Live
HitlistAI

Open Models16 Oct 2026

Hugging Face: BERT tokenization ~50× faster than 0.23

Why it matters

Hugging Face's tokenizers work made BERT (WordPiece) encoding about 50× faster than v0.23 (English chart: 50.8×). Normalization was ~88% of encode time, now ~13%; token ids unchanged. WordPiece still drives ~627M Hub downloads/month. Still a PR at blog time.

Read the source: Hugging Face (opens in a new tab)

All hits from 16 Oct 2026More Open ModelsGet the daily hits by email