tokenizer_config.json

{
  "name_or_path": "models/gpt_personachat_cache",
  "special_tokens_map_file": null,
  "tokenizer_class": "OpenAIGPTTokenizer",
  "unk_token": "<unk>"
}