CODEWITHSUNDEEP
huggingface-transformerscaptionhuggingface
Nina Grundlingh
Nina Grundlingh

Reputation: 1

BLIP encoder-based

I’m wanting to use BLIP for image captioning. How can I ensure that captions are generated by an encoder and not decoder? I’ve been using the huggingface model:https://huggingface.co/docs/transformers/model_doc/blip

Thanks!

I tried setting is_decoder=False in BlipConfigText to configure the model however, I can’t get my model to train.

Upvotes: 0

Views: 311

Answers (0)

Related Questions