-
Notifications
You must be signed in to change notification settings - Fork 625
Add ability to add MIL ops to ET that are not yet in coremltools #12648
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
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/12648
Note: Links to docs will display an error until the docs builds have been completed. ⏳ No Failures, 95 PendingAs of commit d8eb2b6 with merge base 0fbd6d4 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
This PR needs a
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes, this great! I'd definitely recommend to add this! This has the potential to go beyond, e.g. to handle custom ops
Can I get a stamp @cccclai? I need someone from PyTorch to approve to merge |
b37ff5f
to
ec79df7
Compare
This is CoreML support for blockwise/channelwise quantization of linear/embedding layers using torchao APIs. This PR is based on top of #12648, and will be rebased once that lands to show the changes isolated to quantization.
…uantization APIs (#12665) This switches the ANE model to use to_edge_transform_and_lower and torchao quantization APIs. To use to_edge_transform_and_lower, we first need to land: #12629 To use torchao quant APIs, we first need to land #12648 and #12664. This PR contains all of the changes from those PRs because it is rebased on them. I will rebase on main once those PRs land to make this easier to review.
…orch#12648) This introduces torch_ops.py in ET's CoreML compiler. This lets us add critical MIL ops that are not yet in ET's version of coremltools.
This is CoreML support for blockwise/channelwise quantization of linear/embedding layers using torchao APIs. This PR is based on top of pytorch#12648, and will be rebased once that lands to show the changes isolated to quantization.
…uantization APIs (pytorch#12665) This switches the ANE model to use to_edge_transform_and_lower and torchao quantization APIs. To use to_edge_transform_and_lower, we first need to land: pytorch#12629 To use torchao quant APIs, we first need to land pytorch#12648 and pytorch#12664. This PR contains all of the changes from those PRs because it is rebased on them. I will rebase on main once those PRs land to make this easier to review.
This introduces torch_ops.py in ET's CoreML compiler. This lets us add critical MIL ops that are not yet in ET's version of coremltools.