Lightning-AI / Lightning-AI/pytorch-lightning

TPU related documentation updates

Open
#17,492 0 comments 2 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

accelerator: tpu docs fabric pl
Dominant language
Python
Stars
31.4k
Forks
3.8k
Avg merge
6d 7h
Merged PRs (30d)
6

Description

### 📚 Documentation

There's several TPU specific topics to update in the documentation before the v2.1 release

### Update the guide in Trainer with:
- [ ] v4 support
- [ ] gcloud instructions
- [ ] PjRT support

### Write a full guide for running on TPUs with Fabric:

- Installation + gcloud instructions
- How to use the accelerator
- Launching processes
- Where to use mark_step()
- Potential pitfalls and how to avoid them (recompilation etc)
- FSDP
...

Creating this issue to track this.

### Additional context

https://pytorch.org/xla/release/2.0 already contains a lot of useful information to pull from

cc @borda @carmocca @JackCaoG @steventk-g @Liyang90 @justusschock @awaelchli

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

Review the existing Trainer documentation and the PyTorch/XLA 2.0 TPU release guide first, then identify where the Trainer guide and Fabric documentation belong. Done means updating the Trainer guide for v4, gcloud, and PjRT, and adding a complete Fabric TPU guide covering installation, acceleration, process launching, mark_step(), pitfalls, and FSDP.

Written by the indexing model from the issue text.

Assessment

Tech stack
gcp, python
Domain
documentation, machine-learning
Issue type
Documentation
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.