Data Upload

Currently HyperAI supports the following methods for data upload:

1. Upload data in a running "Jupyter Workspace"
2. Use the "Upload" function in HyperAI's "Data Warehouse"
3. Use FileBrowser for data upload

The "Data Warehouse" upload supports both browser upload and bayes command-line tool upload. For small-scale data, uploading through "Jupyter Workspace" can meet the requirements. However, if you need more data management features (create, delete, modify, query, batch directory upload, resumable upload, etc.), you can use FileBrowser for data upload.

Upload Data via "Jupyter Workspace"

After launching the "Jupyter Workspace", you can upload individual files through the "Upload" button in the "File Browser" on the left side.

During the upload of large files (larger than 15MB), a progress bar will be displayed below:

Dragging and dropping files to the left "File Browser" will also trigger file upload.

:::caution Note Uploading files through "Jupyter Workspace" does not support "resumable upload". Any upload failure requires re-uploading the entire file. :::

"Upload" Function in HyperAI's "Data Warehouse"

:::note The upload function for "Datasets" is consistent with the "Upload" function for "Models". Here we use "Datasets" as an example. :::

Click the "+" button next to "Datasets" in the left menu bar, enter the dataset name and description to create an empty dataset.

After the dataset is created, you can click "Upload New Version" in the upper right corner of the dataset page.

The size limit here depends on the current user's remaining space, with a maximum support of "500 GB" file upload.

Multiple File Upload

In the upload interface, you can upload up to 500 files at once, supporting file multi-selection, file drag-and-drop, and other methods to select files for upload.

:::caution Note Uploaded files cannot handle file name conflicts. Therefore, when dragging and dropping folders for upload, if there are files with the same name, only one file will ultimately be saved. :::

Resumable Upload

Dataset upload supports "resumable upload". When an upload is interrupted due to network issues or accidents, if the uploaded file has not been modified and the target dataset has not changed, the data will continue from the previous upload.

:::caution Note Resumable upload is supported for a maximum period of 15 days. If the uploaded dataset is not completed within 15 days from the first upload, the data will be deleted. :::

Upload via bayes Command-line Tool

Detailed instructions are provided in Upload Data with bayes.

Use FileBrowser for Data Upload

FileBrowser is a web-based file manager that allows you to access datasets in HyperAI through a browser and perform operations such as creating, deleting, modifying, querying data, and batch directory upload.

Here is a video tutorial introducing how to use FileBrowser for data upload.

I don't see any Chinese text to translate in your message. You've only provided an HTML iframe code for embedding a Bilibili video player.

If you have Chinese text you'd like me to translate to English, please provide it and I'll be happy to help.