sakura.sakuraAE.sakuraAE.train_story

sakuraAE.train_story(story: list, resume=False, resume_dict=None)

Train the model on the given sets of tasks represented as a list of storylines.

Parameters:

story (list[dict[str, Any]]) – A list of dictionaries should contain the necessary information for training and/or testing, such as hyperparameters, task-specific configurations and checkpoint settings.
resume (bool, optional) – Whether to resume training from a previous state using the information provided in <resume_dict>, defaults to False
resume_dict (dict[str, Any], optional) – A dictionary containing the state information needed to resume training

Note

See also train(), test(), train_hybrid() and train_hybrid_fastload() for details on configurations of different tasks.