single_sample_metrics#

Common metrics to calculate performance on single samples.

Functions

`calculate_per_sample_loss`(model, task_type, ...)	Calculate error per sample for a given model and a dataset.
`per_sample_cross_entropy`(y_true, y_pred[, eps])	Calculate cross entropy on a single sample.
`per_sample_mse`(y_true, y_pred)	Calculate mean square error on a single value.

RegressionUniformModel.predict

calculate_per_sample_loss