Result reproduction issue.
@liusong1222 is already working on this.
Since Sep 6, 2025.
- Dominant language
- Python
- Stars
- 1.7k
- Forks
- 181
- Avg merge
- 1d 3h
- Merged PRs (30d)
- 2
Description
Hello, may I ask what the testing environment for qwen3-14b-eagle3 is? I used your dataset/alpaca/question.jsonl data for testing, but I was not able to reproduce the reported results of speedup 2.16x, τ3.13.
My environment: 2 × RTX 4090, sglang
Test script:
CUDA_VISIBLE_DEVICES=6,7 python3 -m sglang.launch_server
--model /models/Qwen3-14B
--speculative-algorithm EAGLE3
--speculative-draft-model-path /models/EAGLE3-Qwen3-14B
--speculative-num-steps 3
--speculative-eagle-topk 10
--speculative-num-draft-tokens 60
--mem-fraction-static 0.8
--attention-backend fa3
--tp 2
--log-requests-level 3
2>&1 | tee server.log
result：
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Assessment
This issue has not been assessed yet.