tiny/word_deberta-v2-tiny.mlx.json

{
  "tokenizer": {
    "_target_": "transformers.AutoTokenizer.from_pretrained",
    "pretrained_model_name_or_path": "ku-nlp/deberta-v2-tiny-japanese",
    "additional_special_tokens": [
      "[著者]",
      "[読者]",
      "[不特定:人]",
      "[不特定:物]",
      "[NULL]",
      "[NA]",
      "[ROOT]",
      " "
    ],
    "_convert_": "all"
  },
  "predict_cfg": {
    "_target_": "kwja.datamodule.datasets.WordInferenceDataset",
    "max_seq_length": 256,
    "document_split_stride": 2,
    "cohesion_tasks": [
      "pas_analysis",
      "bridging_reference_resolution",
      "coreference_resolution"
    ],
    "exophora_referents": [
      "著者",
      "読者",
      "不特定:人",
      "不特定:物"
    ],
    "restrict_cohesion_target": true,
    "pas_cases": [
      "ガ",
      "ヲ",
      "ニ",
      "ガ２",
      "デ",
      "ト",
      "時間"
    ],
    "br_cases": [
      "ノ"
    ],
    "special_tokens": [
      "[著者]",
      "[読者]",
      "[不特定:人]",
      "[不特定:物]",
      "[NULL]",
      "[NA]",
      "[ROOT]",
      " "
    ],
    "tokenizer": {
      "_target_": "transformers.AutoTokenizer.from_pretrained",
      "pretrained_model_name_or_path": "ku-nlp/deberta-v2-tiny-japanese",
      "additional_special_tokens": [
        "[著者]",
        "[読者]",
        "[不特定:人]",
        "[不特定:物]",
        "[NULL]",
        "[NA]",
        "[ROOT]",
        " "
      ],
      "_convert_": "all"
    },
    "juman_file": null
  },
  "prediction_writer_cfg": {
    "_target_": "kwja.callbacks.WordModuleWriter",
    "ambig_surf_specs": [
      {
        "conjtype": "イ形容詞アウオ段",
        "conjform": "エ基本形"
      },
      {
        "conjtype": "イ形容詞イ段",
        "conjform": "エ基本形"
      },
      {
        "conjtype": "イ形容詞イ段特殊",
        "conjform": "エ基本形"
      }
    ]
  },
  "encoder_repo": "ku-nlp/deberta-v2-tiny-japanese",
  "max_seq_length": 256,
  "document_split_stride": 2,
  "dependency_topk": 4,
  "discourse_threshold": 0.0,
  "vocab_size": 32008,
  "num_cohesion_relations": 9
}