handler.py

from typing import  Dict, List, Any
import base64
from PIL import Image
from io import BytesIO
from diffusers import StableDiffusionControlNetPipeline, ControlNetModel
from diffusers.pipelines.stable_diffusion import StableDiffusionSafetyChecker

import torch


import numpy as np
import cv2
import controlnet_hinter

# set device
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
if device.type != 'cuda':
    raise ValueError("need to run on GPU")
# set mixed precision dtype
dtype = torch.bfloat16 if torch.cuda.get_device_capability()[0] == 8 else torch.float16
    
CONTROLNET_MAPPING = {
    "depth": {
        "model_id": "lllyasviel/sd-controlnet-depth",
        "hinter": controlnet_hinter.hint_depth
    },
}


SD_ID_MAPPING = {
    "dreamshaper":  "stablediffusionapi/dreamshaper-xl",
    "juggernaut":  "stablediffusionapi/juggernaut-xl-v8",
    "realistic-vision":"SG161222/Realistic_Vision_V1.4",
    "rev-animated":"s6yx/ReV_Animated"
}

class EndpointHandler():
    def __init__(self, path=""):
        self.control_type = "depth"
        self.controlnet = ControlNetModel.from_pretrained(CONTROLNET_MAPPING[self.control_type]["model_id"],torch_dtype=dtype).to(device)

        self.stable_diffusion_id = "Lykon/dreamshaper-8"

        print(f"Using stable diffusion model: {self.stable_diffusion_id}")
        
        self.pipe = StableDiffusionControlNetPipeline.from_pretrained(self.stable_diffusion_id,
                                                                      controlnet=self.controlnet, 
                                                                      torch_dtype=dtype,
                                                                      safety_checker=StableDiffusionSafetyChecker.from_pretrained("CompVis/stable-diffusion-safety-checker", torch_dtype=dtype)).to("cuda")

        # Define Generator with seed
        self.generator = torch.Generator(device=device.type).manual_seed(3)

    def __call__(self, data: Any) -> List[List[Dict[str, float]]]:
        """
        :param data: A dictionary contains `prompt` and optional `image_depth_map` field.
        :return: A dictionary with `image` field contains image in base64.
        """
        

        # hyperparamters
        #sd_model = data.pop("sd_model", None)
        prompt = data.pop("inputs", None)
        negative_prompt = data.pop("negative_prompt", None)
        image_depth_map = data.pop("image_depth_map", None)
        steps = data.pop("steps", 25)
        scale = data.pop("scale", 7)
        height = data.pop("height", None)
        width = data.pop("width", None)
        controlnet_conditioning_scale = data.pop("controlnet_conditioning_scale", 1)

        # Check if neither prompt nor image is provided
        if prompt is None:
            return {"error": "Please provide a prompt"}
        
        if(image_depth_map is None):
            return {"error": "Please provide a image_depth_map"}
                
        # process image
        image = self.decode_base64_image(image_depth_map)

        # run inference pipeline
        out = self.pipe(
            prompt=prompt, 
            negative_prompt=negative_prompt,
            image=image,
            num_inference_steps=steps, 
            guidance_scale=scale,
            num_images_per_prompt=1,
            height=height,
            width=width,
            controlnet_conditioning_scale=controlnet_conditioning_scale,
            generator=self.generator
        )
        
        # return first generate PIL image
        return out.images[0]
    
    # helper to decode input image
    def decode_base64_image(self, image_string):
        base64_image = base64.b64decode(image_string)
        buffer = BytesIO(base64_image)
        image = Image.open(buffer)
        return image