Calm2-7B

Training compute

5.5×10²² FLOP

Parameters

7B

Published

Nov 1, 2023

Calm2-7B is an AI model developed by CyberAgent (Japan), first published in November 2023. It works in the language domain, on tasks such as language modeling/generation.

Training it took an estimated 5.5×10²² FLOP of compute (estimation method: operation counting). The model has 7,000,000,000 parameters. It was trained on roughly 1.3T datapoints.

Access: Open weights (unrestricted). Its weights are openly available. Epoch AI rates the confidence of this record as confident.

Full record

Organization: CyberAgent
Country of organization: Japan
Domain: Language
Task: Language modeling/generation
Training compute: 5.5×10²² FLOP
Compute estimation method: Operation counting
Parameters: 7,000,000,000
Dataset size: 1.3T
Model accessibility: Open weights (unrestricted)
Open weights: Yes
Epoch confidence: Confident