config.json

{
  "_name_or_path": "microsoft/markuplm-base",
  "architectures": [
    "MarkupLMForTokenClassification"
  ],
  "attention_probs_dropout_prob": 0.1,
  "bos_token_id": 0,
  "classifier_dropout": null,
  "eos_token_id": 2,
  "gradient_checkpointing": false,
  "hidden_act": "gelu",
  "hidden_dropout_prob": 0.1,
  "hidden_size": 768,
  "id2label": {
    "0": "title",
    "1": "author",
    "2": "date",
    "3": "body",
    "4": "topic",
    "5": "source",
    "6": "other"
  },
  "initializer_range": 0.02,
  "intermediate_size": 3072,
  "label2id": {
    "author": 1,
    "body": 3,
    "date": 2,
    "other": 6,
    "source": 5,
    "title": 0,
    "topic": 4
  },
  "layer_norm_eps": 1e-05,
  "max_depth": 50,
  "max_position_embeddings": 514,
  "max_xpath_subs_unit_embeddings": 1024,
  "max_xpath_tag_unit_embeddings": 256,
  "model_type": "markuplm",
  "num_attention_heads": 12,
  "num_hidden_layers": 12,
  "pad_token_id": 1,
  "position_embedding_type": "absolute",
  "subs_pad_id": 1001,
  "tag_pad_id": 216,
  "torch_dtype": "float32",
  "transformers_version": "4.29.0.dev0",
  "type_vocab_size": 1,
  "use_cache": true,
  "vocab_size": 50267,
  "xpath_unit_hidden_size": 32
}