import gradio as gr
from huggingface_hub import login
import os

hf_token = os.environ.get("HF_TOKEN")
login(token=hf_token)

from diffusers import ControlNetModel, StableDiffusionXLControlNetPipeline, AutoencoderKL
from diffusers.utils import load_image
from PIL import Image
import torch
import numpy as np
import cv2


controlnet = ControlNetModel.from_pretrained(
    "diffusers/controlnet-canny-sdxl-1.0",
    torch_dtype=torch.float16
)
vae = AutoencoderKL.from_pretrained("madebyollin/sdxl-vae-fp16-fix", torch_dtype=torch.float16)
pipe = StableDiffusionXLControlNetPipeline.from_pretrained(
    "stabilityai/stable-diffusion-xl-base-1.0",
    controlnet=controlnet,
    vae=vae,
    torch_dtype=torch.float16, 
    variant="fp16",
    use_safetensors=True
)
pipe.to("cuda")

custom_model = "fffiloni/eugene_jour_general"

# This is where you load your trained weights
pipe.load_lora_weights(custom_model, use_auth_token=True)

#pipe.enable_model_cpu_offload()

def infer(image_in, prompt, controlnet_conditioning_scale, guidance_scale):
    prompt = prompt
    negative_prompt = ""

    image = load_image(image_in)

    #controlnet_conditioning_scale = 0.25  # recommended for good generalization

    image = np.array(image)
    image = cv2.Canny(image, 100, 200)
    image = image[:, :, None]
    image = np.concatenate([image, image, image], axis=2)
    image = Image.fromarray(image)

    lora_scale= 0.9
    
    images = pipe(
        prompt, 
        negative_prompt=negative_prompt, 
        image=image, 
        #controlnet_conditioning_scale=controlnet_conditioning_scale,
        guidance_scale = guidance_scale,
        num_inference_steps=50,
        cross_attention_kwargs={"scale": lora_scale}
        ).images

    images[0].save(f"hug_lab.png")

    return f"hug_lab.png"

with gr.Blocks() as demo:
    with gr.Column():
        image_in = gr.Image(source="upload", type="filepath")
        prompt = gr.Textbox(label="Prompt")
        guidance_scale = gr.Slider(label="Guidance Scale", minimum=1.0, maximum=10.0, step=0.1, value=5.0, type="float")
        controlnet_conditioning_scale = gr.Slider(label="Controlnet conditioning Scale", minimum=0.0, maximum=1.0, step=0.1, value=0.5, type="float")
        submit_btn = gr.Button("Submit")
        result = gr.Image(label="Result")

    submit_btn.click(
        fn = infer,
        inputs = [image_in, prompt, controlnet_conditioning_scale, guidance_scale ],
        outputs = [result]
    )

demo.queue().launch()