{
  "model_cfg": {
    "model_cfg": {
      "embed_dim": 1024,
      "vision_cfg": {
        "image_size": 336,
        "layers": 1,
        "width": 336,
        "patch_size": 14
      },
      "text_cfg": {
        "context_length": 512,
        "vocab_size": 501153,
        "layers": 1
      }
    }
  },
  "preprocess_cfg": {
    "mean": [
      0.48145466,
      0.4578275,
      0.40821073
    ],
    "std": [
      0.26862954,
      0.26130258,
      0.27577711
    ]
  }
}