transformers
66db33dd - Fix mismatching loading in from_pretrained with/without accelerate (#28414)

Commit

1 year ago

Fix mismatching loading in from_pretrained with/without accelerate (#28414) * fix mismatching behavior in from_pretrained with/without accelerate * meaningful refactor * remove added space * add test * fix model on the hub * comment * use tiny model * style

References

#28414 - Fix mismatching loading in from_pretrained with/without accelerate

#29969 - [SigLIP] Add fast tokenizer

#32831 - [Docs] Update resources

#33111 - [Backbone] Remove out_features everywhere

#33174 - [Zero-shot image classification pipeline] Remove tokenizer_kwargs

#39821 - Support MetaCLIP 2

#58 - Add EoMT DINOv3 model

#59 - Fix attention mask handling in EoMT-DINOv3 converter

#41212 - Add EoMT with DINOv3 backbone

#62 - Add initial DEIMv2 model implementation

Author

fxmarty

Parents

002566f3

transformers 66db33dd - Fix mismatching loading in from_pretrained with/without accelerate (#28414)

transformers
66db33dd - Fix mismatching loading in from_pretrained with/without accelerate (#28414)