Statement Curriculum Learning

Training compute

1.8×10²² FLOP

Parameters

774M

Published

Mar 2, 2022

Statement Curriculum Learning is an AI model developed by OpenAI (United States), first published in March 2022. It works in the language and mathematics domain, on tasks such as automated theorem proving, language modeling/generation, question answering and mathematical reasoning.

Training it took an estimated 1.8×10²² FLOP of compute (estimation method: hardware). The model has 774,000,000 parameters. It was trained on roughly 372B datapoints. Training ran on NVIDIA A100.

Access: API access. Its weights are not openly released. The reference paper has 161 citations. Epoch AI rates the confidence of this record as likely.

Full record

Organization: OpenAI
Country of organization: United States
Domain: Language, Mathematics
Task: Automated theorem proving, Language modeling/generation, Question answering, Mathematical reasoning
Training compute: 1.8×10²² FLOP
Compute estimation method: Hardware
Parameters: 774,000,000
Dataset size: 372B
Training hardware: NVIDIA A100
Chip-hours: 48K
Model accessibility: API access
Open weights: No
Citations: 161
Epoch confidence: Likely