GigaToken: ~1000x faster Language model tokenization

200 points | by syrusakbary 3 hours ago

35 comments