Share files between jobs on a volume

How it works

1. 1Volume(name=, mount_path=) in a decorator mounts it; the same name from different functions is the same folder.
2. 2Storage in the dashboard lists volumes with a file browser, previews, upload and download, and where each one is mounted.
3. 3levers9 ls, cp, mv, rm work on VOLUME/path; cp uploads a folder in one go.
4. 4Datasets, notebooks and pipeline artifacts all live on volumes, so any of them can be read by any function.

The code

from levers9 import Volume, job

DATA = Volume(name="datasets", mount_path="/data")

@job(volumes=[DATA])
def count():
    import os
    return len(os.listdir("/data/uploads"))

levers9 volume create datasets
levers9 cp ./recordings levers9://datasets/uploads
levers9 ls datasets/uploads