ESM2 is a family of transformer-based protein language models developed by Meta AI. The 650M parameter variant (esm2_t33_650M_UR50D) is trained on millions of protein sequences and excels at understanding evolutionary patterns in biology. It is hosted on Hugging Face for inference and fine-tuning by researchers in computational biology and bioinformatics.
Esm2 T33 650M UR50D sits in PulseGate's Foundation models & chat category. It focuses on predicting protein properties and structures from amino acid sequences without experimental data. It is built as an open-source project for bioinformaticians. The project is open source (Open Source). Esm2 T33 650M UR50D is available on the web and API.
Behind Esm2 T33 650M UR50D is Meta AI, based in the United States, and it first shipped in 2022. Key capabilities include Masked LM, Protein Modeling, and 650M Parameters. It exposes integrations via a public API.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do