Mastering Computer Vision with TensorFlow 2.x

TensorBoard running on a local machine

TensorBoard can be added to your model training by adding the following lines of code. Check the code supplied on the GitHub page for the exact location:

tensorboard_callback = tf.keras.callbacks.TensorBoard(log_dir=log_dir, histogram_freq=1)history = model.fit(x=x_train, y=y_train, epochs=25, validation_data=(x_test, y_test), callbacks=[tensorboard_callback])

Then, TensorBoard graphs can be visualized after training has started by typing the following in the terminal:

%tensorboard --logdir logs/fit