Convert TF and Numpy ops in whisper_audio_convert.py to Keras Ops #2225
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
mattdangerw
reviewed
Apr 25, 2025
| outputs = WhisperAudioConverter(**self.init_kwargs)(audio_tensor) | ||
|
|
||
| # Verify shape. | ||
| self.assertEqual(outputs.shape, (5, 80)) | ||
| # Verify output. | ||
| expected = [1.1656, 1.0151, -0.8343, -0.8343, -0.8343] | ||
| self.assertAllClose(outputs[:, 0], expected, atol=0.01, rtol=0.01) | ||
| self.assertAllClose(outputs[:, 0], expected, atol=0.01, rtol=0.01) |
There was a problem hiding this comment.
keep newlines at end of files.
|
|
||
| # Convert the tensor to a Ragged Tensor. | ||
| if isinstance(audio, tf.Tensor): | ||
| audio = tf.RaggedTensor.from_tensor(audio) | ||
|
|
||
| # Pad audio. | ||
| audio_shape = audio.shape.as_list() | ||
| audio_shape = list(audio.shape) |
There was a problem hiding this comment.
I believe tf.shape cannot always be listified like this. Maybe call ops.shape?
|
@abheesht17 if you have time can you review this? You probably have the most insight here. |
mattdangerw
reviewed
Apr 25, 2025
| @@ -1,14 +1,10 @@ | |||
| import numpy as np | |||
| import keras.ops as ops | |||
| import tensorflow as tf | |||
There was a problem hiding this comment.
We can never do a bare import of tf like this. Check other files in the library.
| @@ -214,21 +219,21 @@ def call(self, audio): | |||
|
|
|||
| rank_1_input = audio.shape.rank == 1 | |||
| if rank_1_input: | |||
| audio = tf.expand_dims(audio, 0) | |||
| audio = ops.expand_dims(audio, 0) | |||
|
|
|||
| # Convert the tensor to a Ragged Tensor. | |||
| if isinstance(audio, tf.Tensor): | |||
There was a problem hiding this comment.
We could probably switch to a @preprocessing_function annotation and remove this? Something to try at least.
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.
This PR is in continuation to the draft PR .
Converted Numpy and TF ops into Keras ops in whisper_audio_converter.py and in whisper_audio_converter_test.py