config.json

{
  "_name_or_path": "backup/snowy-yogurt-6/",
  "adapt_attn_span": 32,
  "adapt_bs": 384,
  "adapt_dynamic": false,
  "adapt_nb_heads": 12,
  "adapt_reg": false,
  "adapt_span_cache": false,
  "adapt_span_loss_coeff": 0.01,
  "adapt_span_ramp": 16,
  "architectures": [
    "BertModel"
  ],
  "attention_probs_dropout_prob": 0.05,
  "attention_probs_dropout_prob_noise": 0.155,
  "attention_probs_dropout_prob_reference": 0.1,
  "classifier_dropout": null,
  "dynamic_ramp": false,
  "gradient_checkpointing": false,
  "hack": false,
  "headmask": 0.0,
  "hidden_act": "gelu",
  "hidden_dropout_prob": 0.05,
  "hidden_dropout_prob_noise": 0.155,
  "hidden_dropout_prob_reference": 0.1,
  "hidden_size": 768,
  "initializer_range": 0.02,
  "intermediate_size": 3072,
  "layer_norm_eps": 1e-12,
  "layerdrop": 0.0,
  "layerdrop_dynamic": false,
  "layerdrop_manifold": null,
  "layerdrop_noise": false,
  "layerdrop_noise_std": 0.1,
  "max_position_embeddings": 512,
  "mixup": false,
  "mixup_highest_layer": 12,
  "mixup_index": -99,
  "mixup_lowest_layer": 9,
  "mixup_manifold": null,
  "mixup_noise": false,
  "mixup_rate": [
    0.0,
    0.0,
    0.0
  ],
  "mixup_same_sample": false,
  "mixup_samples": [
    0,
    1,
    1
  ],
  "mixup_strategy1": [
    0,
    1,
    0
  ],
  "mixup_strategy2": [
    0,
    0,
    0
  ],
  "model_type": "bert",
  "multi_dropout": false,
  "num_attention_heads": 12,
  "num_hidden_layers": 12,
  "pad_token_id": 0,
  "position_embedding_type": "absolute",
  "single_dropout": false,
  "task_alpha": 1.0,
  "task_beta": 1.05,
  "task_lambda": 0.013,
  "torch_dtype": "float32",
  "transformers_version": "4.10.0.dev0",
  "type_vocab_size": 2,
  "use_cache": true,
  "vocab_size": 30522
}