{
  "_name_or_path": "ltg/norbert4-xlarge",
  "architectures": [
    "GptBertForTokenClassification"
  ],
  "attention_dropout": 0.0,
  "attn_implementation": "flash_attention_2",
  "auto_map": {
    "AutoConfig": "ltg/norbert4-xlarge--configuration_gptbert.GptBertConfig",
    "AutoModel": "ltg/norbert4-xlarge--modeling_gptbert.GptBertModel",
    "AutoModelForCausalLM": "ltg/norbert4-xlarge--modeling_gptbert.GptBertForCausalLM",
    "AutoModelForMaskedLM": "ltg/norbert4-xlarge--modeling_gptbert.GptBertForMaskedLM",
    "AutoModelForMultipleChoice": "ltg/norbert4-xlarge--modeling_gptbert.GptBertForMultipleChoice",
    "AutoModelForQuestionAnswering": "ltg/norbert4-xlarge--modeling_gptbert.GptBertForQuestionAnswering",
    "AutoModelForSequenceClassification": "ltg/norbert4-xlarge--modeling_gptbert.GptBertForSequenceClassification",
    "AutoModelForTokenClassification": "ltg/norbert4-xlarge--modeling_gptbert.GptBertForTokenClassification"
  },
  "bos_token_id": 1,
  "classifier_dropout": 0.2,
  "deterministic_flash_attn": false,
  "embedding_dropout": 0.1,
  "eos_token_id": 2,
  "global_window_length": 8192,
  "hidden_dropout": 0.0,
  "hidden_size": 1536,
  "id2label": {
    "0": "O",
    "1": "B-PER",
    "2": "I-PER",
    "3": "B-ORG",
    "4": "I-ORG",
    "5": "B-LOC",
    "6": "I-LOC",
    "7": "B-GPE_LOC",
    "8": "I-GPE_LOC",
    "9": "B-GPE_ORG",
    "10": "I-GPE_ORG",
    "11": "B-PROD",
    "12": "I-PROD",
    "13": "B-EVT",
    "14": "I-EVT",
    "15": "B-MISC",
    "16": "I-MISC",
    "17": "B-DRV",
    "18": "I-DRV"
  },
  "intermediate_size": 4096,
  "label2id": {
    "B-DRV": 17,
    "B-EVT": 13,
    "B-GPE_LOC": 7,
    "B-GPE_ORG": 9,
    "B-LOC": 5,
    "B-MISC": 15,
    "B-ORG": 3,
    "B-PER": 1,
    "B-PROD": 11,
    "I-DRV": 18,
    "I-EVT": 14,
    "I-GPE_LOC": 8,
    "I-GPE_ORG": 10,
    "I-LOC": 6,
    "I-MISC": 16,
    "I-ORG": 4,
    "I-PER": 2,
    "I-PROD": 12,
    "O": 0
  },
  "layer_norm_eps": 1e-07,
  "local_global_ratio": 4,
  "local_window_length": 256,
  "mask_token_id": 4,
  "max_position_embeddings": 16384,
  "model": "norbert4",
  "num_attention_heads": 24,
  "num_layers": 32,
  "pad_token_id": 3,
  "query_key_head_size": 64,
  "rope_theta": 160000,
  "torch_dtype": "bfloat16",
  "transformers_version": "4.48.0",
  "unk_token_id": 0,
  "use_cache": false,
  "value_head_size": 64,
  "vocab_size": 51200
}