app.py

import gradio as gr
import requests
import io
from PIL import Image
import json
import os

# Load LoRAs from JSON
with open('loras.json', 'r') as f:
    loras = json.load(f)

# API call function
def query(payload, api_url, token):
    headers = {"Authorization": f"Bearer {token}"}
    response = requests.post(api_url, headers=headers, json=payload)
    return io.BytesIO(response.content)

# Gradio UI
with gr.Blocks() as demo:  # Removed the css argument
    title = gr.HTML(
        """<h1><img src="https://i.imgur.com/vT48NAO.png" alt="LoRA"> LoRA the Explorer</h1>""",
        elem_id="title",
    )
    gallery = gr.Gallery(
        value=[(item["image"], item["title"]) for item in loras],
        label="LoRA Gallery",
        allow_preview=False,
        columns=3,
        elem_id="gallery",
        show_share_button=False
    )
    prompt = gr.Textbox(label="Prompt", show_label=False, lines=1, max_lines=1, placeholder="Type a prompt after selecting a LoRA", elem_id="prompt")
    advanced_options = gr.Accordion("Advanced options", open=False)
    weight = gr.Slider(0, 10, value=1, step=0.1, label="LoRA weight")
    result = gr.Image(interactive=False, label="Generated Image", elem_id="result-image")

    # Define the function to run when the button is clicked
    def run_lora(prompt, weight):
        selected_lora = loras[0]  # You may need to adjust this index if you have multiple models
        api_url = f"https://api-inference.huggingface.co/models/{selected_lora['repo']}"
        trigger_word = selected_lora["trigger_word"]
        token = os.getenv("API_TOKEN")  # This will read the API token set in your managed environment
        payload = {"inputs": f"{prompt} {trigger_word}"}
        
        print("Calling query function...")
        image_bytes = query(payload, api_url, token)
        print("Query function executed successfully.")
        return Image.open(image_bytes)

    print("Starting Gradio UI...")
    gr.Interface(
        fn=run_lora,
        inputs=[prompt, weight],
        outputs=[result],
    ).launch()