zlei6/ComfyUI
1
0
Fork 0
mirror of https://github.com/Comfy-Org/ComfyUI.git synced 2026-08-25 18:32:35 +08:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'master' into worksplit-multigpu

Browse Source
This commit is contained in:
Jedrzej Kosinski
2025-02-17 19:35:58 -06:00
parent d2504fb701 31e54b7052
commit 048f4f0b3a
48 changed files with 10819 additions and 8408 deletions
Download Patch File Download Diff File Expand all files Collapse all files

46
comfy_extras/nodes_load_3d.py
View File

@@ -20,9 +20,7 @@ class Load3D():
"width": ("INT", {"default": 1024, "min": 1, "max": 4096, "step": 1}),
"height": ("INT", {"default": 1024, "min": 1, "max": 4096, "step": 1}),
"material": (["original", "normal", "wireframe", "depth"],),
"light_intensity": ("INT", {"default": 10, "min": 1, "max": 20, "step": 1}),
"up_direction": (["original", "-x", "+x", "-y", "+y", "-z", "+z"],),
"fov": ("INT", {"default": 75, "min": 10, "max": 150, "step": 1}),
}}
RETURN_TYPES = ("IMAGE", "MASK", "STRING")
@@ -34,22 +32,14 @@ class Load3D():
CATEGORY = "3d"
def process(self, model_file, image, **kwargs):
if isinstance(image, dict):
image_path = folder_paths.get_annotated_filepath(image['image'])
mask_path = folder_paths.get_annotated_filepath(image['mask'])
image_path = folder_paths.get_annotated_filepath(image['image'])
mask_path = folder_paths.get_annotated_filepath(image['mask'])
load_image_node = nodes.LoadImage()
output_image, ignore_mask = load_image_node.load_image(image=image_path)
ignore_image, output_mask = load_image_node.load_image(image=mask_path)
load_image_node = nodes.LoadImage()
output_image, ignore_mask = load_image_node.load_image(image=image_path)
ignore_image, output_mask = load_image_node.load_image(image=mask_path)
return output_image, output_mask, model_file,
else:
# to avoid the format is not dict which will happen the FE code is not compatibility to core,
# we need to this to double-check, it can be removed after merged FE into the core
image_path = folder_paths.get_annotated_filepath(image)
load_image_node = nodes.LoadImage()
output_image, output_mask = load_image_node.load_image(image=image_path)
return output_image, output_mask, model_file,
return output_image, output_mask, model_file,
class Load3DAnimation():
@classmethod
@@ -66,9 +56,7 @@ class Load3DAnimation():
"width": ("INT", {"default": 1024, "min": 1, "max": 4096, "step": 1}),
"height": ("INT", {"default": 1024, "min": 1, "max": 4096, "step": 1}),
"material": (["original", "normal", "wireframe", "depth"],),
"light_intensity": ("INT", {"default": 10, "min": 1, "max": 20, "step": 1}),
"up_direction": (["original", "-x", "+x", "-y", "+y", "-z", "+z"],),
"fov": ("INT", {"default": 75, "min": 10, "max": 150, "step": 1}),
}}
RETURN_TYPES = ("IMAGE", "MASK", "STRING")
@@ -80,20 +68,14 @@ class Load3DAnimation():
CATEGORY = "3d"
def process(self, model_file, image, **kwargs):
if isinstance(image, dict):
image_path = folder_paths.get_annotated_filepath(image['image'])
mask_path = folder_paths.get_annotated_filepath(image['mask'])
image_path = folder_paths.get_annotated_filepath(image['image'])
mask_path = folder_paths.get_annotated_filepath(image['mask'])
load_image_node = nodes.LoadImage()
output_image, ignore_mask = load_image_node.load_image(image=image_path)
ignore_image, output_mask = load_image_node.load_image(image=mask_path)
load_image_node = nodes.LoadImage()
output_image, ignore_mask = load_image_node.load_image(image=image_path)
ignore_image, output_mask = load_image_node.load_image(image=mask_path)
return output_image, output_mask, model_file,
else:
image_path = folder_paths.get_annotated_filepath(image)
load_image_node = nodes.LoadImage()
output_image, output_mask = load_image_node.load_image(image=image_path)
return output_image, output_mask, model_file,
return output_image, output_mask, model_file,
class Preview3D():
@classmethod
@@ -101,9 +83,7 @@ class Preview3D():
return {"required": {
"model_file": ("STRING", {"default": "", "multiline": False}),
"material": (["original", "normal", "wireframe", "depth"],),
"light_intensity": ("INT", {"default": 10, "min": 1, "max": 20, "step": 1}),
"up_direction": (["original", "-x", "+x", "-y", "+y", "-z", "+z"],),
"fov": ("INT", {"default": 75, "min": 10, "max": 150, "step": 1}),
}}
OUTPUT_NODE = True
@@ -123,9 +103,7 @@ class Preview3DAnimation():
return {"required": {
"model_file": ("STRING", {"default": "", "multiline": False}),
"material": (["original", "normal", "wireframe", "depth"],),
"light_intensity": ("INT", {"default": 10, "min": 1, "max": 20, "step": 1}),
"up_direction": (["original", "-x", "+x", "-y", "+y", "-z", "+z"],),
"fov": ("INT", {"default": 75, "min": 10, "max": 150, "step": 1}),
}}
OUTPUT_NODE = True

