How to use To use Eagle3 with SGLang, first replace the qwen3.py file in SGLang’s directory (sglang/python/sglang/srt/models/) with the qwen3.py file from this project.
The launch command for using Eagle3 with SGLang is:
python3 -m sglang.launch_server --model Qwen/Qwen3-4B-Instruct-2507 --speculative-algorithm EAGLE3 --speculative-draft-model-path Tengyunw/qwen3_4b_eagle3 --speculative-num-steps 6 --speculative-eagle-topk 10 --speculative-num-draft-tokens 32 --mem-fraction 0.9 --cuda-graph-max-bs 2 --dtype bfloat16