{
  "architectures": [
    "Sam3ViTModel"
  ],
  "attention_dropout": 0.0,
  "dtype": "float32",
  "global_attn_indexes": [
    0,
    1
  ],
  "hidden_act": "gelu",
  "hidden_dropout": 0.0,
  "hidden_size": 32,
  "image_size": 224,
  "initializer_range": 0.02,
  "intermediate_size": 64,
  "layer_norm_eps": 1e-06,
  "layer_scale_init_value": null,
  "model_type": "sam3_vit_model",
  "num_attention_heads": 4,
  "num_channels": 3,
  "num_hidden_layers": 2,
  "patch_size": 14,
  "pretrain_image_size": 336,
  "rope_theta": 10000.0,
  "transformers_version": "5.16.0.dev0",
  "window_size": 8
}