104
comfy_extras/nodes_lumina2.py Normal file
View File

@@ -0,0 +1,104 @@
from comfy.comfy_types import IO, ComfyNodeABC, InputTypeDict
import torch
class RenormCFG:
@classmethod
def INPUT_TYPES(s):
return {"required": { "model": ("MODEL",),
"cfg_trunc": ("FLOAT", {"default": 100, "min": 0.0, "max": 100.0, "step": 0.01}),
"renorm_cfg": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 100.0, "step": 0.01}),
}}
RETURN_TYPES = ("MODEL",)
FUNCTION = "patch"
CATEGORY = "advanced/model"
def patch(self, model, cfg_trunc, renorm_cfg):
def renorm_cfg_func(args):
cond_denoised = args["cond_denoised"]
uncond_denoised = args["uncond_denoised"]
cond_scale = args["cond_scale"]
timestep = args["timestep"]
x_orig = args["input"]
in_channels = model.model.diffusion_model.in_channels
if timestep[0] < cfg_trunc:
cond_eps, uncond_eps = cond_denoised[:, :in_channels], uncond_denoised[:, :in_channels]
cond_rest, _ = cond_denoised[:, in_channels:], uncond_denoised[:, in_channels:]
half_eps = uncond_eps + cond_scale * (cond_eps - uncond_eps)
half_rest = cond_rest
if float(renorm_cfg) > 0.0:
ori_pos_norm = torch.linalg.vector_norm(cond_eps
, dim=tuple(range(1, len(cond_eps.shape))), keepdim=True
)
max_new_norm = ori_pos_norm * float(renorm_cfg)
new_pos_norm = torch.linalg.vector_norm(
half_eps, dim=tuple(range(1, len(half_eps.shape))), keepdim=True
)
if new_pos_norm >= max_new_norm:
half_eps = half_eps * (max_new_norm / new_pos_norm)
else:
cond_eps, uncond_eps = cond_denoised[:, :in_channels], uncond_denoised[:, :in_channels]
cond_rest, _ = cond_denoised[:, in_channels:], uncond_denoised[:, in_channels:]
half_eps = cond_eps
half_rest = cond_rest
cfg_result = torch.cat([half_eps, half_rest], dim=1)
# cfg_result = uncond_denoised + (cond_denoised - uncond_denoised) * cond_scale
return x_orig - cfg_result
m = model.clone()
m.set_model_sampler_cfg_function(renorm_cfg_func)
return (m, )
class CLIPTextEncodeLumina2(ComfyNodeABC):
SYSTEM_PROMPT = {
"superior": "You are an assistant designed to generate superior images with the superior "\
"degree of image-text alignment based on textual prompts or user prompts.",
"alignment": "You are an assistant designed to generate high-quality images with the "\
"highest degree of image-text alignment based on textual prompts."
}
SYSTEM_PROMPT_TIP = "Lumina2 provide two types of system prompts:" \
"Superior: You are an assistant designed to generate superior images with the superior "\
"degree of image-text alignment based on textual prompts or user prompts. "\
