xinyun/ComfyUI
1
0
Fork 0
mirror of https://git.datalinker.icu/comfyanonymous/ComfyUI synced 2026-09-13 15:07:13 +08:00
Code Issues Packages Projects Releases Wiki Activity

* Simplify and align with comfy style

Browse Source
This commit is contained in:
Dango233 2024-10-29 11:10:28 +00:00
parent 19631f153b
commit 139df89057
1 changed files with 22 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

40
comfy_extras/nodes_sd3.py
View File

@ -3,7 +3,7 @@ import comfy.sd
import comfy.model_management import comfy.model_management
import nodes import nodes
import torch import torch
import re
class TripleCLIPLoader: class TripleCLIPLoader:
@classmethod @classmethod
def INPUT_TYPES(s): def INPUT_TYPES(s):
@ -106,8 +106,8 @@ class SkipLayerGuidanceSD3:
return {"required": {"model": ("MODEL", ), return {"required": {"model": ("MODEL", ),
"layers": ("STRING", {"default": "7,8,9", "multiline": False}), "layers": ("STRING", {"default": "7,8,9", "multiline": False}),
"scale": ("FLOAT", {"default": 3.0, "min": 0.0, "max": 10.0, "step": 0.1}), "scale": ("FLOAT", {"default": 3.0, "min": 0.0, "max": 10.0, "step": 0.1}),
"start_percent": ("FLOAT", {"default": 0.05, "min": 0.0, "max": 1.0, "step": 0.001}), "start_percent": ("FLOAT", {"default": 0.01, "min": 0.0, "max": 1.0, "step": 0.001}),
"end_percent": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.001}) "end_percent": ("FLOAT", {"default": 0.15, "min": 0.0, "max": 1.0, "step": 0.001})
}} }}
RETURN_TYPES = ("MODEL",) RETURN_TYPES = ("MODEL",)
FUNCTION = "skip_guidance" FUNCTION = "skip_guidance"
@ -122,31 +122,35 @@ class SkipLayerGuidanceSD3:
assert layers.replace(",", "").isdigit(), "Layers must be comma separated integers" assert layers.replace(",", "").isdigit(), "Layers must be comma separated integers"
def skip(args, extra_args): def skip(args, extra_args):
return args return args
model_sampling = model.get_model_object("model_sampling")
def post_cfg_function(args): def post_cfg_function(args):
model = args["model"]
cond_pred = args["cond_denoised"] cond_pred = args["cond_denoised"]
cond = args["cond"] cond = args["cond"]
cfg_result = args["denoised"] cfg_result = args["denoised"]
sigma = args["sigma"] sigma = args["sigma"]
sigma = args["sigma"]
x = args["input"] x = args["input"]
model_options = args["model_options"].copy()
percentage = 1 - (1000 ** sigma[0].item())/1000
if scale > 0 and percentage > start_percent and percentage < end_percent: for layer in layers:
(slg,) = comfy.samplers.calc_cond_batch(m_slg.model, [cond], x, sigma, m_slg.model_options) model_options = comfy.model_patcher.set_model_options_patch_replace(model_options, skip, "dit", "double_block", layer)
model_sampling.percent_to_sigma(start_percent)
sigma_start = model_sampling.percent_to_sigma(start_percent)
sigma_end = model_sampling.percent_to_sigma(end_percent)
sigma_ = sigma[0].item()
if scale > 0 and sigma_ > sigma_end and sigma_ < sigma_start:
(slg,) = comfy.samplers.calc_cond_batch(model, [cond], x, sigma, model_options)
cfg_result = cfg_result + (cond_pred - slg) * scale cfg_result = cfg_result + (cond_pred - slg) * scale
return cfg_result return cfg_result
layers = [int(x) for x in layers.split(",")]
m_post_cfg = model.clone()
m_slg = model.clone()
for layer in layers: layers = re.findall(r'\d+', layers)
m_slg.set_model_patch_replace(skip, "dit", "double_block", layer) layers = [int(i) for i in layers]
m = model.clone()
m.set_model_sampler_post_cfg_function(post_cfg_function)
m_post_cfg.set_model_sampler_post_cfg_function(post_cfg_function) return (m, )
return (m_post_cfg, )
NODE_CLASS_MAPPINGS = { NODE_CLASS_MAPPINGS = {