CODEWITHSUNDEEP
tensorflowgoogle-colaboratorytensorflow2.0
VI11age_ID10T
VI11age_ID10T

Reputation: 161

Tensorflow 2.0 beta GPU running in jupyter notebook, but not in google colab

I am working with tensorflow 2.0 beta, and while i managed to get my GPU working on anaconda through a few youtube tutorials I am unable to get my gpu running in google colab. I know google has the option to enable a gpu from one of their servers but My GTX 1070 is much faster, and i need to run off colab and not just Jupyter exclusively.

So I read the documentation like a good boy and the only thing i think i could have done wrong is my path settings I have screenshots bellow.

I followed several different youtube tutorials faithfully until the final one here gave me a way to install it to jupyter. Which is great, but I also need it to run on google colab as well.

I've been trying this since Friday and it's now tuesday and I'm losing my mind over this. Help me stackoverflow, you're my only hope.

https://i.sstatic.net/FtXs3.jpg

Upvotes: 1

Views: 312

Answers (1)

Stewart_R
Stewart_R

Reputation: 14515

If you can get it running on your own Jupyter server then you can point colab to that local server.

Full instructions here: https://research.google.com/colaboratory/local-runtimes.html but edited highlights are:

  1. install jupyter_http_over_ws:
pip install jupyter_http_over_ws
jupyter serverextension enable --py jupyter_http_over_ws
  1. start your local server allowing colab domain:
jupyter notebook \
  --NotebookApp.allow_origin='https://colab.research.google.com' \
  --port=8888 \
  --NotebookApp.port_retries=0
  1. Click 'connect to local runtime' in colab

Upvotes: 1

Related Questions