AI models
Big Transformer for Back-Translation
Training compute
4.8×10²⁰ FLOP
Published
Aug 28, 2018
Big Transformer for Back-Translation is an AI model developed by Facebook AI Research and Google Brain (United States and France), first published in August 2018. It works in the language domain, on tasks such as translation.
Training it took an estimated 4.8×10²⁰ FLOP of compute (estimation method: hardware). It was trained on roughly 4.5B datapoints. Training ran on 128 NVIDIA Tesla V100 DGXS 16 GB for about 27.67 hours. The compute alone is estimated at $2K in 2023 dollars.
Access: Open weights (unrestricted). Its weights are openly available. The reference paper has 1,155 citations. Epoch AI rates the confidence of this record as likely.
Full record
- Organization
- Facebook AI Research, Google Brain
- Country of organization
- United States, France
- Domain
- Language
- Task
- Translation
- Training compute
- 4.8×10²⁰ FLOP
- Compute estimation method
- Hardware
- Dataset size
- 4.5B
- Training hardware
- NVIDIA Tesla V100 DGXS 16 GB
- Chips used
- 128
- Training time
- 28 h
- Chip-hours
- 3.5K
- Training power draw
- 66.2 kW
- Training cost (2023 USD)
- $2K
- Numerical format
- FP16
- Model accessibility
- Open weights (unrestricted)
- Open weights
- Yes
- Citations
- 1,155
- Epoch confidence
- Likely
More from Facebook AI Research,Google Brain
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