SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

🚨Early-error🚨 config will error out if output_attentions=True and the attn implementation is wrong (#38288)

Browse Source 
* Protect ParallelInterface

* early error out on output attention setting for no wraning in modeling

* modular update

* fixup

* update model tests

* update

* oups

* set model's config

* more cases

* ??

* properly fix

* fixup

* update

* last onces

* update

* fix?

* fix wrong merge commit

* fix hub test

* nits

* wow I am tired

* updates

* fix pipeline!

---------

Co-authored-by: Lysandre <hi@lysand.re>
This commit is contained in:
Arthur
2025-05-23 17:17:38 +02:00
committed by GitHub
parent 896833c183
commit f5d45d89c4
71 changed files with 157 additions and 144 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
tests/models/encoder_decoder/test_modeling_encoder_decoder.py
View File

@@ -457,6 +457,7 @@ class EncoderDecoderMixin:
decoder_attention_mask = decoder_attention_mask[:, :-1]
encoder_model, decoder_model = self.get_encoder_decoder_model(config, decoder_config)
enc_dec_model = EncoderDecoderModel(encoder=encoder_model, decoder=decoder_model)
enc_dec_model.config._attn_implementation = "eager" # model config -> won't work
enc_dec_model.config.output_attentions = True # model config -> won't work
enc_dec_model.to(torch_device)
outputs_encoder_decoder = enc_dec_model(