tokenizer_config.json

{
  "added_tokens_decoder": {
    "0": {
      "content": "<unk>",
      "lstrip": false,
      "normalized": false,
      "rstrip": false,
      "single_word": false,
      "special": true
    },
    "1": {
      "content": "<s>",
      "lstrip": false,
      "normalized": false,
      "rstrip": false,
      "single_word": false,
      "special": true
    },
    "2": {
      "content": "</s>",
      "lstrip": false,
      "normalized": false,
      "rstrip": false,
      "single_word": false,
      "special": true
    },
    "92544": {
      "content": "<table>",
      "lstrip": false,
      "normalized": true,
      "rstrip": false,
      "single_word": false,
      "special": false
    },
    "92545": {
      "content": "</table>",
      "lstrip": false,
      "normalized": true,
      "rstrip": false,
      "single_word": false,
      "special": false
    },
    "92546": {
      "content": "<tr>",
      "lstrip": false,
      "normalized": true,
      "rstrip": false,
      "single_word": false,
      "special": false
    },
    "92547": {
      "content": "</tr>",
      "lstrip": false,
      "normalized": true,
      "rstrip": false,
      "single_word": false,
      "special": false
    },
    "92548": {
      "content": "<td>",
      "lstrip": false,
      "normalized": true,
      "rstrip": false,
      "single_word": false,
      "special": false
    },
    "92549": {
      "content": "</td>",
      "lstrip": false,
      "normalized": true,
      "rstrip": false,
      "single_word": false,
      "special": false
    },
    "92550": {
      "content": "<th>",
      "lstrip": false,
      "normalized": true,
      "rstrip": false,
      "single_word": false,
      "special": false
    },
    "92551": {
      "content": "</th>",
      "lstrip": false,
      "normalized": true,
      "rstrip": false,
      "single_word": false,
      "special": false
    },
    "92552": {
      "content": "colspan=",
      "lstrip": false,
      "normalized": true,
      "rstrip": false,
      "single_word": false,
      "special": false
    },
    "92553": {
      "content": "rowspan=",
      "lstrip": false,
      "normalized": true,
      "rstrip": false,
      "single_word": false,
      "special": false
    },
    "92554": {
      "content": "<thead>",
      "lstrip": false,
      "normalized": true,
      "rstrip": false,
      "single_word": false,
      "special": false
    },
    "92555": {
      "content": "</thead>",
      "lstrip": false,
      "normalized": true,
      "rstrip": false,
      "single_word": false,
      "special": false
    },
    "92556": {
      "content": "<tbody>",
      "lstrip": false,
      "normalized": true,
      "rstrip": false,
      "single_word": false,
      "special": false
    },
    "92557": {
      "content": "</tbody>",
      "lstrip": false,
      "normalized": true,
      "rstrip": false,
      "single_word": false,
      "special": false
    }
  },
  "auto_map": {
    "AutoTokenizer": [
      "tokenization_internlm_xcomposer2.InternLMXComposer2Tokenizer",
      null
    ]
  },
  "bos_token": "<s>",
  "clean_up_tokenization_spaces": false,
  "eos_token": "</s>",
  "model_max_length": 1000000000000000019884624838656,
  "pad_token": "</s>",
  "padding_side": "right",
  "tokenizer_class": "InternLMXComposer2Tokenizer",
  "unk_token": "<unk>",
  "use_fast": false
}