deeppavlov.core.commands

Basic training and inference functions.

deeppavlov.core.commands.infer.build_model(config: Union[str, pathlib.Path, dict], mode: str = 'infer', load_trained: bool = False, download: bool = False, serialized: Optional[bytes] = None) → deeppavlov.core.common.chainer.Chainer[source]

Build and return the model described in corresponding configuration file.

deeppavlov.core.commands.infer.interact_model(config: Union[str, pathlib.Path, dict]) → None[source]

Start interaction with the model described in corresponding configuration file.

deeppavlov.core.commands.infer.predict_on_stream(config: Union[str, pathlib.Path, dict], batch_size: int = 1, file_path: Optional[str] = None) → None[source]

Make a prediction with the component described in corresponding configuration file.

deeppavlov.core.commands.train.get_iterator_from_config(config: dict, data: dict)[source]

Create iterator (from config) for specified data.

deeppavlov.core.commands.train.read_data_by_config(config: dict)[source]

Read data by dataset_reader from specified config.

deeppavlov.core.commands.train.train_evaluate_model_from_config(config: Union[str, pathlib.Path, dict], iterator: Union[deeppavlov.core.data.data_learning_iterator.DataLearningIterator, deeppavlov.core.data.data_fitting_iterator.DataFittingIterator] = None, *, to_train: bool = True, evaluation_targets: Optional[Iterable[str]] = None, to_validate: Optional[bool] = None, download: bool = False, start_epoch_num: Optional[int] = None, recursive: bool = False) → Dict[str, Dict[str, float]][source]

Make training and evaluation of the model described in corresponding configuration file.