get_predictions_once()¶

This async method returns the model’s predictions just once at the moment it is called. It is used when the model is activated with continues_msg in False. In this manner, the server does not continually publish predictions to a topic; instead, it waits for the prediction retrieval service to be called.

Reference¶

Arguments¶

Arguments	Type	Default value	Description
as_dict	`bool`	`False`	If set to `true`, the predictions will be sent as a dictionary; otherwise, they will be returned as a list.
get_timestamp	`bool`	`False`	If set to `true`, a tuple with the predictions and the prediction timestamp will be returned; otherwise, only the predictions will be returned.

Return¶

if as_dict is True and get_timestamp is True:

Type	Description
`tuple`	(`dict`, `tuple`) Dictionary with prediccions and tuple with timestamp (sec, nanosec).

if as_dict is True and get_timestamp is False:

Type	Description
`dict`	Dictionary with prediccions.

if as_dict is False and get_timestamp is True:

Type	Description
`tuple`	(`list`, `tuple`) List with prediccions and tuple with timestamp (sec, nanosec).

if as_dict is False and get_timestamp is False:

Type	Description
`list`	List with prediccions.

Exceptions¶

None

See the complete list of cv exceptions.

Examples¶

...
self.predictor = await self.cv.enable_model(
        name='predictor_name',
        continues_msg=False,
        source='chest',
    )
...
last_predictions, timestamp = \
        await self.predictor.get_predictions_once(get_timestamp=True)
...