CARP

Training compute

10²² FLOP

Parameters

643M

Published

Feb 6, 2024

CARP is an AI model developed by Microsoft Research (United States), first published in February 2024. It works in the biology domain, on tasks such as protein or nucleotide language model (plm/nlm).

Training it took an estimated 10²² FLOP of compute (estimation method: hardware,operation counting). The model has 643,000,000 parameters. It was trained on roughly 1.9B datapoints. Training ran on 128 NVIDIA V100.

Access: Open weights (unrestricted). Its weights are openly available. The reference paper has 138 citations. Epoch AI rates the confidence of this record as confident.

Full record

Organization: Microsoft Research
Country of organization: United States
Domain: Biology
Task: Protein or nucleotide language model (pLM/nLM)
Training compute: 10²² FLOP
Compute estimation method: Hardware, Operation counting
Parameters: 643,000,000
Dataset size: 1.9B
Training hardware: NVIDIA V100
Chips used: 128
Training power draw: 76.0 kW
Model accessibility: Open weights (unrestricted)
Open weights: Yes
Citations: 138
Epoch confidence: Confident