Closed A2va closed 1 year ago
Also I have a question, in cli_lora_pti.py
we can add more than one concept (person/style/object) but how to train two concepts with the same data set ? Do I need to caption files with the concept token ?
Thanks! You can optionally provide captions and use token_map to train two concepts
With this change, it's possible to preprocess files with manual captioning