Read a CIFAR-10 or CIFAR-100 dataset.
Read a CIFAR-10 or CIFAR-100 dataset.
the Hadoop-style path to a CIFAR data file in the given format.
the format, either 'CIFAR-10' or 'CIFAR-100' (binary variant).
the maximum size (in bytes) of data per partition. Default: 10MB.
Extends org.apache.spark.sql.DataFrameReader with CIFAR support.