{
  "model_type": "hunyuan_ocr",
  "src": "C:\\hunyuan\\modle",
  "hidden_size": 1024,
  "num_hidden_layers": 24,
  "num_attention_heads": 16,
  "num_key_value_heads": 8,
  "head_dim": 128,
  "vocab_size": 120818,
  "rope_theta": 10000.0,
  "max_position_embeddings": 32768,
  "intermediate_size": 3584,
  "static_seq_len": 512,
  "params": {
    "vision_encoder_param": "vision_encoder.ncnn.param",
    "vision_encoder_bin": "vision_encoder.ncnn.bin",
    "projector_param": "projector.ncnn.param",
    "projector_bin": "projector.ncnn.bin",
    "embed_param": "embed.ncnn.param",
    "embed_bin": "embed.ncnn.bin",
    "decoder_param": "decoder.ncnn.param",
    "decoder_bin": "decoder.ncnn.bin",
    "lm_head_param": "lm_head.ncnn.param",
    "lm_head_bin": "lm_head.ncnn.bin",
    "perceptron_weights": "perceptron_weights.bin"
  },
  "tokenizer": {
    "vocab_file": "vocab.txt",
    "merges_file": "merges.txt",
    "bos": "<｜hy_begin▁of▁sentence｜>",
    "eos": "<｜hy_place▁holder▁no▁102｜>",
    "pad": "<｜hy_▁pad▁｜>",
    "unk": "",
    "cls": "",
    "sep": "",
    "mask": ""
  },
  "setting": {
    "vision": {}
  }
}