config.json

{
  "_name_or_path": "nyu-visionx/Cambrian-S-7B-S3",
  "add_faster_video": false,
  "add_time_instruction": true,
  "aggregator_embed_dim": 2048,
  "architectures": [
    "LlavaQwenForCausalLM"
  ],
  "attention_dropout": 0.0,
  "bos_token_id": 151643,
  "cam_mix_mlp_lr": null,
  "cam_out_projector_lr": null,
  "cam_out_projector_type": "mlp2x_gelu_ln",
  "cam_token_distill_type": "mse",
  "cam_token_distill_weight": 1.0,
  "cam_token_embed_dim": 2048,
  "cam_token_norm": false,
  "cam_token_projector_lr": null,
  "cam_token_projector_type": "mlp1x_gelu",
  "camera_token_index": 0,
  "camera_tokens_mode": "camera_tokens",
  "camera_tokens_place": "append_to_frame",
  "custom_llm_ce_loss": false,
  "debug_trajectory_check": false,
  "dino_embed_dim": 1024,
  "distill_loss_type": "mse",
  "distill_loss_weight": 1.0,
  "distill_projector_type": "mlp1x_gelu",
  "downstream_head_lr": 0.0001,
  "enable_camera": true,
  "enable_depth": false,
  "enable_point": false,
  "enable_rec_pose_viz": true,
  "eos_token_id": 151645,
  "faster_token_stride": 10,
  "feature_fusion_method": "concat",
  "feature_fusion_stage": "before_projector",
  "feature_fusion_type": "none",
  "feature_projector_hidden_dim": null,
  "feature_projector_type": "mlp",
  "force_sample": true,
  "fusion_attention_heads": 8,
  "head_name": null,
  "hidden_act": "silu",
  "hidden_size": 3584,
  "image_aspect_ratio": "anyres_max_9",
  "image_crop_resolution": null,
  "image_grid_pinpoints": [
    [
      384,
      384
    ],
    [
      384,
      768
    ],
    [
      384,
      1152
    ],
    [
      384,
      1536
    ],
    [
      384,
      1920
    ],
    [
      384,
      2304
    ],
    [
      768,
      384
    ],
    [
      768,
      768
    ],
    [
      768,
      1152
    ],
    [
      768,
      1536
    ],
    [
      768,
      1920
    ],
    [
      768,
      2304
    ],
    [
      1152,
      384
    ],
    [
      1152,
      768
    ],
    [
      1152,
      1152
    ],
    [
      1152,
      1536
    ],
    [
      1152,
      1920
    ],
    [
      1152,
      2304
    ],
    [
      1536,
      384
    ],
    [
      1536,
      768
    ],
    [
      1536,
      1152
    ],
    [
      1536,
      1536
    ],
    [
      1536,
      1920
    ],
    [
      1536,
      2304
    ],
    [
      1920,
      384
    ],
    [
      1920,
      768
    ],
    [
      1920,
      1152
    ],
    [
      1920,
      1536
    ],
    [
      1920,
      1920
    ],
    [
      1920,
      2304
    ],
    [
      2304,
      384
    ],
    [
      2304,
      768
    ],
    [
      2304,
      1152
    ],
    [
      2304,
      1536
    ],
    [
      2304,
      1920
    ],
    [
      2304,
      2304
    ]
  ],
  "image_split_resolution": null,
  "image_token_index": 151646,
  "initializer_range": 0.02,
  "intermediate_size": 18944,
  "load_rec_model": true,
  "max_position_embeddings": 32768,
  "max_window_layers": 28,
  "mm_hidden_size": 1152,
  "mm_img_tok_num": 64,
  "mm_newline_position": "grid",
  "mm_patch_merge_type": "spatial_unpad",
  "mm_projector_lr": null,
  "mm_projector_type": "mlp2x_gelu",
  "mm_resampler_type": null,
  "mm_spatial_pool_mode": "bilinear",
  "mm_spatial_pool_stride": 2,
  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model,mm_rec_head",
  "mm_use_im_patch_token": false,
  "mm_use_im_start_end": false,
  "mm_vision_select_feature": "patch",
  "mm_vision_select_layer": -2,
  "mm_vision_tower": "google/siglip-so400m-patch14-384",
  "mm_vision_tower_lr": 2e-06,
  "model_type": "llava_qwen",
  "num_attention_heads": 28,
  "num_hidden_layers": 28,
  "num_intermediate_layers": 4,
  "num_key_value_heads": 4,
  "pos_skipping_range": 4096,
  "query_loss_frame_idx": 0,
  "query_mode": "query_after_image",
  "random_feature_baseline": false,
  "rec_camera_loss_components": "all",
  "rec_camera_loss_type": "l1",
  "rec_camera_loss_weight": 5.0,
  "rec_camera_loss_weight_focal": 0.5,
  "rec_camera_loss_weight_rot": 1.0,
  "rec_camera_loss_weight_trans": 10.0,
  "rec_depth_gradient_loss_fn": "grad",
  "rec_depth_head_patch_start_idx": 1,
  "rec_depth_loss_weight": 1.0,
  "rec_depth_valid_range": 0.98,
  "rec_dpt_head_type": "dpt_head",
  "rec_embed_dim": 2048,
  "rec_loss_weight": 0.5,
  "rec_model_path": "None",
  "rec_point_gradient_loss_fn": "grad",
  "rec_point_loss_weight": 0.0,
  "rec_point_valid_range": 0.98,
  "rec_pose_viz_frequency": 256,
  "rec_projector_type": "mlp1x_gelu",
  "rms_norm_eps": 1e-06,
  "rope_scaling": null,
  "rope_theta": 1000000.0,
  "sliding_window": 131072,
  "tie_word_embeddings": false,
  "token_num": 2,
  "tokenizer_model_max_length": 32768,
  "tokenizer_padding_side": "right",
  "tokenspace_cam_reference_frame": "none",
  "tokenspace_cam_supervision_type": "none",
  "tokenspace_cam_text_strategy": "none",
  "torch_dtype": "bfloat16",
  "transformers_version": "4.40.0.dev0",
  "unpad_image": false,
  "use_cache": true,
  "use_distill": false,
  "use_feature_fusion": false,
  "use_mm_proj": true,
  "use_point_masks": true,
  "use_pos_skipping": false,
  "use_scale_alignment": true,
  "use_sliding_window": false,
  "use_trajectory_balance": true,
  "use_vggt_cam_input": false,
  "use_vggt_cam_output": false,
  "vggt_cam_fusion_mode": "splice",
  "vggt_cam_projector_lr": null,
  "vggt_cam_projector_type": "mlp2x_gelu",
  "vision_tower_pretrained": null,
  "vocab_size": 151672
}