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

[i1] Optimization to add packed_storage to i1 tensor #19604

Draft
wants to merge 10 commits into
base: main
Choose a base branch
from

Conversation

lialan
Copy link
Contributor

@lialan lialan commented Jan 5, 2025

  • A new optimization pass as part of GlobalOptimization pipeline
  • Finds tensor inputs with i1 element types, and attach packed_storage attribute to it.
  • Provides finer control via commandline option: which ones to pack and which ones to not pack.

lialan added 9 commits January 8, 2025 06:44
* make `packed_storage` as a type of `iree_encoding` attribute, and make type converters accept it.
* `i1` tensors with `#iree_encoding.packed_storage` will be interpreted as packed i1 type, same as specifying `--iree-experimental-packed-i1-storage`.
* `--iree-experimental-packed-i1-storage` are kept for testing purposes. We can drop this option after frontend enables emitting `i1` tensors with attributes.

Signed-off-by: Alan Li <[email protected]>
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