Make the quantized path the main testing path, and introduce a nop quantizer for fp32 cases #7915
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
…antizer for fp32 cases Summary: For a while, testing APIs were `quantize_and_run` and `run_and_verify`, with the former calling the latter. That flow is a bit inconvenient since the quantized and fp32 cases are not consistent, and the names are also inconsistent. This diff changes the two main APIs to become `export_run_and_verify` and `quantize_export_run_and_verify` to be more descriptive. It also changes the calling order; we now use a nop quantizer for the fp32 case, allowing us to use the exact same flow as the quantized cases. The existing `run_and_verify` function is made "private" (as far as python goes at least) and now takes in an `ExportedProgram` instead of the `torch.nn.Module` before. Finally, it removes the `eval()` part of `export_program`, since now everything should go through the quantizer (including as a nop). Reviewed By: zonglinpeng, hsharma35 Differential Revision: D67561806
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/7915
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit b72f4b5 with merge base d68ca28 ( 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: D67561806 |
hsharma35
approved these changes
Jan 23, 2025
YIWENX14
pushed a commit
that referenced
this pull request
Jan 28, 2025
…antizer for fp32 cases Differential Revision: D67561806 Pull Request resolved: #7915
zonglinpeng
pushed a commit
to zonglinpeng/executorch
that referenced
this pull request
Jan 30, 2025
…antizer for fp32 cases Differential Revision: D67561806 Pull Request resolved: pytorch#7915
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.
Summary:
For a while, testing APIs were
quantize_and_runandrun_and_verify, with the former calling the latter. That flow is a bit inconvenient since the quantized and fp32 cases are not consistent, and the names are also inconsistent.This diff changes the two main APIs to become
export_run_and_verifyandquantize_export_run_and_verifyto be more descriptive.It also changes the calling order; we now use a nop quantizer for the fp32 case, allowing us to use the exact same flow as the quantized cases.
The existing
run_and_verifyfunction is made "private" (as far as python goes at least) and now takes in anExportedPrograminstead of thetorch.nn.Modulebefore.Finally, it removes the
eval()part ofexport_program, since now everything should go through the quantizer (including as a nop).Reviewed By: zonglinpeng, hsharma35
Differential Revision: D67561806