Skip to content

[BUG] fixes in kadinsky pipeline #11080

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 8 commits into from
Apr 21, 2025
Merged

[BUG] fixes in kadinsky pipeline #11080

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
4f3bd13
bug fix kadinsky pipeline
ishan-modi Mar 17, 2025
e8e29ac
code quality updates
ishan-modi Mar 17, 2025
b24c6a8
update
ishan-modi Mar 17, 2025
e6fa809
update
ishan-modi Apr 9, 2025
32a5c8a
addressed PR comments
ishan-modi Apr 10, 2025
cdd54f4
addressed PR comments
ishan-modi Apr 10, 2025
410bc5a
update
ishan-modi Apr 10, 2025
237e502
Merge branch 'main' into fixes-issue-11060
ishan-modi Apr 19, 2025
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
12 changes: 5 additions & 7 deletions src/diffusers/pipelines/kandinsky2_2/pipeline_kandinsky2_2_controlnet_img2img.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -135,13 +135,11 @@ def __init__(
scheduler=scheduler,
movq=movq,
)
movq_scale_factor = 2 ** (len(self.movq.config.block_out_channels) - 1)
self.image_processor = VaeImageProcessor(
vae_scale_factor=movq_scale_factor,
vae_latent_channels=self.movq.config.latent_channels,
resample="bicubic",
reducing_gap=1,
)
kwargs = {}
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ohh we can just give them a default value like this

movq_scale_factor = 2 ** (len(self.movq.config.block_out_channels) - 1) if getattr(self, "movq", None) else ..
movq_latent_channels = self.movq.config.latent_channels if getattr(self, "movq", None) else ..

to be consistent with how it is handled in other pipelines, for example https://github.com/huggingface/diffusers/blob/main/src/diffusers/pipelines/stable_diffusion_3/pipeline_stable_diffusion_3.py#L218

if self.movq:
kwargs["vae_scale_factor"] = 2 ** (len(self.movq.config.block_out_channels) - 1)
kwargs["vae_latent_channels"] = self.movq.config.latent_channels
self.image_processor = VaeImageProcessor(resample="bicubic", reducing_gap=1, **kwargs)

# Copied from diffusers.pipelines.kandinsky.pipeline_kandinsky_img2img.KandinskyImg2ImgPipeline.get_timesteps
def get_timesteps(self, num_inference_steps, strength, device):
Expand Down
12 changes: 5 additions & 7 deletions src/diffusers/pipelines/kandinsky2_2/pipeline_kandinsky2_2_img2img.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -107,13 +107,11 @@ def __init__(
scheduler=scheduler,
movq=movq,
)
movq_scale_factor = 2 ** (len(self.movq.config.block_out_channels) - 1)
self.image_processor = VaeImageProcessor(
vae_scale_factor=movq_scale_factor,
vae_latent_channels=self.movq.config.latent_channels,
resample="bicubic",
reducing_gap=1,
)
kwargs = {}
if self.movq:
kwargs["vae_scale_factor"] = 2 ** (len(self.movq.config.block_out_channels) - 1)
kwargs["vae_latent_channels"] = self.movq.config.latent_channels
self.image_processor = VaeImageProcessor(resample="bicubic", reducing_gap=1, **kwargs)

# Copied from diffusers.pipelines.kandinsky.pipeline_kandinsky_img2img.KandinskyImg2ImgPipeline.get_timesteps
def get_timesteps(self, num_inference_steps, strength, device):
Expand Down
12 changes: 5 additions & 7 deletions src/diffusers/pipelines/kandinsky3/pipeline_kandinsky3_img2img.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -76,13 +76,11 @@ def __init__(
self.register_modules(
tokenizer=tokenizer, text_encoder=text_encoder, unet=unet, scheduler=scheduler, movq=movq
)
movq_scale_factor = 2 ** (len(self.movq.config.block_out_channels) - 1)
self.image_processor = VaeImageProcessor(
vae_scale_factor=movq_scale_factor,
vae_latent_channels=self.movq.config.latent_channels,
resample="bicubic",
reducing_gap=1,
)
kwargs = {}
if self.movq:
kwargs["vae_scale_factor"] = 2 ** (len(self.movq.config.block_out_channels) - 1)
kwargs["vae_latent_channels"] = self.movq.config.latent_channels
self.image_processor = VaeImageProcessor(resample="bicubic", reducing_gap=1, **kwargs)

def get_timesteps(self, num_inference_steps, strength, device):
# get the original timestep using init_timestep
Expand Down
Loading