airflow.providers.common.ai.example_dags.example_llm¶

Example DAGs demonstrating LLMOperator and @task.llm usage.

Classes¶

`Entities`	Named entities extracted from a text.
`LogNotifier`	BaseNotifier class for sending notifications.

Functions¶

`example_llm_operator`()
`example_llm_operator_structured`()
`example_llm_operator_agent_params`()
`example_llm_decorator`()
`example_llm_decorator_structured`()
`example_llm_operator_usage_limits`()
`example_llm_operator_templated_usage_limits`()
`example_llm_operator_approval`()

Module Contents¶

class airflow.providers.common.ai.example_dags.example_llm.Entities(/, **data)[source]¶

Bases: pydantic.BaseModel

Named entities extracted from a text.

names: list[str][source]¶

locations: list[str][source]¶

airflow.providers.common.ai.example_dags.example_llm.example_llm_operator()[source]¶

airflow.providers.common.ai.example_dags.example_llm.example_llm_operator_structured()[source]¶

airflow.providers.common.ai.example_dags.example_llm.example_llm_operator_agent_params()[source]¶

airflow.providers.common.ai.example_dags.example_llm.example_llm_decorator()[source]¶

airflow.providers.common.ai.example_dags.example_llm.example_llm_decorator_structured()[source]¶

airflow.providers.common.ai.example_dags.example_llm.example_llm_operator_usage_limits()[source]¶

airflow.providers.common.ai.example_dags.example_llm.example_llm_operator_templated_usage_limits()[source]¶

class airflow.providers.common.ai.example_dags.example_llm.LogNotifier(message)[source]¶

Bases: airflow.providers.common.compat.notifier.BaseNotifier

BaseNotifier class for sending notifications.

It can be used asynchronously (preferred) if async_notify`is implemented and/or synchronously if `notify is implemented.

Currently, the Dag/Task state change callbacks run on the Dag Processor and only support sync usage.

Usage::

# Asynchronous usage await Notifier(context)

# Synchronous usage notifier = Notifier() notifier(context)

template_fields = ('message',)[source]¶

message[source]¶

notify(context)[source]¶

Send a notification (sync).

Implementing this is a requirement for running this notifier in the Dag processor, which is where the on_success_callback and on_failure_callback run.

Parameters:: context – The airflow context

airflow.providers.common.ai.example_dags.example_llm.example_llm_operator_approval()[source]¶