AI models
B2T connection (16L)
Training compute
2.8×10¹⁹ FLOP
Published
Jun 1, 2022
B2T connection (16L) is an AI model developed by LINE Corporation and Tohoku University (Japan), first published in June 2022. It works in the language domain, on tasks such as language modeling/generation and question answering.
Training it took an estimated 2.8×10¹⁹ FLOP of compute (estimation method: hardware,operation counting). It was trained on roughly 103M datapoints. Training ran on 192 NVIDIA P100 for about 7 hours.
Access: Open weights (unrestricted). Its weights are openly available. The reference paper has 18 citations. Epoch AI rates the confidence of this record as confident.
Full record
- Organization
- LINE Corporation, Tohoku University
- Country of organization
- Japan
- Domain
- Language
- Task
- Language modeling/generation, Question answering
- Training compute
- 2.8×10¹⁹ FLOP
- Compute estimation method
- Hardware, Operation counting
- Dataset size
- 103M
- Training hardware
- NVIDIA P100
- Chips used
- 192
- Training time
- 7 h
- Training power draw
- 96.3 kW
- Model accessibility
- Open weights (unrestricted)
- Open weights
- Yes
- Citations
- 18
- 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