English
text generation
maxkm's picture
Update README.md
2fc4b4e verified
metadata
license: mit
language:
  - en
tags:
  - text generation
datasets:
  - fhswf/TinyStoriesV2_cleaned

BPE_GPT2_TinyStoriesV2_cleaned

BPE Tokenizer Model for dataset 'fhswf/TinyStoriesV2_cleaned'

Based on get-neo BPE Tokenizer, but with a smaller vocabulary. Trained with TinyStoriesV2.

  • Vocab Size: 1024
  • 256 Base chars
  • 1 extra Token: <|endoftext|>
  • 767 merges