AI/diffusers
1
0
Fork 0
mirror of https://github.com/huggingface/diffusers.git synced 2026-01-27 17:22:53 +03:00
Code Activity

Remove unnecessary offset in img2img (#1653)

Browse Source 
remove unnecessary offset in img2img
This commit is contained in:
Patrick von Platen
2022-12-10 19:26:25 +01:00
committed by GitHub
parent ea64a7860a
commit fc94c60c83
5 changed files with 10 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
src/diffusers/pipelines/alt_diffusion/pipeline_alt_diffusion_img2img.py
View File

@@ -376,11 +376,9 @@ class AltDiffusionImg2ImgPipeline(DiffusionPipeline):
def get_timesteps(self, num_inference_steps, strength, device):
# get the original timestep using init_timestep
offset = self.scheduler.config.get("steps_offset", 0)
init_timestep = int(num_inference_steps * strength) + offset
init_timestep = min(init_timestep, num_inference_steps)
init_timestep = min(int(num_inference_steps * strength), num_inference_steps)
t_start = max(num_inference_steps - init_timestep + offset, 0)
t_start = max(num_inference_steps - init_timestep, 0)
timesteps = self.scheduler.timesteps[t_start:]
return timesteps, num_inference_steps - t_start

6
src/diffusers/pipelines/stable_diffusion/pipeline_cycle_diffusion.py
View File

@@ -414,11 +414,9 @@ class CycleDiffusionPipeline(DiffusionPipeline):
# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion_img2img.StableDiffusionImg2ImgPipeline.get_timesteps
def get_timesteps(self, num_inference_steps, strength, device):
# get the original timestep using init_timestep
offset = self.scheduler.config.get("steps_offset", 0)
init_timestep = int(num_inference_steps * strength) + offset
init_timestep = min(init_timestep, num_inference_steps)
init_timestep = min(int(num_inference_steps * strength), num_inference_steps)
t_start = max(num_inference_steps - init_timestep + offset, 0)
t_start = max(num_inference_steps - init_timestep, 0)
timesteps = self.scheduler.timesteps[t_start:]
return timesteps, num_inference_steps - t_start

6
src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion_depth2img.py
View File

@@ -323,11 +323,9 @@ class StableDiffusionDepth2ImgPipeline(DiffusionPipeline):
# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion_img2img.StableDiffusionImg2ImgPipeline.get_timesteps
def get_timesteps(self, num_inference_steps, strength, device):
# get the original timestep using init_timestep
offset = self.scheduler.config.get("steps_offset", 0)
init_timestep = int(num_inference_steps * strength) + offset
init_timestep = min(init_timestep, num_inference_steps)
init_timestep = min(int(num_inference_steps * strength), num_inference_steps)
t_start = max(num_inference_steps - init_timestep + offset, 0)
t_start = max(num_inference_steps - init_timestep, 0)
timesteps = self.scheduler.timesteps[t_start:]
return timesteps, num_inference_steps - t_start

6
src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion_img2img.py
View File

@@ -381,11 +381,9 @@ class StableDiffusionImg2ImgPipeline(DiffusionPipeline):
def get_timesteps(self, num_inference_steps, strength, device):
# get the original timestep using init_timestep
offset = self.scheduler.config.get("steps_offset", 0)
init_timestep = int(num_inference_steps * strength) + offset
init_timestep = min(init_timestep, num_inference_steps)
init_timestep = min(int(num_inference_steps * strength), num_inference_steps)
t_start = max(num_inference_steps - init_timestep + offset, 0)
t_start = max(num_inference_steps - init_timestep, 0)
timesteps = self.scheduler.timesteps[t_start:]
return timesteps, num_inference_steps - t_start

6
src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion_inpaint_legacy.py
View File

@@ -396,11 +396,9 @@ class StableDiffusionInpaintPipelineLegacy(DiffusionPipeline):
# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion_img2img.StableDiffusionImg2ImgPipeline.get_timesteps
def get_timesteps(self, num_inference_steps, strength, device):
# get the original timestep using init_timestep
offset = self.scheduler.config.get("steps_offset", 0)
init_timestep = int(num_inference_steps * strength) + offset
init_timestep = min(init_timestep, num_inference_steps)
init_timestep = min(int(num_inference_steps * strength), num_inference_steps)
t_start = max(num_inference_steps - init_timestep + offset, 0)
t_start = max(num_inference_steps - init_timestep, 0)
timesteps = self.scheduler.timesteps[t_start:]
return timesteps, num_inference_steps - t_start