config.json

{
  "architectures": [
    "MyBertForMaskedLM"
  ],
  "attention_probs_dropout_prob": 0.0,
  "auto_map": {
      "AutoConfig": "configuration_mybert.MyBertConfig",
      "AutoModel": "modeling_mybert.MyBertModel",
      "AutoModelForMaskedLM": "modeling_mybert.MyBertForMaskedLM",
      "AutoModelForSequenceClassification": "modeling_mybert.MyBertForSequenceClassification",
      "AutoModelForTokenClassification": "modeling_mybert.MyBertForTokenClassification",
      "AutoModelForQuestionAnswering": "modeling_mybert.MyBertForQuestionAnswering"
  },
  "dtype": "float32",
  "hidden_dropout_prob": 0.0,
  "hidden_size": 768,
  "initializer_range": 0.02,
  "intermediate_size": 2048,
  "layer_norm_eps": 1e-05,
  "max_position_embeddings": 2048,
  "model_type": "mybert",
  "num_attention_heads": 12,
  "num_hidden_layers": 12,
  "pad_token_id": 0,
  "rope_theta": 10000.0,
  "sparse_prediction": true,
  "tie_word_embeddings": true,
  "transformers_version": "5.13.1",
  "use_bias": false,
  "vocab_size": 28996
}