ModelPublicadoExterno

Sauti Tts v2

This is a Swahili text-to-speech (TTS) model fine-tuned on top of the English Chatterbox Multilingual backbone using a LoRA adapter. It was developed by Msingi-AI as part of the Sauti TTS V2 project.

Fuentehuggingface.co/msingiai/sauti_tts_v2/tree/09de6d51288c5bf42931c082bcc9ebcbce68ab87

TipoLoRAVersiónadapter_modelCcalidad

Fuente del modelo

Tipo: LoRA
Tamaño del archivo: 365 MB
Versión: vadapter_model
Actualizado: 25 sept 2026
Descargas: 0
Fuente: Hugging Face

Descripción de la fuente

Sauti TTS V2 (Chatterbox Swahili LoRA)

This is a Swahili text-to-speech (TTS) model fine-tuned on top of the English Chatterbox Multilingual backbone using a LoRA adapter. It was developed by Msingi-AI as part of the Sauti TTS V2 project.

Model Details

Base Model: Chatterbox Multilingual (English backbone + multilingual grapheme tokenizer)
Training Method: LoRA fine-tuning
Language: Swahili (sw)
License: MIT

Training Data

The model was fine-tuned on Google's WaxalNLP swa_tts studio dataset (CC-BY-4.0). The dataset was filtered using a transcript-agreement check, dropping clips where an Automatic Speech Recognition (ASR) judge detected a Character Error Rate (CER) > 10% against the reference text.

Evaluation Results

The model was evaluated against a held-out evaluation set of 48 Swahili sentences across various stress categories (general, code-switching, named entities, numbers/dates). Intelligibility is measured by Word Error Rate (WER) and Character Error Rate (CER) under two ASR judges: zero-shot openai/whisper-large-v3 and a Swahili-fine-tuned Whisper (Jacaranda-Health/ASR-STT). Swahili's agglutinative morphology tends to inflate WER, making CER the headline metric. UTMOS is used as a proxy for audio quality.

Scope	openai/whisper-large-v3 WER / CER	Jacaranda-Health/ASR-STT WER / CER	UTMOS
Overall	0.388 / 0.088	0.139 / 0.038	3.887
Overall (Plain Swahili)	0.384 / 0.086	0.092 / 0.022	3.871
Code Switch	0.400 / 0.094	0.279 / 0.084	3.934
General	0.320 / 0.069	0.148 / 0.036	3.872
Named Entities	0.421 / 0.094	0.027 / 0.005	3.897
Numbers & Dates	0.411 / 0.095	0.100 / 0.026	3.844

Usage

The weights provided here are the LoRA adapter (new_lang_adapter directory). They must be loaded on top of the original Chatterbox backbone using the Chatterbox fine-tuning toolkit.

Fuentes

1 fuente

Hugging FacePrincipalVersión 09de6d51288c5bf42931c082bcc9ebcbce68ab87

huggingface.co/msingiai/sauti_tts_v2/tree/09de6d51288c5bf42931c082bcc9ebcbce68ab87

Verificado 6 sept

Artefactos del modelo

1 artefacto

new_lang_adapter/adapter_model.safetensors

safetensors · 365 MB · SHA-256 19a916c36388…c791 · Hugging Face

Download file

Extractos de fuentes

2 extractos

Readme6 septhuggingface.co/msingiai/sauti_tts_v2

--- language: sw tags: - text-to-speech - tts - swahili - chatterbox - lora license: mit datasets: - google/WaxalNLP --- # Sauti TTS V2 (Chatterbox Swahili LoRA) This is a Swahili text-to-speech (TTS) model fine-tuned on top of the English Chatterbox Multilingual backbone using a LoRA adapter. It was developed by **Msingi-AI** as part of the Sauti TTS V2 project. ## Model Details * **Base Model**: Chatterbox Multilingual (English backbone + multilingual grapheme tokenizer) * **Training Method**: LoRA fine-tuning * **Language**: Swahili (`sw`) * **License**: MIT ## Training Data The model was fine-tuned on Google's [WaxalNLP `swa_tts`](https://huggingface.co/datasets/google/WaxalNLP) studio dataset (CC-BY-4.0). The dataset was filtered using a transcript-agreement check, dropping clips where an Automatic Speech Recognition (ASR) judge detected a Character Error Rate (CER) > 10% against the reference text. ## Evaluation Results The model was evaluated against a held-out evaluation set of 48 Swahili sentences across various stress categories (general, code-switching, named entities, numbers/dates). Intelligibility is measured by Word Error Rate (WER) and Character Error Rate (CER) under two ASR judges: zero-shot `openai/whisper-large-v3` and a Swahili-fine-tuned Whisper (`Jacaranda-Health/ASR-STT`). Swahili's agglutinative morphology tends to inflate WER, making CER the headline metric. UTMOS is used as a proxy for audio quality. | Scope | openai/whisper-large-v3 WER / CER | Jacaranda-Health/ASR-STT WER / CER | UTMOS | |---|---|---|---| | **Overall** | 0.388 / 0.088 | 0.139 / 0.038 | 3.887 | | **Overall (Plain Swahili)** | 0.384 / 0.086 | **0.092 / 0.022** | 3.871 | | **Code Switch** | 0.400 / 0.094 | 0.279 / 0.084 | 3.934 | | **General** | 0.320 / 0.069 | 0.148 / 0.036 | 3.872 | | **Named Entities** | 0.421 / 0.094 | 0.027 / 0.005 | 3.897 | | **Numbers & Dates** | 0.411 / 0.095 | 0.100 / 0.026 | 3.844 | ## Usage The weights provided here are the LoRA adapter (`new_lang_adapter` directory). They must be loaded on top of the original Chatterbox backbone using the [Chatterbox fine-tuning toolkit](https://github.com/gokhaneraslan/chatterbox-finetuning).

Note6 septhuggingface.co/msingiai/sauti_tts_v2

Source context: 0 downloads · 2 likes · Pipeline text-to-speech · Library chatterbox · Repo msingiai/sauti_tts_v2