Mirrored_Repos/InvokeAI
1
0
Fork 0
mirror of https://github.com/invoke-ai/InvokeAI synced 2024-08-30 20:32:17 +00:00
Code Issues Packages Projects Releases Wiki Activity

cut over from karras to model noise schedule for higher steps

Browse Source 
The k_samplers come with a "karras" noise schedule which performs
very well at low step counts but becomes noisy at higher ones.

This commit introduces a threshold (currently 30 steps) at which the
k samplers will switch over from using karras to the older model
noise schedule.
This commit is contained in:
Lincoln Stein 2022-10-22 23:02:50 -04:00
parent a6e7aa8f97
commit 3e48b9ff85
1 changed files with 11 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
ldm/models/diffusion/ksampler.py
View File

@ -12,6 +12,10 @@ from ldm.modules.diffusionmodules.util import (
extract_into_tensor,
)
# at this threshold, the scheduler will stop using the Karras
# noise schedule and start using the model's schedule
STEP_THRESHOLD = 30
def cfg_apply_threshold(result, threshold = 0.0, scale = 0.7):
if threshold <= 0.0:
return result
@ -98,8 +102,13 @@ class KSampler(Sampler):
rho=7.,
device=self.device,
)
self.sigmas = self.model_sigmas
#self.sigmas = self.karras_sigmas
if ddim_num_steps >= STEP_THRESHOLD:
print(f'>> number of steps ({ddim_num_steps}) >= {STEP_THRESHOLD}: using model sigmas')
self.sigmas = self.model_sigmas
else:
print(f'>> number of steps ({ddim_num_steps}) < {STEP_THRESHOLD}: using karras sigmas')
self.sigmas = self.karras_sigmas
# ALERT: We are completely overriding the sample() method in the base class, which
# means that inpainting will not work. To get this to work we need to be able to