Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Update deepspeed example #2587

Merged
merged 2 commits into from
Sep 11, 2023
Merged

Update deepspeed example #2587

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
ae2c96e
Update dependencies in DeepSpeed example
Sep 11, 2023
63e3251
Update dependencies in DeepSpeed example README.md
Sep 11, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
19 changes: 18 additions & 1 deletion examples/large_models/deepspeed/Readme.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,8 +2,25 @@

This document briefs on serving large HuggingFace (HF) models on multiple GPUs using deepspeed. We are using facebook/opt-30b in this example

To run this example we need to have deepspeed installed. This has been added to the requirement.txt which can be bundled during model packaging.
### Pre-requisites

- Install CUDA. Verified to be working with CUDA 11.7.
- Verified to be working with:

```bash
torch 2.0.1+cu117
torch-model-archiver 0.8.2
torch-workflow-archiver 0.2.10
torchaudio 2.0.2+cu117
torchdata 0.6.1
torchserve 0.8.2
torchtext 0.15.2+cpu
torchvision 0.15.2+cu117
transformers 4.33.1
deepspeed 0.10.2
```

To run this example we need to have deepspeed installed. This has been added to the requirement.txt which can be bundled during model packaging.

```bash
pip install deepspeed
Expand Down
Loading