YaYi 2.0

Training compute

4.8×10²³ FLOP

Parameters

30B

Published

Dec 22, 2023

YaYi 2.0 is an AI model developed by Yayi (Wenge) (China), first published in December 2023. It works in the language domain, on tasks such as language modeling/generation.

Training it took an estimated 4.8×10²³ FLOP of compute. The model has 30,000,000,000 parameters. Training ran on 1,000 NVIDIA A800 PCIe 40 GB.

Access: Open weights (restricted use). Its weights are openly available. The reference paper has 8 citations. Epoch AI rates the confidence of this record as likely.

Full record

Organization: Yayi (Wenge)
Country of organization: China
Domain: Language
Task: Language modeling/generation
Training compute: 4.8×10²³ FLOP
Parameters: 30,000,000,000
Training hardware: NVIDIA A800 PCIe 40 GB
Chips used: 1,000
Training power draw: 495.5 kW
Model accessibility: Open weights (restricted use)
Open weights: Yes
Citations: 8
Epoch confidence: Likely