How do you read Tensorboard files programmatically?

Question

How can you write a python script to read Tensorboard log files, extracting the loss and accuracy and other numerical data, without launching the GUI tensorboard --logdir=...?

Answer 1

To read a TFEvent you can get a Python iterator that yields Event protocol buffers.

# This example supposes that the events file contains summaries with a

# summary value tag 'loss'.  These could have been added by calling

# `add_summary()`, passing the output of a scalar summary op created with

# with: `tf.scalar_summary(['loss'], loss_tensor)`.

for e in tf.train.summary_iterator(path_to_events_file):

    for v in e.summary.value:

        if v.tag == 'loss' or v.tag == 'accuracy':

            print(v.simple_value)

more info: summary_iterator. 

For more details on Tensorboard,  Machine Learning Course and Tensorflow Tutorials.