{
  "architectures": [
    "ViSCoP_VisionEncoderModel"
  ],
  "attention_dropout": 0.0,
  "auto_map": {
    "AutoConfig": "configuration_viscop_vision_encoder.ViSCoP_VisionEncoderConfig",
    "AutoModel": "modeling_viscop_vision_encoder.ViSCoP_VisionEncoderModel"
  },
  "hidden_act": "gelu_pytorch_tanh",
  "hidden_size": 1152,
  "interaction_module": "cross_attention",
  "interaction_module_layers": null,
  "intermediate_size": 4304,
  "layer_norm_eps": 1e-06,
  "model_type": "viscop_vision_encoder",
  "num_attention_heads": 16,
  "num_channels": 3,
  "num_hidden_layers": 27,
  "num_visual_probes": 16,
  "patch_size": 14,
  "torch_dtype": "bfloat16",
  "transformers_version": "4.46.3"
}