A Q6 K GGUF quantization of the MOSS-TTS v1.5 backbone (Qwen3-8B), packaged for the moss tts delay llama.cpp inference path instead of the full bf16 PyTorch weights — so MOSS-TTS fits comfortably on memory-constrained machines (e.g. 16–24 GB Apple Silicon). Contents backbone q6 k.gguf — the MOSS-TTS v1.5 backbone quantized to Q6 K , with the token-embedding tensor ( token embd ) kept at F16 ....
--- license: apache-2.0 base_model: OpenMOSS-Team/MOSS-TTS-v1.5 pipeline_tag: text-to-speech language: - en - zh tags: - text-to-speech - tts - gguf - llama.cpp - quantized library_name: gguf --- # MOSS-TTS v1.5 — Q6_K GGUF A **Q6_K GGUF quantization** of the [MOSS-TTS v1.5](https://huggingface.co/OpenMOSS-Team/MOSS-TTS-v1.5) backbone (Qwen3-8B), packaged for the `moss_tts_delay` **llama.cpp** inference path instead of the full bf16 PyTorch weights — so MOSS-TTS fits comfortably on memory-constrained machines (e.g. 16–24 GB Apple Silicon). ## Contents - **`backbone_q6_k.gguf`** — the MOSS-TTS v1.5 backbone quantized to **Q6_K**, with the **token-embedding tensor (`token_embd`) kept at F16**. This is deliberate and important: MOSS-TTS reads per-token text/audio embeddings directly from that table, and quantizing it below F16 degrades output. Quantize everything else, keep `token_embd` at F16 (`llama-quantize --token-embedding-type f16`). - **`extracted/`** — the npy embedding + LM-head tables and tokenizer that the `moss_tts_delay` llama.cpp pipeline loads alongside the GGUF (the audio/codec heads that don't live inside the backbone GGUF). ## Not included (get these from the official sources) - The **ONNX audio codec** (decoder/encoder): [`OpenMOSS-Team/MOSS-Audio-Tokenizer-ONNX`](https://huggingface.co/OpenMOSS-Team/MOSS-Audio-Tokenizer-ONNX). - The full **bf16 weights**: [`OpenMOSS-Team/MOSS-TTS-v1.5`](https://huggingface.co/OpenMOSS-Team/MOSS-TTS-v1.5). ## How it was built Convert the official v1.5 weights to F16 GGUF (`convert_hf_to_gguf.py`), then `llama-quantize --token-embedding-type f16 ... Q6_K`. On an internal speech-quality battery (word-error-rate, speaker-similarity, and mel-cepstral-distortion), **Q6_K matched Q8** while using less memory — making it a good default quant for running MOSS-TTS locally. ## License & credit **Apache-2.0**, inherited from MOSS-TTS v1.5 by the **OpenMOSS team**. This repository only re-packages their model as a quantized GGUF; all model credit belongs to OpenMOSS. Please cite and follow the upstream model's terms.
A Q6 K GGUF quantization of the MOSS-TTS v1.5 backbone (Qwen3-8B), packaged for the moss tts delay llama.cpp inference path instead of the full bf16 PyTorch weights — so MOSS-TTS fits comfortably on memory-constrained machines (e.g. 16–24 GB Apple Silicon).…
