AI models
ProteinBERT
Training compute
6.5×10¹⁹ FLOP
Parameters
16M
Published
Feb 10, 2022
ProteinBERT is an AI model developed by Hebrew University of Jerusalem, Ben-Gurion University of the Negev and Deep Trading (Israel and United States), first published in February 2022. It works in the biology domain, on tasks such as proteins, protein generation and protein representation learning.
Training it took an estimated 6.5×10¹⁹ FLOP of compute (estimation method: hardware). The model has 16,000,000 parameters. It was trained on roughly 37.6B datapoints. Training ran on 1 NVIDIA Quadro RTX 5000 for about 672 hours.
Access: Open weights (unrestricted). Its weights are openly available. The reference paper has 775 citations. Epoch AI rates the confidence of this record as confident.
Full record
- Organization
- Hebrew University of Jerusalem, Ben-Gurion University of the Negev, Deep Trading
- Country of organization
- Israel, United States
- Domain
- Biology
- Task
- Proteins, Protein generation, Protein representation learning
- Training compute
- 6.5×10¹⁹ FLOP
- Compute estimation method
- Hardware
- Parameters
- 16,000,000
- Dataset size
- 37.6B
- Training hardware
- NVIDIA Quadro RTX 5000
- Chips used
- 1
- Training time
- 672 h
- Training power draw
- 254 W
- Model accessibility
- Open weights (unrestricted)
- Open weights
- Yes
- Citations
- 775
- Epoch confidence
- Confident
SourceEpoch AI, 'AI Models'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/data/ai-models. Licensed under CC BY 4.0.
← All ai models