bad-gpt / encoder.py

Commit History

  1. Switch from custom tokenizer to tiktoken

    e3fc82f

  1. Retrieve dataset from dataset instead of text file

    2a3cf64
  2. Clone from github

    3c8279d