{
  "architectures": [
    "AbLang2ForMaskedLM"
  ],
  "attention_probs_dropout_prob": 0.1,
  "bias": true,
  "classifier_dropout": null,
  "hidden_act": "swiglu",
  "hidden_dropout_prob": 0.1,
  "hidden_size": 480,
  "initializer_range": 0.02,
  "intermediate_size": 1920,
  "layer_norm_eps": 1e-12,
  "max_position_embeddings": 512,
  "model_type": "ablang2",
  "num_attention_heads": 20,
  "num_hidden_layers": 12,
  "pad_token_id": 0,
  "pooler_type": "mean",
  "position_embedding_type": "absolute",
  "qk_layernorm": false,
  "residue_scaling": 1.0,
  "torch_dtype": "float32",
  "transformers_version": "4.46.3",
  "type_vocab_size": 2,
  "use_cache": true,
  "use_flash_attn": false,
  "vocab_size": 26
}