HighFocusRecords/Stable-Diffusion-3.5-Large-InferenceAPI

1import gradio as gr2import requests3import io4import random5import os6import time7from PIL import Image8from deep_translator import GoogleTranslator9import json10 11# Project by Nymbo12 13API_URL = "https://api-inference.huggingface.co/models/stabilityai/stable-diffusion-3.5-large"14API_TOKEN = os.getenv("HF_READ_TOKEN")15headers = {"Authorization": f"Bearer {API_TOKEN}"}16timeout = 10017 18# Function to query the API and return the generated image19def query(prompt, is_negative=False, steps=35, cfg_scale=7, sampler="DPM++ 2M Karras", seed=-1, strength=0.7, width=1024, height=1024):20    if prompt == "" or prompt is None:21        return None22 23    key = random.randint(0, 999)24    25    API_TOKEN = random.choice([os.getenv("HF_READ_TOKEN")])26    headers = {"Authorization": f"Bearer {API_TOKEN}"}27    28    # Translate the prompt from Russian to English if necessary29    prompt = GoogleTranslator(source='ru', target='en').translate(prompt)30    print(f'\033[1mGeneration {key} translation:\033[0m {prompt}')31 32    # Add some extra flair to the prompt33    prompt = f"{prompt} | ultra detail, ultra elaboration, ultra quality, perfect."34    print(f'\033[1mGeneration {key}:\033[0m {prompt}')35    36    # Prepare the payload for the API call, including width and height37    payload = {38        "inputs": prompt,39        "is_negative": is_negative,40        "steps": steps,41        "cfg_scale": cfg_scale,42        "seed": seed if seed != -1 else random.randint(1, 1000000000),43        "strength": strength,44        "parameters": {45            "width": width,  # Pass the width to the API46            "height": height  # Pass the height to the API47        }48    }49 50    # Send the request to the API and handle the response51    response = requests.post(API_URL, headers=headers, json=payload, timeout=timeout)52    if response.status_code != 200:53        print(f"Error: Failed to get image. Response status: {response.status_code}")54        print(f"Response content: {response.text}")55        if response.status_code == 503:56            raise gr.Error(f"{response.status_code} : The model is being loaded")57        raise gr.Error(f"{response.status_code}")58    59    try:60        # Convert the response content into an image61        image_bytes = response.content62        image = Image.open(io.BytesIO(image_bytes))63        print(f'\033[1mGeneration {key} completed!\033[0m ({prompt})')64        return image65    except Exception as e:66        print(f"Error when trying to open the image: {e}")67        return None68 69# CSS to style the app70css = """71#app-container {72    max-width: 800px;73    margin-left: auto;74    margin-right: auto;75}76"""77 78# Build the Gradio UI with Blocks79with gr.Blocks(theme='Nymbo/Nymbo_Theme', css=css) as app:80    # Add a title to the app81    gr.HTML("<center><h1>Stable Diffusion 3.5 Large</h1></center>")82    83    # Container for all the UI elements84    with gr.Column(elem_id="app-container"):85        # Add a text input for the main prompt86        with gr.Row():87            with gr.Column(elem_id="prompt-container"):88                with gr.Row():89                    text_prompt = gr.Textbox(label="Prompt", placeholder="Enter a prompt here", lines=2, elem_id="prompt-text-input")90                91                # Accordion for advanced settings92                with gr.Row():93                    with gr.Accordion("Advanced Settings", open=False):94                        negative_prompt = gr.Textbox(label="Negative Prompt", placeholder="What should not be in the image", value="(deformed, distorted, disfigured), poorly drawn, bad anatomy, wrong anatomy, extra limb, missing limb, floating limbs, (mutated hands and fingers), disconnected limbs, mutation, mutated, ugly, disgusting, blurry, amputation, misspellings, typos", lines=3, elem_id="negative-prompt-text-input")95                        with gr.Row():96                            width = gr.Slider(label="Width", value=1024, minimum=64, maximum=1216, step=32)97                            height = gr.Slider(label="Height", value=1024, minimum=64, maximum=1216, step=32)98                        steps = gr.Slider(label="Sampling steps", value=35, minimum=1, maximum=100, step=1)99                        cfg = gr.Slider(label="CFG Scale", value=7, minimum=1, maximum=20, step=1)100                        strength = gr.Slider(label="Strength", value=0.7, minimum=0, maximum=1, step=0.001)101                        seed = gr.Slider(label="Seed", value=-1, minimum=-1, maximum=1000000000, step=1) # Setting the seed to -1 will make it random102                        method = gr.Radio(label="Sampling method", value="DPM++ 2M Karras", choices=["DPM++ 2M Karras", "DPM++ SDE Karras", "Euler", "Euler a", "Heun", "DDIM"])103 104        # Add a button to trigger the image generation105        with gr.Row():106            text_button = gr.Button("Run", variant='primary', elem_id="gen-button")107        108        # Image output area to display the generated image109        with gr.Row():110            image_output = gr.Image(type="pil", label="Image Output", elem_id="gallery")111        112        # Bind the button to the query function with the added width and height inputs113        text_button.click(query, inputs=[text_prompt, negative_prompt, steps, cfg, method, seed, strength, width, height], outputs=image_output)114 115# Launch the Gradio app116app.launch(show_api=True, share=False)