20220119_ESIPJanMtg_Rao.pdf (3.35 MB)
Enabling Community Contribution of Cloud-based Training for AI/ML in Earth and Space Science
presentationposted on 2022-01-13, 18:27 authored by Yuhan RaoYuhan Rao, Christopher Slocum
Interactive notebooks have been used for training and sharing AI/ML applications in ESS community. However, these efforts are usually side products related to a specific research project and developed by the researchers themselves. The notebook development process typically does not directly engage potential users, which may reduce the value and impact of the final notebooks. To enable community contribution and maximize the value of interactive notebooks for AI/ML training, we developed a cloud-based training framework including notebook template, contributing guides, and tools to assess the readability of the notebooks. This presentation was given during the 2022 ESIP January meeting held virtually in January 2022.