Training hyperparams

get(config_name, overriding_params=None)

Class for creating training hyper parameters dictionary, taking defaults from yaml files in src/super_gradients/recipes.

Parameters:

Name	Type	Description	Default
config_name		yaml config filename in recipes (for example coco2017_yolox).	required
overriding_params	Dict	Dict, dictionary like object containing entries to override in the recipe's training hyper parameters dictionary.	None

Source code in V3_2/src/super_gradients/training/training_hyperparams/training_hyperparams.py

def get(config_name, overriding_params: Dict = None) -> Dict:
    """
    Class for creating training hyper parameters dictionary, taking defaults from yaml
     files in src/super_gradients/recipes.

    :param config_name: yaml config filename in recipes (for example coco2017_yolox).
    :param overriding_params: Dict, dictionary like object containing entries to override in the recipe's training
     hyper parameters dictionary.
    """
    if overriding_params is None:
        overriding_params = dict()

    cfg = load_recipe(config_name=config_name)  # This loads the full recipe, not just training_hyperparams
    cfg = hydra.utils.instantiate(cfg)
    training_params = cfg.training_hyperparams

    training_params = override_default_params_without_nones(overriding_params, training_params)
    return training_params