[CLIPGuidedStableDiffusion] take the correct text embeddings (#667)

take the correct text embeddings

[CLIPGuidedStableDiffusion] take the correct text embeddings (#667)
take the correct text embeddings
c16761e9 · Suraj Patil · GitHub · 7f31142c · c16761e9
Unverified Commit c16761e9 authored Sep 28, 2022 by Suraj Patil Committed by GitHub Sep 28, 2022
Show whitespace changes
Inline Side-by-side

Showing with 1 addition and 1 deletion

examples/community/clip_guided_stable_diffusion.py examples/community/clip_guided_stable_diffusion.py +1 -1

No files found.
--- a/examples/community/clip_guided_stable_diffusion.py
+++ b/examples/community/clip_guided_stable_diffusion.py
@@ -284,7 +284,7 @@ class CLIPGuidedStableDiffusion(DiffusionPipeline):
            # perform clip guidance
            if clip_guidance_scale > 0:
                text_embeddings_for_guidance = (
-                    text_embeddings.chunk(2)[0] if do_classifier_free_guidance else text_embeddings
+                    text_embeddings.chunk(2)[1] if do_classifier_free_guidance else text_embeddings
                )
                noise_pred, latents = self.cond_fn(
                    latents,