Whisper: remove redundant assisted generation tests #34814
+3
−165
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
gante
changed the title
|
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. |
Comment on lines
-1464
to
-1469
| elif has_default_max_length: # by default let's always generate 20 new tokens | ||
| if generation_config.max_length == GenerationConfig().max_length: | ||
| generation_config.max_length = generation_config.max_length + input_ids_length | ||
| max_position_embeddings = getattr(self.config, "max_position_embeddings", None) | ||
| if max_position_embeddings is not None: | ||
| generation_config.max_length = min(generation_config.max_length, max_position_embeddings) |
There was a problem hiding this comment.
No, this should be kept as it really helps the pipeline / default usage!
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Outdated suggestions cannot be applied.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.
What does this PR do?
Part of making CI green in preparation for #34807 (test fetcher fetches generation tests)
max_length-related changes. The default max length is20and, due to recent changes, a different length was being returned in some cases. Whisper was one of the affected models, it was defaulting to a length of21.test_assisted_decoding_encoder_decoder_shared_encoder(integration test) -- the point of this test was to test assisted generation with DistilWhisper-like structure, which used to have a custom input (assistant_encoder_outputs). a) That input is no longer present in Whisper b) the DistilWhisper structure didn't take off, so it's overkill to have an abstract test with dummy classes c) we have other DistilWhisper tests.test_model_kwarg_assisted_decoding_encoder_decoder(integration test) -- added when DistiWhisper was added. Meanwhile, assisted generation was modified to work with any encoder-decoder model, and we have mixin tests.