app.py

import gradio as gr
import numpy as np
import random

from diffusers import DiffusionPipeline
import torch

device = "cuda" if torch.cuda.is_available() else "cpu"

# ✅ WAI Illustrious 1.6 model
model_repo_id = "WAI-Illustrious/WAI-Illustrious-1.6"

torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32

pipe = DiffusionPipeline.from_pretrained(model_repo_id, torch_dtype=torch_dtype)
pipe = pipe.to(device)

# ✅ Performance optimizations
if torch.cuda.is_available():
    pipe.enable_xformers_memory_efficient_attention()
    pipe.enable_model_cpu_offload()

MAX_SEED = np.iinfo(np.int32).max
MAX_IMAGE_SIZE = 1024


def infer(
    prompt,
    negative_prompt,
    seed,
    randomize_seed,
    width,
    height,
    guidance_scale,
    num_inference_steps,
    progress=gr.Progress(track_tqdm=True),
):
    if randomize_seed:
        seed = random.randint(0, MAX_SEED)

    generator = torch.Generator().manual_seed(seed)

    # ✅ Generate 4 images
    images = pipe(
        prompt=[prompt] * 4,
        negative_prompt=[negative_prompt] * 4 if negative_prompt else None,
        guidance_scale=guidance_scale,
        num_inference_steps=num_inference_steps,
        width=width,
        height=height,
        generator=generator,
    ).images

    return images, seed


examples = [
    "masterpiece, best quality, anime girl, detailed eyes",
    "1girl, silver hair, fantasy armor, glowing sword",
    "anime landscape, sunset, cinematic lighting",
]

css = """
#col-container {
    margin: 0 auto;
    max-width: 720px;
}
"""

with gr.Blocks(css=css) as demo:
    with gr.Column(elem_id="col-container"):
        gr.Markdown("# WAI Illustrious 1.6 - Text to Image")

        with gr.Row():
            prompt = gr.Text(
                label="Prompt",
                show_label=False,
                max_lines=1,
                placeholder="Enter your prompt",
                container=False,
            )
            run_button = gr.Button("Run", scale=0, variant="primary")

        # ✅ Gallery instead of single image
        result = gr.Gallery(label="Results", show_label=False, columns=2)

        with gr.Accordion("Advanced Settings", open=False):
            negative_prompt = gr.Text(
                label="Negative prompt",
                max_lines=1,
                placeholder="low quality, bad anatomy",
            )

            seed = gr.Slider(
                label="Seed",
                minimum=0,
                maximum=MAX_SEED,
                step=1,
                value=0,
            )

            randomize_seed = gr.Checkbox(label="Randomize seed", value=True)

            with gr.Row():
                width = gr.Slider(
                    label="Width",
                    minimum=256,
                    maximum=MAX_IMAGE_SIZE,
                    step=32,
                    value=1024,
                )

                height = gr.Slider(
                    label="Height",
                    minimum=256,
                    maximum=MAX_IMAGE_SIZE,
                    step=32,
                    value=1024,
                )

            with gr.Row():
                guidance_scale = gr.Slider(
                    label="Guidance scale",
                    minimum=0.0,
                    maximum=10.0,
                    step=0.1,
                    value=5.0,  # ✅ better default
                )

                num_inference_steps = gr.Slider(
                    label="Steps",
                    minimum=1,
                    maximum=50,
                    step=1,
                    value=25,  # ✅ better default
                )

        gr.Examples(examples=examples, inputs=[prompt])

    gr.on(
        triggers=[run_button.click, prompt.submit],
        fn=infer,
        inputs=[
            prompt,
            negative_prompt,
            seed,
            randomize_seed,
            width,
            height,
            guidance_scale,
            num_inference_steps,
        ],
        outputs=[result, seed],
    )

if __name__ == "__main__":
    demo.launch()