Update app.py

app.py

@@ -9,15 +9,21 @@ from torchvision import datasets, transforms, models
 
 MODEL_NAME = 'ResNeXt-101-64x4d'
 DEVICE = "cuda" if torch.cuda.is_available() else "cpu"
+MEAN = [0.485, 0.456, 0.406]
+STD = [0.229, 0.224, 0.225]
 
 if (torch.cuda.is_available()):
 	model = torch.load(MODEL_NAME+'.ptm')
-else
+else:
 	model = torch.load(MODEL_NAME+'.ptm', map_location=torch.device('cpu'))
 
-
 labels = ['Apple__black_rot', 'Apple__healthy', 'Apple__rust', 'Apple__scab', 'Cassava__bacterial_blight', 'Cassava__brown_streak_disease', 'Cassava__green_mottle', 'Cassava__healthy', 'Cassava__mosaic_disease', 'Cherry__healthy', 'Cherry__powdery_mildew', 'Chili__healthy', 'Chili__leaf curl', 'Chili__leaf spot', 'Chili__whitefly', 'Chili__yellowish', 'Coffee__cercospora_leaf_spot', 'Coffee__healthy', 'Coffee__red_spider_mite', 'Coffee__rust', 'Corn__common_rust', 'Corn__gray_leaf_spot', 'Corn__healthy', 'Corn__northern_leaf_blight', 'Cucumber__diseased', 'Cucumber__healthy', 'Gauva__diseased', 'Gauva__healthy', 'Grape__black_measles', 'Grape__black_rot', 'Grape__healthy', 'Grape__leaf_blight_(isariopsis_leaf_spot)', 'Jamun__diseased', 'Jamun__healthy', 'Lemon__diseased', 'Lemon__healthy', 'Mango__diseased', 'Mango__healthy', 'Peach__bacterial_spot', 'Peach__healthy', 'Pepper_bell__bacterial_spot', 'Pepper_bell__healthy', 'Pomegranate__diseased', 'Pomegranate__healthy', 'Potato__early_blight', 'Potato__healthy', 'Potato__late_blight', 'Rice__brown_spot', 'Rice__healthy', 'Rice__hispa', 'Rice__leaf_blast', 'Rice__neck_blast', 'Soybean__bacterial_blight', 'Soybean__caterpillar', 'Soybean__diabrotica_speciosa', 'Soybean__downy_mildew', 'Soybean__healthy', 'Soybean__mosaic_virus', 'Soybean__powdery_mildew', 'Soybean__rust', 'Soybean__southern_blight', 'Strawberry___leaf_scorch', 'Strawberry__healthy', 'Sugarcane__bacterial_blight', 'Sugarcane__healthy', 'Sugarcane__red_rot', 'Sugarcane__red_stripe', 'Sugarcane__rust', 'Tea__algal_leaf', 'Tea__anthracnose', 'Tea__bird_eye_spot', 'Tea__brown_blight', 'Tea__healthy', 'Tea__red_leaf_spot', 'Tomato__bacterial_spot', 'Tomato__early_blight', 'Tomato__healthy', 'Tomato__late_blight', 'Tomato__leaf_mold', 'Tomato__mosaic_virus', 'Tomato__septoria_leaf_spot', 'Tomato__spider_mites_(two_spotted_spider_mite)', 'Tomato__target_spot', 'Tomato__yellow_leaf_curl_virus', 'Wheat__brown_rust', 'Wheat__healthy', 'Wheat__septoria', 'Wheat__yellow_rust']
 
+predictTransform = transforms.Compose([
+	transforms.ToTensor(),
+	transforms.Normalize(mean=MEAN, std=STD)
+])
+
 def predict(img):
     img = predictTransform(img).unsqueeze(0).to(DEVICE)
     with torch.inference_mode():