Lightning-AI / Lightning-AI/lightning-thunder
Advanced Indexing with sequences
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 1.5k
- Forks
- 121
- PR merge metrics
- No merged PRs in 30d
Description
🚀 Feature
This is a follow up to PR #710.
Currently, thunder indexing supports the following (credit to @mruberry):
- basic indexing
- advanced indexing with a sequence of 0D or 1D integer tensors (possibly starting with an ellipsis)
- advanced indexing with a single / or sequence of
list(e.g.a[[1,2]],a[[1,2], [1,2]]) - indexing with a sequence that consists of a single number
and... - combination of basic indexing and advanced indexing that Thunder supports (e.g.
a[slice(1,3), [1,2]],a[None, [2,3]])
I am trying to extend this to support other sequences such as a[(1,2),]. I am seeking for advices and suggestions regarding what other indexing we want to support in Thunder.
cc. @mruberry , @t-vi
Contributor guide
No contributing guide indexed for this repository
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 with PR #710 and the existing indexing behavior listed in this issue. Review how Thunder currently handles integer-tensor, list, scalar, ellipsis, and mixed basic/advanced indexing, then identify which tuple or other sequence cases should be supported. Done requires an agreed scope for additional sequence indexing and corresponding support.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- compilers
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100