End of training

README.md

@@ -1,9 +1,11 @@
 ---
 base_model: Qwen/Qwen3-30B-A3B
+datasets: HectorHe/math7k
 library_name: transformers
 model_name: Qwen3-MOE-sft-math7k
 tags:
 - generated_from_trainer
+- open-r1
 - trl
 - sft
 licence: license
@@ -11,7 +13,7 @@ licence: license
 
 # Model Card for Qwen3-MOE-sft-math7k
 
-This model is a fine-tuned version of [Qwen/Qwen3-30B-A3B](https://huggingface.co/Qwen/Qwen3-30B-A3B).
+This model is a fine-tuned version of [Qwen/Qwen3-30B-A3B](https://huggingface.co/Qwen/Qwen3-30B-A3B) on the [HectorHe/math7k](https://huggingface.co/datasets/HectorHe/math7k) dataset.
 It has been trained using [TRL](https://github.com/huggingface/trl).
 
 ## Quick start

config.json

@@ -32,7 +32,7 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.52.0.dev0",
-  "use_cache": false,
+  "use_cache": true,
   "use_sliding_window": false,
   "vocab_size": 151936
 }

training.log

@@ -2245,3 +2245,5 @@ weight_decay=0.0,
   (lm_head): Linear(in_features=2048, out_features=151936, bias=False)
 )
 2025-08-18 23:22:14 - INFO - __main__ - *** Save model ***
+2025-08-18 23:40:03 - INFO - __main__ - Model saved to data/Qwen3-MOE/sft/math7k
+2025-08-18 23:40:03 - INFO - __main__ - Pushing to hub...