FaceEnhance / face_enhance.py
Rishi Desai
moving to torch 2.5.1 for zerogpu
59fed21
raw
history blame contribute delete
11.2 kB
import os
import random
import sys
from typing import Sequence, Mapping, Any, Union
import torch
import spaces
COMFYUI_PATH = "./ComfyUI"
"""
To avoid loading the models each time, we store them in a global variable.
"""
COMFY_MODELS = None
def get_value_at_index(obj: Union[Sequence, Mapping], index: int) -> Any:
"""Returns the value at the given index of a sequence or mapping.
If the object is a sequence (like list or string), returns the value at the given index.
If the object is a mapping (like a dictionary), returns the value at the index-th key.
Some return a dictionary, in these cases, we look for the "results" key
Args:
obj (Union[Sequence, Mapping]): The object to retrieve the value from.
index (int): The index of the value to retrieve.
Returns:
Any: The value at the given index.
Raises:
IndexError: If the index is out of bounds for the object and the object is not a mapping.
"""
try:
return obj[index]
except KeyError:
return obj["result"][index]
def find_path(name: str, path: str = None) -> str:
"""
Recursively looks at parent folders starting from the given path until it finds the given name.
Returns the path as a Path object if found, or None otherwise.
"""
# If no path is given, use the current working directory
if path is None:
path = os.getcwd()
# Check if the current directory contains the name
if name in os.listdir(path):
path_name = os.path.join(path, name)
print(f"{name} found: {path_name}")
return path_name
# Get the parent directory
parent_directory = os.path.dirname(path)
# If the parent directory is the same as the current directory, we've reached the root and stop the search
if parent_directory == path:
return None
# Recursively call the function with the parent directory
return find_path(name, parent_directory)
def add_comfyui_directory_to_sys_path() -> None:
"""
Add 'ComfyUI' to the sys.path
"""
sys.path.append(COMFYUI_PATH)
def add_extra_model_paths() -> None:
"""
Parse the optional extra_model_paths.yaml file and add the parsed paths to the sys.path.
"""
try:
from test import load_extra_path_config
except ImportError:
print(
"Could not import load_extra_path_config from main.py. Looking in utils.extra_config instead."
)
from utils.extra_config import load_extra_path_config
extra_model_paths = find_path("extra_model_paths.yaml")
if extra_model_paths is not None:
load_extra_path_config(extra_model_paths)
else:
print("Could not find the extra_model_paths config file.")
add_comfyui_directory_to_sys_path()
add_extra_model_paths()
def import_custom_nodes() -> None:
"""Find all custom nodes in the custom_nodes folder and add those node objects to NODE_CLASS_MAPPINGS
This function sets up a new asyncio event loop, initializes the PromptServer,
creates a PromptQueue, and initializes the custom nodes.
"""
import asyncio
import execution
from nodes import init_extra_nodes
import server
# Creating a new event loop and setting it as the default loop
loop = asyncio.new_event_loop()
asyncio.set_event_loop(loop)
# Creating an instance of PromptServer with the loop
server_instance = server.PromptServer(loop)
execution.PromptQueue(server_instance)
# Initializing custom nodes
init_extra_nodes()
from nodes import (
LoadImage,
SaveImage,
NODE_CLASS_MAPPINGS,
CLIPTextEncode,
VAELoader,
VAEEncode,
DualCLIPLoader,
VAEDecode,
UNETLoader,
ControlNetLoader,
ControlNetApplyAdvanced,
)
@torch.inference_mode()
def load_models():
dualcliploader = DualCLIPLoader()
dualcliploader_94 = dualcliploader.load_clip(
clip_name1="t5xxl_fp16.safetensors",
clip_name2="clip_l.safetensors",
type="flux",
device="default",
)
vaeloader = VAELoader()
vaeloader_95 = vaeloader.load_vae(vae_name="ae.safetensors")
pulidfluxmodelloader = NODE_CLASS_MAPPINGS["PulidFluxModelLoader"]()
pulidfluxmodelloader_44 = pulidfluxmodelloader.load_model(
pulid_file="pulid_flux_v0.9.1.safetensors"
)
pulidfluxevacliploader = NODE_CLASS_MAPPINGS["PulidFluxEvaClipLoader"]()
pulidfluxevacliploader_45 = pulidfluxevacliploader.load_eva_clip()
pulidfluxinsightfaceloader = NODE_CLASS_MAPPINGS["PulidFluxInsightFaceLoader"]()
pulidfluxinsightfaceloader_46 = pulidfluxinsightfaceloader.load_insightface(
provider="CUDA"
)
controlnetloader = ControlNetLoader()
controlnetloader_49 = controlnetloader.load_controlnet(
control_net_name="Flux_Dev_ControlNet_Union_Pro_ShakkerLabs.safetensors"
)
unetloader = UNETLoader()
unetloader_93 = unetloader.load_unet(
unet_name="flux1-dev.safetensors", weight_dtype="default"
)
return {
"dualcliploader_94": dualcliploader_94,
"vaeloader_95": vaeloader_95,
"pulidfluxmodelloader_44": pulidfluxmodelloader_44,
"pulidfluxevacliploader_45": pulidfluxevacliploader_45,
"pulidfluxinsightfaceloader_46": pulidfluxinsightfaceloader_46,
"controlnetloader_49": controlnetloader_49,
"unetloader_93": unetloader_93
}
def initialize_models():
global COMFY_MODELS
if COMFY_MODELS is None:
import_custom_nodes() # Ensure NODE_CLASS_MAPPINGS is initialized
COMFY_MODELS = load_models()
initialize_models()
def main(
face_image: str,
input_image: str,
output_image: str,
dist_image: str = None,
positive_prompt: str = "",
id_weight: float = 0.75,
):
global COMFY_MODELS
if COMFY_MODELS is None:
raise ValueError("Models must be initialized before calling main(). Call initialize_models() first.")
