From 685a3d4a9219ba2cf454cd3a46ac86103ae42265 Mon Sep 17 00:00:00 2001 From: Vyacheslav Moskalev Date: Sat, 9 Mar 2024 20:28:19 +0700 Subject: [PATCH] Fix CFG denoiser. --- py/smZ_cfg_denoiser.py | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/py/smZ_cfg_denoiser.py b/py/smZ_cfg_denoiser.py index 3fb8628..4751fcc 100644 --- a/py/smZ_cfg_denoiser.py +++ b/py/smZ_cfg_denoiser.py @@ -253,10 +253,8 @@ class CFGNoisePredictor(torch.nn.Module): self.orig = comfy.samplers.CFGNoisePredictor(model) #CFGNoisePredictorOrig(model) self.inner_model = model self.inner_model2 = CFGDenoiser(model.apply_model) - self.inner_model2.num_timesteps = model.num_timesteps self.inner_model2.device = self.ksampler.device if hasattr(self.ksampler, "device") else None self.s_min_uncond = 0.0 - self.alphas_cumprod = model.alphas_cumprod self.c_adm = None self.init_cond = None self.init_uncond = None @@ -308,12 +306,13 @@ def set_model_k(self: KSampler): self.model_denoise = CFGNoisePredictor(self.model) # main change if ((getattr(self.model, "parameterization", "") == "v") or (getattr(self.model, "model_type", -1) == model_base.ModelType.V_PREDICTION)): - self.model_wrap = wrap_model(self.model_denoise, quantize=True) + self.model_wrap = wrap_model(self.model_denoise) self.model_wrap.parameterization = getattr(self.model, "parameterization", "v") else: - self.model_wrap = wrap_model(self.model_denoise, quantize=True) + self.model_wrap = wrap_model(self.model_denoise) self.model_wrap.parameterization = getattr(self.model, "parameterization", "eps") - self.model_k = KSamplerX0Inpaint(self.model_wrap) + sigmas = self.calculate_sigmas(self.steps) + self.model_k = KSamplerX0Inpaint(self.model_wrap, sigmas) class SDKSampler(comfy.samplers.KSampler): def __init__(self, *args, **kwargs):