You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When an Experiment is stopped by the user, we may want to indicate that the partially-run experiment exists and can be resumed.
This functionality depends on the ability to detect this, which I'm not currently sure is possible. It seems that partial experiments show up like queued ones, but that may just be due to how the Webview is reading them.
For the UI, one suggested implementation to change the current "Run Experiment" button's text to "Resume Experiment" when a cancelled partial run exists.
It's also been mentioned that we may also want the ability to remove this partial experiment and restart from the last complete one, but that's will require a more complex solution that builds off a simpler solution to this ticket, so it may warrant its own Issue.
The text was updated successfully, but these errors were encountered:
When this issue was made, I was under the impression that an experiment stopped at a checkpoint will be written at the current checkpoint. If I currently understand correctly, that interpretation was incorrect and stopping an experiment will not change anything that would change the output of dvc exp show.
If my new understanding is correct, this issue is actually not applicable, since there would be no partial experiment to resume.
Related to #230
When an Experiment is stopped by the user, we may want to indicate that the partially-run experiment exists and can be resumed.
This functionality depends on the ability to detect this, which I'm not currently sure is possible. It seems that partial experiments show up like queued ones, but that may just be due to how the Webview is reading them.
For the UI, one suggested implementation to change the current "Run Experiment" button's text to "Resume Experiment" when a cancelled partial run exists.
It's also been mentioned that we may also want the ability to remove this partial experiment and restart from the last complete one, but that's will require a more complex solution that builds off a simpler solution to this ticket, so it may warrant its own Issue.
The text was updated successfully, but these errors were encountered: