-
Notifications
You must be signed in to change notification settings - Fork 320
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[feat] Add activeloop deeplake plugin #2594
[feat] Add activeloop deeplake plugin #2594
Conversation
Hi @drahnreb Thanks for the contribution, it's exciting. 🙌 |
@drahnreb just removed the duplicated workflow. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@drahnreb, awesome job! Thank you for taking care of this.
Left a couple of comments/questions.
""" | ||
AIM_NAME = 'deeplake.dataset' | ||
|
||
def __init__(self, dataset: deeplake.Dataset, auto_commit: bool = True, auto_save_view: bool = True): |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thinking loud here; what are the reasonable defaults here? Could the automatic commit/save have some performance implications unexpected for the user?
CC @gorarakelyan
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes, it could for saving views. Committing should be fairly fast.
Not doing any, you won‘t have actual traceability for the run. Attaching the dataset meta information is almost useless.
Defaulting both to False
will disable commit/save with warnings and leave more flexibility to take care of this before runs (as it should be done in the first place).
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
resolved with 78bdb80
@@ -4,6 +4,11 @@ | |||
|
|||
@CustomObject.alias('hub.dataset') | |||
class HubDataset(CustomObject): | |||
from aim.utils.deprecation import deprecation_warning | |||
deprecation_warning(remove_version='3.17', msg='Using HubDataset is deprecated!\ |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Is this really a deprecation? What if one continues to use the older version of deeplake?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
That‘s a valid question. The name change happened only roughly half a year ago. But seemingly without downwards compatibility nor official deprecation warning. I should have split the deprecation from the deeplake implementation. Also, since both implementations work in parallel.
Renamed the PR, since the PR naming check failed. |
5f56826
to
f68b398
Compare
Rebased and removed the deprecation from this PR. Will add another dedicated PR for it to discuss. The hub package is stale and unmaintained, up to you to decide how long to keep it supported in aim. |
f68b398
to
78bdb80
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Approving PR as all the comments are addressed.
Will add CHANGELOG.md entry and merge changes
# Conflicts: # CHANGELOG.md
hub
is nowdeeplake
I added a
DeeplakeDataset
that covers the oldHubDataset
behavior.Including:
HubDataset