Explanation:
To utilize GPUs effectively, installing GPU drivers is essential if they're not already present. Deep Learning VM images come with these drivers pre-installed. Switching to Pytorch would necessitate significant recoding and wouldn't solve the GPU driver issue. Upgrading Python doesn't resolve missing drivers, and altering the service account's role doesn't address the driver installation requirement. For more details, visit Google Cloud's documentation on GPU drivers.
Ultimate access to all questions.
No comments yet.
You've set up a Compute Engine instance with an attached GPU, yet the GPU isn't utilized during Tensorflow model training. Which two actions could you take to enable GPU usage for your model training?
A
Switch from Tensorflow to Pytorch
B
Upgrade Python 3 on the virtual machine
C
Opt for a Deep Learning VM image
D
Install the necessary GPU drivers
E
Assign the Owner basic role to the VM's service account