AsBatch adds a batch size to a clone opt.

AsType explicitly sets the type of the input.

NameAsBatch takes an input name and converts it to its batch name.

NewInput returns a new input.

NewOpts returns a new set of options for a model.

NewPseudoHuberLoss return a new huber loss.

NewSequential returns a new sequential model.

ValuesFrom returns the value as an array of gorgonia values.

WithBatchSize sets the batch size for the model.

WithGraphLogger adds a logger to the model which will print out the graph operations as they occur.

WithLogger adds a logger to the model.

WithLoss uses a specific loss function with the model.

WithMetrics sets the metrics that the model should track.

WithOptimizer uses a specific optimizer function.

WithoutTracker uses no tracking with the model.

WithTracker adds a tracker to the model, if not provided one will be created.

TrainBatchLossMetric is the metric for batch training loss.

TrainLossMetric is the metric for training loss.

AllMetrics are all metrics.

CrossEntropy loss.

MSE is standard mean squared error loss.

PseudoCrossEntropy loss.

PseudoHuber is the Huber loss function.

CrossEntropyLoss is standard cross entropy loss.

Input into the model.

InputLayer is an input layer to be used in a chain.

MSELoss is mean squared error loss.

Opts are optsion for a model.

PseudoCrossEntropyLoss is standard cross entropy loss.

PseudoHuberLoss is a loss that is less sensetive to outliers.

Sequential model.

InputOr is a sum type of input or inputs.

Loss is the loss of a model.

Model is a prediction model.

ValueOr is a sum type that represents a gorgonia.Value or []gorgonia.Value.

CloneOpt are clone options.

InputOpt is an input option.

Inputs is a slice of input.

Metric tracked by the model.

Metrics is a set of metric.

Opt is a model option.

Values is a slice of value.