Rita-XLarge — это мощная генеративная нейросеть, обученная на 280 миллионах белковых последовательностей. Этот ИИ значительно ускоряет проектирование новых белков, открывая невероятные возможности для современной биоинженерии и медицины.
In this work we introduce RITA: a suite of autoregressive generative models for protein sequences, with up to 1.2 billion parameters, trained on over 280 million protein sequences belonging to the UniRef-100 database. Such generative models hold the promise of greatly accelerating protein design. We conduct the first systematic study of how capabilities evolve with model size for autoregressive transformers in the protein domain: we evaluate RITA models in next amino acid pre- diction, zero-shot fitness, and enzyme function prediction, showing benefits from increased scale. We release the RITA models openly, to the benefit of the research community.