pii-redactor-tokenizer-v1

Byte-level BPE tokenizer for the PII_MODEL browser-local redaction model.

Configuration

  • tokenizer type: byte-level BPE
  • vocab size: 64000
  • special tokens: 21
  • target corpus characters: 5000000000
  • accepted corpus characters: 4376978638
  • accepted documents: 5474812

This tokenizer is intended for span-level PII detection and placeholder-aware denoising.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support