Popular repositories Loading

1. onpolicydistillation onpolicydistillation Public

   🛠️ Apply on-policy distillation to enhance Qwen3-0.6b's performance on GSM8K by learning from its own outputs, reducing bias during inference.

   Jupyter Notebook 2

2. trtgr trtgr Public
3. 5-yt6yuy 5-yt6yuy Public