DRAFT : function to allow adding arbitrary files to the run object#1387
DRAFT : function to allow adding arbitrary files to the run object#1387SubhadityaMukherjee wants to merge 3 commits into
Conversation
LennartPurucker
left a comment
LennartPurucker
left a comment
There was a problem hiding this comment.
Heyho, coming back to this after a long time.
What are the use cases here exactly? What kind of files do we want users to upload?
I guess the website already supports this, so we could support this as well but I would try to avoid adding new code that is not used.
Also, the current way to support it seems very hacky. I recommend instead to add an official list-like parameter in the init (None/empty by default), and which one can then fill through such a function. And then _get_file_elements has code to check if this parameter is not empty and adds its content to the file elements.
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## develop #1387 +/- ##
===========================================
- Coverage 84.09% 83.71% -0.39%
===========================================
Files 38 38
Lines 5231 5255 +24
===========================================
Hits 4399 4399
- Misses 832 856 +24 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
PGijsbers
left a comment
PGijsbers
left a comment
There was a problem hiding this comment.
Some notes.
@LennartPurucker the use case would be sharing models with a run.
| ) | ||
| return description | ||
|
|
||
| def _check_file_size(self, file: Any, max_file_size_mb: int = 100) -> bool: |
There was a problem hiding this comment.
Have a look at https://docs.python.org/3/library/pathlib.html#pathlib.Path.stat :)
| return False | ||
| return True | ||
|
|
||
| def add_file_to_run(self, file: Any, name: str, max_file_size_mb: int = 100) -> None: |
There was a problem hiding this comment.
I agree with Lennart we should just have proper support if we add this, e.g., simply keep a files: list[Path] property and update the _get_file_elements function accordingly.
4 participants
Reference Issue
What does this PR implement/fix? Explain your changes.
This adds a function to the run class that allows adding an arbitrary file of a limited size to be uploaded along with the run. Something similar has been implemented in OpenML pytorch and I thought it would be nice to eventually have it as part of OpenML python as well.
How should this PR be tested?
Adding a random file and continuing the run as usual is enough. I will add tests and test it after the servers are back online.
Any other comments?
This is my first PR here so please be gentle 🙈