Lightning-AI / Lightning-AI/litgpt

Pretrain with Llama3.1-70b

Open
#1,828 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

help wanted question
Dominant language
Python
Stars
13.7k
Forks
1.5k
Avg merge
15h 37m
Merged PRs (30d)
1

Description

I am using litgpt to pretrain Llama3.1-70b with 4 nodes each having 8 h100 Gpus but still I am getting Cuda out of memory error. Also what pytorch version should I use? I am using global batch_size 8, and mini batch size 1

Any suggestions for this case?

Contributor guide

Open the contributing guide

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. Start by reviewing the Llama3.1-70b pretraining setup and the reported global and mini batch sizes across four 8×H100 nodes; done would require resolving the CUDA out-of-memory error and identifying the compatible PyTorch version.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
distributed-systems, machine-learning
Issue type
Bug
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
15/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.