Tutorial 90 - Setting up your free Google Colab account for deep learning

Colab is a free Jupyter notebook environment that runs entirely in the cloud.
Out of the box, Colab supports many machine learning libraries without the need to install.
The primary reason for you to use Colab is: Free Cloud service with free GPU.
Colab uses Google Drive for storing your notebooks
Make sure you have a Google account
Make sure you log on to your Google Drive account at least once.
Access Colab at: colab.research.google.com/
You need to select GPU as part of your runtime if you’d like to use it for machine learning.
To verify the hardware specifications of your virtual machine, here are a few commands you can type in your Colab Notebook.
from psutil import *
#Number of CPUs
cpu_count()
#CPU model and speed
!lscpu |grep 'Model name'
#Available Hard disk space
!df -h / | awk '{print $4}'
#Usable memory
!free -h --si | awk '/Mem:/{print $2}'
#GPU type (make sure you select GPU in the runtime)
!nvidia-smi -L
#Full GPU specifications
!nvidia-smi

[@tilkesh]

Thank you very much.

[@surflaweb]

I didn't know those comands. Thanks man.

[@ZEISS_arivis]

Glad to help

[@shokenlee7315]

Excellent videos

[@ZEISS_arivis]

Thanks

[@user-nv2xy8dx5b]

thank you very much

[@anganamondal3646]

Quite helpful

[@ZEISS_arivis]

Glad you think so!

[@yahaisha]

good..tq

[@annicetrazafindratovolahy1512]

Can we use our local driver when training model in colab? I mean, if my dataset is located in my local driver and its size is 100GB, how do I use this dataset located in my computer using colab GPU?

[@ZEISS_arivis]

You cannot use cloud Colab resources while pointing to your local drive. You can either upload your data to Google Drive or run a local Colab instance that uses your local resources and can access local files. But this means you will not have GPU access if your local system does not have one.