hpcaitech/Elixir

Elixir: Train a Large Language Model on a Small GPU Cluster

29
/ 100
Experimental

This project helps machine learning engineers train very large language models efficiently, even with a smaller cluster of GPUs. It takes your model and optimizer configurations and automatically determines the most memory-efficient way to distribute parameters and manage memory across CPUs and GPUs. This is for ML engineers and researchers working on large-scale model training.

No commits in the last 6 months.

Use this if you need to train extremely large language models but are constrained by the memory capacity of your existing GPU cluster.

Not ideal if you are working with smaller models or already have access to a very large, high-memory GPU cluster.

large-language-models distributed-training GPU-optimization deep-learning-infrastructure ML-resource-management
No License Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 6 / 25
Maturity 8 / 25
Community 15 / 25

How are scores calculated?

Stars

15

Forks

5

Language

Python

License

Last pushed

Jun 08, 2023

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/transformers/hpcaitech/Elixir"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.