Dependencies

When you are defining your agents, we've made it easy to inject essential dependencies like the a path to the underlying asset, frame iterators, etc. Similarly, you can also inject your own custom dependencies, if you need to.

The way you inject dependencies is by type-annotating your agent function variables with the Depends class.

from typing_extensions import Annotated

from encord.core.dependencies import Depends
# or from fastapi import Depends # if you are building a fastapi app

from encord.{module}.dependencies import dep_single_frame

def my_agent(frame: Annotated[np.ndarray, Depends(dep_single_frame)]):
    # the frame will be available here.

The {module} depends on which type of agent you're building. Please see the references section for more details on available agents.

Custom dependencies

Custom dependencies are also easy. Just define how to load them in a function and depend on that function. The function it self can also depend on other dependencies.

Here is an example:

def my_custom_dependency(label_row: LabelRowV2) -> dict:
    # e.g., look up additional data in own db
    return db.query("whatever")

@runner.stage(stage="<my_stage_name>")
def by_custom_data(
    custom_data: Annotated[dict, Depends(my_custom_dependency)]
) -> str:
    # `custom_data` automatically injected here.
    # ... do your thing
    # then, return name of task pathway.