AI models
FastSpeech 2
Training compute
2.3×10¹⁸ FLOP
Parameters
27M
Published
Aug 8, 2022
FastSpeech 2 is an AI model developed by Zhejiang University (ZJU) and Microsoft Research Asia (China), first published in August 2022. It works in the speech domain, on tasks such as text-to-speech (tts).
Training it took an estimated 2.3×10¹⁸ FLOP of compute (estimation method: hardware). The model has 27,000,000 parameters. Training ran on 1 NVIDIA V100.
Access: Unreleased. Its weights are not openly released. Epoch AI rates the confidence of this record as likely.
Full record
- Organization
- Zhejiang University (ZJU), Microsoft Research Asia
- Country of organization
- China
- Domain
- Speech
- Task
- Text-to-speech (TTS)
- Training compute
- 2.3×10¹⁸ FLOP
- Compute estimation method
- Hardware
- Parameters
- 27,000,000
- Training hardware
- NVIDIA V100
- Chips used
- 1
- Chip-hours
- 17.02
- Training power draw
- 330 W
- Model accessibility
- Unreleased
- Open weights
- No
- Epoch confidence
- Likely
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