config.json

{
  "adapter_config": {
    "attention_dropout": 0.0,
    "float32_attention": true,
    "head_dim": 72,
    "hidden_act": "silu",
    "hidden_size": 1152,
    "image_feature_dropout": 0.0,
    "image_padding_embed": null,
    "initializer_range": 0.02,
    "intermediate_size": 18944,
    "model_type": "",
    "num_attention_heads": 16,
    "num_key_value_heads": 16,
    "residual_dropout": 0.0,
    "text_hidden_size": 3584,
    "vit_layers": [
      -3,
      -9
    ]
  },
  "architectures": [
    "MolmoActForActionReasoning"
  ],
  "auto_map": {
    "AutoConfig": "configuration_molmoact.MolmoActConfig",
    "AutoModelForImageTextToText": "modeling_molmoact.MolmoActForActionReasoning"
  },
  "image_patch_id": 152066,
  "initializer_range": 0.02,
  "llm_config": {
    "additional_vocab_size": 128,
    "attention_dropout": 0.0,
    "embedding_dropout": 0.0,
    "head_dim": 128,
    "hidden_act": "silu",
    "hidden_size": 3584,
    "initializer_range": 0.02,
    "intermediate_size": 18944,
    "layer_norm_eps": 1e-06,
    "max_position_embeddings": 4096,
    "model_type": "molmoact_llm",
    "norm_after": false,
    "num_attention_heads": 28,
    "num_hidden_layers": 28,
    "num_key_value_heads": 4,
    "qk_norm_type": "olmo",
    "qkv_bias": true,
    "residual_dropout": 0.0,
    "rope_scaling": null,
    "rope_theta": 1000000.0,
    "use_cache": true,
    "use_qk_norm": false,
    "vocab_size": 152064
  },
  "model_type": "molmoact",
  "n_action_bins": 256,
  "norm_stats": {
    "fractal20220817_data": {
      "action": {
        "mask": [
          true,
          true,
          true,
          true,
          true,
          true,
          false
        ],
        "max": [
          2.9984593391418457,
          22.09052848815918,
          2.7507524490356445,
          1.570636510848999,
          1.5321086645126343,
          1.5691522359848022,
          1.0
        ],
        "mean": [
          0.006987582892179489,
          0.006265917327255011,
          -0.01262515690177679,
          0.04333311319351196,
          -0.005756212864071131,
          0.0009130256366916001,
          0.5354204773902893
        ],
        "min": [
          -2.0204520225524902,
          -5.497899532318115,
          -2.031663417816162,
          -1.569917917251587,
          -1.569892168045044,
          -1.570419430732727,
          0.0
        ],
        "q01": [
          -0.22453527510166169,
          -0.14820013284683228,
          -0.231589707583189,
          -0.3517994859814644,
          -0.4193011274933815,
          -0.43643461108207704,
          0.0
        ],
        "q99": [
          0.17824687153100965,
          0.14938379630446405,
          0.21842354819178575,
          0.5892666035890578,
          0.35272657424211445,
          0.44796681255102094,
          1.0
        ],
        "std": [
          0.0692116990685463,
          0.05970962345600128,
          0.07353084534406662,
          0.15610496699810028,
          0.13164450228214264,
          0.14593800902366638,
          0.497110515832901
        ]
      },
      "num_trajectories": 87212,
      "num_transitions": 3786400,
      "proprio": {
        "max": [
          0.0,
          0.0,
          0.0,
          0.0,
          0.0,
          0.0,
          0.0
        ],
        "mean": [
          0.0,
          0.0,
          0.0,
          0.0,
          0.0,
          0.0,
          0.0
        ],
        "min": [
          0.0,
          0.0,
          0.0,
          0.0,
          0.0,
          0.0,
          0.0
        ],
        "q01": [
          0.0,
          0.0,
          0.0,
          0.0,
          0.0,
          0.0,
          0.0
        ],
        "q99": [
          0.0,
          0.0,
          0.0,
          0.0,
          0.0,
          0.0,
          0.0
        ],
        "std": [
          0.0,
          0.0,
          0.0,
          0.0,
          0.0,
          0.0,
          0.0
        ]
      }
    }
  },
  "tie_word_embeddings": false,
  "torch_dtype": "bfloat16",
  "transformers_version": "4.52.3",
  "use_cache": true,
  "vit_config": {
    "attention_dropout": 0.0,
    "float32_attention": true,
    "head_dim": 72,
    "hidden_act": "gelu_pytorch_tanh",
    "hidden_size": 1152,
    "image_default_input_size": [
      378,
      378
    ],
    "image_num_pos": 729,
    "image_patch_size": 14,
    "initializer_range": 0.02,
    "intermediate_size": 4304,
    "layer_norm_eps": 1e-06,
    "model_type": "molmoact_vit",
    "num_attention_heads": 16,
    "num_hidden_layers": 27,
    "num_key_value_heads": 16,
    "patch_bias": true,
    "pre_layernorm": false,
    "residual_dropout": 0.0,
    "use_cls_token": false
  }
}