Update app.py from anycoder

app.py

@@ -1,167 +0,0 @@
-import gradio as gr
-import numpy as np
-import time
-from PIL import Image
-import torch
-from torchvision import transforms
-import torchvision.models as models
-import os
-
-# Model loading with caching for efficient GPU usage
-class ImageProcessor:
-    def __init__(self):
-        self.model = None
-        self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-        self.transform = transforms.Compose([
-            transforms.Resize(256),
-            transforms.CenterCrop(224),
-            transforms.ToTensor(),
-            transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
-        ])
-        self.class_names = None
-        self.load_model()
-    
-    def load_model(self):
-        """Load model with efficient GPU memory management"""
-        if self.model is None:
-            # Use a smaller, efficient model for GPU optimization
-            self.model = models.resnet18(pretrained=True)
-            self.model = self.model.to(self.device)
-            self.model.eval()
-            
-            # Load class names (cached)
-            try:
-                # Try to load from common locations
-                class_names_path = os.path.join(os.path.dirname(__file__), "imagenet_classes.txt")
-                if os.path.exists(class_names_path):
-                    with open(class_names_path) as f:
-                        self.class_names = [line.strip() for line in f.readlines()]
-                else:
-                    # Fallback to default class names
-                    self.class_names = [f"class_{i}" for i in range(1000)]
-            except Exception:
-                self.class_names = [f"class_{i}" for i in range(1000)]
-    
-    def process_image(self, image):
-        """Process image with GPU optimization"""
-        start_time = time.time()
-        
-        # Convert PIL Image to tensor
-        img_tensor = self.transform(image).unsqueeze(0).to(self.device)
-        
-        # Run inference with GPU optimization
-        with torch.no_grad():
-            outputs = self.model(img_tensor)
-        
-        # Get predictions
-        probabilities = torch.nn.functional.softmax(outputs[0], dim=0)
-        top5_prob, top5_catid = torch.topk(probabilities, 5)
-        
-        # Format results
-        results = []
-        for i in range(top5_prob.size(0)):
-            class_name = self.class_names[top5_catid[i]]
-            results.append({
-                "class": class_name,
-                "probability": float(top5_prob[i])
-            })
-        
-        processing_time = time.time() - start_time
-        return results, processing_time
-
-# Initialize processor (loaded once when the app starts)
-processor = ImageProcessor()
-
-def predict(image):
-    """Main prediction function with error handling and loading states"""
-    if image is None:
-        raise gr.Error("Please upload an image first!")
-    
-    try:
-        # Show loading state
-        yield "Processing image...", None, None
-        
-        # Process the image
-        results, processing_time = processor.process_image(image)
-        
-        # Format output
-        output_text = f"Processing time: {processing_time:.2f} seconds\n\nTop 5 predictions:\n"
-        for result in results:
-            output_text += f"- {result['class']}: {result['probability']:.2%}\n"
-        
-        yield None, output_text, None
-        
-    except Exception as e:
-        raise gr.Error(f"Error processing image: {str(e)}")
-
-# Create custom theme for modern UI
-custom_theme = gr.themes.Soft(
-    primary_hue="blue",
-    secondary_hue="indigo",
-    neutral_hue="slate",
-    font=gr.themes.GoogleFont("Inter"),
-    text_size="lg",
-    spacing_size="lg",
-    radius_size="md"
-).set(
-    button_primary_background_fill="*primary_600",
-    button_primary_background_fill_hover="*primary_700",
-    block_title_text_weight="600",
-    input_background_fill="*surface_light",
-    input_border_color="*border_subtle",
-    input_border_width="1px",
-    input_border_radius="8px",
-)
-
-with gr.Blocks() as demo:
-    gr.Markdown("""
-    # 🚀 AI Image Classifier
-    
-    Upload an image to get instant predictions using a pre-trained ResNet model. 
-    Optimized for efficient GPU usage with smart container management.
-    
-    Built with anycoder - [Deploy on Hugging Face Spaces](https://huggingface.co/spaces/akhaliq/anycoder)
-    """)
-    
-    with gr.Row():
-        with gr.Column():
-            image_input = gr.Image(
-                type="pil",
-                label="Upload Image",
-                height=400,
-                interactive=True
-            )
-            upload_btn = gr.Button("Process Image", variant="primary", scale=1)
-        
-        with gr.Column():
-            output_text = gr.Textbox(
-                label="Predictions",
-                lines=10,
-                interactive=False,
-                placeholder="Results will appear here..."
-            )
-            processing_status = gr.Textbox(
-                label="Status",
-                interactive=False,
-                visible=False
-            )
-    
-    # Event listeners with proper Gradio 6 syntax
-    upload_btn.click(
-        fn=predict,
-        inputs=[image_input],
-        outputs=[processing_status, output_text],
-        api_visibility="public"
-    )
-
-# Launch with Modal-optimized parameters
-demo.launch(
-    theme=custom_theme,
-    footer_links=[{"label": "Built with anycoder", "url": "https://huggingface.co/spaces/akhaliq/anycoder"}],
-    show_error=True,
-    server_name="0.0.0.0",
-    server_port=7860,
-    max_threads=100,
-    analytics_enabled=False,
-    share=False  # Disable share for production deployment
-)