Config: allow saving with generative parameters if untouched #33886
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
commented
Oct 2, 2024
gante
commented
Oct 2, 2024
regisss
approved these changes
Oct 2, 2024
|
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. |
gante
force-pushed
the
save_whisper_config
branch
from
98bf416 to
d8aec4d
Compare
gante
force-pushed
the
save_whisper_config
branch
from
d8aec4d to
29e63cb
Compare
gante
added
the
WIP
|
This approach is opening a can of worms 😠 Going to fix this issue another way |
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Outdated suggestions cannot be applied.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?
In a recent PR (#32659), we started raising exceptions if we were to store a model config with generative parameters. If we were saving a model, which saves the config, the generative parameters were moved to
GenerationConfigbefore saving thePreTrainedConfig, working around the issue when saving a model.However, a common use case was uncovered. If a user loads a
PreTrainedConfigwith generative parameters and tries to save it, the user has done nothing wrong and yet they see an exception. This PR lowers the exception to a warning in that situation, preventing the code from crashing :)(Thank you @regisss for flagging this issue and @ydshieh for the testing suggestion)