Preprocess_input for pytorch

Is there any function in pytorch equivalent to preprocess_input in the below code

from keras.applications.resnet50 import ResNet50, preprocess_input

def people_counter(img_path):
    # extract bottleneck features
    bottleneck_feature = ResNet50(weights='imagenet', include_top=False).predict(preprocess_input(path_to_tensor(img_path)))
    
    # obtain predicted vector
    predicted_vector = model.predict(bottleneck_feature)
    
    # return dog breed that is predicted by the model
    return classes[np.argmax(predicted_vector)]

Sure, see the tutorial documentation here on data loading. Particularly, look at the Transforms section which discusses doing some preprocessing with your dataset and loaders.