xinyun/ComfyUI
1
0
Fork 0
mirror of https://git.datalinker.icu/comfyanonymous/ComfyUI synced 2026-08-16 01:36:41 +08:00
Code Issues Packages Projects Releases Wiki Activity

restore first stochastic step.

Browse Source
This commit is contained in:
Balladie 2025-08-22 00:38:22 +09:00
parent 8c6c3e5e04
commit e82f886351
1 changed files with 8 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
comfy/k_diffusion/sampling.py
View File

@ -1364,17 +1364,17 @@ def res_multistep(model, x, sigmas, extra_args=None, callback=None, disable=None
for i in trange(len(sigmas) - 1, disable=disable): for i in trange(len(sigmas) - 1, disable=disable):
denoised = model(x, sigmas[i] * s_in, **extra_args) denoised = model(x, sigmas[i] * s_in, **extra_args)
# sigma_down, sigma_up = get_ancestral_step(sigmas[i], sigmas[i + 1], eta=eta) sigma_down, sigma_up = get_ancestral_step(sigmas[i], sigmas[i + 1], eta=eta)
if callback is not None: if callback is not None:
callback({"x": x, "i": i, "sigma": sigmas[i], "sigma_hat": sigmas[i], "denoised": denoised}) callback({"x": x, "i": i, "sigma": sigmas[i], "sigma_hat": sigmas[i], "denoised": denoised})
if sigmas[i + 1] == 0 or old_denoised is None: if sigma_down == 0 or old_denoised is None:
# Euler method # Euler method
if cfg_pp: if cfg_pp:
d = to_d(x, sigmas[i], uncond_denoised) d = to_d(x, sigmas[i], uncond_denoised)
x = denoised + d * sigmas[i + 1] x = denoised + d * sigma_down
else: else:
d = to_d(x, sigmas[i], denoised) d = to_d(x, sigmas[i], denoised)
dt = sigmas[i + 1] - sigmas[i] dt = sigma_down - sigmas[i]
x = x + d * dt x = x + d * dt
else: else:
# Second order multistep method in https://arxiv.org/pdf/2308.02157 # Second order multistep method in https://arxiv.org/pdf/2308.02157
@ -1395,8 +1395,10 @@ def res_multistep(model, x, sigmas, extra_args=None, callback=None, disable=None
else: else:
x = sigmas[i + 1] / sigmas[i] * (-h * eta).exp() * x + alpha_next * h_eta * (b1 * denoised + b2 * old_denoised) x = sigmas[i + 1] / sigmas[i] * (-h * eta).exp() * x + alpha_next * h_eta * (b1 * denoised + b2 * old_denoised)
# Noise addition # Noise addition
sigma_up = sigmas[i + 1] * (-2 * h * eta).expm1().neg().sqrt() if sigmas[i + 1] > 0:
if old_denoised is not None:
sigma_up = sigmas[i + 1] * (-2 * h * eta).expm1().neg().sqrt()
x = x + noise_sampler(sigmas[i], sigmas[i + 1]) * s_noise * sigma_up x = x + noise_sampler(sigmas[i], sigmas[i + 1]) * s_noise * sigma_up
if cfg_pp: if cfg_pp: