tokenizer/added_tokens.json

{
  "<point line characters>": 49408,
  "<|endoftext|>": 49407,
  "<|startoftext|>": 49406
}