Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[tk5] Implement slice analysis and sufficient op coverage for a softmax kernel. #222

Merged
merged 5 commits into from
Dec 6, 2023

Conversation

stellaraccident
Copy link
Contributor

@stellaraccident stellaraccident commented Dec 6, 2023

  • Adds arithmetic binary broadcast.
  • Adds f32 arithmetic.
  • Handles torch ops exp, max, sum (only reduce to scalar at the moment).
  • General python signature validation and unpacking.
  • Some initial attribute propagation work needed to handle unsigned properly.
  • A start at full generality for symbolic dimensions and slice analysis. Needs to be reworked before too long but this gets us basic functionality on static kernels while leaving room for dynamic to emerge.

@stellaraccident stellaraccident changed the title [tk5] Implement slice analysis and sufficient typing to load/store softmax vectors [tk5] Implement slice analysis and sufficient op coverage for a softmax kernel. Dec 6, 2023
@stellaraccident stellaraccident marked this pull request as ready for review December 6, 2023 23:53
@stellaraccident stellaraccident merged commit 247396c into main Dec 6, 2023
3 checks passed
@stellaraccident stellaraccident deleted the tk5 branch December 6, 2023 23:54
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant