[Question] fp8 flash attention
Open
question
- Dominant language
- Python
- Stars
- 7.4k
- Forks
- 742
- Avg merge
- 1d 1h
- Merged PRs (30d)
- 104
Description
### Required prerequisites
- [x] I have read the documentation .
- [x] I have searched the [Issue Tracker](https://github.com/tile-ai/tilelang/issues) that this hasn't already been reported. (comment there if it has.)
### Questions
How can I implement flash attention with fp8 ?
Contributor guide
Research direction
Start with the TileLang documentation linked in the prerequisites. The issue names no files, tests, examples, or acceptance criteria, so first clarify the intended FP8 flash-attention scope and what would count as done before attempting implementation.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Quiet
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100