import gradio as gr
import numpy as np
import random

import spaces #[uncomment to use ZeroGPU]
from diffusers import DiffusionPipeline
import torch
from InstantID import InstantIDModelLoader, InstantIDFaceAnalysis, ApplyInstantID

# from InstantID import InstantIDModelLoader, InstantIDFaceAnalysis, ApplyInstantID

device = "cuda" if torch.cuda.is_available() else "cpu"
model_repo_id = "emilianJR/epiCRealism"  # Replace to the model you would like to use

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

class InstantIDPipeline(DiffusionPipeline):
    def __init__(self, *args, **kwargs):
        super().__init__(*args, **kwargs)
        self.instantid_model = None
        self.face_analyzer = None
        
    def load_instantid(self, model_path, device="cuda"):
        # Initialize InstantID components
        self.instantid_model = InstantIDModelLoader.load_model(model_path)[0]
        self.face_analyzer = InstantIDFaceAnalysis.load_insight_face(
            "CUDA" if device == "cuda" else "CPU"
        )[0]
        
    def __call__(
        self,
        prompt,
        negative_prompt=None,
        image=None,
        num_inference_steps=50,
        guidance_scale=7.5,
        width=512,
        height=512,
        generator=None,
    ):
        # Regular diffusion pipeline call if no face image is provided
        if image is None:
            return super().__call__(
                prompt=prompt,
                negative_prompt=negative_prompt,
                num_inference_steps=num_inference_steps,
                guidance_scale=guidance_scale,
                width=width,
                height=height,
                generator=generator,
            )
            
        # Apply InstantID processing when image is provided
        face_features = self.face_analyzer.get(image)
        if not face_features:
            raise ValueError("No face detected in the provided image")
            
        # Process with InstantID
        # Reference implementation from InstantID/InstantID.py:
        # Lines referencing ApplyInstantID class implementation:

# Initialize custom pipeline
pipe = InstantIDPipeline.from_pretrained(model_repo_id, torch_dtype=torch_dtype)
pipe = pipe.to(device)

# Load InstantID components if available
try:
    instantid_model_path = "models/instantid/instantid.safetensors"  # Adjust path as needed
    pipe.load_instantid(instantid_model_path, device)
except Exception as e:
    print(f"Warning: InstantID initialization failed: {e}")

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


@spaces.GPU #[uncomment to use ZeroGPU]
def infer(
    prompt,
    negative_prompt,
    seed,
    randomize_seed,
    width,
    height,
    guidance_scale,
    num_inference_steps,
    face_image=None,  # New parameter
    progress=gr.Progress(track_tqdm=True),
):
    if randomize_seed:
        seed = random.randint(0, MAX_SEED)

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

    image = pipe(
        prompt=prompt,
        negative_prompt=negative_prompt,
        guidance_scale=guidance_scale,
        num_inference_steps=num_inference_steps,
        width=width,
        height=height,
        generator=generator,
        image=face_image,  # Pass the face image if provided
    ).images[0]

    return image, seed


examples = [
    "Astronaut in a classroom, cold color palette, muted colors, detailed, 8k",
    "An astronaut riding a green horse",
    "A delicious ceviche cheesecake slice",
]

css = """
#col-container {
    margin: 0 auto;
    max-width: 640px;
}
"""

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

        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")

        # Add face image input
        face_image = gr.Image(
            label="Reference Face Image (Optional)",
            type="numpy",
            visible=True
        )

        result = gr.Image(label="Result", show_label=False)

        with gr.Accordion("Advanced Settings", open=False):
            negative_prompt = gr.Text(
                label="Negative prompt",
                max_lines=1,
                placeholder="Enter a negative prompt",
                visible=False,
            )

            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,  # Replace with defaults that work for your model
                )

                height = gr.Slider(
                    label="Height",
                    minimum=256,
                    maximum=MAX_IMAGE_SIZE,
                    step=32,
                    value=1024,  # Replace with defaults that work for your model
                )

            with gr.Row():
                guidance_scale = gr.Slider(
                    label="Guidance scale",
                    minimum=0.0,
                    maximum=10.0,
                    step=0.1,
                    value=0.0,  # Replace with defaults that work for your model
                )

                num_inference_steps = gr.Slider(
                    label="Number of inference steps",
                    minimum=1,
                    maximum=50,
                    step=10,
                    value=2,  # Replace with defaults that work for your model
                )

        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,
            face_image,  # Add face image to inputs
        ],
        outputs=[result, seed],
    )

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