app ready

app.py

@@ -1,7 +1,79 @@
 import gradio as gr
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+import torchvision.transforms as transforms
+from PIL import Image
+import numpy as np
 
-def greet(name):
-    return "Hello " + name + "!!"
+# Define the same model architecture
+class ConvNet(nn.Module):
+    def __init__(self):
+        super().__init__()
+        self.conv1 = nn.Conv2d(3, 32, 3)
+        self.pool = nn.MaxPool2d(2, 2)
+        self.conv2 = nn.Conv2d(32, 64, 3)
+        self.conv3 = nn.Conv2d(64, 64, 3)
+        self.fc1 = nn.Linear(64 * 4 * 4, 64)
+        self.fc2 = nn.Linear(64, 10)
+    
+    def forward(self, x):
+        x = F.relu(self.conv1(x))
+        x = self.pool(x)
+        x = F.relu(self.conv2(x))
+        x = self.pool(x)
+        x = F.relu(self.conv3(x))
+        x = torch.flatten(x, 1)
+        x = F.relu(self.fc1(x))
+        x = self.fc2(x)
+        return x
 
-demo = gr.Interface(fn=greet, inputs="text", outputs="text")
-demo.launch()
+# Initialize model and load weights
+model = ConvNet()
+model.load_state_dict(torch.load('cnn.pth', map_location=torch.device('cpu')))
+model.eval()
+
+# Define classes
+classes = ('plane', 'car', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck')
+
+# Define preprocessing
+transform = transforms.Compose([
+    transforms.Resize((32, 32)),
+    transforms.ToTensor(),
+    transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))
+])
+
+def predict(img):
+    # Convert to PIL Image if needed
+    if not isinstance(img, Image.Image):
+        img = Image.fromarray(img)
+    
+    # Preprocess the image
+    img = transform(img).unsqueeze(0)
+    
+    # Get predictions
+    with torch.no_grad():
+        outputs = model(img)
+        probabilities = F.softmax(outputs, dim=1)
+        
+    # Get top 3 predictions
+    probs, indices = torch.topk(probabilities, 3)
+    predictions = []
+    for prob, idx in zip(probs[0], indices[0]):
+        predictions.append((classes[idx], float(prob)))
+    
+    # Format the results
+    return {pred[0]: pred[1] for pred in predictions}
+
+# Create Gradio interface
+iface = gr.Interface(
+    fn=predict,
+    inputs=gr.Image(type="pil"),
+    outputs=gr.Label(num_top_classes=3),
+    examples=[["example1.jpg"], ["example2.jpg"]],  # Optional: Add example images
+    title="CIFAR-10 Image Classifier",
+    description="Upload an image to classify it into one of these categories: plane, car, bird, cat, deer, dog, frog, horse, ship, or truck"
+)
+
+# Launch the app
+iface.launch()