"Alignment: You are an assistant designed to generate high-quality images with the highest "\
"degree of image-text alignment based on textual prompts."
@classmethod
def INPUT_TYPES(s) -> InputTypeDict:
return {
"required": {
"system_prompt": (list(CLIPTextEncodeLumina2.SYSTEM_PROMPT.keys()), {"tooltip": CLIPTextEncodeLumina2.SYSTEM_PROMPT_TIP}),
"user_prompt": (IO.STRING, {"multiline": True, "dynamicPrompts": True, "tooltip": "The text to be encoded."}),
"clip": (IO.CLIP, {"tooltip": "The CLIP model used for encoding the text."})
}
}
RETURN_TYPES = (IO.CONDITIONING,)
OUTPUT_TOOLTIPS = ("A conditioning containing the embedded text used to guide the diffusion model.",)
FUNCTION = "encode"
CATEGORY = "conditioning"
DESCRIPTION = "Encodes a system prompt and a user prompt using a CLIP model into an embedding that can be used to guide the diffusion model towards generating specific images."
def encode(self, clip, user_prompt, system_prompt):
if clip is None:
raise RuntimeError("ERROR: clip input is invalid: None\n\nIf the clip is from a checkpoint loader node your checkpoint does not contain a valid clip or text encoder model.")
system_prompt = CLIPTextEncodeLumina2.SYSTEM_PROMPT[system_prompt]
prompt = f'{system_prompt} <Prompt Start> {user_prompt}'
tokens = clip.tokenize(prompt)
return (clip.encode_from_tokens_scheduled(tokens), )
NODE_CLASS_MAPPINGS = {
"CLIPTextEncodeLumina2": CLIPTextEncodeLumina2,
"RenormCFG": RenormCFG
}
NODE_DISPLAY_NAME_MAPPINGS = {
"CLIPTextEncodeLumina2": "CLIP Text Encode for Lumina2",
}

21
comfy_extras/nodes_model_advanced.py
View File

@@ -3,6 +3,8 @@ import comfy.model_sampling
import comfy.latent_formats
import nodes
import torch
import node_helpers
class LCM(comfy.model_sampling.EPS):
def calculate_denoised(self, sigma, model_output, model_input):
@@ -294,6 +296,24 @@ class RescaleCFG:
m.set_model_sampler_cfg_function(rescale_cfg)
return (m, )
class ModelComputeDtype:
@classmethod
def INPUT_TYPES(s):
return {"required": { "model": ("MODEL",),
"dtype": (["default", "fp32", "fp16", "bf16"],),
}}
RETURN_TYPES = ("MODEL",)
FUNCTION = "patch"
CATEGORY = "advanced/debug/model"
def patch(self, model, dtype):
m = model.clone()
m.set_model_compute_dtype(node_helpers.string_to_torch_dtype(dtype))
return (m, )
NODE_CLASS_MAPPINGS = {
"ModelSamplingDiscrete": ModelSamplingDiscrete,
"ModelSamplingContinuousEDM": ModelSamplingContinuousEDM,
@@ -303,4 +323,5 @@ NODE_CLASS_MAPPINGS = {
"ModelSamplingAuraFlow": ModelSamplingAuraFlow,
"ModelSamplingFlux": ModelSamplingFlux,
"RescaleCFG": RescaleCFG,
"ModelComputeDtype": ModelComputeDtype,
}