OnRunMetric

Class wandb.automations.OnRunMetric

A run metric satisfies a user-defined condition.

*,
event_type: Literal[RUN_METRIC_THRESHOLD, RUN_METRIC_CHANGE, RUN_METRIC_ZSCORE],
scope: wandb.automations.scopes.ProjectScope,
filter: Annotated[wandb.automations.events.RunMetricFilter, Json, BeforeValidator(func=<function ensure_json>, json_schema_input_type=PydanticUndefined), PlainSerializer(func=<function to_json>, return_type=PydanticUndefined, when_used='always')]

Args

Literal[RUN_METRIC_THRESHOLD, RUN_METRIC_CHANGE, RUN_METRIC_ZSCORE]

wandb.automations.scopes.ProjectScope

The scope of the event: must be a project.

Annotated[wandb.automations.events.RunMetricFilter, Json, BeforeValidator(func= None, json_schema_input_type=PydanticUndefined), PlainSerializer(func= None, return_type=PydanticUndefined, when_used='always')]

Run and/or metric condition(s) that must be satisfied for this event to trigger.

Examples

Define an event that triggers for any run in project “my-project” when the average of the last 5 values of metric “my-metric” exceeds 123.45:

from wandb import Api
from wandb.automations import OnRunMetric, RunEvent

api = Api()
project = api.project(name="my-project")

event = OnRunMetric(
    scope=project,
    filter=RunEvent.metric("my-metric").avg(5).gt(123.45),
)

W&B Models

W&B Weave

Serverless Inference

Serverless Training

Class wandb.automations.OnRunMetric

Args

Examples

​Class wandb.automations.OnRunMetric

​Args

​Examples

Class wandb.automations.OnRunMetric

Args

Examples