Petals: Collaborative Inference and Fine-tuning of Large Models

TL;DR

Petals enables collaborative inference and fine-tuning of large language models by combining resources from multiple parties, making it more accessible and flexible for research and interactive applications.

Contribution

This work introduces Petals, a system that allows collaborative inference and fine-tuning of large models, outperforming offloading techniques and providing access to model internals.

Findings

Inference of BLOOM-176B on consumer GPUs at 1 step/sec

Petals outperforms RAM offloading for large models

Supports training and sharing custom model extensions

Abstract

Many NLP tasks benefit from using large language models (LLMs) that often have more than 100 billion parameters. With the release of BLOOM-176B and OPT-175B, everyone can download pretrained models of this scale. Still, using these models requires high-end hardware unavailable to many researchers. In some cases, LLMs can be used more affordably via RAM offloading or hosted APIs. However, these techniques have innate limitations: offloading is too slow for interactive inference, while APIs are not flexible enough for research that requires access to weights, attention or logits. In this work, we propose Petals - a system for inference and fine-tuning of large models collaboratively by joining the resources of multiple parties. We demonstrate that this strategy outperforms offloading for very large models, running inference of BLOOM-176B on consumer GPUs with $\approx$ 1 step per second, which is enough for many interactive LLM applications. Unlike most inference APIs, Petals also natively exposes hidden states of served models, allowing to train and share custom model extensions based on efficient fine-tuning methods.