Enable ROCm devices to correctly use keras.layers.CuDNNGRU and keras.layers.CuDNNLSTM #21443
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
fchollet
reviewed
Jul 14, 2025
.github/workflows/format.yml
Outdated
| @@ -20,7 +20,7 @@ jobs: | |||
| python -m pip install --upgrade pip setuptools | |||
| echo "::set-output name=dir::$(pip cache dir)" | |||
| - name: pip cache | |||
| uses: actions/cache@v2 | |||
| uses: actions/cache@v4 | |||
There was a problem hiding this comment.
Please remove these 2 files from the PR
There was a problem hiding this comment.
Reverted to the original state before the changes.
| self.bias[self.units * 5 :], | ||
| ] | ||
|
|
||
| if tf.sysconfig.get_build_info()["is_cuda_build"]: |
There was a problem hiding this comment.
The default layout is compatible with ROCm; here, the CUDA layout is adapted by checking whether the device is CUDA.
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Outdated suggestions cannot be applied.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.
Using keras.layers.CuDNNGRU and keras.layers.CuDNNLSTM on ROCm devices may introduce accuracy issues due to incorrect weight layout. This modification ensures compatibility with both CUDA and ROCm weight layouts.