ai.djl.training.dataset

Interface Sampler

All Known Implementing Classes:

BatchSampler

public interface Sampler

An interface for sampling data items from a RandomAccessDataset.

A Sampler implementation returns an iterator of batches for the RandomAccessDataset. Instead of returning the actual items, it returns the item indices. Different samplers can have different ways of sampling such as sampling with or without replacement.

Many of the samplers may also make use of Sampler.SubSamplers which sample not in batches but in individual data item indices.

Nested Class Summary

Nested Classes

Modifier and Type	Interface and Description
static interface	Sampler.SubSampler An interface that samples a single data item at a time.

Method Summary

Modifier and Type	Method and Description
int	getBatchSize() Returns the batch size of the Sampler.
java.util.Iterator<java.util.List<java.lang.Long>>	sample(RandomAccessDataset dataset) Fetches an iterator that iterates through the given RandomAccessDataset in mini-batches of indices.

Method Detail

sample

java.util.Iterator<java.util.List<java.lang.Long>> sample(RandomAccessDataset dataset)

Fetches an iterator that iterates through the given RandomAccessDataset in mini-batches of indices.

Parameters:

dataset - the RandomAccessDataset to sample from

Returns:

an iterator that iterates through the given RandomAccessDataset in mini-batches of indices

getBatchSize

int getBatchSize()

Returns the batch size of the Sampler.

Returns:

the batch size of the Sampler, -1 if batch size is not fixed