Automating Code Adaptation for MLOps -- A Benchmarking Study on LLMs

TL;DR

This study benchmarks the ability of GPT-3.5-turbo and WizardCoder models to automate MLOps functionalities in ML code, demonstrating GPT-3.5-turbo's superior performance across various tasks such as code adaptation and translation.

Contribution

It introduces a benchmarking framework for evaluating LLMs on MLOps tasks and shows GPT-3.5-turbo's significant advantages over WizardCoder in these functionalities.

Findings

GPT-3.5-turbo achieves 55% Pass@3 in model optimization.

GPT-3.5-turbo attains 100% accuracy in experiment tracking.

GPT-3.5-turbo outperforms WizardCoder in multiple MLOps tasks.

Abstract

This paper explores the possibilities of the current generation of Large Language Models for incorporating Machine Learning Operations (MLOps) functionalities into ML training code bases. We evaluate the performance of OpenAI (gpt-3.5-turbo) and WizardCoder (open-source, 15B parameters) models on the automated accomplishment of various MLOps functionalities in different settings. We perform a benchmarking study that assesses the ability of these models to: (1) adapt existing code samples (Inlining) with component-specific MLOps functionality such as MLflow and Weights & Biases for experiment tracking, Optuna for hyperparameter optimization etc., and (2) perform the task of Translation from one component of an MLOps functionality to another, e.g., translating existing GitPython library based version control code to Data Version Control library based. We also propose three different approaches that involve teaching LLMs to comprehend the API documentation of the components as a reference while accomplishing the Translation tasks. In our evaluations, the gpt-3.5-turbo model significantly outperforms WizardCoder by achieving impressive Pass@3 accuracy in model optimization (55% compared to 0% by WizardCoder), experiment tracking (100%, compared to 62.5% by WizardCoder), model registration (92% compared to 42% by WizardCoder) and hyperparameter optimization (83% compared to 58% by WizardCoder) on average, in their best possible settings, showcasing its superior code adaptability performance in complex MLOps tasks.