gpytorchwrapper.src.models.model_load

Functions

load_model(config, model_dump, train_x, train_y)

Load a model from a config file and the dumped model.
gpytorchwrapper.src.models.model_load.load_model(config: Config, model_dump: dict, train_x: Tensor, train_y: Tensor) → tuple[GP, Likelihood][source]

Load a model from a config file and the dumped model. The model and likelihood objects are set to training mode with double precision.

Parameters:

  • config (Config) – Object specifying the configuration used for training.

  • model_dump (dict) – The unpickled dumped model file

  • train_x (Tensor) – The input tensor used for training

  • train_y (Tensor) – The output tensor used for training

Returns:

return – The loaded GP and likelihood objects

Return type:

tuple of GP and Likelihood