Esrgan.py

import spaces
import os
import torch
from PIL import Image
from RealESRGAN import RealESRGAN
import gradio as gr
from huggingface_hub import HfApi
import datetime

device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')

model2 = RealESRGAN(device, scale=2)
model2.load_weights('weights/RealESRGAN_x2.pth', download=True)
model4 = RealESRGAN(device, scale=4)
model4.load_weights('weights/RealESRGAN_x4.pth', download=True)
model8 = RealESRGAN(device, scale=8)
model8.load_weights('weights/RealESRGAN_x8.pth', download=True)

def upload_to_hf(image_path, folder, filename):
    api = HfApi()
    api.upload_file(
        path_or_fileobj=image_path,
        path_in_repo=f"{folder}/{filename}",
        repo_id='DamarJati/esr-dev', 
        repo_type='dataset',
        token=os.getenv('HF_TOKEN') 
    )

@spaces.GPU()
def inference(image, size):
    global model2, model4, model8
    if image is None:
        raise gr.Error("Image not uploaded")

    width, height = image.size
    if width >= 5000 or height >= 5000:
        raise gr.Error("The image is too large.")

    if torch.cuda.is_available():
        torch.cuda.empty_cache()

    folder = ''
    result = None
    if size == '2x':
        try:
            result = model2.predict(image.convert('RGB'))
        except torch.cuda.OutOfMemoryError as e:
            print(e)
            model2 = RealESRGAN(device, scale=2)
            model2.load_weights('weights/RealESRGAN_x2.pth', download=False)
            result = model2.predict(image.convert('RGB'))
        folder = '2'
    elif size == '4x':
        try:
            result = model4.predict(image.convert('RGB'))
        except torch.cuda.OutOfMemoryError as e:
            print(e)
            model4 = RealESRGAN(device, scale=4)
            model4.load_weights('weights/RealESRGAN_x4.pth', download=False)
            result = model4.predict(image.convert('RGB'))
        folder = '4'
    else:
        try:
            result = model8.predict(image.convert('RGB'))
        except torch.cuda.OutOfMemoryError as e:
            print(e)
            model8 = RealESRGAN(device, scale=8)
            model8.load_weights('weights/RealESRGAN_x8.pth', download=False)
            result = model8.predict(image.convert('RGB'))
        folder = '8'

    # Generate a timestamp-based filename
    timestamp = datetime.datetime.now().strftime("%H%M%S%f%d%m%Y")
    filename = f"{timestamp}.png"

    # Save the original and upscaled images to local temporary files
    original_filename = f"original_{filename}"
    upscaled_filename = f"{folder}_{filename}"
    image.save(original_filename)
    result.save(upscaled_filename)

    # Upload the original image and upscaled image to Hugging Face Datasets
    upload_to_hf(original_filename, "original", filename)
    upload_to_hf(upscaled_filename, folder, filename)

    print(f"Image size ({device}): {size} ... OK")
    return result

title = "Real ESRGAN UpScale: 2x 4x 8x"
description = "AI-powered image resolution enhancement .<br>Donation: https://ko-fi.com/Damarjati"

gr.Interface(
    inference,
    [gr.Image(type="pil"), gr.Radio(['2x', '4x', '8x'], type="value", value='2x', label='Resolution model')],
    gr.Image(type="pil", label="Output"),
    title=title,
    description=description,
    examples=[['example0.jpg', "2x"]],
    allow_flagging='never',
    cache_examples=False
).queue(api_open=True).launch(show_error=True, show_api=True)