NSML: A Machine Learning Platform That Enables You to Focus on Your Models

TL;DR

NSML is a machine learning platform designed to automate manual tasks like GPU management and model comparison, allowing researchers to focus more on developing models.

Contribution

The paper introduces NSML, a system that automates resource management and model tracking, addressing key challenges faced by researchers using ML libraries.

Findings

Substantial efficiency improvements in model development

Effective automation of GPU allocation and model comparison

Enhanced visualization and management of training processes

Abstract

Machine learning libraries such as TensorFlow and PyTorch simplify model implementation. However, researchers are still required to perform a non-trivial amount of manual tasks such as GPU allocation, training status tracking, and comparison of models with different hyperparameter settings. We propose a system to handle these tasks and help researchers focus on models. We present the requirements of the system based on a collection of discussions from an online study group comprising 25k members. These include automatic GPU allocation, learning status visualization, handling model parameter snapshots as well as hyperparameter modification during learning, and comparison of performance metrics between models via a leaderboard. We describe the system architecture that fulfills these requirements and present a proof-of-concept implementation, NAVER Smart Machine Learning (NSML). We test the system and confirm substantial efficiency improvements for model development.