Keras: use Tensorboard with train_on_batch()

user_4391129 0
Viewed 10790

For the keras functions fit() and fit_generator() there is the possibility of tensorboard visualization by passing a keras.callbacks.TensorBoard object to the functions. For the train_on_batch() function there obviously are no callback available. Are there other options in keras to create a Tensorboard in this case?

tensorboard keras
2 Answers

A possible way to create the TensorBoard callback, and drive it manually:

# This example shows how to use keras TensorBoard callback
# with model.train_on_batch

import tensorflow.keras as keras

# Setup the model
model = keras.models.Sequential()
model.add(...) # Add your layers
model.compile(...) # Compile as usual

batch_size=256

# Create the TensorBoard callback,
# which we will drive manually
tensorboard = keras.callbacks.TensorBoard(
  log_dir='/tmp/my_tf_logs',
  histogram_freq=0,
  batch_size=batch_size,
  write_graph=True,
  write_grads=True
)
tensorboard.set_model(model)

# Transform train_on_batch return value
# to dict expected by on_batch_end callback
def named_logs(model, logs):
  result = {}
  for l in zip(model.metrics_names, logs):
    result[l[0]] = l[1]
  return result

# Run training batches, notify tensorboard at the end of each epoch
for batch_id in range(1000):
  x_train,y_train = create_training_data(batch_size)
  logs = model.train_on_batch(x_train, y_train)
  tensorboard.on_epoch_end(batch_id, named_logs(model, logs))

tensorboard.on_train_end(None)
user_1162730
Related