{
  "backbone": {
    "attention_head_dim": 64,
    "cross_attention_dim": 768,
    "in_channels": 1024,
    "num_attention_heads": 16,
    "num_layers": 16
  },
  "backbone_cls": "tsr.models.transformer.transformer_1d.Transformer1D",
  "cond_image_size": 512,
  "decoder": {
    "activation": "silu",
    "in_channels": 120,
    "n_hidden_layers": 9,
    "n_neurons": 64
  },
  "decoder_cls": "tsr.models.network_utils.NeRFMLP",
  "image_tokenizer": {
    "pretrained_model_name_or_path": "facebook/dino-vitb16"
  },
  "image_tokenizer_cls": "tsr.models.tokenizers.image.DINOSingleImageTokenizer",
  "post_processor": {
    "in_channels": 1024,
    "out_channels": 40
  },
  "post_processor_cls": "tsr.models.network_utils.TriplaneUpsampleNetwork",
  "renderer": {
    "density_activation": "exp",
    "density_bias": -1.0,
    "feature_reduction": "concat",
    "num_samples_per_ray": 128,
    "radius": 0.87
  },
  "renderer_cls": "tsr.models.nerf_renderer.TriplaneNeRFRenderer",
  "tokenizer": {
    "num_channels": 1024,
    "plane_size": 32
  },
  "tokenizer_cls": "tsr.models.tokenizers.triplane.Triplane1DTokenizer"
}