Skip to content

Fix pipeline dtype unexpected change when using SDXL reference community pipelines in float16 mode #10670

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Jan 28, 2025
Merged

Fix pipeline dtype unexpected change when using SDXL reference community pipelines in float16 mode #10670

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 7 additions & 1 deletion examples/community/stable_diffusion_xl_controlnet_reference.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -193,7 +193,8 @@ class StableDiffusionXLControlNetReferencePipeline(StableDiffusionXLControlNetPi

def prepare_ref_latents(self, refimage, batch_size, dtype, device, generator, do_classifier_free_guidance):
refimage = refimage.to(device=device)
if self.vae.dtype == torch.float16 and self.vae.config.force_upcast:
needs_upcasting = self.vae.dtype == torch.float16 and self.vae.config.force_upcast
if needs_upcasting:
self.upcast_vae()
refimage = refimage.to(next(iter(self.vae.post_quant_conv.parameters())).dtype)
if refimage.dtype != self.vae.dtype:
Expand Down Expand Up @@ -223,6 +224,11 @@ def prepare_ref_latents(self, refimage, batch_size, dtype, device, generator, do

# aligning device to prevent device errors when concating it with the latent model input
ref_image_latents = ref_image_latents.to(device=device, dtype=dtype)

# cast back to fp16 if needed
if needs_upcasting:
self.vae.to(dtype=torch.float16)

return ref_image_latents

def prepare_ref_image(
Expand Down
8 changes: 7 additions & 1 deletion examples/community/stable_diffusion_xl_reference.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -139,7 +139,8 @@ def retrieve_timesteps(
class StableDiffusionXLReferencePipeline(StableDiffusionXLPipeline):
def prepare_ref_latents(self, refimage, batch_size, dtype, device, generator, do_classifier_free_guidance):
refimage = refimage.to(device=device)
if self.vae.dtype == torch.float16 and self.vae.config.force_upcast:
needs_upcasting = self.vae.dtype == torch.float16 and self.vae.config.force_upcast
if needs_upcasting:
self.upcast_vae()
refimage = refimage.to(next(iter(self.vae.post_quant_conv.parameters())).dtype)
if refimage.dtype != self.vae.dtype:
Expand Down Expand Up @@ -169,6 +170,11 @@ def prepare_ref_latents(self, refimage, batch_size, dtype, device, generator, do

# aligning device to prevent device errors when concating it with the latent model input
ref_image_latents = ref_image_latents.to(device=device, dtype=dtype)

# cast back to fp16 if needed
if needs_upcasting:
self.vae.to(dtype=torch.float16)

return ref_image_latents

def prepare_ref_image(
Expand Down