app.py

import spaces
import random
import torch
import cv2
import gradio as gr
import numpy as np
from huggingface_hub import snapshot_download
from transformers import CLIPVisionModelWithProjection, CLIPImageProcessor, pipeline
from diffusers.utils import load_image
from kolors.pipelines.pipeline_controlnet_xl_kolors_img2img import StableDiffusionXLControlNetImg2ImgPipeline
from kolors.models.modeling_chatglm import ChatGLMModel
from kolors.models.tokenization_chatglm import ChatGLMTokenizer
from kolors.models.controlnet import ControlNetModel
from diffusers import AutoencoderKL
from kolors.models.unet_2d_condition import UNet2DConditionModel
from diffusers import EulerDiscreteScheduler
from PIL import Image, ImageDraw, ImageFont
from annotator.midas import MidasDetector
from annotator.dwpose import DWposeDetector
from annotator.util import resize_image, HWC3
import os

device = "cuda"
ckpt_dir = snapshot_download(repo_id="Kwai-Kolors/Kolors")
ckpt_dir_depth = snapshot_download(repo_id="Kwai-Kolors/Kolors-ControlNet-Depth")
ckpt_dir_canny = snapshot_download(repo_id="Kwai-Kolors/Kolors-ControlNet-Canny")
ckpt_dir_pose = snapshot_download(repo_id="Kwai-Kolors/Kolors-ControlNet-Pose")

# Add translation pipeline
translator = pipeline("translation", model="Helsinki-NLP/opus-mt-ko-en")

text_encoder = ChatGLMModel.from_pretrained(f'{ckpt_dir}/text_encoder', torch_dtype=torch.float16).half().to(device)
tokenizer = ChatGLMTokenizer.from_pretrained(f'{ckpt_dir}/text_encoder')
vae = AutoencoderKL.from_pretrained(f"{ckpt_dir}/vae", revision=None).half().to(device)
scheduler = EulerDiscreteScheduler.from_pretrained(f"{ckpt_dir}/scheduler")
unet = UNet2DConditionModel.from_pretrained(f"{ckpt_dir}/unet", revision=None).half().to(device)
controlnet_depth = ControlNetModel.from_pretrained(f"{ckpt_dir_depth}", revision=None).half().to(device)
controlnet_canny = ControlNetModel.from_pretrained(f"{ckpt_dir_canny}", revision=None).half().to(device)
controlnet_pose = ControlNetModel.from_pretrained(f"{ckpt_dir_pose}", revision=None).half().to(device)

pipe_depth = StableDiffusionXLControlNetImg2ImgPipeline(
    vae=vae,
    controlnet=controlnet_depth,
    text_encoder=text_encoder,
    tokenizer=tokenizer,
    unet=unet,
    scheduler=scheduler,
    force_zeros_for_empty_prompt=False
)

pipe_canny = StableDiffusionXLControlNetImg2ImgPipeline(
    vae=vae,
    controlnet=controlnet_canny,
    text_encoder=text_encoder,
    tokenizer=tokenizer,
    unet=unet,
    scheduler=scheduler,
    force_zeros_for_empty_prompt=False
)

pipe_pose = StableDiffusionXLControlNetImg2ImgPipeline(
    vae=vae,
    controlnet=controlnet_pose,
    text_encoder=text_encoder,
    tokenizer=tokenizer,
    unet=unet,
    scheduler=scheduler,
    force_zeros_for_empty_prompt=False
)

@spaces.GPU
def translate_korean_to_english(text):
    if any(ord(char) >= 0xAC00 and ord(char) <= 0xD7A3 for char in text):  # Check if Korean characters are present
        translated = translator(text, max_length=512)[0]['translation_text']
        return translated
    return text

