{
  "family": "moss_tts_local",
  "display_name": "MOSS-TTS-Local",
  "description": "Flagship MOSS-TTS model for high-fidelity 31-language and code-switched speech, zero-shot voice cloning, long-form generation, and fine-grained Pinyin, phoneme, and duration control.",
  "category": "tts",
  "status": "supported",
  "tasks": [
    "tts",
    "clone"
  ],
  "modes": [
    "offline"
  ],
  "languages": [
    "ar",
    "cs",
    "da",
    "de",
    "el",
    "en",
    "es",
    "fa",
    "fi",
    "fr",
    "he",
    "hi",
    "hu",
    "it",
    "ja",
    "ko",
    "mk",
    "ms",
    "nl",
    "pl",
    "pt",
    "ro",
    "ru",
    "sv",
    "sw",
    "th",
    "tl",
    "tr",
    "vi",
    "yue",
    "zh"
  ],
  "capabilities": {
    "clone": [
      "speaker_reference"
    ]
  },
  "options": {
    "request": [
      {
        "name": "max_tokens",
        "type": "int",
        "description": "Maximum generated audio frames; default 4096.",
        "required": false,
        "min": 0,
        "default": 4096
      },
      {
        "name": "do_sample",
        "type": "bool",
        "description": "Enable stochastic audio-token sampling; default true.",
        "required": false,
        "default": true
      },
      {
        "name": "temperature",
        "type": "float",
        "description": "Audio-token sampling temperature; default 1.7.",
        "required": false,
        "min": 0.0,
        "default": 1.7
      },
      {
        "name": "top_p",
        "type": "float",
        "description": "Audio-token nucleus sampling limit; default 0.8.",
        "required": false,
        "min": 0.0,
        "max": 1.0,
        "default": 0.8
      },
      {
        "name": "top_k",
        "type": "int",
        "description": "Audio-token top-k sampling limit; default 25.",
        "required": false,
        "min": 0,
        "default": 25
      },
      {
        "name": "repetition_penalty",
        "type": "float",
        "description": "Audio-token repetition penalty; default 1.0.",
        "required": false,
        "min": 0.0,
        "default": 1.0
      },
      {
        "name": "seed",
        "type": "int",
        "description": "Sampling seed; omitted value uses a random seed.",
        "required": false,
        "min": 0
      },
      {
        "name": "text_chunk_size",
        "type": "int",
        "description": "Long-form text chunk size in Unicode codepoints; default 2048.",
        "required": false,
        "min": 1,
        "default": 2048
      },
      {
        "name": "text_temperature",
        "type": "float",
        "description": "Text-gate sampling temperature; default 1.0 for Local and 1.5 for Nano.",
        "required": false,
        "default": 1.0
      },
      {
        "name": "text_top_p",
        "type": "float",
        "description": "Text-gate nucleus sampling limit; default 1.0.",
        "required": false,
        "default": 1.0
      },
      {
        "name": "text_top_k",
        "type": "int",
        "description": "Text-gate top-k sampling limit; default 50.",
        "required": false,
        "default": 50
      },
      {
        "name": "text_chunk_mode",
        "type": "enum",
        "description": "Long-form text chunking mode; default word_budget. Text chunk size defaults to 2048 Unicode codepoints.",
        "preset": "text_chunk_mode_full",
        "required": false,
        "default": "word_budget"
      }
    ],
    "session": [
      {
        "name": "weight_type",
        "type": "enum",
        "description": "Backbone weight storage type.",
        "values": [
          "auto",
          "native",
          "f32",
          "f16",
          "bf16",
          "q8_0"
        ],
        "required": false,
        "default": "auto"
      },
      {
        "name": "reference_cache_slots",
        "type": "int",
        "description": "Prepared reference-voice cache slots; default 1, set 0 to disable.",
        "required": false,
        "min": 0,
        "default": 1
      }
    ],
    "load": []
  },
  "runtime": {
    "tags": [
      "gguf"
    ]
  },
  "sources": [
    {
      "format": "gguf",
      "roots": {
        "model": ".",
        "weights": "$gguf"
      },
      "files": {
        "config": "model:config.json",
        "tokenizer_json": "model:tokenizer.json",
        "tokenizer_config": "model:tokenizer_config.json",
        "tokenizer_vocab": "model:vocab.json",
        "tokenizer_merges": "model:merges.txt",
        "audio_tokenizer_config": "model:audio_tokenizer/config.json"
      },
      "tensors": {
        "model_weights": {
          "source": "weights:",
          "prefix": "model_weights"
        },
        "audio_tokenizer_weights": {
          "source": "weights:",
          "prefix": "audio_tokenizer_weights"
        }
      }
    },
    {
      "format": "safetensors",
      "roots": {
        "model": ".",
        "audio_tokenizer": "audio_tokenizer"
      },
      "files": {
        "config": "model:config.json",
        "tokenizer_json": "model:tokenizer.json",
        "tokenizer_config": "model:tokenizer_config.json",
        "tokenizer_vocab": "model:vocab.json",
        "tokenizer_merges": "model:merges.txt",
        "audio_tokenizer_config": "audio_tokenizer:config.json"
      },
      "tensors": {
        "model_weights": "model:model.safetensors",
        "audio_tokenizer_weights": "audio_tokenizer:model.safetensors.index.json"
      }
    }
  ],
  "package_defaults": {
    "download": {
      "kind": "huggingface_snapshot",
      "repo": "audio-cpp/audio.cpp-gguf",
      "revision": "main",
      "gated": false
    }
  },
  "packages": [
    {
      "id": "moss_tts_local_v1_5_q8_0",
      "display_name": "MOSS-TTS-Local v1.5 Q8_0 GGUF",
      "default": true,
      "format": "gguf",
      "precision": "q8_0",
      "target_directory": "MOSS-TTS-Local-v1.5-GGUF",
      "files": [
        "MOSS-TTS-Local-v1.5-GGUF/moss-tts-local-v1.5-q8_0.gguf"
      ],
      "strip_prefix": "MOSS-TTS-Local-v1.5-GGUF"
    },
    {
      "id": "moss_tts_local_v1_5_bf16",
      "display_name": "MOSS-TTS-Local v1.5 BF16 GGUF",
      "format": "gguf",
      "precision": "bf16",
      "target_directory": "MOSS-TTS-Local-v1.5-GGUF",
      "files": [
        "MOSS-TTS-Local-v1.5-GGUF/moss-tts-local-v1.5-bf16.gguf"
      ],
      "strip_prefix": "MOSS-TTS-Local-v1.5-GGUF"
    }
  ]
}
