lm-sys / lm-sys/FastChat

How to let vllm_worker to support load-8bits?

Open
#2,999 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
39.5k
Forks
4.8k
PR merge metrics
No merged PRs in 30d

Description

We want to reduce the usage of GPU memory by loading 8 bit model while faster the feedback by vllm_worker.py. However, it seems vllm_worker not support load-8bits option. Will it have any way to do this?

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

Start by reading vllm_worker.py and trace how model-loading options are parsed and passed to the worker. Determine whether 8-bit loading can be supported through the existing path; done means the worker accepts the requested option, loads an 8-bit model, and verifies reduced GPU memory usage without breaking normal loading.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.