openml.extensions.mxnet.Config

class openml.extensions.mxnet.Config

Represents the configuration of the OpenML MXNet Extension

batch_size = None

batch_size represents the processing batch size for training

criterion_gen(self, task: openml.tasks.task.OpenMLTask) → mxnet.gluon.loss.Loss

loss_gen returns the loss criterion based on the task type

epoch_count = None

epoch_count represents the number of epochs the model should be trained for

initializer_gen(self, task: openml.tasks.task.OpenMLTask) → 'Optional[mxnet.init.Initializer]'

initializer_gen returns the initializer to be used for a given OpenML task

metric_gen(self, task: openml.tasks.task.OpenMLTask) → mxnet.metric.EvalMetric

metric_gen returns the metric to be used for the given task

optimizer_gen(self, lr_scheduler: mxnet.lr_scheduler.LRScheduler, task: openml.tasks.task.OpenMLTask) → mxnet.optimizer.Optimizer

optimizer_gen returns the optimizer to be used for a given OpenMLTask

predict(self, output: Union[mxnet.ndarray.NDArray, mxnet.symbol.Symbol], task: openml.tasks.task.OpenMLTask) → 'backend_type'

predict turns the outputs of the model into actual predictions

predict_proba(self, output: Union[mxnet.ndarray.NDArray, mxnet.symbol.Symbol]) → Union[mxnet.ndarray.NDArray, mxnet.symbol.Symbol]

predict_proba turns the outputs of the model into probabilities for each class

progress_callback(self, fold: int, rep: int, epoch: int, step: int, loss: mxnet.ndarray.NDArray, metric: mxnet.metric.EvalMetric)

progress_callback is called when a training step is finished, in order to report

the current progress

sanitize(self, output: mxnet.ndarray.NDArray) → mxnet.ndarray.NDArray

sanitize sanitizes the input data in order to ensure that models can be trained safely

scheduler_gen(self, task: openml.tasks.task.OpenMLTask) → mxnet.lr_scheduler.LRScheduler

scheduler_gen returns the scheduler to be used for a given task