Hubi-TTS

Installation

pip install torch torchaudio transformers huggingface_hub

Inference

import torch
from transformers import AutoModel

device = "cuda" if torch.cuda.is_available() else "cpu"

# Load model
model = AutoModel.from_pretrained("ayeeyokoris/hubi-tts", trust_remote_code=True).to(device)

# 1. Standard TTS
model.tts("Salaam,sidee tahay!", save_path="output.wav")

# 2. Voice Cloning (with a reference audio prompt)
# model.tts("Speaking with a cloned voice.", prompt_audio="reference.wav", save_path="cloned.wav")

Citation

@misc{hubi_tts,
  author = {ayeeyokoris},
  title = {Hubi-TTS: Fast Text-to-Speech Somali Model},
  year = {2026},
  publisher = {Hugging Face},
  journal = {Hugging Face Hub},
  howpublished = {\url{https://huggingface.co/ayeeyokoris/hubi-tts}}
}