Democratizing AI: Open-source Scalable LLM Training on GPU-based Supercomputers

Training and fine-tuning large language models (LLMs) with hundreds of billions to trillions of parameters requires tens of thousands of GPUs, and a highly scalable software stack. In this work, we present a novel four-dimensional hybrid parallel algorithm implemented in a highly scalable, portable,...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:SC24: International Conference for High Performance Computing, Networking, Storage and Analysis S. 1 - 14
Hauptverfasser: Singh, Siddharth, Singhania, Prajwal, Ranjan, Aditya, Kirchenbauer, John, Geiping, Jonas, Wen, Yuxin, Jain, Neel, Hans, Abhimanyu, Shu, Manli, Tomar, Aditya, Goldstein, Tom, Bhatele, Abhinav
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 17.11.2024
Schlagworte:
asynchrony
collective communication
Computational modeling
GPGPUs
High performance computing
Kernel
Large language models
Optimization
Parallel algorithms
parallel training
Supercomputers
Training
Training data
Transformers
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Training and fine-tuning large language models (LLMs) with hundreds of billions to trillions of parameters requires tens of thousands of GPUs, and a highly scalable software stack. In this work, we present a novel four-dimensional hybrid parallel algorithm implemented in a highly scalable, portable, open-source framework called AxoNn. We describe several performance optimizations in AxoNN to improve matrix multiply kernel performance, overlap non-blocking collectives with computation, and performance modeling to choose performance optimal configurations. These have resulted in unprecedented scaling and peak flop/s (bf16) for training of GPT-style transformer models on Perlmutter (620.1 Petaflop/s), Frontier (1.381 Exaflop/s) and Alps (1.423 Exaflop/s). While the abilities of LLMs improve with the number of trainable parameters, so do privacy and copyright risks caused by memorization of training data, which can cause disclosure of sensitive or private information at inference time. We highlight this side effect of scale through experiments that explore "catastrophic memorization," where models are sufficiently large to memorize training data in a single pass, and present an approach to prevent it. As part of this study, we demonstrate fine-tuning of a 405-billion parameter LLM using AxoNN on Frontier.
DOI:10.1109/SC41406.2024.00010