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 remote-tracking branch 'origin/master' into merge-master-into-worksplit-multigpu

Browse Source 
Amp-Thread-ID: https://ampcode.com/threads/T-019e4352-d45e-75bc-8ed7-ed3a7f6d129a
Co-authored-by: Amp <amp@ampcode.com>

# Conflicts:
#	comfy/ldm/sam3/detector.py
#	comfy/ldm/sam3/tracker.py
#	comfy/model_base.py
#	comfy/quant_ops.py
#	comfy/supported_models.py
#	comfy_api_nodes/apis/bytedance.py
#	comfy_api_nodes/nodes_bytedance.py
#	comfy_api_nodes/nodes_openai.py
#	comfy_extras/frame_interpolation_models/film_net.py
#	comfy_extras/frame_interpolation_models/ifnet.py
#	comfy_extras/nodes_ace.py
#	comfy_extras/nodes_frame_interpolation.py
#	comfy_extras/nodes_lt_audio.py
#	comfy_extras/nodes_sam3.py
#	comfy_extras/nodes_video_model.py
#	folder_paths.py
#	nodes.py
#	requirements.txt
This commit is contained in:
Jedrzej Kosinski
2026-05-19 21:39:34 -07:00
parent 819c7c0702 72e3f6081c
commit ff766e5cfa
263 changed files with 75214 additions and 3395 deletions
Download Patch File Download Diff File Expand all files Collapse all files

228
nodes.py
View File

@@ -32,7 +32,7 @@ import comfy.controlnet
from comfy.comfy_types import IO, ComfyNodeABC, InputTypeDict, FileLocator
from comfy_api.internal import register_versions, ComfyAPIWithVersion
from comfy_api.version_list import supported_versions
from comfy_api.latest import io, ComfyExtension
from comfy_api.latest import io, ComfyExtension, InputImpl
import comfy.clip_vision
@@ -330,7 +330,7 @@ class VAEDecodeTiled:
RETURN_TYPES = ("IMAGE",)
FUNCTION = "decode"
CATEGORY = "_for_testing"
CATEGORY = "experimental"
def decode(self, vae, samples, tile_size, overlap=64, temporal_size=64, temporal_overlap=8):
if tile_size < overlap * 4:
@@ -377,7 +377,7 @@ class VAEEncodeTiled:
RETURN_TYPES = ("LATENT",)
FUNCTION = "encode"
CATEGORY = "_for_testing"
CATEGORY = "experimental"
def encode(self, vae, pixels, tile_size, overlap, temporal_size=64, temporal_overlap=8):
t = vae.encode_tiled(pixels, tile_x=tile_size, tile_y=tile_size, overlap=overlap, tile_t=temporal_size, overlap_t=temporal_overlap)
@@ -493,7 +493,7 @@ class SaveLatent:
OUTPUT_NODE = True
CATEGORY = "_for_testing"
CATEGORY = "experimental"
def save(self, samples, filename_prefix="ComfyUI", prompt=None, extra_pnginfo=None):
full_output_folder, filename, counter, subfolder, filename_prefix = folder_paths.get_save_image_path(filename_prefix, self.output_dir)
@@ -538,7 +538,7 @@ class LoadLatent:
files = [f for f in os.listdir(input_dir) if os.path.isfile(os.path.join(input_dir, f)) and f.endswith(".latent")]
return {"required": {"latent": [sorted(files), ]}, }
CATEGORY = "_for_testing"
CATEGORY = "experimental"
RETURN_TYPES = ("LATENT", )
FUNCTION = "load"
@@ -758,7 +758,7 @@ class LoraLoader:
FUNCTION = "load_lora"
CATEGORY = "loaders"
DESCRIPTION = "LoRAs are used to modify diffusion and CLIP models, altering the way in which latents are denoised such as applying styles. Multiple LoRA nodes can be linked together."
DESCRIPTION = "This LoRA loader is used to modify both diffusion and CLIP models, altering the way in which latents are denoised such as applying styles. Multiple LoRA nodes can be linked together."
