I am accustomed to creating new deep learning architectures for different problems, but which framework (Keras, Pytorch, TensorFlow) to choose is often harder.

Since there’s an uncertainty in it, it’s good to know the fundamental operations on those framework’s fundamental units (NumPy, Torch, Tensor).

In this post, I have performed a handful of the same operations across the 3 frameworks, also tried my hands on visualization for most of them.

This is a beginner-friendly post, so let’s get started.

1. Installation

pip install numpy
	pip install tensorflow
	pip install torch
view raw
pips.py hosted with ❤ by GitHub

2. Version Check

import numpy as np
	import tensorflow as tf
	import torch

	print(np.__version__)
	print(tf.__version__)
	print(torch.__version__)

	#### OUTPUT ###

	2.3.0 
	1.18.5 
	1.6.0+cu101
view raw
VC.py hosted with ❤ by GitHub

3. Array Initialization ~ 1-D, 2-D, 3-D

Scalar and 1-D Array

Scalar, 1-D, 2-D arrays

#torch #tensor #numpy #deep-learning #tensorflow #deep learning