Yenes/My-Model

1import torch
2import torch.nn as nn
3import torchvision.transforms as transforms
4import gradio as gr
5from PIL import Image
6
7class ConvModel(nn.Module):
8    def __init__(self):
9        super().__init__()
10        self.cnn1 = nn.Sequential(
11            nn.Conv2d(3, 16, kernel_size=3, padding=1),
12            nn.ReLU(),
13            nn.MaxPool2d(2)
14        )
15        self.cnn2 = nn.Sequential(
16            nn.Conv2d(16, 32, kernel_size=3, padding=1),
17            nn.ReLU(),
18            nn.MaxPool2d(2)
19        )
20        self.fc = nn.Sequential(
21            nn.Flatten(),
22            nn.Linear(32 * 56 * 56, 2)
23        )
24
25    def forward(self, x):
26        x = self.cnn1(x)
27        x = self.cnn2(x)
28        x = self.fc(x)
29        return x
30
31model = ConvModel()
32model.load_state_dict(torch.load("conv_model.pth", map_location="cpu"))
33model.eval()
34
35class_names=['NORMAL', 'PNEUMONIA']
36
37transform = transforms.Compose([
38    transforms.Resize((224, 224)),
39    transforms.ToTensor(),
40    transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
41])
42
43def predict(img):
44    img = transform(img).unsqueeze(0)
45    with torch.inference_mode():
46        pred_probs = torch.softmax(model(img), dim=1)
47
48    pred_labels_and_probs = {class_names[i]: float(pred_probs[0][i]) for i in range(len(class_names))}
49    return pred_labels_and_probs
50
51
52title = "Zatürre Bulucu"
53description = "Gönderilen fotoğrafa göre Sağlıklı mı yoksa Zatürre mi olduğunu tahmin eder."
54
55demo = gr.Interface(
56    fn=predict,
57    inputs=gr.Image(type="pil"),
58    outputs=[gr.Label(num_top_classes=2, label="Predictions")],
59    title=title,
60    description=description
61)
62
63demo.launch(debug=False, share=True)