Core

Core Building blocks for AutoML Tasks

AutoModel

Base model that defines hyperparameter search methods and initializes Ray. All other tasks are implementation of AutoModel.

Parameters:

Name	Type	Description	Default
datamodule	flash.DataModule	DataModule from Flash or PyTorch Lightning	required
max_epochs	[int]	Maximum number of epochs for which model will train	required
max_steps	Optional[int]	Maximum number of steps for each epoch. Defaults None.	required
optimization_metric	str	Value on which hyperparameter search will run.	required
n_trials	int	Number of trials for HPO	required
suggested_conf	Dict	Any extra suggested configuration	required
timeout	int	HPO will stop after timeout	required
prune	bool	Whether to stop unpromising training.	required
tune_confs	Dict	raytune configurations. See more at Ray docs.	required
best_trial	bool	If true model will be loaded with best weights from HPO otherwise	required

hp_tune(self, name=None, ray_config=None, trainer_config=None, mode=None, gpu=0, cpu=None, resume=False)

Search Hyperparameter and builds model with the best params

    automodel = AutoClassifier(data)  # implements `AutoModel`
    automodel.hp_tune(name="gflow-example", gpu=1)

Parameters:

Name	Type	Description	Default
name	Optional[str]	name of the experiment.	None
ray_config	Optional[dict]	configuration passed to ray.tune.run(...)	None
trainer_config	Optional[dict]	configuration passed to pl.trainer.fit(...)	None
mode	Optional[str]	Whether to maximize or mimimize the optimization_metric.	None
gpu	Optional[float]	Amount of GPU resource per trial.	0
cpu	Optional[float]	CPU cores per trial	None
resume	bool	Whether to resume the training or not.	False

---

AutoClassifier

Implements AutoModel for classification tasks.

build_model(self, config)

Every Task implementing AutoClassifier has to implement a build model method that can build torch.nn.Module from dictionary config and return the model.

---

Last update: August 26, 2021