Skip to content
ModelListedExternal
Download

Cey 25 Mms Tts Mozilla Kdh

Fine-tuned facebook/mms-tts-kdh on Mozilla Data Collective data for Tem / Kotokoli (kdh).

Sourcehuggingface.co/Umbaji/cey-25-mms-tts-mozilla-kdh/tree/5a81ed7a8ee3cdbd5b75bca829742c0a4fd760e5
KindOtherBase modelfacebook/mms-tts-kdhVersionRevision 5a81ed7

Model source

Kind
Other
File size
139 MB
Base model
facebook/mms-tts-kdh
Version
vRevision 5a81ed7
Updated
Sep 21, 2026
Downloads
11
Tasks
Audio
Source
Hugging Face

Source description

MMS-TTS Tem / Kotokoli — Mozilla Fine-tuned

Fine-tuned facebook/mms-tts-kdh on Mozilla Data Collective data for Tem / Kotokoli (kdh).

Training Statistics

MetricValue
Training samples13110
Validation samples2314
Best val mel-L13.4433
Dataset sourceMozilla Data Collective

Usage

from transformers import VitsModel, VitsTokenizer
import torch, torchaudio

model = VitsModel.from_pretrained("Umbaji/cey-25-mms-tts-mozilla-kdh")
tokenizer = VitsTokenizer.from_pretrained("Umbaji/cey-25-mms-tts-mozilla-kdh")

inputs = tokenizer("your text here", return_tensors="pt")
with torch.no_grad():
    waveform = model(**inputs).waveform[0]

torchaudio.save("out.wav", waveform.unsqueeze(0), model.config.sampling_rate)

Training

  • Loss: Mel-spectrogram L1
  • Optimizer: AdamW (lr=2e-4, betas=(0.8, 0.99))
  • Scheduler: ExponentialLR γ=0.999
  • Epochs: 1 | Effective batch: 16

Fine-tuned: 2026-06-12 — Mozilla Data Collective + Eyaa-Tom

Sources

1 source
Hugging FacePrimaryVersion 5a81ed7a8ee3cdbd5b75bca829742c0a4fd760e5
huggingface.co/Umbaji/cey-25-mms-tts-mozilla-kdh/tree/5a81ed7a8ee3cdbd5b75bca829742c0a4fd760e5

Verified Sep 21

Model artifacts

1 artifact

model.safetensors

safetensors · 139 MB · SHA-256 039dca955598…36c8 · Hugging Face

Download file

Source excerpts

2 excerpts
ReadmeSep 21huggingface.co/Umbaji/cey-25-mms-tts-mozilla-kdh

--- language: - kdh tags: - text-to-speech - mms-tts - vits - kdh - west-africa - mozilla-data-collective - finetuned license: cc-by-nc-4.0 base_model: facebook/mms-tts-kdh pipeline_tag: text-to-speech --- # MMS-TTS Tem / Kotokoli — Mozilla Fine-tuned Fine-tuned [facebook/mms-tts-kdh](https://huggingface.co/facebook/mms-tts-kdh) on Mozilla Data Collective data for **Tem / Kotokoli** (`kdh`). ## Training Statistics | Metric | Value | |--------|-------| | Training samples | 13110 | | Validation samples | 2314 | | Best val mel-L1 | 3.4433 | | Dataset source | Mozilla Data Collective | ## Usage ```python from transformers import VitsModel, VitsTokenizer import torch, torchaudio model = VitsModel.from_pretrained("Umbaji/cey-25-mms-tts-mozilla-kdh") tokenizer = VitsTokenizer.from_pretrained("Umbaji/cey-25-mms-tts-mozilla-kdh") inputs = tokenizer("your text here", return_tensors="pt") with torch.no_grad(): waveform = model(**inputs).waveform[0] torchaudio.save("out.wav", waveform.unsqueeze(0), model.config.sampling_rate) ``` ## Training - **Loss**: Mel-spectrogram L1 - **Optimizer**: AdamW (lr=2e-4, betas=(0.8, 0.99)) - **Scheduler**: ExponentialLR γ=0.999 - **Epochs**: 1 | **Effective batch**: 16 *Fine-tuned: 2026-06-12 — Mozilla Data Collective + Eyaa-Tom*

NoteSep 21huggingface.co/Umbaji/cey-25-mms-tts-mozilla-kdh

Source context: 11 downloads · 0 likes · Pipeline text-to-speech · Repo Umbaji/cey-25-mms-tts-mozilla-kdh