Gear The working directory
every time "implement" After starting to run, its runtime environment will bind its directory /hyperai/home, As a container "working directory" . The contents of this directory will be in "implement" Saved after closing. And it can be done through the "working directory" Tab View.
Display of key documents
"implement" Some key files in will be displayed on the container's page. If there are multiple items in a container "implement" So the latest "implement" The files will be displayed.
Display default README.md
The container will automatically search after being closed /hyperai/home/README.md file. And display its content on the container's page. This document needs to follow markdown grammar.
Display default Jupyter file
As shown above README.md The logic is similar. The container will automatically search after being closed /hyperai/home Under the directory .ipynb file. And display its content on the container's page.
When the container is closed. The system will follow "index.ipynb", "main.ipynb", "default.ipynb" The order is in /hyperai/home Search for files in the root directory. The first file found in order will be displayed on the container's page. If none of the above files are found, randomly obtain one from the root directory .ipynb Display documents.
See "working directory" content
If you need to view "implement" The result. There are two ways to view it:
- For those still running "Jupyter working space" , You can directly view or download it in the editor. It can also be done at "Container page" of "working directory" Tab View, however** "working directory" The content of the tab is periodically extracted from the running "implement" In sync**of. Therefore, in operation "implement" his "working directory" The file content in the tab may not be the latest data
- about "Python script" Only in "working directory" Tab View. If "implement" Still running, please note that its content may not be the latest data
File Preview
The container's "working directory" Some files in the support preview. The supported file types include .txt .py .ipynb .csv .md And multiple types of images.
Jupyter Preview of files:
Python File Preview:
Download file
stay "implement" After closing, "working directory" The data in the tab is /hyperai/home The final data content in the directory, and "data warehouse " equally. You can browse through the page. Preview and download operations. As shown below, click "Download the current file" You can take the whole "working directory" Download folder. Of course, you can also click the button on the right side of the file to download a single file.
Create a working directory as a data warehouse version
Currently, container support will "working directory" Or one of the folders can be created as a new version of the data warehouse. This way, we can "working directory" Manage the models generated or cleaned data separately. Convenient for future use.
Bind the working directory as data
Already closed "implement" The working directory can be bound to a new one "implement" in: