INNER CODE UNIT · Python

create_tensorboard_callback

mrdbourke/tensorflow-deep-learning · extras/helper_functions.py:137

def create_tensorboard_callback(dir_name, experiment_name):
  """
  Creates a TensorBoard callback instand to store log files.

  Stores log files with the filepath:
    "dir_name/experiment_name/current_datetime/"

  Args:
    dir_name: target directory to store TensorBoard log files
    experiment_name: name of experiment directory (e.g. efficientnet_model_1)
  """
  log_dir = dir_name + "/" + experiment_name + "/" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")
  tensorboard_callback = tf.keras.callbacks.TensorBoard(
      log_dir=log_dir
  )
  print(f"Saving TensorBoard log files to: {log_dir}")
  return tensorboard_callback

View source record →