tensorboard.Embedding

tensorboard.Embedding(self, name, vectors, metadata=dict())

One embedding of a Projector log directory.

Attributes

Name	Type	Description
name	str	The tensor_name it was logged under.
vectors	numpy.ndarray	One row per point, shape (points, dimensions).
metadata	dict of str to list of str	Each metadata column, one value per point, such as {"label": [...]}. A metadata file with one column and no header is called label.