.. _l-properties: Properties ========== .. contents:: :local: :depth: 1 Summary +++++++ +------------------------------------------------------------------------------------------------------------------+-------------------------------------+-------------------------------------------------------------------------+ | property | class parent | truncated documentation | +==================================================================================================================+=====================================+=========================================================================+ | :meth:`data_np ` | OrtDataLoader | Returns a tuple of the datasets in numpy. | +------------------------------------------------------------------------------------------------------------------+-------------------------------------+-------------------------------------------------------------------------+ | :meth:`data_ort ` | OrtDataLoader | Returns a tuple of the datasets in onnxruntime C_OrtValue. | +------------------------------------------------------------------------------------------------------------------+-------------------------------------+-------------------------------------------------------------------------+ | :meth:`needs_grad ` | OrtGradientForwardBackwardOptimizer | Returns the True if the gradient update needs to retain past gradients. | +------------------------------------------------------------------------------------------------------------------+-------------------------------------+-------------------------------------------------------------------------+ | :meth:`needs_grad ` | BaseLearningRate | Returns the True if the gradient update needs to retain past gradients. | +------------------------------------------------------------------------------------------------------------------+-------------------------------------+-------------------------------------------------------------------------+ | :meth:`needs_grad ` | LearningRateSGD | Returns the True if the gradient update needs to retain past gradients. | +------------------------------------------------------------------------------------------------------------------+-------------------------------------+-------------------------------------------------------------------------+ | :meth:`needs_grad ` | LearningRateSGDNesterov | Returns the True if the gradient update needs to retain past gradients. | +------------------------------------------------------------------------------------------------------------------+-------------------------------------+-------------------------------------------------------------------------+ | :meth:`saved_tensors ` | OrtGradientForwardBackwardFunction | Returns saved tensors during forward step. | +------------------------------------------------------------------------------------------------------------------+-------------------------------------+-------------------------------------------------------------------------+ | :meth:`trained_coef_ ` | OrtGradientForwardBackwardOptimizer | Returns the trained coefficients a dictionary. | +------------------------------------------------------------------------------------------------------------------+-------------------------------------+-------------------------------------------------------------------------+ | :meth:`value ` | BaseLearningRate | Returns the current learning rate. | +------------------------------------------------------------------------------------------------------------------+-------------------------------------+-------------------------------------------------------------------------+ | :meth:`value ` | LearningRateSGD | Returns the current learning rate. | +------------------------------------------------------------------------------------------------------------------+-------------------------------------+-------------------------------------------------------------------------+ | :meth:`value ` | LearningRateSGDNesterov | Returns the current learning rate. | +------------------------------------------------------------------------------------------------------------------+-------------------------------------+-------------------------------------------------------------------------+