Tensorshow

Tensorshow is a python module for inspecting TFRecords.

Tensorshow can convert a TFRecord to a pandas dataframe.

import tensorshow

# The column labels of `df` are the features of the tf.train.example protobufs.
df = tensorshow.dataframe_from('path/to/tfrecord')

Tensorshow can be used as a command line utility. It will convert a tfrecord to an html file on the command line.

python tensorshow --tfrecord=/Users/joel/train.tfrecord --html_file=Users/joel/out.html

Images stored as byte strings will be automatically detected and displayed as images rather than text. The out.html file looks like this when you open it with a browser.

Tensorshow can be used in a jupyter notebook to preview a TFRecord. The head function will show the first 5 tf.train.examples by default and the random function will show five random tf.train.examples from the tfrecord.