SEARCH_ALIASES = ["lora", "load lora", "apply lora", "lora loader", "lora model"]
def load_lora(self, model, clip, lora_name, strength_model, strength_clip):
@@ -767,17 +767,19 @@ class LoraLoader:
lora_path = folder_paths.get_full_path_or_raise("loras", lora_name)
lora = None
lora_metadata = None
if self.loaded_lora is not None:
if self.loaded_lora[0] == lora_path:
lora = self.loaded_lora[1]
lora_metadata = self.loaded_lora[2] if len(self.loaded_lora) > 2 else None
else:
self.loaded_lora = None
if lora is None:
lora = comfy.utils.load_torch_file(lora_path, safe_load=True)
self.loaded_lora = (lora_path, lora)
lora, lora_metadata = comfy.utils.load_torch_file(lora_path, safe_load=True, return_metadata=True)
self.loaded_lora = (lora_path, lora, lora_metadata)
model_lora, clip_lora = comfy.sd.load_lora_for_models(model, clip, lora, strength_model, strength_clip)
model_lora, clip_lora = comfy.sd.load_lora_for_models(model, clip, lora, strength_model, strength_clip, lora_metadata=lora_metadata)
return (model_lora, clip_lora)
class LoraLoaderModelOnly(LoraLoader):
@@ -788,6 +790,7 @@ class LoraLoaderModelOnly(LoraLoader):
"strength_model": ("FLOAT", {"default": 1.0, "min": -100.0, "max": 100.0, "step": 0.01}),
}}
RETURN_TYPES = ("MODEL",)
DESCRIPTION = "This LoRAs loader is used to modify the diffusion model, altering the way in which latents are denoised such as applying styles. Multiple LoRA nodes can be linked together."
FUNCTION = "load_lora_model_only"
def load_lora_model_only(self, model, lora_name, strength_model):
@@ -795,50 +798,26 @@ class LoraLoaderModelOnly(LoraLoader):
class VAELoader:
video_taes = ["taehv", "lighttaew2_2", "lighttaew2_1", "lighttaehy1_5", "taeltx_2"]
image_taes = ["taesd", "taesdxl", "taesd3", "taef1"]
image_taes = ["taesd", "taesdxl", "taesd3", "taef1", "taef2"]
@staticmethod
def vae_list(s):
vaes = folder_paths.get_filename_list("vae")
approx_vaes = folder_paths.get_filename_list("vae_approx")
sdxl_taesd_enc = False
sdxl_taesd_dec = False
sd1_taesd_enc = False
sd1_taesd_dec = False
sd3_taesd_enc = False
sd3_taesd_dec = False
f1_taesd_enc = False
f1_taesd_dec = False
have_img_encoder, have_img_decoder = set(), set()
for v in approx_vaes:
if v.startswith("taesd_decoder."):
sd1_taesd_dec = True
elif v.startswith("taesd_encoder."):
sd1_taesd_enc = True
elif v.startswith("taesdxl_decoder."):
sdxl_taesd_dec = True
elif v.startswith("taesdxl_encoder."):
sdxl_taesd_enc = True
elif v.startswith("taesd3_decoder."):
sd3_taesd_dec = True
elif v.startswith("taesd3_encoder."):
sd3_taesd_enc = True
elif v.startswith("taef1_encoder."):
f1_taesd_dec = True
elif v.startswith("taef1_decoder."):
f1_taesd_enc = True
else:
parts = v.split("_", 1)
if len(parts) != 2 or parts[0] not in s.image_taes:
for tae in s.video_taes:
if v.startswith(tae):
vaes.append(v)
if sd1_taesd_dec and sd1_taesd_enc:
vaes.append("taesd")
if sdxl_taesd_dec and sdxl_taesd_enc:
vaes.append("taesdxl")
if sd3_taesd_dec and sd3_taesd_enc:
vaes.append("taesd3")
if f1_taesd_dec and f1_taesd_enc:
vaes.append("taef1")
break
continue
