Promote Supermask out of prototype #1729
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/1729
Note: Links to docs will display an error until the docs builds have been completed. ⏳ No Failures, 1 PendingAs of commit 343f3fe with merge base 7b37eb0 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
jcaip
added
sparsity
topic: bc-breaking
cpuhrsch
approved these changes
Feb 18, 2025
liangel-02
pushed a commit
that referenced
this pull request
Aug 25, 2025
This PR promotes Supermask and block sparsity from prototype -> `torchao.sparsity`, instead of the `apply_supermask` function which was previously closely coupled with SAM. It adds a new public API for `SupermaskLinear`, which users can use to add Supermask to their models for training with ``` sparsify_(model, lambda x: SupermaskLinear.from_linear(x, block_size=64, sparsity_level=0.9) ``` To accelerate for inference, we convert the `SupermaskLinear` model back into a `nn.Linear`, which simplifies the Supermask logic: ``` sparsify_(model, lambda x: SupermaskLinear.to_linear(x, sparsity_level=0.9) ``` **bc-breaking** The previous prototype APIs, `torchao.sparsity.prototype.superblock.supermask` and `torchao.prototype.sparsity.superblock.supermask` have been deprecated. You can use `torchao.sparsity.supermask` instead.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR promotes Supermask and block sparsity from prototype ->
torchao.sparsity, instead of theapply_supermaskfunction which was previously closely coupled with SAM.It adds a new public API for
SupermaskLinear, which users can use to add Supermask to their models for training withTo accelerate for inference, we convert the
SupermaskLinearmodel back into ann.Linear, which simplifies the Supermask logic:bc-breaking
The previous prototype APIs,
torchao.sparsity.prototype.superblock.supermaskandtorchao.prototype.sparsity.superblock.supermaskhave been deprecated.You can use
torchao.sparsity.supermaskinstead.