@spaces.GPU
def process_canny_condition(image, canny_threods=[100,200]):
    np_image = image.copy()
    np_image = cv2.Canny(np_image, canny_threods[0], canny_threods[1])
    np_image = np_image[:, :, None]
    np_image = np.concatenate([np_image, np_image, np_image], axis=2)
    np_image = HWC3(np_image)
    return Image.fromarray(np_image)

model_midas = MidasDetector()
@spaces.GPU
def process_depth_condition_midas(img, res = 1024):
    h,w,_ = img.shape
    img = resize_image(HWC3(img), res)
    result = HWC3(model_midas(img))
    result = cv2.resize(result, (w,h))
    return Image.fromarray(result)

model_dwpose = DWposeDetector()
@spaces.GPU
def process_dwpose_condition(image, res=1024):
    h,w,_ = image.shape
    img = resize_image(HWC3(image), res)
    out_res, out_img = model_dwpose(image) 
    result = HWC3(out_img)
    result = cv2.resize(result, (w,h))
    return Image.fromarray(result)

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

@spaces.GPU
def infer_depth(prompt, 
          image = None, 
          negative_prompt = "nsfw, facial shadows, low resolution, jpeg artifacts, blurry, bad quality, dark face, neon lights", 
          seed = 397886929, 
          randomize_seed = False,
          guidance_scale = 6.0, 
          num_inference_steps = 50,
          controlnet_conditioning_scale = 0.7,
          control_guidance_end = 0.9,
          strength = 1.0
        ):
    prompt = translate_korean_to_english(prompt)
    negative_prompt = translate_korean_to_english(negative_prompt)
    
    if randomize_seed:
        seed = random.randint(0, MAX_SEED)
    generator = torch.Generator().manual_seed(seed)
    init_image = resize_image(image, MAX_IMAGE_SIZE)
    pipe = pipe_depth.to("cuda")
    condi_img = process_depth_condition_midas(np.array(init_image), MAX_IMAGE_SIZE)
    image = pipe(
        prompt=prompt,
        image=init_image,
        controlnet_conditioning_scale=controlnet_conditioning_scale,
        control_guidance_end=control_guidance_end, 
        strength=strength, 
        control_image=condi_img,
        negative_prompt=negative_prompt, 
        num_inference_steps=num_inference_steps, 
        guidance_scale=guidance_scale,
        num_images_per_prompt=1,
        generator=generator,
    ).images[0]
    return [condi_img, image], seed

@spaces.GPU
def infer_canny(prompt, 
          image = None, 
          negative_prompt = "nsfw, facial shadows, low resolution, jpeg artifacts, blurry, bad quality, dark face, neon lights", 
          seed = 397886929, 
          randomize_seed = False,
          guidance_scale = 6.0, 
          num_inference_steps = 50,
          controlnet_conditioning_scale = 0.7,
          control_guidance_end = 0.9,
          strength = 1.0
        ):
    prompt = translate_korean_to_english(prompt)
    negative_prompt = translate_korean_to_english(negative_prompt)
    
    if randomize_seed:
        seed = random.randint(0, MAX_SEED)
    generator = torch.Generator().manual_seed(seed)
    init_image = resize_image(image, MAX_IMAGE_SIZE)
    pipe = pipe_canny.to("cuda")
    condi_img = process_canny_condition(np.array(init_image))
    image = pipe(
        prompt=prompt,
        image=init_image,
        controlnet_conditioning_scale=controlnet_conditioning_scale,
        control_guidance_end=control_guidance_end, 
        strength=strength, 
        control_image=condi_img,
        negative_prompt=negative_prompt, 
        num_inference_steps=num_inference_steps, 
        guidance_scale=guidance_scale,
        num_images_per_prompt=1,
        generator=generator,
    ).images[0]
    return [condi_img, image], seed