with torch.inference_mode():
dualcliploader_94 = COMFY_MODELS["dualcliploader_94"]
vaeloader_95 = COMFY_MODELS["vaeloader_95"]
pulidfluxmodelloader_44 = COMFY_MODELS["pulidfluxmodelloader_44"]
pulidfluxevacliploader_45 = COMFY_MODELS["pulidfluxevacliploader_45"]
pulidfluxinsightfaceloader_46 = COMFY_MODELS["pulidfluxinsightfaceloader_46"]
controlnetloader_49 = COMFY_MODELS["controlnetloader_49"]
unetloader_93 = COMFY_MODELS["unetloader_93"]
cliptextencode = CLIPTextEncode()
cliptextencode_23 = cliptextencode.encode(
text="", clip=get_value_at_index(dualcliploader_94, 0)
)
loadimage = LoadImage()
loadimage_24 = loadimage.load_image(image=face_image)
loadimage_40 = loadimage.load_image(image=input_image)
vaeencode = VAEEncode()
vaeencode_35 = vaeencode.encode(
pixels=get_value_at_index(loadimage_40, 0),
vae=get_value_at_index(vaeloader_95, 0),
)
randomnoise = NODE_CLASS_MAPPINGS["RandomNoise"]()
randomnoise_39 = randomnoise.get_noise(noise_seed=random.randint(1, 2**64))
cliptextencode_42 = cliptextencode.encode(
text=positive_prompt, clip=get_value_at_index(dualcliploader_94, 0)
)
ksamplerselect = NODE_CLASS_MAPPINGS["KSamplerSelect"]()
ksamplerselect_50 = ksamplerselect.get_sampler(sampler_name="euler")
applypulidflux = NODE_CLASS_MAPPINGS["ApplyPulidFlux"]()
setunioncontrolnettype = NODE_CLASS_MAPPINGS["SetUnionControlNetType"]()
controlnetapplyadvanced = ControlNetApplyAdvanced()
basicguider = NODE_CLASS_MAPPINGS["BasicGuider"]()
basicscheduler = NODE_CLASS_MAPPINGS["BasicScheduler"]()
samplercustomadvanced = NODE_CLASS_MAPPINGS["SamplerCustomAdvanced"]()
vaedecode = VAEDecode()
applypulidflux_133 = applypulidflux.apply_pulid_flux(
weight=id_weight,
start_at=0.10000000000000002,
end_at=1,
fusion="mean",
fusion_weight_max=1,
fusion_weight_min=0,
train_step=1000,
use_gray=True,
model=get_value_at_index(unetloader_93, 0),
pulid_flux=get_value_at_index(pulidfluxmodelloader_44, 0),
eva_clip=get_value_at_index(pulidfluxevacliploader_45, 0),
face_analysis=get_value_at_index(pulidfluxinsightfaceloader_46, 0),
image=get_value_at_index(loadimage_24, 0),
unique_id=1674270197144619516,
)
setunioncontrolnettype_41 = setunioncontrolnettype.set_controlnet_type(
type="tile", control_net=get_value_at_index(controlnetloader_49, 0)
)
controlnetapplyadvanced_37 = controlnetapplyadvanced.apply_controlnet(
strength=1,
start_percent=0.1,
end_percent=0.8,
positive=get_value_at_index(cliptextencode_42, 0),
negative=get_value_at_index(cliptextencode_23, 0),
control_net=get_value_at_index(setunioncontrolnettype_41, 0),
image=get_value_at_index(loadimage_40, 0),
vae=get_value_at_index(vaeloader_95, 0),
)
basicguider_122 = basicguider.get_guider(
model=get_value_at_index(applypulidflux_133, 0),
conditioning=get_value_at_index(controlnetapplyadvanced_37, 0),
)
basicscheduler_131 = basicscheduler.get_sigmas(
scheduler="beta",
steps=28,
denoise=0.75,
model=get_value_at_index(applypulidflux_133, 0),
)
samplercustomadvanced_1 = samplercustomadvanced.sample(
noise=get_value_at_index(randomnoise_39, 0),
guider=get_value_at_index(basicguider_122, 0),
sampler=get_value_at_index(ksamplerselect_50, 0),
sigmas=get_value_at_index(basicscheduler_131, 0),
latent_image=get_value_at_index(vaeencode_35, 0),
)
vaedecode_114 = vaedecode.decode(
samples=get_value_at_index(samplercustomadvanced_1, 0),
vae=get_value_at_index(vaeloader_95, 0),
)
save_comfy_images(get_value_at_index(vaedecode_114, 0), [output_image])
def save_comfy_images(images, output_dirs):
# images is a PyTorch tensor with shape [batch_size, height, width, channels]
import numpy as np
from PIL import Image
for idx, image in enumerate(images):
# Create the output directory if it doesn't exist
output_dir = os.path.dirname(output_dirs[idx])
if output_dir and not os.path.exists(output_dir):
os.makedirs(output_dir, exist_ok=True)
numpy_image = 255. * image.cpu().numpy()
numpy_image = np.clip(numpy_image, 0, 255).astype(np.uint8)
pil_image = Image.fromarray(numpy_image)
pil_image.save(output_dirs[idx])
@spaces.GPU
def face_enhance(face_image: str, input_image: str, output_image: str, dist_image: str = None, positive_prompt: str = "", id_weight: float = 0.75):
initialize_models() # Ensure models are loaded
main(face_image, input_image, output_image, dist_image, positive_prompt, id_weight)
if __name__ == "__main__":
pass