AI models
Emu1 (BAAI)
Training compute
2.7×10²¹ FLOP
Parameters
14B
Published
Jul 11, 2023
Emu1 (BAAI) is an AI model developed by Beijing Academy of Artificial Intelligence / BAAI, Tsinghua University and Peking University (China), first published in July 2023. It works in the vision, multimodal and language domain, on tasks such as image generation, text autocompletion, text-to-image and 3 more.
Training it took an estimated 2.7×10²¹ FLOP of compute (estimation method: hardware). The model has 14,000,000,000 parameters. Training ran on 128 NVIDIA A100 SXM4 80 GB for about 48 hours.
Access: Open weights (non-commercial). Its weights are openly available. It is built on top of Stable Diffusion (LDM-KL-8-G),EVA-01,LLaMA-13B. Epoch AI rates the confidence of this record as confident.
Full record
- Organization
- Beijing Academy of Artificial Intelligence / BAAI, Tsinghua University, Peking University
- Country of organization
- China
- Domain
- Vision, Multimodal, Language
- Task
- Image generation, Text autocompletion, Text-to-image, Visual question answering, Image captioning, Language modeling/generation
- Training compute
- 2.7×10²¹ FLOP
- Compute estimation method
- Hardware
- Parameters
- 14,000,000,000
- Training hardware
- NVIDIA A100 SXM4 80 GB
- Chips used
- 128
- Training time
- 48 h
- Chip-hours
- 6.1K
- Training power draw
- 101.8 kW
- Model accessibility
- Open weights (non-commercial)
- Open weights
- Yes
- Base model
- Stable Diffusion (LDM-KL-8-G), EVA-01, LLaMA-13B
- Epoch confidence
- Confident
More from Beijing Academy of Artificial Intelligence / BAAI,Tsinghua University,Peking University
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