EVA-01
EVA-01 is an AI model developed by Beijing Academy of Artificial Intelligence / BAAI, Huazhong University of Science and Technology, Zhejiang University (ZJU) and Beijing Institute of Technology (China), first published in November 2022. It works in the vision domain, on tasks such as image classification, object detection, semantic segmentation and video classification.
Training it took an estimated 1.5×10²² FLOP of compute (estimation method: hardware). The model has 1,011,000,000 parameters. It was trained on roughly 7.6B datapoints. Training ran on 128 NVIDIA A100 SXM4 40 GB for about 348 hours. The compute alone is estimated at $29K in 2023 dollars.
Access: Open weights (unrestricted). Its weights are openly available. The reference paper has 982 citations. Epoch AI rates the confidence of this record as confident.
- Organization
- Beijing Academy of Artificial Intelligence / BAAI, Huazhong University of Science and Technology, Zhejiang University (ZJU), Beijing Institute of Technology
- Country of organization
- China
- Domain
- Vision
- Task
- Image classification, Object detection, Semantic segmentation, Video classification
- Training compute
- 1.5×10²² FLOP
- Compute estimation method
- Hardware
- Parameters
- 1,011,000,000
- Dataset size
- 7.6B
- Training hardware
- NVIDIA A100 SXM4 40 GB
- Chips used
- 128
- Training time
- 348 h
- Chip-hours
- 44.5K
- Training power draw
- 102.4 kW
- Training cost (2023 USD)
- $29K
- Numerical format
- FP16
- Model accessibility
- Open weights (unrestricted)
- Open weights
- Yes
- Citations
- 982
- Epoch confidence
- Confident