if parts[1].startswith("encoder."):
have_img_encoder.add(parts[0])
elif parts[1].startswith("decoder."):
have_img_decoder.add(parts[0])
vaes += [k for k in have_img_decoder if k in have_img_encoder]
vaes.append("pixel_space")
return vaes
@@ -901,6 +880,11 @@ class VAELoader:
else:
vae_path = folder_paths.get_full_path_or_raise("vae", vae_name)
sd, metadata = comfy.utils.load_torch_file(vae_path, return_metadata=True)
if vae_name == "taef2":
if metadata is None:
metadata = {"tae_latent_channels": 128}
else:
metadata["tae_latent_channels"] = 128
resolved = comfy.model_management.resolve_gpu_device_option(device)
vae = comfy.sd.VAE(sd=sd, metadata=metadata, device=resolved)
vae.throw_exception_if_invalid()
@@ -1066,7 +1050,7 @@ class CLIPLoader:
@classmethod
def INPUT_TYPES(s):
return {"required": { "clip_name": (folder_paths.get_filename_list("text_encoders"), ),
"type": (["stable_diffusion", "stable_cascade", "sd3", "stable_audio", "mochi", "ltxv", "pixart", "cosmos", "lumina2", "wan", "hidream", "chroma", "ace", "omnigen2", "qwen_image", "hunyuan_image", "flux2", "ovis", "longcat_image"], ),
"type": (["stable_diffusion", "stable_cascade", "sd3", "stable_audio", "mochi", "ltxv", "pixart", "cosmos", "lumina2", "wan", "hidream", "chroma", "ace", "omnigen2", "qwen_image", "hunyuan_image", "flux2", "ovis", "longcat_image", "cogvideox"], ),
},
"optional": {
"device": (comfy.model_management.get_gpu_device_options(), {"advanced": True}),
@@ -1076,7 +1060,7 @@ class CLIPLoader:
CATEGORY = "advanced/loaders"
DESCRIPTION = "[Recipes]\n\nstable_diffusion: clip-l\nstable_cascade: clip-g\nsd3: t5 xxl/ clip-g / clip-l\nstable_audio: t5 base\nmochi: t5 xxl\ncosmos: old t5 xxl\nlumina2: gemma 2 2B\nwan: umt5 xxl\n hidream: llama-3.1 (Recommend) or t5\nomnigen2: qwen vl 2.5 3B"
DESCRIPTION = "[Recipes]\n\nstable_diffusion: clip-l\nstable_cascade: clip-g\nsd3: t5 xxl/ clip-g / clip-l\nstable_audio: t5 base\nmochi: t5 xxl\ncogvideox: t5 xxl (226-token padding)\ncosmos: old t5 xxl\nlumina2: gemma 2 2B\nwan: umt5 xxl\n hidream: llama-3.1 (Recommend) or t5\nomnigen2: qwen vl 2.5 3B"
@classmethod
def VALIDATE_INPUTS(cls, device="default"):
@@ -1345,7 +1329,7 @@ class LatentFromBatch:
@classmethod
def INPUT_TYPES(s):
return {"required": { "samples": ("LATENT",),
"batch_index": ("INT", {"default": 0, "min": 0, "max": 63}),
"batch_index": ("INT", {"default": 0, "min": -MAX_RESOLUTION, "max": MAX_RESOLUTION}),
"length": ("INT", {"default": 1, "min": 1, "max": 64}),
}}
RETURN_TYPES = ("LATENT",)
@@ -1356,7 +1340,9 @@ class LatentFromBatch:
def frombatch(self, samples, batch_index, length):
s = samples.copy()
s_in = samples["samples"]
batch_index = min(s_in.shape[0] - 1, batch_index)
if batch_index < 0:
batch_index += s_in.shape[0]
batch_index = max(0, min(s_in.shape[0] - 1, batch_index))
length = min(s_in.shape[0] - batch_index, length)
s["samples"] = s_in[batch_index:batch_index + length].clone()
if "noise_mask" in samples:
@@ -1567,7 +1553,7 @@ class LatentBlend:
RETURN_TYPES = ("LATENT",)
FUNCTION = "blend"
CATEGORY = "_for_testing"
