In this repo will commit my day to day learning of the pytorch starting from march 4th onwards.
-> March 4th:
1. Pytorch is a framewrok or library that runs deep learning algorithms. It is simple and yet faster than tensorflow or keras.
2. Pytorch is capable of connecting with GPU to enhance the training time.
3. Pytorch data structure is torch which is equivalent to numpy as they say. It is computational faster as well to do the operations. Also numpy can't run on the GPU that's where torch was introduced to run on the GPU!