I'm having a problem with the easynegative embedding and as the only thing I found about this is another comment, i'm trying my luck too:
>> Notice: embeddings/easynegative.pt was trained on a model with an incompatible token dimension: 768 vs 8
The embedding does not show in the list and does not seem to do anything when I try to use it anyway. (No ">> Preparing tokens for textual inversion")