ertk.train.ExperimentConfig

class ertk.train.ExperimentConfig(name: str = 'default', data: ~ertk.dataset.dataset.DataLoadConfig = '???', model: ~ertk.train.ModelConfig = '???', eval: ~ertk.train.EvalConfig | None = None, evals: ~typing.Dict[str, ~ertk.train.EvalConfig] = <factory>, training: ~ertk.train.TrainConfig = <factory>, results: str = '', metrics: ~typing.List[str] = <factory>)

Bases: ERTKConfig

Class to hold experiment configuration.

__init__(name: str = 'default', data: ~ertk.dataset.dataset.DataLoadConfig = '???', model: ~ertk.train.ModelConfig = '???', eval: ~ertk.train.EvalConfig | None = None, evals: ~typing.Dict[str, ~ertk.train.EvalConfig] = <factory>, training: ~ertk.train.TrainConfig = <factory>, results: str = '', metrics: ~typing.List[str] = <factory>) → None

Methods

__init__([name, data, model, eval, evals, ...])
from_file(path)	Create config from YAML file and optionlly override some values.

Inherited Methods

default()	Create default config.
from_config(config)	Create config object from any compatible config.
merge_with_args(args)	Merge config with command-line arguments.
merge_with_config(config)	Merge other config into this config.
to_dictconfig()	Convert config to DictConfig.
to_file(path)	Write config to YAML file.
to_string()	Generate YAML string representation of config.

Attributes

data	Data loading configuration.
eval	Evaluation configuration.
model	Model configuration.
name	Experiment name.
results	Path to output results.
evals	Dictionary of evaluation configurations.
training	Training configuration.
metrics

data: DataLoadConfig = '???'

Data loading configuration.

eval: EvalConfig | None = None

Evaluation configuration.

evals: Dict[str, EvalConfig]

Dictionary of evaluation configurations.

classmethod from_file(path: PathLike | str) → T

Create config from YAML file and optionlly override some values.

Parameters:

path: os.Pathlike or str

The path to YAML file containing config.

override: collection of str, optional

Argument overrides in the form of key=value pairs.

Returns:

ERKConfig

The resulting config.

model: ModelConfig = '???'

Model configuration.

name: str = 'default'

Experiment name.

results: str = ''

Path to output results.

training: TrainConfig

Training configuration.