Libraries like Sklearn and Tensorflow are great as they simplify the development of machine learning models for us. Nonetheless from a learner's perspective, the problem with using these APIs is that they take away a great chunk of the learning about what's really going on under the hood. To build intuitions and really understand the nuts and bolts about different models and architectures, there's no better way than trying to implement them from scratch, which is what this repo is trying to do.
Note that this repo is primarily for learning purpose and does not aim to produce the most optimal and efficient implementation of the algorithms.