S4

Training compute

7.8×10¹⁹ FLOP

Parameters

249M

Published

Oct 31, 2021

S4 is an AI model developed by Stanford University (United States), first published in October 2021. It works in the language domain, on tasks such as language modeling/generation.

Training it took an estimated 7.8×10¹⁹ FLOP of compute. The model has 249,000,000 parameters. It was trained on roughly 103M datapoints. Training ran on 8 NVIDIA A100.

Access: Open weights (unrestricted). Its weights are openly available. The reference paper has 3,576 citations. Epoch AI rates the confidence of this record as likely.

Full record

Organization: Stanford University
Country of organization: United States
Domain: Language
Task: Language modeling/generation
Training compute: 7.8×10¹⁹ FLOP
Parameters: 249,000,000
Dataset size: 103M
Training hardware: NVIDIA A100
Chips used: 8
Training power draw: 6.5 kW
Numerical format: FP16
Model accessibility: Open weights (unrestricted)
Open weights: Yes
Citations: 3,576
Epoch confidence: Likely