Public Resources Overview
Provides some common public datasets, pre-trained models, and public tutorials
HyperAI provides three types of public resources:
- Public Datasets
- Public Tutorials
- Public Models
Public Datasets
Training machine learning models typically requires a large amount of labeled data. Currently, there are numerous public datasets available on the market (such as PASCAL_VOC) for machine learning practitioners to use directly. Some of these datasets are quite large in scale. To facilitate users to directly use them on the HyperAI platform (rather than downloading and uploading them themselves), HyperAI periodically provides these common datasets to HyperAI users in the form of "Public Datasets".
:::note When you bind data from public datasets to a container in read-only mode, the space used by the dataset will not count against your personal storage quota :::
When "Creating a New Container", you can bind public datasets to the container:
Public Tutorials
Public tutorials are mainly some tutorials available on the market to help users familiarize themselves with using HyperAI. Through "cloning", users can directly run the tutorials, making it convenient for them to learn the relevant content.
Public Models
Similar to public datasets, there are numerous pre-trained models available on the market that can be used for transfer learning. These pre-trained models are typically trained using large-scale datasets and substantial computational resources. Similarly, these pre-trained models can also be bound to containers when creating new containers.
Permissions
Public resources do not support batch downloading. Users must access the content through data binding.