Lightning-AI / Lightning-AI/pytorch-lightning
Feature request: FSDP native strategy for TPUs
Open
Nobody has claimed this yet.
accelerator: tpu
feature
strategy: fsdp
- Dominant language
- Python
- Stars
- 31.4k
- Forks
- 3.8k
- Avg merge
- 6d 7h
- Merged PRs (30d)
- 6
Description
A recent contribution to the pytorch_xla repo allows using FSDP in PyTorch XLA for sharding Module parameters across data-parallel workers. https://github.com/pytorch/xla/pull/3431.
cc @borda @kaushikb11 @rohitgr7 @SeanNaren @awaelchli @akihironitta
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 PyTorch/XLA pull request 3431 to understand the FSDP support for sharding module parameters across data-parallel workers. Then inspect how strategies are organized in pytorch-lightning and define the integration needed for native FSDP on TPUs; done means the strategy is supported and covered by appropriate tests.
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
- 25/100