Automated Configuration Synthesis for Machine Learning Models: A git-Based Requirement and Architecture Management System

TL;DR

This paper presents a tool that automatically generates runtime configurations for AI software systems from textual requirements stored in git, enabling traceability and optimization of configurations based on architectural descriptions.

Contribution

It introduces a novel approach combining git-based requirements management with architecture-driven configuration synthesis for AI systems.

Findings

Enables traceable configuration generation from textual requirements.

Supports dynamic property adjustment with rationale documentation.

Facilitates runtime optimization of ML models.

Abstract

This work introduces a tool for generating runtime configurations automatically from textual requirements stored as artifacts in git repositories (a.k.a. T-Reqs) alongside the software code. The tool leverages T-Reqs-modelled architectural description to identify relevant configuration properties for the deployment of artificial intelligence (AI)-enabled software systems. This enables traceable configuration generation, taking into account both functional and non-functional requirements. The resulting configuration specification also includes the dynamic properties that need to be adjusted and the rationale behind their adjustment. We show that this intermediary format can be directly used by the system or adapted for specific targets, for example in order to achieve runtime optimisations in term of ML model size before deployment.