We will use the same FeaturesDataset class that we created for ResNet and use it to create data loaders for the train and validation dataset in the following code:
# Create dataset for train and validation convolution features
trn_feat_dset = FeaturesDataset(trn_features,trn_labels)
val_feat_dset = FeaturesDataset(val_features,val_labels)
# Create data loaders for batching the train and validation datasets
trn_feat_loader = DataLoader(trn_feat_dset,batch_size=64,shuffle=True,drop_last=True)
val_feat_loader = DataLoader(val_feat_dset,batch_size=64)
Time to create the model and train it.