chronos_model

Chronos foundation model implementation for time series forecasting.

ChronosModel

chronos_model.ChronosModel · inherits BaseModel

Chronos foundation model wrapper for time series forecasting.

This class provides a unified interface for the Amazon Chronos model, which is a large language model specifically designed for time series forecasting.

Attributes: * model_size: Size of the Chronos model ('tiny', 'mini', 'small', 'base', 'large') * context_length: Number of past time steps used as context * num_samples: Number of predictive samples to generate

__init__(self, params, settings)

Initialize the Chronos model wrapper.

Parameter	Type	Default	Description
params	Dict[str, Any]	-	Model parameters dictionary
settings	Dict[str, Any]	-	Settings dictionary containing device, python_version, etc.

train(self, y_context, y_target, timestamps_context, timestamps_target, **kwargs) -> ChronosModel

Initialize the Chronos model (no training required for foundation models).

Parameters:

Parameter	Type	Default	Description
y_context	np.ndarray	-	Past target values (not used for training, for compatibility)
y_target	np.ndarray	-	Future target values (not used for training, for compatibility)
timestamps_context	np.ndarray	-	Timestamps for y_context (not used)
timestamps_target	np.ndarray	-	Timestamps for y_target (not used)
**kwargs		-	Additional keyword arguments

Returns: Self (The model instance).

predict(self, y_context, timestamps_context, timestamps_target, **kwargs) -> np.ndarray

Make predictions using the trained Chronos model.

Parameters:

Parameter	Type	Default	Description
y_context	np.ndarray	-	Recent target values for context
timestamps_context	np.ndarray	-	Timestamps for context data
timestamps_target	np.ndarray	-	Timestamps for target data
**kwargs		-	Additional keyword arguments

Returns: np.ndarray (Model predictions).

Raises: * ValueError: If model is not fitted or required data is missing.