Workaround transformers overwriting model_type when saving dpr models #765
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
This reverts commit 4af4b3d.
julian-risch
changed the title
Timoeller
approved these changes
Jun 8, 2021
There was a problem hiding this comment.
Nice one.
I like the parametrization and documentation of the tests (since they can be quite confusing).
You even fixed the typo in the docs about cty_encoder : )
I added a comment that would make the solution even nicer (I believe). I would be fine with merging already now.
…geModel parent class
|
❤️ ;) |
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Outdated suggestions cannot be applied.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.
For DPR models, transformers overwrites the model_type parameter specified in a model's config when saving the model:
https://github.com/huggingface/transformers/blob/64e78564a519cda2b4408803e2781c604e1e3bdd/src/transformers/configuration_utils.py#L626
For DPR models with a non-BERT tokenizer, e.g., CamembertTokenizer as in haystack issue #1046 this prevents loading the correct tokenizer when the model is loaded again. For example, the model_type camembert is specified in the config when loading the model from the model hub, but the model_type dpr is saved.
To overcome this problem, we set the model_type of transformers.DPRConfig on-the-fly to the model_type as specified in the model's config file.
This PR corresponds to haystack PR #1060