AI models
Transformer + Simple Recurrent Unit
Training compute
1.1×10¹⁹ FLOP
Parameters
90M
Published
Sep 17, 2018
Transformer + Simple Recurrent Unit is an AI model developed by ASAPP, Cornell University, Google and Princeton University (United States), first published in September 2018. It works in the language domain, on tasks such as translation.
Training it took an estimated 1.1×10¹⁹ FLOP of compute (estimation method: hardware). The model has 90,000,000 parameters. It was trained on roughly 112.5M datapoints. Training ran on 8 NVIDIA V100. The compute alone is estimated at $45 in 2023 dollars.
Access: Unreleased. Its weights are not openly released. The reference paper has 306 citations. Epoch AI rates the confidence of this record as confident.
Full record
- Organization
- ASAPP, Cornell University, Google, Princeton University
- Country of organization
- United States
- Domain
- Language
- Task
- Translation
- Training compute
- 1.1×10¹⁹ FLOP
- Compute estimation method
- Hardware
- Parameters
- 90,000,000
- Dataset size
- 112.5M
- Training hardware
- NVIDIA V100
- Chips used
- 8
- Training power draw
- 5.0 kW
- Training cost (2023 USD)
- $45
- Model accessibility
- Unreleased
- Open weights
- No
- Citations
- 306
- 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