soldni's picture
Create README.md
37388b3
metadata
license: apache-2.0
datasets:
  - EleutherAI/pile
language:
  - en
tags:
  - tokenizer

A copy of Eleuther AI's gpt-neox-20b, with three special tokens added to mask PII:

  • |||EMAIL_ADDRESS|||
  • |||PHONE_NUMBER|||
  • |||IP_ADDRESS|||