app.py

import os
import gradio as gr
import base64
from random import randint
from all_models import models
from io import BytesIO
from PIL import Image
from fastapi import FastAPI, Request
from deep_translator import GoogleTranslator

css_code = """
#custom_textbox {
    width: 100%;
    min-height: 150px;
}
#custom_gen_button {
    background: #4CAF50 !important;
    color: white !important;
}
#custom_stop_button {
    background: #F44336 !important;
    color: white !important;
}
#custom_image {
    width: 100%;
    max-height: 768px;
}
"""

# Initialize translator
translator = GoogleTranslator(source='auto', target='en')

# Load models
models_load = {}
for model in models:
    try:
        models_load[model] = gr.load(f'models/{model}')
    except Exception as error:
        models_load[model] = gr.Interface(lambda txt: None, ['text'], ['image'])

app = FastAPI()

def gen_image(model_str, prompt):
    if model_str == 'NA':
        return None
    # Translate prompt to English {noise} {klir}
    translated_prompt = translator.translate(prompt)
    noise = str(randint(0, 4294967296))
    klir = '| ultra detail, ultra elaboration, ultra quality, perfect'
    return models_load[model_str](f'{translated_prompt}')

def image_to_base64(image):
    buffered = BytesIO()
    if isinstance(image, str):  # if it's a file path
        img = Image.open(image)
        img.save(buffered, format="JPEG")
    else:  # if it's a PIL Image
        image.save(buffered, format="JPEG")
    return base64.b64encode(buffered.getvalue()).decode()

# API endpoint
@app.post("/generate")
async def api_generate(request: Request):
    data = await request.json()
    model = data.get('model', models[0])
    prompt = data.get('prompt', '')
    
    if model not in models:
        return {"error": "Model not found"}
    
    # Translate prompt to English for API endpoint too
    translated_prompt = translator.translate(prompt)
    image = gen_image(model, translated_prompt)
    if image is None:
        return {"error": "Image generation failed"}
    
    base64_str = image_to_base64(image)
    
    return {
        "status": "success",
        "model": model,
        "original_prompt": prompt,
        "translated_prompt": translated_prompt,
        "image_base64": base64_str,
        "image_format": "jpeg"
    }

# Gradio Interface
def make_me():
    with gr.Row():
        # Left Column (50% width)
        with gr.Column(scale=1, min_width=400):
            txt_input = gr.Textbox(
                label='Your prompt:', 
                lines=4, 
                container=False, 
                elem_id="custom_textbox", 
                placeholder="Enter your prompt here..."
            )
            
            model_dropdown = gr.Dropdown(
                models, 
                label="Select LoRA Model", 
                value=models[0] if models else None,
                container=False
            )
            
            with gr.Row():
                gen_button = gr.Button(
                    'Generate Image', 
                    elem_id="custom_gen_button",
                    variant='primary'
                )
                stop_button = gr.Button(
                    'Stop', 
                    variant='stop',
                    elem_id="custom_stop_button",
                    interactive=False
                )
        
        # Right Column (50% width)
        with gr.Column(scale=1, min_width=400):
            output_image = gr.Image(
                label="Generated Image", 
                elem_id="custom_image", 
                show_label=True, 
                interactive=False
            )
            
            json_output = gr.JSON(
                label="API Response",
                container=False
            )
            
    # Functionality remains the same
    def generate_wrapper(model_str, prompt):
        # Translate prompt to English
        translated_prompt = translator.translate(prompt)
        image = gen_image(model_str, translated_prompt)
        if image is None:
            return None, {"error": "Generation failed"}
        
        base64_str = image_to_base64(image)
        response = {
            "status": "success",
            "model": model_str,
            "original_prompt": prompt,
            "translated_prompt": translated_prompt,
            "image_base64": base64_str,
            "image_format": "jpeg"
        }
        return image, response
    
    def on_generate_click():
        return gr.Button(interactive=False), gr.Button(interactive=True)
    
    def on_stop_click():
        return gr.Button(interactive=True), gr.Button(interactive=False)
    
    gen_event = gen_button.click(
        on_generate_click,
        inputs=None,
        outputs=[gen_button, stop_button]
    ).then(
        generate_wrapper,
        [model_dropdown, txt_input],
        [output_image, json_output]
    ).then(
        on_stop_click,
        inputs=None,
        outputs=[gen_button, stop_button]
    )
    
    stop_button.click(
        on_stop_click,
        inputs=None,
        outputs=[gen_button, stop_button],
        cancels=[gen_event]
    )

# Create Gradio app
with gr.Blocks(css=css_code, title="Image Generation App") as demo: 
    gr.Markdown("# Image Generation Tool")
    make_me()

# Enable queue before mounting
demo.queue()

# Mount Gradio app to FastAPI
app = gr.mount_gradio_app(app, demo, path="/")

if __name__ == "__main__":
    import uvicorn
    uvicorn.run(app, host="0.0.0.0", port=7860)