tokenizer/added_tokens.json

{
  "<function populate_text_encoder at 0x7f968a2e7d30>": 49408
}