Paths and setup

Train - h5

Val - h5

Test - h5

Hyperparameters

Num_classes = 5

Batch_size = 64

Epochs = 50

Load Feature Function

Read Feature (X)

Read Labels (y)

One hot Encoding Labels

Load Train/Val/Test Features and Build classifier on top of ViT Features

Build Model

Compile Model

Setup Call backs

1. model checkpoints

2. Early stopping

3. Reduce Ron Plateau

Train model with Training Data

Validate on X_Val every epoch and use callbacks

Evaluate model on test Data

Accuracy