lm-sys / lm-sys/FastChat

Launching Embedding Module Separately Due to Insufficient VRAMs

Open
#1,971 0 comments 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

I've been going through the FastChat [documentation](https://github.com/lm-sys/FastChat/blob/main/docs/langchain_integration.md) and noticed the command to launch chat-completion/completion/embeddings using:

`python3 -m fastchat.serve.model_worker --model-names "gpt-3.5-turbo,text-davinci-003,text-embedding-ada-002" --model-path lmsys/vicuna-7b-v1.3`

However, due to VRAM limitations, I'm experiencing difficulties running the entire stack simultaneously. I'm interested in finding a way to solely launch the embedding module, independent of the language model.

Could you guide me on how to perform this operation? Your assistance is greatly appreciated.

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 with docs/langchain_integration.md and the documented fastchat.serve.model_worker command. Check whether the existing serving entry point supports launching only the embedding module with reduced VRAM usage; done means the standalone procedure is documented clearly.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
ai, documentation
Issue type
Documentation
Difficulty
2/5
Estimated time
1-3 hours
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.