ModelListedExternal

Moss Nano Ptbr 4voices

MOSS-TTS-Nano (100M, CPU) fine-tunado em 4 vozes PT-BR (3M+1F, refs reais CML-TTS, 5k clips/voz via Qwen3-TTS-1.7B clone, 3 epocas). WER por voz: 05=0.10, 07=0.08, 11=0.05, 12=0.03 (Whisper-large-v3). Modo...

Sourcehuggingface.co/marcosremar2/moss-nano-ptbr-4voices/tree/6e9a04bed84c8f061b7d345303dc2994322abc3e

KindText modelBase modelOpenMOSS-Team/MOSS-TTS-NanoVersionpytorch_model

Model source

Kind: Text model
File size: 272 MB
Base model: OpenMOSS-Team/MOSS-TTS-Nano
Version: vpytorch_model
Updated: Sep 19, 2026
Downloads: 53
Tasks: Audio
Source: Hugging Face

Source description

moss-nano-ptbr-4voices

MOSS-TTS-Nano (100M, CPU) fine-tunado em 4 vozes PT-BR (3M+1F, refs reais CML-TTS, ~5k clips/voz via Qwen3-TTS-1.7B clone, 3 epocas). WER por voz: 05=0.10, 07=0.08, 11=0.05, 12=0.03 (Whisper-large-v3). Modo voice_clone: passe o audio de referencia da voz desejada.

Sources

1 source

Hugging FacePrimaryVersion 6e9a04bed84c8f061b7d345303dc2994322abc3e

huggingface.co/marcosremar2/moss-nano-ptbr-4voices/tree/6e9a04bed84c8f061b7d345303dc2994322abc3e

Verified Sep 19

Model artifacts

1 artifact

pytorch_model.bin

bin · 272 MB · SHA-256 ffe0e9cfe617…18ae · Hugging Face

Download

Source excerpts

2 excerpts

ReadmeSep 19huggingface.co/marcosremar2/moss-nano-ptbr-4voices

--- license: apache-2.0 language: - pt pipeline_tag: text-to-speech base_model: OpenMOSS-Team/MOSS-TTS-Nano tags: - tts - portuguese --- # moss-nano-ptbr-4voices MOSS-TTS-Nano (100M, CPU) fine-tunado em 4 vozes PT-BR (3M+1F, refs reais CML-TTS, ~5k clips/voz via Qwen3-TTS-1.7B clone, 3 epocas). **WER por voz: 05=0.10, 07=0.08, 11=0.05, 12=0.03** (Whisper-large-v3). Modo voice_clone: passe o audio de referencia da voz desejada.

NoteSep 19huggingface.co/marcosremar2/moss-nano-ptbr-4voices

Source context: 53 downloads · 0 likes · Pipeline text-to-speech · Repo marcosremar2/moss-nano-ptbr-4voices