This commit is contained in:
Jukka Seppänen 2024-11-24 22:27:27 +02:00
commit 1ade29084e

View File

@ -507,6 +507,7 @@ class CogVideoXPipeline(DiffusionPipeline, CogVideoXLoraLoaderMixin):
image_cond_latents = image_cond_latents.repeat(1, latents.shape[1], 1, 1, 1)
else:
logger.info(f"Received {image_cond_latents.shape[1]} image conditioning frames")
image_cond_latents = image_cond_latents.to(self.vae_dtype)
# 6. Prepare extra step kwargs. TODO: Logic should ideally just be moved out of the pipeline
extra_step_kwargs = self.prepare_extra_step_kwargs(generator, eta)