Headline about the
Pharmaceutical Industry

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat.

Protein Sequence Generation

Our protein sequence generator demo leverages the power of large language models to generate novel protein structures. Given that protein sequences are analogous to natural language, LLMs can effectively capture the relationships between amino acid sequences, structures, and functions.

For this demo, we utilize the Phi 3 model due to its compact size and competitive performance. Users can specify the desired properties of the protein they wish to design, and the LLM will generate the corresponding protein structure. The predicted structure can then be visualized using AlphaFold, providing a clear representation of the designed protein. This technology has potential applications across various industries, including drug development, nanotechnology, and biocatalysis.

WHITEPAPER

Evaluating Protein Language Modeling on ET-SoC-1

This whitepaper investigates techniques for customizing open-source large language models (e.g., LLama3 and Phi3) for Protein Language Modeling. These modifications enable users to create novel protein sequences using natural language input. We demonstrate how we’ve expanded Protein Language models to produce broader variations within protein families. The paper also outlines methods for compiling and running these enhanced models on Esperanto’s ET-SoC-1 platform. This facilitates a comparative study of performance and efficiency between conventional GPUs and this innovative RISC-V-based general-purpose machine learning accelerator.

DOWNLOAD WHITEPAPER