jfrick
/

scholarly-code-tokenizer

Publish code-focused 32K tokenizer and promote it to the default

0ebdb4f verified 11 days ago

107 Bytes

{
  "pad_token": "<|pad|>",
  "bos_token": "<|bos|>",
  "eos_token": "<|eos|>",
  "unk_token": "<|unk|>"
}