{
  "add_pooling_layer": true,
  "architectures": [
    "ViTNepaForPreTraining"
  ],
  "attention_probs_dropout_prob": 0.0,
  "drop_path_prob": 0.0,
  "dtype": "float32",
  "hidden_act": "gelu",
  "hidden_dropout_prob": 0.0,
  "hidden_size": 768,
  "image_size": 224,
  "initializer_range": 0.02,
  "intermediate_size": 3072,
  "is_causal": true,
  "layer_norm_eps": 1e-12,
  "layerscale_value": 1e-05,
  "model_type": "vit_nepa",
  "num_attention_heads": 12,
  "num_channels": 3,
  "num_hidden_layers": 12,
  "num_prompt_token": 0,
  "patch_size": 14,
  "pos_embed_jitter": null,
  "pos_embed_rescale": 2.0,
  "pos_embed_shift": null,
  "qk_norm": true,
  "qk_norm_affine": false,
  "qk_norm_bias": false,
  "qkv_bias": true,
  "rope_theta": 100.0,
  "transformers_version": "4.56.2",
  "use_gated_mlp": false
}