neptune-tensorboard 1.0.0rc2

Neptune.ai Tensorboard integration library

neptune-tensorboard

Note

This integration is still being updated for the main Neptune client library. It is currently only available for the Neptune legacy API.

---

Overview

neptune-tensorboard integrates TensorBoard with Neptune to give you the best of both worlds. Enjoy tracking from TensorBoard with the organization and collaboration of Neptune.

With neptune-tensorboard you can have your TensorBoard experiment runs hosted in a beautiful knowledge repo that lets you invite and manage project contributors.

All you need to do to convert your past runs from TensorBoard logdir is run:

neptune tensorboard /path/to/logdir --project USER_NAME/PROJECT_NAME

You can connect Neptune to your TensorBoard and log all future experiments by adding the following to your scripts:

import neptune
import neptune_tensorboard as neptune_tb

neptune.init(api_token='YOUR_TOKEN', project_qualified_name='USER_NAME/PROJECT_NAME') # credentials
neptune_tb.integrate_with_tensorflow()

neptune.create_experiment()

You will have your experiments hosted on Neptune and easily shareable with the world.

Documentation

See neptune-tensorboard docs for more info.

Get started

Register

Go to neptune.ai and sign up.

It is completely free for individuals and non-organizations, and you can invite others to join your team!

Get your API token

In the bottom-left corner, click your user menu and select Get your API token.

Set NEPTUNE_API_TOKEN environment variable

Go to your console and run:

export NEPTUNE_API_TOKEN='your_long_api_token'

Create your first project

Click All projects → New project. Choose a name for it and whether you want it public or private.

Install lib

pip install neptune-tensorboard

Sync your TensorBoard logdir with Neptune

neptune tensorboard /path/to/logdir --project USER_NAME/PROJECT_NAME

Connect Neptune to TensorBoard to log future runs

You can connect Neptune to your TensorBoard and log all future experiments by adding the following to your scripts:

import neptune
import neptune_tensorboard as neptune_tb

neptune.init(api_token='YOUR_TOKEN', project_qualified_name='USER_NAME/PROJECT_NAME')  # credentials
neptune_tb.integrate_with_tensorflow()

neptune.create_experiment()

Explore and Share

You can now explore and organize your experiments in Neptune, and share it with anyone:

• by sending a link to your project, experiment or chart if it is public
• or invite people to your project if you want to keep it private!

Getting help

If you get stuck, don't worry. We are here to help.

The best order of communication is:

• GitHub issues
• Email

Contributing

If you see something that you don't like, you are more than welcome to contribute!

There are many options:

• Submit a feature request or a bug here, on Github
• Submit a pull request that deals with an open feature request or bug
• Spread the word about Neptune in your community