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
{{ message }}
This repository has been archived by the owner on Sep 18, 2024. It is now read-only.
Short summary about the issue/question:
When the type of metrics is a dict, there is no intermediate result. The final result use "default" key to show the result, and what about the intermediate result? Shouldn't these two be consistent?
Brief what process you are following:
Start an experiment and use the "dict" type metrics. For example:
result =
{
"default": 60.65,
"loss": 1.138
}
Then, we can see the point of the final result, can not see the intermediate result, like these:
How to reproduce it:
Start an experiment and use the "dict" type metrics.
nni Environment:
nni version: 0.5.2
nni mode(local|pai|remote): pai
OS: Ubuntu
python version: 3.5.2
is conda or virtualenv used?: no
is running in docker?: yes
Anything else we need to know:
The text was updated successfully, but these errors were encountered:
PurityFan
changed the title
no intermediate result when our metrics is a dictionary
no intermediate result when metrics is a dictionary
Mar 10, 2019
Short summary about the issue/question:
When the type of metrics is a dict, there is no intermediate result. The final result use "default" key to show the result, and what about the intermediate result? Shouldn't these two be consistent?
Brief what process you are following:
Start an experiment and use the "dict" type metrics. For example:
result =
{
"default": 60.65,
"loss": 1.138
}
Then, we can see the point of the final result, can not see the intermediate result, like these:
How to reproduce it:
Start an experiment and use the "dict" type metrics.
nni Environment:
Anything else we need to know:
The text was updated successfully, but these errors were encountered: