Local downloader w cache

[mads-oestergaard]

What does this PR do?

This PR adds a file format (local:) that users can prefix local paths with to get the files cached by lightning.

Motivation:
I would like to store data on a network drive that our on-prem compute nodes can mount and use, but I would also like to cache the chunks to not over-burden the network connection. With this PR, instead of doing this (which also works):

from lightning.data import StreamingDataset

dataset = StreamingDataset(input_dir="/data/shared-drive/some-data")

one can do

from lightning.data import StreamingDataset

dataset = StreamingDataset(input_dir="local:/data/shared-drive/some-data")

which will be the same, but with caching.

Before submitting

• Was this discussed/agreed via a GitHub issue? (not for typos and docs)
• Did you read the contributor guideline, Pull Request section?
• Did you make sure your PR does only one thing, instead of bundling different changes together?
• Did you make sure to update the documentation with your changes? (if necessary)
• Did you write any new necessary tests? (not for typos and docs)
• Did you verify new and existing tests pass locally with your changes?
• Did you list all the breaking changes introduced by this pull request?
• Did you update the CHANGELOG? (not for typos, docs, test updates, or minor internal changes/refactors)

PR review

Anyone in the community is welcome to review the PR.
Before you start reviewing, make sure you have read the review guidelines. In short, see the following bullet-list:

Reviewer checklist

• Is this pull request ready for review? (if not, please submit in draft mode)
• Check that all items from Before submitting are resolved
• Make sure the title is self-explanatory and the description concisely explains the PR
• Add labels and milestones (and optionally projects) to the PR so it can be classified

[mads-oestergaard]

@tchaton there you go :-)

[tchaton]

mads-oestergaard

You rock @mads-oestergaard.