lm-sys / lm-sys/FastChat

Is there a way to combine data parallel and model parallel?

Open
#1,524 3 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 would like to inquire about the possibility of combining data parallelism and model parallelism in the context of training llm. I found that the model parallel only support 1 batch while data parallel can not distribute one model to many cards. If I have 1000 1080ti cards and I want train a 65B model in a big batch size, what should I do?

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

The issue names no files, tests, or entry points; begin by clarifying whether a concrete FastChat change is wanted for combining data and model parallelism. A complete issue should identify the relevant training path and a test or reproducible configuration, plus define what successful large-batch 65B training looks like.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
distributed-systems, machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.