Add support for using AffineQuantizedTensor with weights_only=True
#630
Conversation
This file contains 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/630
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 009d28e with merge base 8bba8ed ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
jerryzh168
force-pushed
the
add-aqt-to-safe-global
branch
3 times, most recently
from
a8b7ee6
to
42ccee6
Compare
Summary: `torch.load(file, weights_only=True)` is safer so ideally we can use that, by default it does not work with tensor subclasses, since now we have https://pytorch.org/docs/main/notes/serialization.html#torch.serialization.add_safe_globals we can add all tensor subclass classes and special types to globals so that these can work with `weights_only=True` Test Plan: python test/dtypes/test_affine_quantized.py Reviewers: Subscribers: Tasks: Tags:
jerryzh168
force-pushed
the
add-aqt-to-safe-global
branch
from
42ccee6
to
009d28e
Compare
| @@ -466,3 +481,7 @@ def int8_dynamic_activation_int8_semi_sparse_weight(): | |||
| """ | |||
| from torchao.dtypes import SemiSparseLayoutType | |||
| return int8_dynamic_activation_int8_weight(layout_type=SemiSparseLayoutType()) | |||
|
|
|||
|
|
|||
| if TORCH_VERSION_AFTER_2_5: | |||
There was a problem hiding this comment.
A n00b qn: it wasn't super clear to me why these functions are used during unpickling?
but I trust that you added them because they appeared in the error message :)
There was a problem hiding this comment.
yeah that's correct, these functions are serialized as well when we do dynamic quantization
msaroufim
approved these changes
Aug 8, 2024
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:
torch.load(file, weights_only=True)is safer so ideally we can use that, by default it does not work with tensor subclasses, since now we have https://pytorch.org/docs/main/notes/serialization.html#torch.serialization.add_safe_globals we can add all tensor subclass classes and special types to globals so that these can work withweights_only=TrueTest Plan:
python test/dtypes/test_affine_quantized.py
Reviewers:
Subscribers:
Tasks:
Tags: