Starting a recipe for ESC50 #1605
Merged
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
merge wip with train_nmf
Collaborator
|
@ycemsubakan can you comply with the new recipe testing? We need to be able to test this recipe as well :-) |
Collaborator
|
hey @TParcollet @mravanelli, just finished fixing the recipe testing. now everything is ready for review! 😃 |
anautsch
reviewed
Mar 20, 2023
anautsch
reviewed
Mar 20, 2023
Collaborator
anautsch
left a comment
There was a problem hiding this comment.
This PR looks quite ready. I put some comments regarding docstring polishing. Lmk if I should go more nitty gritty - or if it is ok for your taste.
The recipe tests worked out just fine!
(1/6) Running test for ESC50_row_2...
... 56.49s
(2/6) Running test for ESC50_row_3...
... 27.63s
(3/6) Running test for ESC50_row_4...
... 9.81s
(4/6) Running test for ESC50_row_5...
... 16.64s
(5/6) Running test for ESC50_row_6...
... 13.01s
(6/6) Running test for ESC50_row_7...
... 12.71s
TEST PASSED
The README train calls run (fast epochs); the dataset is git-available. So, the data preparation works (which is outside of the recipe testing scope).
I like the +5,030 −0 changes :)
Three files were added to speechbrain/lobes/models. They contain docstring examples (which run as of PR workflows).
Adel-Moumen
reviewed
Mar 21, 2023
Adel-Moumen
reviewed
Mar 21, 2023
Adel-Moumen
reviewed
Mar 21, 2023
Adel-Moumen
reviewed
Mar 21, 2023
Adel-Moumen
reviewed
Mar 21, 2023
Adel-Moumen
reviewed
Mar 21, 2023
Adel-Moumen
reviewed
Mar 21, 2023
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
6 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Outdated suggestions cannot be applied.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.
Starting a recipe for ESC50. Not ready to merge yet, but wanted to start the work!
I used the UrbanSound8k recipe as the base.
I am seeing the ecapa-tdnn network is able to get ~50 % accuracy on test, and ~60 % accuracy on valid, when trained on folds [1, 2, 3] and fold 4 is used validation set.