MLBench is a framework for benchmarking distributed machine learning algorithms. The main goals of MLBench are to provide a fair benchmarking suite for software and hardware systems and to provide reference implementations for the most common distributed machine learning applications. In this report, I will summarize my contributions to the framework. Additionally, I will explain the things I learned and the obstacles I faced while using and contributing to the project. During my semester project, I realized that it is not straightforward for new contributors, especially students, to join the project. To this end, I will try to provide a description of the project structure, and explain the process of local development, testing and contribution. I hope that these instructions will be useful for future contributors and will ease the development process. Additionally, I will provide my opinion on what is the appropriate way to export each section of my report.