manager.trainer.get_last_model

manager.trainer.get_last_model(model_name, model_dir=None)

Get the latest trained model from the cache.

Parameters

Name	Type	Description	Default
model_name	str	Name of the model (e.g., ‘lgbm’, ‘xgb’).	required
model_dir	Optional[Union[str, Path]]	Directory where models are stored. If None, defaults to the library’s cache home.	None

Returns

Name	Type	Description
	int	A tuple (iteration, model_instance). If no model is found on disk
	Any	(cache directory missing, no matching files, or no parseable
	tuple[int, Any]	iteration numbers), returns (-1, None) — a legitimate
	tuple[int, Any]	“fresh install, no model yet” state.

Raises

Name	Type	Description
	OSError	If the latest model file exists on disk but cannot be deserialised (corrupt joblib, version mismatch, etc.).

Examples

import tempfile
from spotforecast2_safe.manager.trainer import get_last_model

# Empty directory — no model files yet.
with tempfile.TemporaryDirectory() as tmpdir:
    iteration, model = get_last_model("lgbm", model_dir=tmpdir)
    print(iteration, model)
    assert iteration == -1
    assert model is None

-1 None