Finegrain/refiners
1
0
Fork 0
mirror of https://github.com/finegrain-ai/refiners.git synced 2024-11-21 21:58:47 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Fix Value dimension in ImageCrossAttention

Browse source
This commit is contained in:
hugojarkoff 2024-01-17 15:53:50 +01:00 committed by Benjamin Trom
parent 7839c54ae8
commit a6a9c8b972
1 changed files with 1 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
src/refiners/foundationals/latent_diffusion/image_prompt.py
View file

@ -252,7 +252,7 @@ class ImageCrossAttention(fl.Chain):
fl.Chain(
fl.UseContext(context="ip_adapter", key="clip_image_embedding"),
fl.Linear(
in_features=text_cross_attention.key_embedding_dim,
in_features=text_cross_attention.value_embedding_dim,
out_features=text_cross_attention.inner_dim,
bias=text_cross_attention.use_bias,
device=text_cross_attention.device,