[data] fix bug in docs enable_true_multi_threading#60515
Conversation
Signed-off-by: iamjustinhsu <jhsu@anyscale.com>
There was a problem hiding this comment.
Code Review
The pull request includes a crucial documentation fix for the enable_true_multi_threading parameter in the ActorPoolStrategy.__init__ method. This correction clarifies that setting enable_true_multi_threading=False limits actor tasks to one per actor, which aligns with the parameter's default behavior and significantly improves the accuracy of the API documentation.
python/ray/data/_internal/compute.py
Outdated
| enable_true_multi_threading: If enable_true_multi_threading=False, no more than 1 actor task | ||
| runs per actor. Otherwise, respects the `max_concurrency` argument. |
There was a problem hiding this comment.
I think "runs" is ambiguous here, and I'm worried this docstring might confuse users if users set enable_true_multi_threading=False but Ray Data launches multiple active tasks per actor (which is the default behavior). Is there a way we could remove this ambiguity?
I don't have a good alternative off the top of my head.
There was a problem hiding this comment.
yea it's kinda of an implementation detail, but as far as the user is concerned, their UDF will "behave" as if one is running at a time per actor, even if that's not what's going on under the hood. I left a larger explanation in the source code and (docs) that goes into more detail for advanced users. What are ur thoughts on just pointing them to that direction as well? I updated my PR to reflect that without the link
Signed-off-by: iamjustinhsu <jhsu@anyscale.com>
Co-authored-by: Balaji Veeramani <bveeramani@berkeley.edu> Signed-off-by: iamjustinhsu <140442892+iamjustinhsu@users.noreply.github.com>
## Description This should say `False` ## Related issues ## Additional information --------- Signed-off-by: iamjustinhsu <jhsu@anyscale.com> Signed-off-by: iamjustinhsu <140442892+iamjustinhsu@users.noreply.github.com> Co-authored-by: Balaji Veeramani <bveeramani@berkeley.edu> Signed-off-by: jinbum-kim <jinbum9958@gmail.com>
## Description This should say `False` ## Related issues ## Additional information --------- Signed-off-by: iamjustinhsu <jhsu@anyscale.com> Signed-off-by: iamjustinhsu <140442892+iamjustinhsu@users.noreply.github.com> Co-authored-by: Balaji Veeramani <bveeramani@berkeley.edu>
## Description This should say `False` ## Related issues ## Additional information --------- Signed-off-by: iamjustinhsu <jhsu@anyscale.com> Signed-off-by: iamjustinhsu <140442892+iamjustinhsu@users.noreply.github.com> Co-authored-by: Balaji Veeramani <bveeramani@berkeley.edu> Signed-off-by: 400Ping <jiekaichang@apache.org>
## Description This should say `False` ## Related issues ## Additional information --------- Signed-off-by: iamjustinhsu <jhsu@anyscale.com> Signed-off-by: iamjustinhsu <140442892+iamjustinhsu@users.noreply.github.com> Co-authored-by: Balaji Veeramani <bveeramani@berkeley.edu> Signed-off-by: Adel Nour <ans9868@nyu.edu>
## Description This should say `False` ## Related issues ## Additional information --------- Signed-off-by: iamjustinhsu <jhsu@anyscale.com> Signed-off-by: iamjustinhsu <140442892+iamjustinhsu@users.noreply.github.com> Co-authored-by: Balaji Veeramani <bveeramani@berkeley.edu> Signed-off-by: peterxcli <peterxcli@gmail.com>
## Description This should say `False` ## Related issues ## Additional information --------- Signed-off-by: iamjustinhsu <jhsu@anyscale.com> Signed-off-by: iamjustinhsu <140442892+iamjustinhsu@users.noreply.github.com> Co-authored-by: Balaji Veeramani <bveeramani@berkeley.edu> Signed-off-by: peterxcli <peterxcli@gmail.com>
Description
This should say
FalseRelated issues
Additional information