Attribute

Python package: neptune-fetcher

Helper for specifying an attribute and picking a metric aggregation function.

When fetching experiments or runs, use the Attribute constructor to filter and sort the returned entries.

Parameters

Name	Type	Default	Description
name	str	-	An attribute name to match exactly.
aggregation	Literal["last", "min", "max", "average", "variance"], optional	None	Aggregation function to apply when specifying a metric of type FloatSeries. Defaults to "last", that is, the last logged value.
type	Literal["float", "int", "string", "bool", "datetime", "float_series", "string_set"], optional	None	Attribute type. Specify it to resolve ambiguity, in case some of the project's runs contain attributes that have the same name but are of a different type.

Examples

Select a metric and pick variance as the aggregation:

import neptune_fetcher.alpha as npt
from neptune_fetcher.alpha.filters import Attribute, Filter


val_loss_variance = Attribute(
    name="val/loss",
    aggregation="variance",
)

Construct a filter around the attribute with the Filter class, then pass it to a fetching or listing method:

tiny_val_loss_variance = Filter.lt(val_loss_variance, 0.01)
npt.fetch_experiments_table(experiments=tiny_val_loss_variance)