cyto_dl.api.cyto_dl_model.cyto_dl_base_model module

class cyto_dl.api.cyto_dl_model.cyto_dl_base_model.CytoDLBaseModel(cfg: DictConfig)

Bases: ABC

A CytoDLBaseModel is used to configure, train, and run predictions on a cyto-dl model.

Not intended for direct use by clients.

Please see the classmethod factory methods instead.

classmethod from_default_config(spatial_dims: int)

Returns a model from the default config.

Parameters:

spatial_dims – dimensions for the model (e.g. 2)

classmethod from_existing_config(config_filepath: Path)

Returns a model from an existing config.

Parameters:

config_filepath – path to a .yaml config file that will be used as the basis

for this CytoDLBaseModel (must be generated by the CytoDLBaseModel subclass that wants to use it).

get_config() → DictConfig

get_experiment_name() → str

predict(manifest_path: str | Path, output_dir: str | Path, checkpoint: Path) → None

save_config(path: Path) → None

set_experiment_name(name: str) → None

train(max_epochs: int, manifest_path: str | Path, output_dir: str | Path, checkpoint: Path | None = None) → None