BioNeMo Framework: a modular, high-performance library for AI model development in drug discovery
It addresses the computational scaling challenges in drug discovery AI for researchers and developers, though it is incremental as it builds on existing methods with a new framework.
The paper introduces the BioNeMo Framework, a modular, high-performance library designed to facilitate training of AI models for drug discovery, achieving training of a 3B parameter protein language model on over 1 trillion tokens in 4.2 days using 256 GPUs.
Artificial Intelligence models encoding biology and chemistry are opening new routes to high-throughput and high-quality in-silico drug development. However, their training increasingly relies on computational scale, with recent protein language models (pLM) training on hundreds of graphical processing units (GPUs). We introduce the BioNeMo Framework to facilitate the training of computational biology and chemistry AI models across hundreds of GPUs. Its modular design allows the integration of individual components, such as data loaders, into existing workflows and is open to community contributions. We detail technical features of the BioNeMo Framework through use cases such as pLM pre-training and fine-tuning. On 256 NVIDIA A100s, BioNeMo Framework trains a three billion parameter BERT-based pLM on over one trillion tokens in 4.2 days. The BioNeMo Framework is open-source and free for everyone to use.