GigaToken: ~1000x faster Language model tokenization

(github.com)

230 points | by syrusakbary 4 hours ago ago

45 comments