@spaces.GPU
def infer_pose(prompt, 
          image = None, 
          negative_prompt = "nsfw, facial shadows, low resolution, jpeg artifacts, blurry, bad quality, dark face, neon lights", 
          seed = 66, 
          randomize_seed = False,
          guidance_scale = 6.0, 
          num_inference_steps = 50,
          controlnet_conditioning_scale = 0.7,
          control_guidance_end = 0.9,
          strength = 1.0
        ):
    prompt = translate_korean_to_english(prompt)
    negative_prompt = translate_korean_to_english(negative_prompt)
    
    if randomize_seed:
        seed = random.randint(0, MAX_SEED)
    generator = torch.Generator().manual_seed(seed)
    init_image = resize_image(image, MAX_IMAGE_SIZE)
    pipe = pipe_pose.to("cuda")
    condi_img = process_dwpose_condition(np.array(init_image), MAX_IMAGE_SIZE)
    image = pipe(
        prompt=prompt,
        image=init_image,
        controlnet_conditioning_scale=controlnet_conditioning_scale,
        control_guidance_end=control_guidance_end, 
        strength=strength, 
        control_image=condi_img,
        negative_prompt=negative_prompt, 
        num_inference_steps=num_inference_steps, 
        guidance_scale=guidance_scale,
        num_images_per_prompt=1,
        generator=generator,
    ).images[0]
    return [condi_img, image], seed

css = """
footer {
    visibility: hidden;
}
"""


def load_description(fp):
    with open(fp, 'r', encoding='utf-8') as f:
        content = f.read()
    return content

# Add the text_to_image function
def text_to_image(text, size, position):
    width, height = 1024, 576
    image = Image.new("RGB", (width, height), "white")
    draw = ImageDraw.Draw(image)
    
    font_files = ["Arial_Unicode.ttf"]
    font = None
    for font_file in font_files:
        font_path = os.path.join(os.path.dirname(__file__), font_file)
        if os.path.exists(font_path):
            try:
                font = ImageFont.truetype(font_path, size=size)
                print(f"Using font: {font_file}")
                break
            except IOError:
                print(f"Error loading font: {font_file}")
    if font is None:
        print("No suitable font found. Using default font.")
        font = ImageFont.load_default()

    lines = text.split('\n')
    max_line_width = 0
    total_height = 0
    line_heights = []
    for line in lines:
        left, top, right, bottom = draw.textbbox((0, 0), line, font=font)
        line_width = right - left
        line_height = bottom - top
        line_heights.append(line_height)
        max_line_width = max(max_line_width, line_width)
        total_height += line_height

    position_mapping = {
        "top-left": (10, 10),
        "top-center": ((width - max_line_width) / 2, 10),
        "top-right": (width - max_line_width - 10, 10),
        "middle-left": (10, (height - total_height) / 2),
        "middle-center": ((width - max_line_width) / 2, (height - total_height) / 2),
        "middle-right": (width - max_line_width - 10, (height - total_height) / 2),
        "bottom-left": (10, height - total_height - 10),
        "bottom-center": ((width - max_line_width) / 2, height - total_height - 10),
        "bottom-right": (width - max_line_width - 10, height - total_height - 10),
    }

    x, y = position_mapping.get(position, ((width - max_line_width) / 2, height - total_height - 10))
    for i, line in enumerate(lines):
        draw.text((x, y), line, fill="black", font=font)
        y += line_heights[i]

    return image

with gr.Blocks(theme="Nymbo/Nymbo_Theme", css=css) as Kolors:
    with gr.Row():
        with gr.Column(elem_id="col-left"):
            with gr.Row():
                prompt = gr.Textbox(
                    label="Prompt",
                    placeholder="Enter your prompt",
                    lines=2
                )
            with gr.Row():
                image_input_type = gr.Radio(["Upload Image", "Generate Text Image"], label="Input Type", value="Upload Image")

            with gr.Row():
                image = gr.Image(label="Image", type="pil", visible=True)
                with gr.Column(visible=False) as text_image_inputs:
                    text_input = gr.Textbox(label="Enter Text", lines=5, placeholder="Type your text here...")
                    font_size = gr.Radio([48, 72, 96, 144], label="Font Size", value=72)
                    text_position = gr.Dropdown(
                        ["top-left", "top-center", "top-right", "middle-left", "middle-center", "middle-right", "bottom-left", "bottom-center", "bottom-right"],
                        label="Text Position",
                        value="middle-center"
                    )
                    generate_text_image = gr.Button("Generate Text Image")

