Image Management
"Image Management" is available under the training and inference business scope. Click "Image Management" to display all framework images by default. Users can view public, group, and personal images. Users can view information such as available and used images. Image list information includes: framework name, image name, sharing attribute, image size, usage count, last used time, image tag, uploader, creation time, remarks...
Container Image List
It is recommended to use AIStation platform pre-installed container images for jobs, including PyTorch, TensorFlow, MXNet, and other AI environment images. You can also upload your own image tar packages to the user directory and use the image loading feature to import them. Note: Most images are based on Ubuntu; check tags or remarks for other versions. Image type...
Image Build
Create via Dockerfile or import external images to pull non-INSPIR official images. For example, use `docker pull nvcr.io/nvidia/pytorch:21.07-py3` to create a development environment with this image and log in to the Shell terminal. This document provides instructions for installing Jupyter and SS... in an Ubuntu 20.04 environment.