#!/usr/bin/env python
#patch 0.01
import os
import random
import uuid
import gradio as gr
import numpy as np
from PIL import Image
import torch
from diffusers import StableDiffusionInstructPix2PixPipeline, EulerAncestralDiscreteScheduler

huggingface_token = os.getenv("HUGGINGFACE_TOKEN")

#DESCRIPTIONx = """## STABLE INSTRUCT 📦

#"""

examples = [
    ["assets/4.png", "Change the color of the jacket to white."],
    ["assets/1.png", "Change the picture to black and white."],
    ["assets/2.png", "Add the chocolate topping to the ice cream."],
    ["assets/3.png", "Make the burger look spicy."],
]

model_id = "timbrooks/instruct-pix2pix"
pipe = StableDiffusionInstructPix2PixPipeline.from_pretrained(model_id, torch_dtype=torch.float16, safety_checker=None)
pipe.scheduler = EulerAncestralDiscreteScheduler.from_config(pipe.scheduler.config)

DESCRIPTION = """
"""

MAX_SEED = np.iinfo(np.int32).max
CACHE_EXAMPLES = False
MAX_IMAGE_SIZE = int(os.getenv("MAX_IMAGE_SIZE", "4096"))
USE_TORCH_COMPILE = os.getenv("USE_TORCH_COMPILE", "0") == "1"
ENABLE_CPU_OFFLOAD = os.getenv("ENABLE_CPU_OFFLOAD", "0") == "1"

device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")

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

def randomize_seed_fn(seed: int, randomize_seed: bool) -> int:
    if randomize_seed:
        seed = random.randint(0, MAX_SEED)
    return seed

def img2img_generate(
    prompt: str,
    init_image: gr.Image,
    negative_prompt: str = "",
    use_negative_prompt: bool = False,
    seed: int = 0,
    guidance_scale: float = 7,
    randomize_seed: bool = False,
    num_inference_steps=30,
    strength: float = 0.8,
    NUM_IMAGES_PER_PROMPT=1,
    use_resolution_binning: bool = True,
    progress=gr.Progress(track_tqdm=True),
):
    pipe.to(device)
    seed = int(randomize_seed_fn(seed, randomize_seed))
    generator = torch.Generator().manual_seed(seed)
    
    if not use_negative_prompt:
        negative_prompt = None  # type: ignore
    
    init_image = init_image.resize((768, 768))

    
    output = pipe(
        prompt=prompt,
        image=init_image,
        negative_prompt=negative_prompt,
        guidance_scale=guidance_scale,
        num_inference_steps=num_inference_steps,
        generator=generator,
        strength=strength,
        num_images_per_prompt=NUM_IMAGES_PER_PROMPT,
        output_type="pil",
    ).images

    return output

css = '''
.gradio-container{max-width: 800px !important}
h1{text-align:center}
'''
with gr.Blocks(css=css, theme="bethecloud/storj_theme") as demo:
  #  gr.Markdown(DESCRIPTIONx)
    with gr.Group():
        with gr.Row(equal_height=True):
            with gr.Column(scale=1):
                img2img_prompt = gr.Text(
                    label="Instruct",
                    show_label=False,
                    max_lines=1,
                    placeholder="Enter your instruction",
                    container=False,
                )
                init_image = gr.Image(label="Image", type="pil")
                with gr.Row():
                    img2img_run_button = gr.Button("Generate", variant="primary")
            with gr.Column(scale=1):
                img2img_output = gr.Gallery(label="Result", elem_id="gallery")
        with gr.Accordion("Advanced options", open=False, visible=False):
            with gr.Row():
                img2img_use_negative_prompt = gr.Checkbox(label="Use negative prompt", value=True)
                img2img_negative_prompt = gr.Text(
                    label="Negative prompt",
                    max_lines=1,
                    value="(deformed, distorted, disfigured:1.3), poorly drawn, bad anatomy, wrong anatomy, extra limb, missing limb, floating limbs, (mutated hands and fingers:1.4), disconnected limbs, mutation, mutated, ugly, disgusting, blurry, amputation",
                    visible=True,
                )
            img2img_seed = gr.Slider(
                label="Seed",
                minimum=0,
                maximum=MAX_SEED,
                step=1,
                value=0,
            )
            img2img_steps = gr.Slider(
                label="Steps",
                minimum=0,
                maximum=60,
                step=1,
                value=25,
            )
            img2img_number_image = gr.Slider(
                label="No.of.Images",
                minimum=1,
                maximum=4,
                step=1,
                value=1,
            )
            img2img_randomize_seed = gr.Checkbox(label="Randomize seed", value=True)
            with gr.Row():
                img2img_guidance_scale = gr.Slider(
                    label="Guidance Scale",
                    minimum=0.1,
                    maximum=10,
                    step=0.1,
                    value=5.0,
                )
                strength = gr.Slider(label="Confidence", minimum=0.0, maximum=1.0, step=0.01, value=0.8)

    gr.Examples(
        examples=examples,
        inputs=[init_image, img2img_prompt],
        outputs=img2img_output,
        fn=img2img_generate,
        cache_examples=CACHE_EXAMPLES,
    )

    img2img_use_negative_prompt.change(
        fn=lambda x: gr.update(visible=x),
        inputs=img2img_use_negative_prompt,
        outputs=img2img_negative_prompt,
        api_name=False,
    )

    gr.on(
        triggers=[
            img2img_prompt.submit,
            img2img_negative_prompt.submit,
            img2img_run_button.click,
        ],
        fn=img2img_generate,
        inputs=[
            img2img_prompt,
            init_image,
            img2img_negative_prompt,
            img2img_use_negative_prompt,
            img2img_seed,
            img2img_guidance_scale,
            img2img_randomize_seed,
            img2img_steps,
            strength,
            img2img_number_image,
        ],
        outputs=[img2img_output],
        api_name="image-to-image",
    )

    #gr.Markdown("⚠️ users are accountable for the content they generate and are responsible for ensuring it meets appropriate ethical standards")

if __name__ == "__main__":
    demo.queue().launch(show_api=False, debug=False)