sgl-project / sgl-project/SpecForge

During training, the accuracy is below 1 and exhibits significant fluctuations.

Open
#365 0 comments 1 reaction 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
1.2k
Forks
347
Avg merge
4d 1h
Merged PRs (30d)
41

Description

Image When training qwen2.5-72B, the accuracy in the logs consistently remained below 1.0 and fluctuated significantly—sometimes around 0.2x, other times around 0.8x. After training, when I tested with SGLang=0.5.5, the acceleration effect was only 1.3X. Could you please advise what the issue might be?

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

No files, tests, or entry points are named. Start by reproducing Qwen2.5-72B training and inspect the accuracy logs alongside the reported SGLang 0.5.5 serving result; the issue is complete when the cause of the fluctuations and low acceleration is identified or documented with actionable reproduction details.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
machine-learning, performance
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.