def load_data_c(self, config):
"""
Load the data for your ML pipeline. Will be fed into `train`.
:param config: object constructed with all the relevant arguments for `load_data`
:type config: ```Union[dict, Config, Any]```
:return: a call to .load_data with the config as params
:rtype: ```Union[Tuple[tf.data.Dataset, tf.data.Dataset], Tuple[np.ndarray, np.ndarray]]```
"""
return self.load_data(**to_d(config))
def train_c(self, config):
"""
Run the training loop for your ML pipeline.
:param config: object constructed with all the relevant arguments for `train`
:type config: ```Union[dict, Config, Any]```
:return: a call to .train with the config as params
:rtype: ```train```
"""
return self.train(**to_d(config))
def load_model_c(self, config):
"""
Load the model.
Takes a model object, or a pipeline that downloads & configures before returning a model object.
:param config: object constructed with all the relevant arguments for `load_model`
:type config: ```Union[dict, Config, Any]```
:return: a call to .load_model with the config as params
:rtype: ```load_model```
"""
return self.load_model(**to_d(config))
def test_to_d(self) -> None:
"""
Tests whether `to_d` creates the right dictionary
"""
self.assertDictEqual(to_d({}), {})
self.assertListEqual(*map(
sorted,
(
to_d(ml_params.utils).keys(),
(ml_params.utils.__all__ + [
"deepcopy",
"environ",
"getmembers",
"parse_to_argv_gen",
"partial",
"path",
"version_info",
"itemgetter",
]),
),
))
def test_properties(self) -> None:
"""
Tests whether `BaseTrainer` has the right properties
"""
self.assertListEqual(
sorted(to_d(BaseTrainer).keys()),
[
"data",
"load_data",
"load_data_c",
"load_model",
"load_model_c",
"model",
"train",
"train_c",
],
)