FWXT
Results
2
comments of
FWXT
hi @shimmyshimmer , **Seems there is a bug when using lora finetune Qwen3, after push merged model to hub, I can't load saved merged model from hf correctly but random...
+1 encounter same issue, qwen2.5 coder 7b seems no speed up using https://huggingface.co/yuhuili/EAGLE-Qwen2-7B-Instruct as draft model. Running on a100. Vicuna series can speed up 3x.