config.json

{
  "architectures": [
    "SpatialLMLlamaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 128000,
  "eos_token_id": [
    128001,
    128008,
    128009
  ],
  "head_dim": 64,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "intermediate_size": 8192,
  "max_position_embeddings": 131072,
  "mlp_bias": false,
  "model_type": "spatiallm_llama",
  "num_attention_heads": 32,
  "num_hidden_layers": 16,
  "num_key_value_heads": 8,
  "point_backbone": "sonata",
  "point_config": {
    "enc_channels": [
      48,
      96,
      192,
      384,
      512
    ],
    "enc_depths": [
      3,
      3,
      3,
      12,
      3
    ],
    "enc_mode": true,
    "enc_num_head": [
      3,
      6,
      12,
      24,
      32
    ],
    "enc_patch_size": [
      1024,
      1024,
      1024,
      1024,
      1024
    ],
    "in_channels": 6,
    "mask_token": true,
    "mlp_ratio": 4,
    "order": [
      "z",
      "z-trans",
      "hilbert",
      "hilbert-trans"
    ],
    "stride": [
      2,
      2,
      2,
      2
    ],
    "num_bins": 1280
  },
  "projector": "mlp",
  "point_end_token_id": 128012,
  "point_start_token_id": 128011,
  "point_token_id": 128013,
  "pretraining_tp": 1,
  "rms_norm_eps": 1e-05,
  "rope_scaling": {
    "factor": 32.0,
    "high_freq_factor": 4.0,
    "low_freq_factor": 1.0,
    "original_max_position_embeddings": 8192,
    "rope_type": "llama3"
  },
  "rope_theta": 500000.0,
  "tie_word_embeddings": true,
  "torch_dtype": "float32",
  "transformers_version": "4.46.1",
  "use_cache": false,
  "vocab_size": 128256
}