Move the quantization API to OSS #3997
Conversation
Summary: As titled. Differential Revision: D58680953
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/3997
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: ✅ You can merge normally! (1 Unrelated Failure)As of commit 19fa9ed with merge base bcabc37 ( BROKEN TRUNK - The following job failed but were present on the merge base:👉 Rebase onto the `viable/strict` branch to avoid these failures
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
|
This pull request was exported from Phabricator. Differential Revision: D58680953 |
| # pyre-ignore[33]: `InputsArgsType` cannot alias to `Any`. | ||
| InputsArgsType = Tuple[Any, ...] |
There was a problem hiding this comment.
Perhaps object instead to fix this error message?
| # pyre-ignore[33]: `InputsArgsType` cannot alias to `Any`. | |
| InputsArgsType = Tuple[Any, ...] | |
| InputsArgsType = Tuple[object, ...] |
| # Calibrate | ||
| prepared_model(*inputs) |
There was a problem hiding this comment.
You'll need to calibrate with more than one sample normally for good PTQ accuracy. Is this API meant more for testing?
There was a problem hiding this comment.
yup this is all dummy for now. We don't even look at quantization accuracy, only implementation accuracy (ref outputs are from the converted forward model).
|
This pull request has been merged in d11e8e1. |
Summary: As titled.
Differential Revision: D58680953