config.json

{
  "_name_or_path": "largetav/",
  "architectures": [
    "BertForDiacritization"
  ],
  "attention_probs_dropout_prob": 0.1,
  "classifier_dropout": null,
  "hidden_act": "gelu",
  "hidden_dropout_prob": 0.1,
  "hidden_size": 1024,
  "initializer_range": 0.02,
  "intermediate_size": 4096,
  "layer_norm_eps": 1e-12,
  "mat_lect_token": "<MAT_LECT>",
  "max_position_embeddings": 2048,
  "model_type": "bert",
  "nikud_classes": [
    "",
    "<MAT_LECT>",
    "\u05bc",
    "\u05b0",
    "\u05b1",
    "\u05b2",
    "\u05b3",
    "\u05b4",
    "\u05b5",
    "\u05b6",
    "\u05b7",
    "\u05b8",
    "\u05b9",
    "\u05ba",
    "\u05bb",
    "\u05bc\u05b0",
    "\u05bc\u05b1",
    "\u05bc\u05b2",
    "\u05bc\u05b3",
    "\u05bc\u05b4",
    "\u05bc\u05b5",
    "\u05bc\u05b6",
    "\u05bc\u05b7",
    "\u05bc\u05b8",
    "\u05bc\u05b9",
    "\u05bc\u05ba",
    "\u05bc\u05bb",
    "\u05c7",
    "\u05bc\u05c7"
  ],
  "num_attention_heads": 16,
  "num_hidden_layers": 24,
  "pad_token_id": 0,
  "position_embedding_type": "absolute",
  "shin_classes": [
    "\u05c1",
    "\u05c2"
  ],
  "torch_dtype": "float32",
  "transformers_version": "4.42.4",
  "type_vocab_size": 2,
  "use_cache": true,
  "vocab_size": 1024
}