{
  "_name_or_path": "vsearch/vdr-cross-modal",
  "architectures": [
    "Retriever"
  ],
  "device": null,
  "encoder_p": {
    "heads": 12,
    "layers": 12,
    "patch_size": 32,
    "resolution": 224,
    "tokenizer_id": "bert-base-uncased",
    "topk": 768,
    "type": "vdr_crossmodal_image",
    "width": 768
  },
  "encoder_q": {
    "max_len": 512,
    "model_id": "bert-base-uncased",
    "topk": 768,
    "type": "vdr_crossmodal_text"
  },
  "shared_encoder": false,
  "torch_dtype": "float32",
  "transformers_version": "4.34.0",
  "type": null
}