What is Google Colab?
Introduction to Google Colab.
If you’re following me on Twitter or LinkedIn, you must have seen some code snippets that seems like a Jupyter notebooks but looks way cooler. You can see in the comments, people asking me about the editor profusely. Well, here’s a whole article on it
What’s Colab?
Google Colaboratory is a hosted Jupyter notebook environment that is free to use and requires no setup. It’s a free cloud service and it supports free GPU!
You can simply improve your Python programming language coding skills OR develop deep learning applications using popular libraries such as Keras, TensorFlow, PyTorch, and OpenCV.
Now all these libraries are already installed for you. So there’s no need to go through the hassle of installing the libraries. Just get right onto coding!
The most important feature that distinguishes Colab from other free cloud services is that it provides GPU and is totally free. You need no membership or a credit card to use it. Colab provides you a free CPU and a GPU.
Why is it Free?
Google released it to the public with the aim of improving the Machine Learning education and research. Thank you Google!
Why GPUs?
If you’re into Machine Learning and Deep Learning, you already know the answer. For Newbies- GPUs are essential because of the amount of data you need to operate on. The mathematical calculation for Deep Learning on a CPU can take months! But these calculations can be assigned to GPUs for faster computation. Training a Neural Network on a CPU in not recommended, a GPU is needed for high computation.
As we all know GPUs are expensive and so are Cloud Platforms like AWS and GCP. So where can we get a GPU? Colab comes to the rescue! Colab provides you a free Nvidia Tesla K80 GPU which is priced around 1.7 lakh Rupees and is a absolute beast in performance. You can use this GPU to train your Neural Network on the cloud and enjoy the hassle free and fast performance and focus on the problems than waiting for hours to train and then tune or debug your model.
You can also use colab for Kaggle or other competitions where you can download Gigabytes of datasets directly to the cloud within seconds, irrelative to your internet connection and not on your local machine.
TPU support is also available on colab. It appears that Google Colab is now offering TPUs as well, although it’s not stable as of now. So you can play around with that as well.
Where to Save?
Your Ipython notebook gets saved on your Google Drive. You can also commit the notebook on your GitHub from colab itself.
Bonus
Unlike your Jupyter notebooks, colab is also available in dark theme. So if you’re a late night coder like me and you love Sublime text, you’re golden!
