app.py

import gradio as gr
import requests
import json
import io
import random
import os
import torch
from transformers import pipeline
from PIL import Image

API_BASE_URL = "https://api-inference.huggingface.co/models/"
MODEL_LIST = [
    "openskyml/dalle-3-xl",
    "Linaqruf/animagine-xl-2.0",
    "Lykon/dreamshaper-7",
    "Linaqruf/animagine-xl",
    "runwayml/stable-diffusion-v1-5",
    "stabilityai/stable-diffusion-xl-base-1.0",
    "prompthero/openjourney-v4",
    "nerijs/pixel-art-xl",
    "Linaqruf/anything-v3.0",
    "playgroundai/playground-v2-1024px-aesthetic",
    "ilovecutiee/fantastical-art-lora",
    "segmind/SSD-1B",
    "segmind/Segmind-Vega",
    "stablediffusionapi/anything-v5",
    "stablediffusionapi/realistic-vision-v51",
    "hakurei/waifu-diffusion"
]
API_TOKEN = os.getenv("HF_READ_TOKEN")  # Make sure to set your Hugging Face token
HEADERS = {"Authorization": f"Bearer {API_TOKEN}"}

pipe = pipeline("text-generation", model="isek-ai/SDPrompt-RetNet-300M", trust_remote_code=True)

def select_model(model_name):
    if model_name in MODEL_LIST:
        return f"{API_BASE_URL}{model_name}"

def extend_prompt(input_text):
    if not input_text.strip():
        gr.Warning("Input text is empty!")
        return None
        
    seed = random.randint(1, 1000000)
    torch.manual_seed(seed)
    
    output = pipe(input_text, max_length=(len(input_text) + random.randint(60, 90)), num_return_sequences=4)

    return output

def generate_image(prompt, selected_model, is_negative=False, steps=1, cfg_scale=6, seed=None):
    if not prompt.strip():
        raise gr.Error("Cannot generate image: Input text is empty!")
        
    model_url = select_model(selected_model)
    API_URL = f"{model_url}"

    payload = {
        "inputs": prompt,
        "is_negative": is_negative,
        "steps": steps,
        "cfg_scale": cfg_scale,
        "seed": seed if seed is not None else random.randint(-1, 2147483647)
    }

    try:
        response = requests.post(API_URL, headers=HEADERS, json=payload)
        response.raise_for_status()
    
        image_bytes = io.BytesIO(response.content)
        image = Image.open(image_bytes)
        return image
        
    except requests.exceptions.RequestException as e:
        raise gr.Error(e)

with gr.Blocks(theme="soft") as playground:
    gr.HTML(
        """
        <div style="text-align: center; margin: 0 auto;">
          <div style="display: inline-flex; align-items: center; gap: 0.8rem; font-size: 1.75rem;">
            <h1 style="font-weight: 900; margin-bottom: 7px; margin-top: 5px;">🎨🤖 Play with SD Models</h1> 
          </div>
          <p style="margin-bottom: 10px; font-size: 94%; line-height: 23px;">
            Explore and create your AI art with Stable Diffusion models!
          </p>
        </div>
        """
    )

    with gr.Row():
        image_output = gr.Image(type="pil", label="Output Image", elem_id="gallery")
        with gr.Column(elem_id="prompt-container"):
            text_prompt = gr.Textbox(label="Prompt", placeholder="a cute cat", lines=1, elem_id="prompt-text-input")
            model_dropdown = gr.Dropdown(label="Model", choices=MODEL_LIST, elem_id="model-dropdown", value="runwayml/stable-diffusion-v1-5")
            gen_button = gr.Button("Generate", variant='primary', elem_id="gen-button")
            extend_button = gr.Button("Extend Prompt", variant='primary', elem_id="extend-button")

    with gr.Accordion("Advanced settings", open=False):
        negative_prompt = gr.Textbox(label="Negative Prompt", value="text, blurry, fuzziness", lines=1, elem_id="negative-prompt-text-input")

    gen_button.click(generate_image, inputs=[text_prompt, model_dropdown, negative_prompt], outputs=image_output)
    extend_button.click(extend_prompt, inputs=text_prompt, outputs=text_prompt)

playground.launch(show_api=False)