Method and apparatus with data loading

An apparatus with data loading, the apparatus comprising:

• one or more processors configured to:
• based on sizes of a plurality of data files in a training dataset, divide the training dataset into a plurality of sub-sets;
• load some data files in each sub-set into a plurality of processors;
• extract candidate packing combinations based on a maximum data file size and a batch size of distributed training;
• determine a packing combination among the candidate packing combinations based on a comparison between a ratio of a number of data files corresponding to each of the candidate packing combinations and a ratio of a number of data files between the plurality of sub-sets;
• determine packed data files by packing the one or more data files according to the packing combination;
• and
• reallocate the packed data files to the processors in the same group.