/kitoken

Fast and versatile tokenizer for language-models, supporting BPE and Unigram tokenization and usable in native and WASM environments

Primary LanguageRustBSD 2-Clause "Simplified" LicenseBSD-2-Clause

Stargazers