Lightning-AI / Lightning-AI/lightning-thunder

Split saved for backward information based on types

Open
#959 0 comments 1 reaction 1 assignee View on GitHub

@IvanYashchuk is already working on this.

Since Aug 19, 2024.

autograd
Dominant language
Python
Stars
1.5k
Forks
121
No merged PRs in 30d

Description

I didn't expect that there could be a pass somewhere in the pipeline that changes saved non-tensor information. Do you think it would be more clear in the trace to have three buckets instead of current two:

  1. saved tensors
  2. saved non-tensor proxies
  3. any other Python object

Originally posted by @IvanYashchuk in https://github.com/Lightning-AI/lightning-thunder/pull/937#discussion_r1713723890

Contributor guide

No contributing guide indexed for this repository

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.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.