app.py

import os
import random
import sys
from typing import Sequence, Mapping, Any, Union
import torch
import gradio as gr
from huggingface_hub import hf_hub_download
import spaces
from comfy import model_management

hf_hub_download(repo_id="black-forest-labs/FLUX.1-Redux-dev", filename="flux1-redux-dev.safetensors", local_dir="models/style_models")
hf_hub_download(repo_id="black-forest-labs/FLUX.1-Depth-dev", filename="flux1-depth-dev.safetensors", local_dir="models/diffusion_models")
hf_hub_download(repo_id="Comfy-Org/sigclip_vision_384", filename="sigclip_vision_patch14_384.safetensors", local_dir="models/clip_vision")
hf_hub_download(repo_id="Kijai/DepthAnythingV2-safetensors", filename="depth_anything_v2_vitl_fp32.safetensors", local_dir="models/depthanything")
hf_hub_download(repo_id="black-forest-labs/FLUX.1-dev", filename="ae.safetensors", local_dir="models/vae/FLUX1")
hf_hub_download(repo_id="comfyanonymous/flux_text_encoders", filename="clip_l.safetensors", local_dir="models/text_encoders")
hf_hub_download(repo_id="comfyanonymous/flux_text_encoders", filename="t5xxl_fp16.safetensors", local_dir="models/text_encoders/t5")

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
    """
    comfyui_path = find_path("ComfyUI")
    if comfyui_path is not None and os.path.isdir(comfyui_path):
        sys.path.append(comfyui_path)
        print(f"'{comfyui_path}' added to sys.path")


def add_extra_model_paths() -> None:
    """
    Parse the optional extra_model_paths.yaml file and add the parsed paths to the sys.path.
    """
    # Ensure custom_nodes directory exists
    custom_nodes_path = os.path.join(os.getcwd(), "custom_nodes")
    if not os.path.exists(custom_nodes_path):
        os.makedirs(custom_nodes_path)
        print(f"Created custom_nodes directory at: {custom_nodes_path}")

    try:
        from main 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()

# Initialize nodes before using them
import_custom_nodes()

# Now import and use NODE_CLASS_MAPPINGS
from nodes import NODE_CLASS_MAPPINGS

# Create instances of the nodes we'll use
try:
    # Load required models
    dualcliploader = NODE_CLASS_MAPPINGS["DualCLIPLoader"]()
    vaeloader = NODE_CLASS_MAPPINGS["VAELoader"]()
    unetloader = NODE_CLASS_MAPPINGS["UNETLoader"]()
    clipvisionloader = NODE_CLASS_MAPPINGS["CLIPVisionLoader"]()
    stylemodelloader = NODE_CLASS_MAPPINGS["StyleModelLoader"]()
    
    # Image processing nodes
    loadimage = NODE_CLASS_MAPPINGS["LoadImage"]()
    imagescale = NODE_CLASS_MAPPINGS["ImageScale"]()
    vaedecode = NODE_CLASS_MAPPINGS["VAEDecode"]()
    vaeencode = NODE_CLASS_MAPPINGS["VAEEncode"]()
    saveimage = NODE_CLASS_MAPPINGS["SaveImage"]()
    
    # Conditioning and sampling nodes
    cliptextencode = NODE_CLASS_MAPPINGS["CLIPTextEncode"]()
    ksampler = NODE_CLASS_MAPPINGS["KSampler"]()
    emptylatentimage = NODE_CLASS_MAPPINGS["EmptyLatentImage"]()
    
except KeyError as e:
    print(f"Error: Could not find node {e} in NODE_CLASS_MAPPINGS")
    print("Available nodes:", list(NODE_CLASS_MAPPINGS.keys()))
    raise

# Load all the models that need a safetensors file
model_loaders = [
    dualcliploader.load_clip(
        clip_name1="t5/t5xxl_fp16.safetensors",
        clip_name2="clip_l.safetensors",
        type="flux",
    ),
    vaeloader.load_vae("vae/FLUX1/ae.safetensors"),
    unetloader.load_unet("diffusion_models/flux1-depth-dev.safetensors"),
    clipvisionloader.load_clip("clip_vision/sigclip_vision_patch14_384.safetensors"),
    stylemodelloader.load_style_model("style_models/flux1-redux-dev.safetensors")
]

# Check which models are valid
valid_models = [
    model for model in model_loaders
    if model is not None and len(model) > 0
]

@spaces.GPU(duration=60)
def generate_image(prompt, structure_image, style_image, depth_strength, style_strength):
    with torch.inference_mode():
        # Set up image dimensions
        width = 1024
        height = 1024
        
        # Load and process the input images
        loaded_structure = loadimage.load_image(structure_image)
        loaded_style = loadimage.load_image(style_image)
        
        # Scale images if needed
        scaled_structure = imagescale.upscale(loaded_structure, width, height, "lanczos", "center")
        scaled_style = imagescale.upscale(loaded_style, width, height, "lanczos", "center")
        
        # Create empty latent
        latent = emptylatentimage.generate(width, height, 1)
        
        # Encode the prompt
        conditioning = cliptextencode.encode(
            clip=get_value_at_index(dualcliploader.load_clip(
                clip_name1="t5/t5xxl_fp16.safetensors",
                clip_name2="clip_l.safetensors",
                type="flux",
            ), 0),
            text=prompt
        )
        
        # Sample the image
        sampled = ksampler.sample(
            model=get_value_at_index(unetloader.load_unet("diffusion_models/flux1-depth-dev.safetensors"), 0),
            positive=conditioning,
            negative=None,
            latent=latent,
            seed=random.randint(1, 2**32),
            steps=20,
            cfg=7.5,
            sampler_name="euler",
            scheduler="normal",
            denoise=1.0,
        )
        
        # Decode the latent to image
        decoded = vaedecode.decode(
            samples=sampled,
            vae=get_value_at_index(vaeloader.load_vae("vae/FLUX1/ae.safetensors"), 0)
        )
        
        # Save the final image
        saved = saveimage.save_images(decoded)
        return saved

if __name__ == "__main__":
    # Comment out the main() call
    
    # Start your Gradio app
    with gr.Blocks() as app:
        # Add a title
        gr.Markdown("# FLUX Style Shaping")

        with gr.Row():
            with gr.Column():
                # Add an input
                prompt_input = gr.Textbox(label="Prompt", placeholder="Enter your prompt here...")
                # Add a `Row` to include the groups side by side 
                with gr.Row():
                    # First group includes structure image and depth strength
                    with gr.Group():
                        structure_image = gr.Image(label="Structure Image", type="filepath")
                        depth_strength = gr.Slider(minimum=0, maximum=50, value=15, label="Depth Strength")
                    # Second group includes style image and style strength
                    with gr.Group():
                        style_image = gr.Image(label="Style Image", type="filepath")
                        style_strength = gr.Slider(minimum=0, maximum=1, value=0.5, label="Style Strength")
                
                # The generate button
                generate_btn = gr.Button("Generate")
            
            with gr.Column():
                # The output image
                output_image = gr.Image(label="Generated Image")

            # When clicking the button, it will trigger the `generate_image` function, with the respective inputs
            # and the output an image
            generate_btn.click(
                fn=generate_image,
                inputs=[prompt_input, structure_image, style_image, depth_strength, style_strength],
                outputs=[output_image]
            )
        app.launch(share=True)