LASER: Language Model Regression for Semi-Structured Workflow Resource and Runtime Estimation

TL;DR

LASER fine-tunes large language models to accurately predict resource usage and runtime for semi-structured cloud workflow jobs, improving efficiency and surpassing existing methods.

Contribution

The paper introduces LASER, a novel framework that leverages LLMs with specialized encoding and decoding techniques for multi-target regression on complex workflow configurations.

Findings

LASER outperforms human experts and SOTA tabular ML baselines.

Scientific notation encoding improves regression accuracy across multiple magnitudes.

Constrained decoding reduces inference latency by over 30%.

Abstract

Accurate prediction of resource consumption and runtime for cloud workflow jobs is critical for scheduling efficiency, yet remains challenging due to the semi-structured nature of job configurations -- comprising shell commands, tool-specific parameters, dependency graphs, and hierarchical metadata. Traditional ML approaches require brittle feature engineering to flatten this rich information into fixed-size vectors, losing critical semantic context. We present LASER, a framework that fine-tunes LLMs on serialized workflow job configurations for multi-target resource and runtime regression. To address the challenges of numerical regression via generation, we introduce scientific notation output encoding for targets spanning multiple orders of magnitude, and constrained decoding with prefix filling to enforce output validity while reducing inference latency by over 30%. We further show that full-attention fine-tuning improves accuracy over sliding-window LLMs on long job contexts. Validated on large-scale chip design workloads, and GHARuntime, a new public benchmark derived from 580,000+ GitHub Actions runs across 27,000+ repositories, LASER outperforms human experts and SOTA tabular ML baselines, with clear model- and data-scaling behavior, establishing a new paradigm for LLM-based regression on semi-structured workflow data.