tokenizer_config.json

{
  "add_prefix_space": true,
  "backend": "tokenizers",
  "bos_token": "<bos>",
  "clean_up_tokenization_spaces": false,
  "eos_token": "<eos>",
  "is_local": false,
  "local_files_only": false,
  "model_max_length": 1000000000,
  "pad_token": "<eos>",
  "tokenizer_class": "PreTrainedTokenizerFast",
  "unk_token": "<unk>",
  "vocab_size": 4096,
  "additional_special_tokens": [
    "<|im_start|>",
    "<|im_end|>"
  ]
}