tensorboard.Session

tensorboard.Session(self, name, hparams=dict(), metrics=dict())

One training session of a sweep.

Attributes

Name	Type	Description
name	str	The session’s run, such as "session_3".
hparams	dict	Each hyperparameter’s value: a number, a string or a bool.
metrics	dict of str to float	Each metric’s last logged value, named tag for one logged in the session’s own run and group/tag for one logged in a run under it, such as validation/loss.