CATEGORY = "experimental"
def blend(self, samples1, samples2, blend_factor:float, blend_mode: str="normal"):
@@ -1644,7 +1630,7 @@ class SetLatentNoiseMask:
def common_ksampler(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent, denoise=1.0, disable_noise=False, start_step=None, last_step=None, force_full_denoise=False):
latent_image = latent["samples"]
latent_image = comfy.sample.fix_empty_latent_channels(model, latent_image, latent.get("downscale_ratio_spacial", None))
latent_image = comfy.sample.fix_empty_latent_channels(model, latent_image, latent.get("downscale_ratio_spacial", None), latent.get("downscale_ratio_temporal", None))
if disable_noise:
noise = torch.zeros(latent_image.size(), dtype=latent_image.dtype, layout=latent_image.layout, device="cpu")
@@ -1663,6 +1649,7 @@ def common_ksampler(model, seed, steps, cfg, sampler_name, scheduler, positive,
force_full_denoise=force_full_denoise, noise_mask=noise_mask, callback=callback, disable_pbar=disable_pbar, seed=seed)
out = latent.copy()
out.pop("downscale_ratio_spacial", None)
out.pop("downscale_ratio_temporal", None)
out["samples"] = samples
return (out, )
@@ -1818,22 +1805,27 @@ class LoadImage:
RETURN_TYPES = ("IMAGE", "MASK")
FUNCTION = "load_image"
def load_image(self, image):
image_path = folder_paths.get_annotated_filepath(image)
dtype = comfy.model_management.intermediate_dtype()
device = comfy.model_management.intermediate_device()
components = InputImpl.VideoFromFile(image_path).get_components()
if components.images.shape[0] > 0:
return (components.images.to(device=device, dtype=dtype), (1.0 - components.alpha[..., -1]).to(device=device, dtype=dtype) if components.alpha is not None else torch.zeros((components.images.shape[0], 64, 64), dtype=dtype, device=device))
# This code is left here to handle animated webp which pyav does not support loading
img = node_helpers.pillow(Image.open, image_path)
output_images = []
output_masks = []
w, h = None, None
dtype = comfy.model_management.intermediate_dtype()
for i in ImageSequence.Iterator(img):
i = node_helpers.pillow(ImageOps.exif_transpose, i)
if i.mode == 'I':
i = i.point(lambda i: i * (1 / 255))
image = i.convert("RGB")
if len(output_images) == 0:
@@ -1848,25 +1840,15 @@ class LoadImage:
if 'A' in i.getbands():
mask = np.array(i.getchannel('A')).astype(np.float32) / 255.0
mask = 1. - torch.from_numpy(mask)
elif i.mode == 'P' and 'transparency' in i.info:
mask = np.array(i.convert('RGBA').getchannel('A')).astype(np.float32) / 255.0
mask = 1. - torch.from_numpy(mask)
else:
mask = torch.zeros((64,64), dtype=torch.float32, device="cpu")
mask = torch.zeros((64, 64), dtype=torch.float32, device="cpu")
output_images.append(image.to(dtype=dtype))
output_masks.append(mask.unsqueeze(0).to(dtype=dtype))
if img.format == "MPO":
break # ignore all frames except the first one for MPO format
output_image = torch.cat(output_images, dim=0)
output_mask = torch.cat(output_masks, dim=0)
if len(output_images) > 1:
output_image = torch.cat(output_images, dim=0)
output_mask = torch.cat(output_masks, dim=0)
