config.json

{
  "_name_or_path": "../models/flexibert-mini/",
  "architectures": [
    "FlexiBERTModel"
  ],
  "attention_heads_list": [
    2,
    2,
    4,
    4
  ],
  "attention_probs_dropout_prob": 0.1,
  "attention_type": [
    "sa",
    "sa",
    "l",
    "l"
  ],
  "bos_token_id": 0,
  "conv_kernel_size": 9,
  "eos_token_id": 2,
  "ff_dim_list": [
    [
      512,
      512,
      512
    ],
    [
      512,
      512,
      512
    ],
    [
      1024
    ],
    [
      1024
    ]
  ],
  "from_model_dict_hetero": false,
  "gradient_checkpointing": false,
  "head_ratio": 2,
  "hidden_act": "gelu",
  "hidden_dim_list": [
    256,
    256,
    128,
    128
  ],
  "hidden_dropout_prob": 0.1,
  "hidden_size": 768,
  "initializer_range": 0.02,
  "intermediate_size": 3072,
  "layer_norm_eps": 1e-12,
  "max_position_embeddings": 512,
  "model_type": "flexibert",
  "num_attention_heads": 12,
  "num_groups": 1,
  "num_hidden_layers": 4,
  "pad_token_id": 0,
  "position_embedding_type": "relative_key",
  "similarity_list": [
    "sdp",
    "sdp",
    "dct",
    "dct"
  ],
  "torch_dtype": "float32",
  "transformers_version": "4.25.0.dev0",
  "type_vocab_size": 2,
  "use_cache": true,
  "vocab_size": 50265
}