retrieval-augmented-diffusion-models icon indicating copy to clipboard operation
retrieval-augmented-diffusion-models copied to clipboard
verified publisher icon CompVis

Combining the CLIP embeddings of text and KNNs: conditional_retrieval_encoder

Open MaximClouser opened this issue 1 year ago • 0 comments

What is the idea behind the unimplemented conditional_retrieval_encoder here? Would this be another encoder to combine the CLIP embs of the original text query and KNNs?

MaximClouser avatar Sep 18 '24 21:09 MaximClouser