clip_lora_vision_encoder & visual_projevtor with youtube dataset.

clip lora textencoder&textprojection about youtube dataset

Upload 3 files

Upload config.json

CREATE : xclip text encoder

initial commit