Log scores and summary information
Use the log_configs()
method to log summary information:
from neptune_scale import Run
if __name__ == "__main__":
run = Run(experiment_name=...)
run.log_configs(
{
"summary/scores/f1": 0.89,
"summary/epoch_average_seconds": 48.2,
"summary/interrupted" = False,
},
)
- Use descriptive namespaces and attributes to clearly label the summary metadata.
- You can assign any the following types to an attribute:
String
,Integer
,Float
,Datetime
, orBoolean
. - To log a complex dictionary, see Logging nested dictionaries.
Find your logged summary information in the Attributes section of the run.
tip
You can display the summary metadata in the runs table and custom dashboards.