Xcodec fix #42095
Open
Xcodec fix #42095
+54
−4
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
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
Contributor
|
[For maintainers] Suggested jobs to run (before merge) run-slow: xcodec |
Cyrilvallez
reviewed
Nov 11, 2025
Comment on lines
+399
to
+431
| @lru_cache | ||
| def _get_conv1d_layers(self, module): | ||
| """ | ||
| Recursively iterate to fetch all Conv1d layers. | ||
| """ | ||
|
|
||
| def get_conv1d_layers_recursive(module: nn.Module): | ||
| params_list = [] | ||
|
|
||
| if isinstance(module, nn.Conv1d): | ||
| params_list.append(module) | ||
|
|
||
| # Recursively check all child modules | ||
| for child in module.children(): | ||
| params_list.extend(get_conv1d_layers_recursive(child)) | ||
|
|
||
| return params_list | ||
|
|
||
| return tuple(get_conv1d_layers_recursive(module)) | ||
|
|
||
| def _get_conv1d_output_lengths(self, input_length, module=None): | ||
| """ | ||
| For a given module, compute the output length that would be obtained after all Conv1d layers. | ||
| """ | ||
| if module is None: | ||
| module = self | ||
|
|
||
| conv1d_layers = self._get_conv1d_layers(module) | ||
|
|
||
| for layer in conv1d_layers: | ||
| input_length = conv1d_output_length(layer, input_length) | ||
|
|
||
| return input_length |
Member
There was a problem hiding this comment.
Is this something we can do earlier, in the config? It would avoid having to do these recursions on modules etc!
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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Outdated suggestions cannot be applied.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.
What does this PR do?
Before a deeper clean (#42039) on xcodec and to break things down a bit, this PR fixes some inefficient logic in X-codec, responsible for unnecessary memory spikes.
Makes me think that the whole
_get_output_lengthwhen applying conv1d, which is repeated a lot throughout audio models, could benefit from some standardisation. In that objective, I've added a new audio util that I'll propagate in a subsequent PR.