1,13 +1,8 @@
 {
-    "eval_loss": 0.012912314385175705,
-    "eval_runtime": 196.6453,
-    "eval_samples": 99,
-    "eval_samples_per_second": 0.503,
-    "eval_steps_per_second": 0.015,
     "total_flos": 0.0,
-    "train_loss": 69.57985635299504,
-    "train_runtime": 92793.5148,
-    "train_samples": 20000,
-    "train_samples_per_second": 0.216,
-    "train_steps_per_second": 0.027
+    "train_loss": 0.026552865276898957,
+    "train_runtime": 9272.812,
+    "train_samples": 93733,
+    "train_samples_per_second": 0.043,
+    "train_steps_per_second": 0.005
 }

generation_config.json

@@ -10,5 +10,5 @@
   "temperature": 0.7,
   "top_k": 20,
   "top_p": 0.8,
-  "transformers_version": "4.49.0.dev0"
+  "transformers_version": "4.51.2"
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 69.57985635299504,
-    "train_runtime": 92793.5148,
-    "train_samples": 20000,
-    "train_samples_per_second": 0.216,
-    "train_steps_per_second": 0.027
+    "train_loss": 0.026552865276898957,
+    "train_runtime": 9272.812,
+    "train_samples": 93733,
+    "train_samples_per_second": 0.043,
+    "train_steps_per_second": 0.005
 }