ESM2-15B
ESM2-15B is an AI model developed by Meta AI, New York University (NYU), Stanford University and Massachusetts Institute of Technology (MIT) (United States), first published in July 2022. It works in the biology domain, on tasks such as proteins, protein or nucleotide language model (plm/nlm) and protein folding prediction.
Training it took an estimated 7.4×10²² FLOP of compute (estimation method: hardware,third-party estimation). The model has 15,000,000,000 parameters. It was trained on roughly 15.4B datapoints. Training ran on 512 NVIDIA V100 for about 1.4K hours. The compute alone is estimated at $163K in 2023 dollars.
Access: Open weights (unrestricted). Its weights are openly available. The reference paper has 636 citations. Epoch AI rates the confidence of this record as confident.
- Organization
- Meta AI, New York University (NYU), Stanford University, Massachusetts Institute of Technology (MIT)
- Country of organization
- United States
- Domain
- Biology
- Task
- Proteins, Protein or nucleotide language model (pLM/nLM), Protein folding prediction
- Training compute
- 7.4×10²² FLOP
- Compute estimation method
- Hardware, Third-party estimation
- Parameters
- 15,000,000,000
- Dataset size
- 15.4B
- Training hardware
- NVIDIA V100
- Chips used
- 512
- Training time
- 1,440 h
- Training power draw
- 308.0 kW
- Training cost (2023 USD)
- $163K
- Model accessibility
- Open weights (unrestricted)
- Open weights
- Yes
- Citations
- 636
- Epoch confidence
- Confident