Dave Berenbaum
Dave Berenbaum
Now that dvclive is decoupled from dvc, it should be pretty trivial to log data in formats expected by other ML tracking tools like mlflow. It would be great to...
Values from previous steps are kept in the following steps. Take the following example: ``` import dvclive live = dvclive.Live() while (epoch := live.get_step()) < 10: live.log("epoch", epoch) if not...
Seems odd to do this. If there are multiple things to log per iteration, you dump multiple times? - Seems better to dump only on `next_step()` - Could also make...
> It seems odd to pass Live(report="html") in no-step scenarios and have that kwarg do nothing. Not sure if it should be part of this PR, but currently, I can...
Currently, dvclive without dvc produces just a tsv output tracking metrics progress. This doesn't seem like enough to make it worthwhile for users. There are at least a few drawbacks...
Users may want to save metrics every epoch but only save models every few epochs. Can we support this in dvclive?
Right now, it's possible to add a new version of any artifact as long as there is no existing artifact version at that Git commit. For example, in https://github.com/iterative/example-gto: ```console...
Fixes #8188 Docs PR: https://github.com/iterative/dvc.org/pull/3972
Docstrings are needed for complete documentation of dvcfs. This is part of https://github.com/iterative/dvc.org/issues/3927.
Would it make the most sense then to only show files as not in cache (or whatever terminology that we end up using) if they are not present in either...