ROCm OCP FP8 Support #1677
Open
ROCm OCP FP8 Support #1677
+22
−6
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/1677
Note: Links to docs will display an error until the docs builds have been completed. ❌ 1 New FailureAs of commit 438f2c5 with merge base ea7910e ( NEW FAILURE - The following job has failed:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
petrex
added
float8
topic: improvement
vkuzo
reviewed
Feb 6, 2025
| for arch in mxArchName: | ||
| if arch in archName: | ||
| return True | ||
| return False | ||
|
|
||
|
|
||
| def is_MI350(): |
There was a problem hiding this comment.
optional: if there is any public info (spec sheet, etc) we can link to about this hardware from the docblock, I think that would be awesome! same for below.
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Outdated suggestions cannot be applied.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.
TLDR: Quick fix for ROCm device check. OCP FP8 support status update.
This pull request includes changes to improve the handling of imports, update configurations, and add new utility functions in the
torchaolibrary. The most important changes include removing comments to avoid circular imports, updating the configuration for supported float8 types, and adding utility functions to check for specific GPU architectures.refer to : pytorch/pytorch#146632
Configuration updates:
torchao/float8/config.py: Updated the configuration for selecting the preferred float8 type pair to include support for OCP F8 variants in MI350/Navi4.New utility functions:
torchao/utils.py: Added new utility functionsis_MI350andis_Navi4to check for specific GPU architectures.Improvements to import handling:
torchao/dtypes/uintx/marlin_qqq_tensor.py: Removed comments to avoid circular imports in the__tensor_unflatten__andfrom_plainmethods. [1] [2]torchao/dtypes/uintx/marlin_sparse_layout.py: Removed comments to avoid circular imports in the__tensor_unflatten__andfrom_plainmethods. [1] [2]