else:
output_image = output_images[0]
output_mask = output_masks[0]
return (output_image, output_mask)
return (output_image.to(device=device, dtype=dtype), output_mask.to(device=device, dtype=dtype))
@classmethod
def IS_CHANGED(s, image):
@@ -1883,57 +1865,49 @@ class LoadImage:
return True
class LoadImageMask:
class LoadImageMask(LoadImage):
ESSENTIALS_CATEGORY = "Image Tools"
SEARCH_ALIASES = ["import mask", "alpha mask", "channel mask"]
_color_channels = ["alpha", "red", "green", "blue"]
@classmethod
def INPUT_TYPES(s):
input_dir = folder_paths.get_input_directory()
files = [f for f in os.listdir(input_dir) if os.path.isfile(os.path.join(input_dir, f))]
return {"required":
{"image": (sorted(files), {"image_upload": True}),
"channel": (s._color_channels, ), }
}
CATEGORY = "mask"
types = super().INPUT_TYPES()
return {
"required": {
**types["required"],
"channel": (s._color_channels, )
}
}
CATEGORY = "image"
RETURN_TYPES = ("MASK",)
FUNCTION = "load_image"
def load_image(self, image, channel):
image_path = folder_paths.get_annotated_filepath(image)
i = node_helpers.pillow(Image.open, image_path)
i = node_helpers.pillow(ImageOps.exif_transpose, i)
if i.getbands() != ("R", "G", "B", "A"):
if i.mode == 'I':
i = i.point(lambda i: i * (1 / 255))
i = i.convert("RGBA")
mask = None
FUNCTION = "load_image_mask"
def load_image_mask(self, image, channel):
image_tensor, mask_tensor = super().load_image(image)
c = channel[0].upper()
if c in i.getbands():
mask = np.array(i.getchannel(c)).astype(np.float32) / 255.0
mask = torch.from_numpy(mask)
if c == 'A':
mask = 1. - mask
if c == 'A':
return (mask_tensor,)
channel_idx = {'R': 0, 'G': 1, 'B': 2}.get(c, 0)
if channel_idx < image_tensor.shape[-1]:
return (image_tensor[..., channel_idx].clone(),)
else:
mask = torch.zeros((64,64), dtype=torch.float32, device="cpu")
return (mask.unsqueeze(0),)
empty_mask = torch.zeros(
image_tensor.shape[:-1],
dtype=image_tensor.dtype,
device=image_tensor.device
)
return (empty_mask,)
@classmethod
def IS_CHANGED(s, image, channel):
image_path = folder_paths.get_annotated_filepath(image)
m = hashlib.sha256()
with open(image_path, 'rb') as f:
m.update(f.read())
return m.digest().hex()
@classmethod
def VALIDATE_INPUTS(s, image):
if not folder_paths.exists_annotated_filepath(image):
return "Invalid image file: {}".format(image)
return True
return super().IS_CHANGED(image)
class LoadImageOutput(LoadImage):
@@ -2024,7 +1998,7 @@ class ImageInvert:
RETURN_TYPES = ("IMAGE",)
FUNCTION = "invert"
CATEGORY = "image"
CATEGORY = "image/color"
def invert(self, image):
s = 1.0 - image
@@ -2040,7 +2014,7 @@ class ImageBatch:
RETURN_TYPES = ("IMAGE",)
FUNCTION = "batch"
CATEGORY = "image"
CATEGORY = "image/batch"
DEPRECATED = True
def batch(self, image1, image2):
@@ -2097,7 +2071,7 @@ class ImagePadForOutpaint:
RETURN_TYPES = ("IMAGE", "MASK")
FUNCTION = "expand_image"
CATEGORY = "image"
CATEGORY = "image/transform"
def expand_image(self, image, left, top, right, bottom, feathering):
d1, d2, d3, d4 = image.size()
