{
  "architectures": [
    "ProSSTForMaskedLM"
  ],
  "attention_probs_dropout_prob": 0.1,
  "auto_map": {
    "AutoConfig": "configuration_prosst.ProSSTConfig",
    "AutoModelForMaskedLM": "modeling_prosst.ProSSTForMaskedLM"
  },
  "hidden_act": "gelu",
  "hidden_dropout_prob": 0.1,
  "hidden_size": 768,
  "initializer_range": 0.02,
  "intermediate_size": 3072,
  "layer_norm_eps": 1e-07,
  "mask_token_id": 24,
  "max_position_embeddings": -1,
  "max_relative_positions": 1024,
  "mlm_probability": 0.15,
  "model_type": "ProSST",
  "num_attention_heads": 12,
  "num_hidden_layers": 12,
  "pad_token_id": 0,
  "pooler_dropout": 0.1,
  "pooler_hidden_act": "gelu",
  "pooler_hidden_size": 768,
  "pooling_head": "mean",
  "pos_att_type": [
    "aa2pos",
    "pos2aa",
    "ss2aa"
  ],
  "position_biased_input": false,
  "position_embedding_type": "relative",
  "relative_attention": true,
  "scale_hidden": 1,
  "ss_vocab_size": 2051,
  "token_dropout": true,
  "torch_dtype": "float32",
  "transformers_version": "4.38.2",
  "type_vocab_size": 0,
  "vocab_size": 25
}