            with gr.Accordion("Advanced Settings", open=False):
                negative_prompt = gr.Textbox(
                    label="Negative prompt",
                    placeholder="Enter a negative prompt",
                    visible=True,
                    value="nsfw, facial shadows, low resolution, jpeg artifacts, blurry, bad quality, dark face, neon lights"
                )
                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():
                    guidance_scale = gr.Slider(
                        label="Guidance scale",
                        minimum=0.0,
                        maximum=10.0,
                        step=0.1,
                        value=6.0,
                    )
                    num_inference_steps = gr.Slider(
                        label="Number of inference steps",
                        minimum=10,
                        maximum=50,
                        step=1,
                        value=30,
                    )
                with gr.Row():
                    controlnet_conditioning_scale = gr.Slider(
                        label="Controlnet Conditioning Scale",
                        minimum=0.0,
                        maximum=1.0,
                        step=0.1,
                        value=0.7,
                    )
                    control_guidance_end = gr.Slider(
                        label="Control Guidance End",
                        minimum=0.0,
                        maximum=1.0,
                        step=0.1,
                        value=0.9,
                    )
                with gr.Row():
                    strength = gr.Slider(
                        label="Strength",
                        minimum=0.0,
                        maximum=1.0,
                        step=0.1,
                        value=1.0,
                    )
            with gr.Row():
                canny_button = gr.Button("Canny", elem_id="button")
                depth_button = gr.Button("Depth", elem_id="button")
                pose_button = gr.Button("Pose", elem_id="button")

        with gr.Column(elem_id="col-right"):
            result = gr.Gallery(label="Result", show_label=False, columns=2)
            seed_used = gr.Number(label="Seed Used")

    def toggle_image_input(choice):
        return {
            image: gr.update(visible=choice == "Upload Image"),
            text_image_inputs: gr.update(visible=choice == "Generate Text Image")
        }

    image_input_type.change(toggle_image_input, image_input_type, [image, text_image_inputs])

    def generate_and_use_text_image(text, size, position):
        text_image = text_to_image(text, size, position)
        return text_image

    generate_text_image.click(
        generate_and_use_text_image,
        inputs=[text_input, font_size, text_position],
        outputs=image
    )

    with gr.Row():
        gr.Examples(
                fn = infer_canny,
                examples = canny_examples,
                inputs = [prompt, image],
                outputs = [result, seed_used],
                label = "Canny"
            )
    with gr.Row():
        gr.Examples(
                fn = infer_depth,
                examples = depth_examples,
                inputs = [prompt, image],
                outputs = [result, seed_used],
                label = "Depth"
            )

    with gr.Row():
        gr.Examples(
                fn = infer_pose,
                examples = pose_examples,
                inputs = [prompt, image],
                outputs = [result, seed_used],
                label = "Pose"
            )

    canny_button.click(
        fn = infer_canny,
        inputs = [prompt, image, negative_prompt, seed, randomize_seed, guidance_scale, num_inference_steps, controlnet_conditioning_scale, control_guidance_end, strength],
        outputs = [result, seed_used]
    )

    depth_button.click(
        fn = infer_depth,
        inputs = [prompt, image, negative_prompt, seed, randomize_seed, guidance_scale, num_inference_steps, controlnet_conditioning_scale, control_guidance_end, strength],
        outputs = [result, seed_used]
    )

    pose_button.click(
        fn = infer_pose,
        inputs = [prompt, image, negative_prompt, seed, randomize_seed, guidance_scale, num_inference_steps, controlnet_conditioning_scale, control_guidance_end, strength],
        outputs = [result, seed_used]
    )

Kolors.queue().launch(debug=True)