diffusers
diffusers copied to clipboard
simplyfy AttentionBlock
This simplifies AttentionBlock by always making q,k,v a 3D tensors like we do in CrossAttention. This way we can also leverage sliced attention and xformers attention in this block.
The documentation is not available anymore as the PR was closed or merged.
I assume all the model slow tests are passing? Merging and will check then