{
  "family": "citrinet_asr",
  "display_name": "Citrinet ASR",
  "description": "NVIDIA CitriNet-family end-to-end English ASR model using a convolutional CTC architecture optimized for transcribing speech segments to text.",
  "category": "asr",
  "status": "supported",
  "tasks": [
    "asr"
  ],
  "modes": [
    "offline"
  ],
  "languages": [
    "en"
  ],
  "capabilities": {},
  "options": {
    "request": [],
    "session": [
      {
        "name": "weight_type",
        "type": "enum",
        "description": "Citrinet ASR weight storage type; default native.",
        "preset": "weight_type_full",
        "required": false,
        "default": "native"
      }
    ],
    "load": []
  },
  "runtime": {
    "tags": [
      "gguf"
    ]
  },
  "sources": [
    {
      "format": "gguf",
      "roots": {
        "model": ".",
        "weights": "$gguf"
      },
      "files": {
        "config": "model:citrinet_256_config.json",
        "tokenizer": "model:citrinet_256_tokenizer.model"
      },
      "tensors": {
        "weights": "weights:"
      }
    },
    {
      "format": "safetensors",
      "roots": {
        "model": "."
      },
      "files": {
        "config": "model:citrinet_256_config.json",
        "tokenizer": "model:citrinet_256_tokenizer.model"
      },
      "tensors": {
        "weights": "model:citrinet_256.safetensors"
      }
    }
  ],
  "package_defaults": {
    "download": {
      "kind": "huggingface_snapshot",
      "repo": "audio-cpp/audio.cpp-gguf",
      "revision": "main",
      "gated": false
    }
  },
  "packages": [
    {
      "id": "citrinet_asr_q8_0",
      "display_name": "Citrinet ASR Q8_0 GGUF",
      "default": true,
      "format": "gguf",
      "precision": "q8_0",
      "target_directory": "Citrinet-ASR-GGUF",
      "files": [
        "Citrinet-ASR-GGUF/citrinet-asr-q8_0.gguf"
      ],
      "strip_prefix": "Citrinet-ASR-GGUF"
    }
  ]
}