The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.
Refactor so that zsnr can be set in the sampling_settings.
comfyanonymous2a18e98ccf08
diff --git a/comfy/model_sampling.py b/comfy/model_sampling.pyindex 4a0f2db60..8b4e095d9 100644--- a/comfy/model_sampling.py+++ b/comfy/model_sampling.py@@ -2,6 +2,25 @@ import torchfrom comfy.ldm.modules.diffusionmodules.util import make_beta_scheduleimport math+def rescale_zero_terminal_snr_sigmas(sigmas):+ alphas_cumprod = 1 / ((sigmas * sigmas) + 1)+ alphas_bar_sqrt = alphas_cumprod.sqrt()++ # Store old values.+ alphas_bar_sqrt_0 = alphas_bar_sqrt[0].clone()+ alphas_bar_sqrt_T = alphas_bar_sqrt[-1].clone()++ # Shift so the last timestep is zero.+ alphas_bar_sqrt -= (alphas_bar_sqrt_T)++ # Scale so the first timestep is back to the old value.+ alphas_bar_sqrt *= alphas_bar_sqrt_0 / (alphas_bar_sqrt_0 - alphas_bar_sqrt_T)++ # Convert alphas_bar_sqrt to betas+ alphas_bar = alphas_bar_sqrt**2 # Revert sqrt+ alphas_bar[-1] = 4.8973451890853435e-08+ return ((1 - alphas_bar) / alphas_bar) ** 0.5+class EPS:def calculate_input(self, sigma, noise):sigma = sigma.view(sigma.shape[:1] + (1,) * (noise.ndim - 1))@@ -48,7 +67,7 @@ class CONST:return latent / (1.0 - sigma)class ModelSamplingDiscrete(torch.nn.Module):- def __init__(self, model_config=None):+ def __init__(self, model_config=None, zsnr=None):super().__init__()if model_config is not None:@@ -61,11 +80,14 @@ class ModelSamplingDiscrete(torch.nn.Module):linear_end = sampling_settings.get("linear_end", 0.012)timesteps = sampling_settings.get("timesteps", 1000)- self._register_schedule(given_betas=None, beta_schedule=beta_schedule, timesteps=timesteps, linear_start=linear_start, linear_end=linear_end, cosine_s=8e-3)+ if zsnr is None:+ zsnr = sampling_settings.get("zsnr", False)++ self._register_schedule(given_betas=None, beta_schedule=beta_schedule, timesteps=timesteps, linear_start=linear_start, linear_end=linear_end, cosine_s=8e-3, zsnr=zsnr)self.sigma_data = 1.0def _register_schedule(self, given_betas=None, beta_schedule="linear", timesteps=1000,- linear_start=1e-4, linear_end=2e-2, cosine_s=8e-3):+ linear_start=1e-4, linear_end=2e-2, cosine_s=8e-3, zsnr=False):if given_betas is not None:betas = given_betaselse:@@ -83,6 +105,9 @@ class ModelSamplingDiscrete(torch.nn.Module):# self.register_buffer('alphas_cumprod_prev', torch.tensor(alphas_cumprod_prev, dtype=torch.float32))sigmas = ((1 - alphas_cumprod) / alphas_cumprod) ** 0.5+ if zsnr:+ sigmas = rescale_zero_terminal_snr_sigmas(sigmas)+self.set_sigmas(sigmas)def set_sigmas(self, sigmas):diff --git a/comfy_extras/nodes_model_advanced.py b/comfy_extras/nodes_model_advanced.pyindex 918e6085a..ed14b61ac 100644--- a/comfy_extras/nodes_model_advanced.py+++ b/comfy_extras/nodes_model_advanced.py@@ -51,25 +51,6 @@ class ModelSamplingDiscreteDistilled(comfy.model_sampling.ModelSamplingDiscrete)return log_sigma.exp().to(timestep.device)-def rescale_zero_terminal_snr_sigmas(sigmas):- alphas_cumprod = 1 / ((sigmas * sigmas) + 1)- alphas_bar_sqrt = alphas_cumprod.sqrt()-- # Store old values.- alphas_bar_sqrt_0 = alphas_bar_sqrt[0].clone()- alphas_bar_sqrt_T = alphas_bar_sqrt[-1].clone()-- # Shift so the last timestep is zero.- alphas_bar_sqrt -= (alphas_bar_sqrt_T)-- # Scale so the first timestep is back to the old value.- alphas_bar_sqrt *= alphas_bar_sqrt_0 / (alphas_bar_sqrt_0 - alphas_bar_sqrt_T)-- # Convert alphas_bar_sqrt to betas- alphas_bar = alphas_bar_sqrt**2 # Revert sqrt- alphas_bar[-1] = 4.8973451890853435e-08- return ((1 - alphas_bar) / alphas_bar) ** 0.5-class ModelSamplingDiscrete:@classmethoddef INPUT_TYPES(s):@@ -100,9 +81,7 @@ class ModelSamplingDiscrete:class ModelSamplingAdvanced(sampling_base, sampling_type):pass- model_sampling = ModelSamplingAdvanced(model.model.model_config)- if zsnr:- model_sampling.set_sigmas(rescale_zero_terminal_snr_sigmas(model_sampling.sigmas))+ model_sampling = ModelSamplingAdvanced(model.model.model_config, zsnr=zsnr)m.add_object_patch("model_sampling", model_sampling)return (m, )