Update conversion script to transformers v4.2.0 #166
Closed
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
Signed-off-by: Adam Wawrzyński <adamwawrzynski@wp.pl>
Signed-off-by: Adam Wawrzyński <adamwawrzynski@wp.pl>
Signed-off-by: Adam Wawrzyński <adamwawrzynski@wp.pl>
|
Hi Adam, i use your script to convert a model to longformer, which works well. But when i load the saved model, i get the warning: Some weights of the model checkpoint at output/xlm-roberta-longformer-unsupervised/0_TransformerWrapper were not used when initializing RobertaLongModel: followed by a list of all weights and biases in the attention layers. Also the performance of the model is drastically worse than before saving and reloading it. Do you know how to fix this? |
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Outdated suggestions cannot be applied.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.
I have updated conversion script from
transformers3.0.2 to 4.2.0. Any suggestions and comments are welcommed.Signed-off-by: Adam Wawrzyński adamwawrzynski@wp.pl