@@ -2231,6 +2205,8 @@ NODE_DISPLAY_NAME_MAPPINGS = {
"StyleModelLoader": "Load Style Model",
"CLIPVisionLoader": "Load CLIP Vision",
"UNETLoader": "Load Diffusion Model",
"unCLIPCheckpointLoader": "Load unCLIP Checkpoint",
"GLIGENLoader": "Load GLIGEN Model",
# Conditioning
"CLIPVisionEncode": "CLIP Vision Encode",
"StyleModelApply": "Apply Style Model",
@@ -2242,7 +2218,7 @@ NODE_DISPLAY_NAME_MAPPINGS = {
"ConditioningSetArea": "Conditioning (Set Area)",
"ConditioningSetAreaPercentage": "Conditioning (Set Area with Percentage)",
"ConditioningSetMask": "Conditioning (Set Mask)",
"ControlNetApply": "Apply ControlNet (OLD)",
"ControlNetApply": "Apply ControlNet (DEPRECATED)",
"ControlNetApplyAdvanced": "Apply ControlNet",
# Latent
"VAEEncodeForInpaint": "VAE Encode (for Inpainting)",
@@ -2260,6 +2236,7 @@ NODE_DISPLAY_NAME_MAPPINGS = {
"LatentFromBatch" : "Latent From Batch",
"RepeatLatentBatch": "Repeat Latent Batch",
# Image
"EmptyImage": "Empty Image",
"SaveImage": "Save Image",
"PreviewImage": "Preview Image",
"LoadImage": "Load Image",
@@ -2267,18 +2244,18 @@ NODE_DISPLAY_NAME_MAPPINGS = {
"LoadImageOutput": "Load Image (from Outputs)",
"ImageScale": "Upscale Image",
"ImageScaleBy": "Upscale Image By",
"ImageInvert": "Invert Image",
"ImageInvert": "Invert Image Colors",
"ImagePadForOutpaint": "Pad Image for Outpainting",
"ImageBatch": "Batch Images",
"ImageCrop": "Image Crop",
"ImageStitch": "Image Stitch",
"ImageBlend": "Image Blend",
"ImageBlur": "Image Blur",
"ImageQuantize": "Image Quantize",
"ImageSharpen": "Image Sharpen",
"ImageBatch": "Batch Images (DEPRECATED)",
"ImageCrop": "Crop Image",
"ImageStitch": "Stitch Images",
"ImageBlend": "Blend Images",
"ImageBlur": "Blur Image",
"ImageQuantize": "Quantize Image",
"ImageSharpen": "Sharpen Image",
"ImageScaleToTotalPixels": "Scale Image to Total Pixels",
"GetImageSize": "Get Image Size",
# _for_testing
# experimental
"VAEDecodeTiled": "VAE Decode (Tiled)",
"VAEEncodeTiled": "VAE Encode (Tiled)",
}
@@ -2400,7 +2377,7 @@ async def load_custom_node(module_path: str, ignore=set(), module_parent="custom
logging.warning(f"Error while calling comfy_entrypoint in {module_path}: {e}")
return False
else:
logging.warning(f"Skip {module_path} module for custom nodes due to the lack of NODE_CLASS_MAPPINGS or NODES_LIST (need one).")
logging.warning(f"Skip {module_path} module for custom nodes due to the lack of NODE_CLASS_MAPPINGS or comfy_entrypoint (need one).")
return False
except Exception as e:
logging.warning(traceback.format_exc())
@@ -2551,6 +2528,7 @@ async def init_builtin_extra_nodes():
"nodes_nop.py",
"nodes_kandinsky5.py",
"nodes_wanmove.py",
"nodes_ar_video.py",
"nodes_image_compare.py",
"nodes_zimage.py",
"nodes_glsl.py",
@@ -2565,9 +2543,15 @@ async def init_builtin_extra_nodes():
"nodes_number_convert.py",
"nodes_painter.py",
"nodes_curve.py",
"nodes_bg_removal.py",
"nodes_rtdetr.py",
"nodes_frame_interpolation.py",
"nodes_sam3.py"
"nodes_sam3.py",
"nodes_void.py",
"nodes_wandancer.py",
"nodes_hidream_o1.py",
"nodes_save_3d.py",
"nodes_moge.py",
]
import_failed = []