You can deploy this model and start earning money today!

ravenxone/
mlx-tada-3b

Name: mlx-tada-3b
Availability: InStock

meta-llama/Llama-3.2-3B

finetune

llama3.2

Views

No views yet

TADA: A Generative Framework for Speech Modeling via Text-Acoustic Dual Alignment

Paper

PyPI

Blog

A unified speech-language model that synchronizes speech and text into a single, cohesive stream via 1:1 alignment.

MLX-TADA-3B

Pre-converted MLX weights for TADA (Text-Acoustic Dual Alignment) speech synthesis on Apple Silicon.

Built on Llama 3.2 3B. Multilingual: English, French, German, Spanish, Portuguese, Italian, Polish, Japanese, Chinese, Arabic.

Component	File	Size
LLM + VibeVoice head	`model/weights.safetensors`	7.6 GB
Aligner	`aligner/weights.safetensors`	852 MB
Decoder (DAC)	`decoder/weights.safetensors`	226 MB
Encoder	`encoder/weights.safetensors`	178 MB
Total		~8.9 GB

All weights are stored in bfloat16 safetensors format.

Prerequisites

TADA models are built on Meta Llama 3.2. You must request access to the Llama models before using TADA:

Visit meta-llama/Llama-3.2-1B or meta-llama/Llama-3.2-3B and accept the license agreement

Quick Start

pip install mlx-tada

Or install from source:

bash

git clone https://github.com/HumeAI/tada.git
cd tada/apple
uv venv && uv pip install -e .

Download a reference audio clip:

curl -O "https://storage.googleapis.com/hume_reference_speakers/ljspeech.wav"

Python

python

from mlx_tada import TadaForCausalLM, save_wav

model = TadaForCausalLM.from_pretrained("HumeAI/mlx-tada-3b", quantize=4)
ref = model.load_reference("ljspeech.wav")
out = model.generate("Hello, this is a test of TADA speech synthesis.", ref)
save_wav(out.audio, "output.wav")

Offline Use

To download the weights locally for offline inference:

python

from huggingface_hub import snapshot_download
snapshot_download("HumeAI/mlx-tada-3b", local_dir="./weights/3b")

Then load from the local path:

model = TadaForCausalLM.from_weights("./weights/3b", quantize=4)

CLI

bash

python -m mlx_tada.generate \
  --weights ./weights/3b \
  --audio ljspeech.wav \
  --text "Hello, this is a test of TADA speech synthesis." \
  --quantize 4 \
  --output output.wav

Hardware Requirements

Precision	Memory
bfloat16 (default)	~17 GB
4-bit quantized	~6 GB

4-bit quantization is recommended for most Apple Silicon devices — it is roughly 10x faster with 60% less memory and minimal quality loss. Machines with 16 GB unified memory should use quantized mode.

TADA GitHub — source code, PyTorch inference, training
TADA Paper — arxiv
HumeAI/tada-3b-ml — PyTorch weights
HumeAI/mlx-tada-1b — 1B English-only MLX weights
HumeAI/tada-codec — shared encoder, decoder, aligner weights

📚 Citation

If you use this project in your research, please cite our paper:

bibtex

@article{dang2026tada,
  title={TADA: A Generative Framework for Speech Modeling via Text-Acoustic Dual Alignment},
  author={Dang, Trung and Rao, Sharath and Gupta, Ananya and Gagne, Christopher and Tzirakis, Panagiotis and Baird, Alice and Cłapa, Jakub Piotr and Chin, Peter and Cowen, Alan},
  journal={arXiv preprint arXiv:2602.23068},
  year={2026}
}

Contact

Hume AI is an empathic AI research company. We research the datasets, tools, and models needed to give empathy to AI models to serve human wellbeing. If you're interested in any of our product or research collaborations, please reach out to us at hello@hume.ai

Acknowledgements

This project is built using Llama 3.2.

Llama 3.2 is licensed under the Llama 3.2 Community License

ravenxone/mlx-tada-3b