Tensorboard not found as magic function in jupyter

Question

I want to run tensorboard in jupyter using the latest tensorflow 2.0.0a0. With the tensorboard version 1.13.1, and python 3.6.

using

... %tensorboard --logdir {logs_base_dir}

I get the error :

UsageError: Line magic function %tensorboard not found

Do you have an idea what the problem could be? It seems that all versions are up to date and the command seems correct too.

Thanks

Answer 1

UPDATE

For newer TF versions (tensorflow>=1.14.0 & tensorflow != 2.0.0a0 - newer than TF2.0-alpha) load the extension like this

%load_ext tensorboard

OLD ANSWER

The extension needs to be loaded first:

%load_ext tensorboard.notebook
%tensorboard --logdir {logs_base_dir}

Answer 2

If you are using TF 2.0 you can execute the code below from your notebook environment, prior to your magic command:

%load_ext tensorboard

Answer 3

That's how I solved it

%load_ext tensorboard
%tensorboard --logdir /content/drive/MyDrive/Dog\ Vision/logs

After --logdir, this is my path directory /content/drive/MyDrive/Dog\ Vision/logs. It should be different for you.

Answer 4

Year 2020

I needed to pip uninstall both tensorflow and tensorboard

Then install tf-nightly

In order to get this to work

%load_ext tensorboard.notebook
%tensorboard --logdir {logs_base_dir}