app.py

import os
import gradio as gr
import torch
from diffusers import DiffusionPipeline, AutoencoderTiny
import random
import spaces
from PIL import Image
from fastapi import FastAPI
from fastapi.responses import FileResponse

# Initialize the base model
dtype = torch.bfloat16
device = "cuda" if torch.cuda.is_available() else "cpu"
base_model = "black-forest-labs/FLUX.1-dev"
taef1 = AutoencoderTiny.from_pretrained("madebyollin/taef1", torch_dtype=dtype).to(device)
pipe = DiffusionPipeline.from_pretrained(base_model, torch_dtype=dtype, vae=taef1).to(device)
MAX_SEED = 2**32-1

# Create a FastAPI app
app = FastAPI()

@spaces.GPU()
def generate_image(prompt, width, height, lora_path, trigger_word, hash_value):
    # Load LoRA weights
    pipe.load_lora_weights(lora_path)
    
    # Combine prompt with trigger word
    full_prompt = f"{trigger_word} {prompt}"
    
    # Set up generation parameters
    seed = random.randint(0, MAX_SEED)
    generator = torch.Generator(device="cuda").manual_seed(seed)
    
    # Generate image
    image = pipe(
        prompt=full_prompt,
        num_inference_steps=28,
        guidance_scale=3.5,
        width=width,
        height=height,
        generator=generator,
    ).images[0]
    
    # Create ./tmp/ directory if it doesn't exist
    os.makedirs('./tmp', exist_ok=True)
    
    # Save the image with the provided hash
    image_path = f'./tmp/{hash_value}.png'
    image.save(image_path)
    
    return image, image_path

def run_lora(prompt, width, height, lora_path, trigger_word, hash_value):
    image, image_path = generate_image(prompt, width, height, lora_path, trigger_word, hash_value)
    return image, image_path

# Set up the Gradio interface
with gr.Blocks() as gradio_app:
    gr.Markdown("# LoRA Image Generator")
    
    with gr.Row():
        prompt = gr.Textbox(label="Prompt", lines=3, placeholder="Enter your prompt here")
    
    with gr.Row():
        width = gr.Slider(label="Width", minimum=256, maximum=1024, step=64, value=512)
        height = gr.Slider(label="Height", minimum=256, maximum=1024, step=64, value=512)
    
    with gr.Row():
        lora_path = gr.Textbox(label="LoRA Path", value="SebastianBodza/Flux_Aquarell_Watercolor_v2")
        trigger_word = gr.Textbox(label="Trigger Word", value="AQUACOLTOK")
        hash_value = gr.Textbox(label="Hash", placeholder="Enter a unique hash for the image")
    
    generate_button = gr.Button("Generate Image")
    
    output_image = gr.Image(label="Generated Image")
    output_path = gr.Textbox(label="Saved Image Path")
    
    generate_button.click(
        fn=run_lora,
        inputs=[prompt, width, height, lora_path, trigger_word, hash_value],
        outputs=[output_image, output_path]
    )

# FastAPI endpoint for downloading images
@app.get("/download/{hash_value}")
async def download_image(hash_value: str):
    image_path = f"./tmp/{hash_value}.png"
    if os.path.exists(image_path):
        return FileResponse(image_path, media_type="image/png", filename=f"{hash_value}.png")
    return {"error": "Image not found"}

# Launch both Gradio and FastAPI
if __name__ == "__main__":
    import uvicorn
    from threading import Thread
    
    def run_gradio():
        gradio_app.queue()
        gradio_app.launch()
    
    # Start Gradio in a separate thread
    Thread(target=run_gradio, daemon=True).start()
    
    # Run FastAPI with uvicorn
    uvicorn.run(app, host="0.0.0.0", port=8000)