config.json

{
  "arch_specifier": "no-align+fused-gelu-mlp",
  "architectures": [
    "Emma-X"
  ],
  "auto_map": {
    "AutoConfig": "configuration_prismatic.OpenVLAConfig",
    "AutoModelForVision2Seq": "modeling_prismatic.EmmaxForActionPrediction"
  },
  "hf_llm_id": "meta-llama/Llama-2-7b-hf",
  "image_resize_strategy": "resize-naive",
  "image_sizes": [
    224,
    224
  ],
  "llm_backbone_id": "llama2-7b-pure",
  "llm_max_length": 2048,
  "model_type": "openvla",
  "n_action_bins": 256,
  "norm_stats": {
    "bridge_orig": {
      "action": {
        "mask": [
          true,
          true,
          true,
          true,
          true,
          true,
          false
        ],
        "max": [
          0.41691166162490845,
          0.25864794850349426,
          0.21218234300613403,
          3.122201919555664,
          1.8618112802505493,
          6.280478477478027,
          1.0
        ],
        "mean": [
          0.0002334193413844332,
          0.0001300490548601374,
          -0.0001276246621273458,
          -0.00015565502690151334,
          -0.0004039333143737167,
          0.0002355769247515127,
          0.5764579772949219
        ],
        "min": [
          -0.4007510244846344,
          -0.13874775171279907,
          -0.22553899884223938,
          -3.2010786533355713,
          -1.8618112802505493,
          -6.279075622558594,
          0.0
        ],
        "q01": [
          -0.02872725307941437,
          -0.04170349963009357,
          -0.026093858778476715,
          -0.08092105075716972,
          -0.09288699507713317,
          -0.20718276381492615,
          0.0
        ],
        "q99": [
          0.028309678435325586,
          0.040855254605412394,
          0.040161586627364146,
          0.08192047759890528,
          0.07792850524187081,
          0.20382574498653397,
          1.0
        ],
        "std": [
          0.009765916503965855,
          0.013689138926565647,
          0.012667354196310043,
          0.02853417582809925,
          0.0306379534304142,
          0.07691461592912674,
          0.49737000465393066
        ]
      },
      "num_trajectories": 60064,
      "num_transitions": 2135463,
      "proprio": {
        "max": [
          0.5862360596656799,
          0.4034728705883026,
          0.36494991183280945,
          1.514088749885559,
          1.570796251296997,
          3.1415255069732666,
          0.0,
          1.1154625415802002
        ],
        "mean": [
          0.3094092905521393,
          0.030575834214687347,
          0.06454048305749893,
          0.006824127864092588,
          -0.07762698084115982,
          0.10757855325937271,
          0.0,
          0.7084035277366638
        ],
        "min": [
          -0.04167502000927925,
          -0.3945816159248352,
          -0.15537554025650024,
          -3.141592502593994,
          -1.4992541074752808,
          -3.14153790473938,
          0.0,
          0.04637829214334488
        ],
        "q01": [
          0.17111587673425674,
          -0.16998695254325866,
          -0.05544630073010921,
          -0.366876106262207,
          -0.5443069756031036,
          -1.3536006283760071,
          0.0,
          0.052190229296684265
        ],
        "q99": [
          0.45320980012416834,
          0.23518154799938193,
          0.1951873075962065,
          0.3806115746498103,
          0.2789784955978382,
          1.8410426235198971,
          0.0,
          1.0105689764022827
        ],
        "std": [
          0.060528464615345,
          0.09188621491193771,
          0.05159851908683777,
          0.13182717561721802,
          0.17031119763851166,
          0.5767312049865723,
          0.0,
          0.3519800305366516
        ]
      }
    }
  },
  "output_projector_states": false,
  "pad_to_multiple_of": 64,
  "pad_token_id": 32000,
  "text_config": {
    "model_type": "llama",
    "pad_token_id": 32000,
    "torch_dtype": "bfloat16",
    "vocab_size": 32064
  },
  "timm_model_ids": [
    "vit_large_patch14_reg4_dinov2.lvd142m",
    "vit_so400m_patch14_siglip_224"
  ],
  "timm_override_act_layers": [
    null,
    null
  ],
  "torch_dtype": "bfloat16",
  "transformers_version": "4.40.1",
  "use_fused_vision_backbone": true,
  "vision_backbone_id": "dinosiglip-vit-so-224px"
}