Tensorflow on Google Cloud Platform

Question

Which Google Cloud Platform service is the easiest for running Tensorflow?

Answer 1

There are a few ways to run Tensorflow:

• CloudML
• Datalab
• and It can also be done manually

For Manually using Tensorflow:

1. Create a project
2. Open the Cloud Shell (a button at the top)
3. List machine types: gcloud compute machine-types list. You can change the machine type I used in the next command.
4. Create an instance:

gcloud compute instances create tf \
  --image container-vm \
  --zone europe-west1-c \
  --machine-type n1-standard-2

1. Run sudo docker run -d -p 8888:8888 --name tf b.gcr.io/tensorflow-udacity/assignments:0.5.0 (change the image name to the desired one)
2. Find your instance in the dashboard and edit the default network.
3. Add a firewall rule to allow your IP as well as protocol and port tcp:8888.
4. Find the External IP of the instance from the dashboard. Open IP:8888 on your browser. Done!
5. When you are finished, delete the created cluster to avoid charges.