CODEWITHSUNDEEP
tensorflowkerastensorboard
talipini talipini
talipini talipini

Reputation: 183

How do I add text summary to Tensorboard on Keras?

How do I add text summary to Tensorboard when using Keras?

I have setup a Tensorboard callback, but I am lost on how to add a text summary.

For example, I would like to add a text summary of different parameters used in the run into the Tensorboard so that for documentation and not lost when I revisit the run log.

One option seems to be to include all the parameter details to the logfile dir name, but that looks exhausting.

How can I better solve this problem?

Upvotes: 3

Views: 3079

Answers (3)

ironbar
ironbar

Reputation: 11

This works on tf 2.1 2.2 and 2.3

class TensorBoardExtended(TensorBoard):
    """
    Extended Tensorboard log that allows to add text

    By default logs:
    - host
    - gpus available

    Parameters
    -------------
    text_dict_to_log : dict
        Dictionary with key, value string that will be logged with Tensorboard
    kwargs : dict
        All the other parameters that are fed to Tensorboard
    """
    def __init__(self, text_dict_to_log=None, **kwargs):
        super().__init__(**kwargs)
        self.text_dict_to_log = text_dict_to_log

    def on_train_begin(self, logs=None):
        # pylint: disable= E1101
        super().on_train_begin(logs=logs)

        try:
            writer = self._get_writer('train')
        except AttributeError: # this is due to differences between tf21, 22 and 23
            writer = self._train_writer

        with writer.as_default():
            for key, value in self.text_dict_to_log.items():
                tf.summary.text(key, tf.convert_to_tensor(value), step=0)

Upvotes: 0

talipini talipini
talipini talipini

Reputation: 183

I have not looked if they removed or moved the default writer. But you can create your own writer and it will show up as a text summary in the Tensorboard. Since TF2.0 defaults to Eager mode, no need for session.

dir_name = os.path.join("your_log_dir", "Text Summary")
writer = tf.summary.create_file_writer(logdir=dir_name)

with writer.as_default():
    tf.summary.text(name="Run_Settings", data= self.settings_str, step=0)

Upvotes: 1

talipini talipini
talipini talipini

Reputation: 183

For anyone else looking for this, I ended up writing a custom callback extending Keras Tensorboard. I will probably log more than summary later and we can extend this to log more stuff on different events. I simplified from another question - to add plot

from keras.callbacks import TensorBoard
import tensorflow as tf

class LoggingTensorBoard(TensorBoard):    

    def __init__(self, log_dir, settings_str_to_log, **kwargs):
        super(LoggingTensorBoard, self).__init__(log_dir, **kwargs)

        self.settings_str = settings_str_to_log

    def on_train_begin(self, logs=None):
        TensorBoard.on_train_begin(self, logs=logs)

        tensor =  tf.convert_to_tensor(self.settings_str)
        summary = tf.summary.text ("Run Settings", tensor)

        with  tf.Session() as sess:
            s = sess.run(summary)
            self.writer.add_summary(s)

Create an instance of this Tensorboard and add to model.fit callback and your settings_str_to_log will show up on the Text Tab in Tensorboard.

Upvotes: 5

Related Questions