import os
import random
import uuid

import gradio as gr
import numpy as np
import spaces
import torch
from diffusers import DiffusionPipeline

MAX_SEED = np.iinfo(np.int32).max
CACHE_EXAMPLES = torch.cuda.is_available() and os.getenv("CACHE_EXAMPLES", "1") == "1"
MAX_IMAGE_SIZE = int(os.getenv("MAX_IMAGE_SIZE", "1536"))

device = torch.device("cuda:0")

pipe = DiffusionPipeline.from_pretrained(
    "playgroundai/playground-v2.5-1024px-aesthetic",
    torch_dtype=torch.float16,
    use_safetensors=True,
    add_watermarker=False,
    variant="fp16"
)
pipe.to(device)
print("Loaded on Device!")


def save_image(img):
    unique_name = str(uuid.uuid4()) + ".png"
    img.save(unique_name)
    return unique_name


@spaces.GPU(enable_queue=True)
def generate(
        prompt: str,
        progress=gr.Progress(track_tqdm=True),
):
    seed = random.randint(0, 2147483647)
    pipe.to(device)
    generator = torch.Generator().manual_seed(seed)

    images = pipe(
        prompt=prompt,
        negative_prompt=None,
        width=1024,
        height=1024,
        guidance_scale=3,
        num_inference_steps=25,
        generator=generator,
        num_images_per_prompt=1,
        use_resolution_binning=True,
        output_type="pil",
    ).images

    image_paths = [save_image(img) for img in images]
    return image_paths

with gr.Blocks() as demo:
    gr.Markdown("# Blossom Playground v2.5")
    with gr.Group():
        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)
        result = gr.Gallery(label="Result", columns=1, show_label=False)

    gr.on(
        triggers=[
            prompt.submit,
            run_button.click,
        ],
        fn=generate,
        inputs=[
            prompt,
        ],
        outputs=[result],
        api_name="run",
    )

if __name__ == "__main__":
    demo.queue(max_size=20).launch()