Verify Distributed Deep Learning Model Implementation Refinement with Iterative Relation Inference

TL;DR

This paper presents GraphGuard, a static analysis tool that uses iterative rewriting to verify that distributed deep learning models correctly refine their sequential counterparts, helping to identify bugs in large-scale model implementations.

Contribution

It introduces a scalable static verification method for distributed deep learning models that can detect bugs by checking model refinement through iterative rewriting.

Findings

Successfully applied to GPT and Llama-3 models.

Scales to large models and deployments.

Provides actionable bug localization output.

Abstract

Distributed machine learning training and inference is common today because today's large models require more memory and compute than can be provided by a single GPU. Distributed models are generally produced by programmers who take a sequential model specification and apply several distribution strategies to distribute state and computation across GPUs. Unfortunately, bugs can be introduced in the process, and a distributed model implementation's outputs might differ from the sequential model's outputs. In this paper, we describe an approach to statically identify such bugs by checking model refinement, that is, can the sequential model's outputs be reconstructed from the distributed model's outputs? Our approach, implemented in GraphGuard, uses iterative rewriting to prove model refinement. Our approach can scale to today's large models and deployments: we evaluate it using GPT and Llama-3. Further, it provides actionable output that aids in bug localization.