Google Cloud is giving a shot in the arm to AI Hub, a one-stop shop for novice and expert AI practitioners. Among upgrades introduced today are a new home page to access public and private files and the ability to share notebooks or AI models the same quick and simple way you can share a Google Doc.
AI Hub launched in beta in April at the Cloud Next developer conference and was first introduced in alpha a year ago following the introduction of Kubeflow Pipelines, a way to share and copy popular workflows with Kubernetes containers.
AI Hub shares popular AI models like Google’s BERT, ResNet, and Inception alongside training materials and Kubeflow workflows.
Earlier this year, Google Cloud also introduced AI Lab for data scientists to collaborate and create AI models together.
Other updates introduced today include a new taxonomy for labeling documents related to AI projects, the ability to favorite assets such as training materials or popular models, and new content from partners like Nvidia.
More than 140 Jupyter notebooks can be deployed as AI Hub notebooks for things like 3D style transfer, image classification, recommendation systems, and time-series forecasting.