NVIDIA-NeMo / NVIDIA-NeMo/Automodel
Integrate the new PyTorch replicate API for Expert data parallel
Open
Nobody has claimed this yet.
enhancement
- Dominant language
- Python
- Stars
- 963
- Forks
- 318
- Avg merge
- 3d 20h
- Merged PRs (30d)
- 143
Description
Describe the solution you'd like
Based on https://github.com/pytorch/torchtitan/pull/1714
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reading the linked torchtitan pull request 1714 to understand the new PyTorch replicate API and its expected Expert data-parallel integration. Then locate Automodel's Expert data-parallel entry points; done means the API is integrated and the relevant training path works with the new behavior.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- 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