app.py

'''
import os
from gradio_client import Client, handle_file
from PIL import Image

# Initialize Gradio client
client = Client("http://localhost:7860")

# Path configuration
input_dir = "Aesthetics_X_Phone_720p_Images_Rec_Captioned_16_9"
output_dir = "processed_output_v2_single_thread"
os.makedirs(output_dir, exist_ok=True)

def process_single_image(png_path):
    """Process a single image and save results"""
    try:
        # Get base filename without extension
        base_name = os.path.splitext(os.path.basename(png_path))[0]
        
        # Corresponding text file path
        txt_path = os.path.join(input_dir, f"{base_name}.txt")
        
        print(f"Processing: {png_path}...", end=" ", flush=True)
        
        # Process image through API (returns WEBP path)
        webp_result = client.predict(
            img=handle_file(png_path),
            model_selection="v2",
            api_name="/predict"
        )
        
        # Output paths
        output_image_path = os.path.join(output_dir, f"{base_name}.png")
        output_text_path = os.path.join(output_dir, f"{base_name}.txt")
        
        # Convert WEBP to PNG
        with Image.open(webp_result) as img:
            img.save(output_image_path, "PNG")
        
        # Copy corresponding text file if exists
        if os.path.exists(txt_path):
            with open(txt_path, 'r', encoding='utf-8') as src, \
                 open(output_text_path, 'w', encoding='utf-8') as dst:
                dst.write(src.read())
        
        print("Done")
        return True
    
    except Exception as e:
        print(f"Failed: {str(e)}")
        return False

def main():
    # Get all PNG files in input directory
    png_files = sorted([
        os.path.join(input_dir, f) 
        for f in os.listdir(input_dir) 
        if f.lower().endswith('.png')
    ])
    
    print(f"Found {len(png_files)} PNG files to process")
    
    # Process files one by one
    success_count = 0
    for i, png_path in enumerate(png_files, 1):
        print(f"\n[{i}/{len(png_files)}] ", end="")
        if process_single_image(png_path):
            success_count += 1
    
    print(f"\nProcessing complete! Success: {success_count}/{len(png_files)}")

if __name__ == "__main__":
    main()
'''

from aura_sr import AuraSR
import gradio as gr
import spaces


class ZeroGPUAuraSR(AuraSR):
    @classmethod
    def from_pretrained(cls, model_id: str = "fal-ai/AuraSR", use_safetensors: bool = True):
        import json
        import torch
        from pathlib import Path
        from huggingface_hub import snapshot_download

        # Check if model_id is a local file
        if Path(model_id).is_file():
            local_file = Path(model_id)
            if local_file.suffix == '.safetensors':
                use_safetensors = True
            elif local_file.suffix == '.ckpt':
                use_safetensors = False
            else:
                raise ValueError(f"Unsupported file format: {local_file.suffix}. Please use .safetensors or .ckpt files.")
            
            # For local files, we need to provide the config separately
            config_path = local_file.with_name('config.json')
            if not config_path.exists():
                raise FileNotFoundError(
                    f"Config file not found: {config_path}. "
                    f"When loading from a local file, ensure that 'config.json' "
                    f"is present in the same directory as '{local_file.name}'. "
                    f"If you're trying to load a model from Hugging Face, "
                    f"please provide the model ID instead of a file path."
                )
            
            config = json.loads(config_path.read_text())
            hf_model_path = local_file.parent
        else:
            hf_model_path = Path(snapshot_download(model_id))
            config = json.loads((hf_model_path / "config.json").read_text())

        model = cls(config)

        if use_safetensors:
            try:
                from safetensors.torch import load_file
                checkpoint = load_file(hf_model_path / "model.safetensors" if not Path(model_id).is_file() else model_id)
            except ImportError:
                raise ImportError(
                    "The safetensors library is not installed. "
                    "Please install it with `pip install safetensors` "
                    "or use `use_safetensors=False` to load the model with PyTorch."
                )
        else:
            checkpoint = torch.load(hf_model_path / "model.ckpt" if not Path(model_id).is_file() else model_id)

        model.upsampler.load_state_dict(checkpoint, strict=True)
        return model



aura_sr = ZeroGPUAuraSR.from_pretrained("fal/AuraSR-v2")
aura_sr_v1 = ZeroGPUAuraSR.from_pretrained("fal-ai/AuraSR")


@spaces.GPU()
def predict(img, model_selection):
    return {'v1': aura_sr_v1, 'v2': aura_sr}.get(model_selection).upscale_4x(img)


demo = gr.Interface(
    predict,
    inputs=[gr.Image(), gr.Dropdown(value='v2', choices=['v1', 'v2'])],
    outputs=gr.Image()
)


demo.launch(share = True)