import torch
from torchvision import models, transforms
from PIL import Image
import gradio as gr

# Mendefinisikan nama kelas
class_names = [
    "calculus",
    "caries",
    "gingivitis",
    "hypodontia",
    "mouth_ulcer",
    "tooth_discoloration"
]

# Mengatur jumlah kelas
num_classes = len(class_names)

# Membuat dan mengkonfigurasi model
model = models.resnet50(pretrained=False)
model.fc = torch.nn.Linear(model.fc.in_features, num_classes)

# Memuat bobot model (sesuaikan path jika diperlukan)
model.load_state_dict(torch.load('best_model.pth', map_location=torch.device('cpu')))
model.eval()

# Mengatur transformasi preprocessing
preprocess = transforms.Compose([
    transforms.Resize((224, 224)),
    transforms.ToTensor(),
    transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
])

# Fungsi untuk melakukan prediksi
def predict_image(image, model, preprocess, class_names):
    processed_image = preprocess(image).unsqueeze(0)

    with torch.no_grad():
        outputs = model(processed_image)
        _, predicted = torch.max(outputs, 1)
        predicted_class = class_names[predicted.item()]

    return predicted_class

# Membuat interface Gradio
iface = gr.Interface(
    fn=predict_image, 
    inputs=[gr.inputs.Image(type='pil')],
    outputs=gr.outputs.Label(num_top_classes=1),
    title="Klasifikasi Gambar Medis",
    description="Upload gambar untuk memprediksi kelasnya."
)

# Menjalankan aplikasi Gradio
iface.launch()