---
viewer: false
---

# MOSS-TTS model payload

This dataset was generated by `MossTtsHFModelsUploader` for `PHRunner.Kaggle.Service.MossTtsGguf`.

## Source

- Hugging Face repository: `niobures/MOSS-TTS-v1.5`
- Revision: `main`
- Kaggle dataset: `kaggle-pool-account/moss-tts-v1-5-python-models`
- Layout: `hf-transformers`
- Required backend: `hf`
- Default model file: `model.safetensors.index.json`

## Files

- `README.md` - support - 13.42 KiB - SHA256 `20581413205fbc34a6cdee97cbbfa8c1564fe1569a0a76e0ed466d3b21fff30b`
- `__init__.py` - support - 92 B - SHA256 `9786b23774caee0a13cd8bd8f2e61527bc240fa8712599de6ed0d711ee7a3902`
- `added_tokens.json` - support - 704 B - SHA256 `9dfb40d3a3f8654c8af17771d767a494e973b9a33d5c3aeffa1b21f77db8e6d3`
- `chat_template.jinja` - support - 352 B - SHA256 `894e3ace884e212ae34e1f46063686e7ebcc8b2fa0153af4579dfa5685f63184`
- `config.json` - support - 2.22 KiB - SHA256 `214fc997d98f51ab57925a5939afc6280e76044198b664221622e70d098ed06e`
- `configuration_moss_tts.py` - support - 5.45 KiB - SHA256 `257842f2fa90124307408f36d55b74e135ee32b7afacbe30280964f7feb329d1`
- `inference_utils.py` - support - 4.97 KiB - SHA256 `12cdf3260114e44c8cba816782f743779ab621e0ba4047e59a3f3579cfcb0e26`
- `merges.txt` - support - 1.59 MiB - SHA256 `8831e4f1a044471340f7c0a83d7bd71306a5b867e95fd870f74d0c5308a904d5`
- `model-00001-of-00004.safetensors` - support - 4.59 GiB - SHA256 `749b82d07cebea77dd15af365e2efeb1dc0dd54537f7010795fdf42f520743a9`
- `model-00002-of-00004.safetensors` - support - 4.58 GiB - SHA256 `4d891c8adb8bf9c135e8a44cd9f003f1dbf8b09278ced48d5af1fe7d989efb39`
- `model-00003-of-00004.safetensors` - support - 4.64 GiB - SHA256 `111013e05174e67748443eafdd4d64c051292d0e5fce2cac23a16988b661eea1`
- `model-00004-of-00004.safetensors` - support - 2 GiB - SHA256 `951ee88cf85996ee5c39441ace6e50f52894adad5da42ad2314b9a9db7f940b8`
- `model.safetensors.index.json` - support - 39.47 KiB - SHA256 `021ef3f74a92ee33fc076dde096d4fa9b7b45c34c9bdbfe730bf79368ac86bb2`
- `modeling_moss_tts.py` - support - 24.74 KiB - SHA256 `360ceee27c67831d820ff5324f8687f4408db1a91aef53683934544879aa90ca`
- `processing_moss_tts.py` - support - 34.7 KiB - SHA256 `e591ff156e8fadcae34db5613a2221fab7ec5d814bfc0ec001e8ec4ef2782425`
- `processor_config.json` - support - 145 B - SHA256 `6821d4805f10fb8b5cc743c021215097e03f1a25f7c220dfc4fb0d62b5dc97e0`
- `special_tokens_map.json` - support - 631 B - SHA256 `55cbd0665670e196ce105db3f0d787e5eb35bce235fcb2bd50acbdd60c1e39c6`
- `tokenizer.json` - support - 10.89 MiB - SHA256 `cb3c8fa82993d515469c2800cc455bff4aaa3c4fed9da1f2b0c0668c304f335a`
- `tokenizer_config.json` - support - 5.37 KiB - SHA256 `87b1f37a4a73e4c4b61637a1446763e8bdea274347f0418cd5c68ead488c8ea4`
- `tts_robust_normalizer_single_script.py` - support - 14.2 KiB - SHA256 `b7c1fab4fd7bdc8c779df99f6af310fd8cd0746b759db684bf147e0c182b3183`
- `vocab.json` - support - 2.65 MiB - SHA256 `ca10d7e9fb3ed18575dd1e277a2579c16d108e32f27439684afa0e10b1440910`

## Compatibility contract

This payload uses the Python llama.cpp + ONNX-tokenizer layout. The companion assets dataset must be generated from the same OpenMOSS source family, normally with:

```text
MossTtsAssetsUploader --gguf-repo niobures/MOSS-TTS-GGUF --audio-tokenizer-repo niobures/MOSS-Audio-Tokenizer-ONNX
```