Tokenization

Splitting raw text into tokens (words, subwords, or characters) that a model can process numerically.