KevinAHM/pocket-tts-onnx

Pocket TTS ONNX

ONNX export of Pocket TTS for lightweight text-to-speech with zero-shot voice cloning.

Model Description

Pocket TTS is a compact text-to-speech model from Kyutai that supports voice cloning from short audio samples. This ONNX export provides:

• Zero-shot voice cloning from any audio reference
• INT8 quantized models for fast CPU inference
• Streaming support with adaptive chunking for real-time playback
• Temperature control for generation diversity
• Dual model architecture for flexible flow matching

Architecture

This export uses a dual model split for the Flow LM:

flow_lm_main   - Transformer/conditioner (produces conditioning vectors)
flow_lm_flow   - Flow network only (Euler integration for latent sampling)

This architecture enables:

• Temperature control: Adjust generation diversity via noise scaling
• Variable LSD steps: Trade off speed vs quality
• External flow loop: Full control over the sampling process

Performance

Benchmarked on a 16-core CPU, lsd_steps=1, measuring flow loop + mimi decoder (excluding voice encoder).

Precision	Flow Main	Flow Net	Decoder	Total Size	RTFx (CPU)
INT8	76 MB	10 MB	23 MB	~200 MB	~4.0x
FP32	303 MB	39 MB	42 MB	~475 MB	~2.8x
PyTorch FP32 (reference)	—	—	—	—	~4.0x

RTFx = Real-time factor (>1.0 means faster than real-time). ONNX INT8 matches PyTorch FP32 throughput.

Optimized Inference

Thread count is automatically tuned to avoid over-subscription on the small sequential matmuls in the autoregressive loop (intra_op_num_threads=min(cpu_count, 4), inter_op_num_threads=1). This alone provides a ~2x speedup over default ORT settings on multi-core machines.

Decoding Strategy

• Offline (generate): Uses threaded parallel decoding — the mimi decoder runs in a background thread, decoding 12-frame chunks while the flow loop generates the next frames. This overlaps generation and decoding for maximum throughput.
• Streaming (stream): Uses adaptive chunking (starts at 2 frames). This ensures instant start (low TTFB) while scaling up chunk sizes for throughput.

Usage

from pocket_tts_onnx import PocketTTSOnnx

# Load model (INT8 by default)
tts = PocketTTSOnnx()

# Generate speech with voice cloning
audio = tts.generate(
    text="Hello, this is a test of voice cloning.",
    voice="reference_sample.wav"
)

# Save output
tts.save_audio(audio, "output.wav")

Temperature Control

Adjust generation diversity with the temperature parameter:

# More deterministic (lower temperature)
tts = PocketTTSOnnx(temperature=0.3)

# Default balance
tts = PocketTTSOnnx(temperature=0.7)

# More diverse/expressive (higher temperature)
tts = PocketTTSOnnx(temperature=1.0)

LSD Steps

Trade off speed vs quality with lsd_steps:

# Default (10 steps)
tts = PocketTTSOnnx(lsd_steps=10)

# Faster (fewer steps, lower quality)
tts = PocketTTSOnnx(lsd_steps=1)

Streaming Mode

For real-time applications with low time-to-first-audio:

for chunk in tts.stream("Hello world!", voice="reference_sample.wav"):
    play_audio(chunk)  # Process each chunk as it arrives

Command Line

python generate.py "Hello, this is a test." reference_sample.wav output.wav

Files

pocket-tts-onnx/
├── onnx/
│   ├── flow_lm_main.onnx          # 303 MB - Flow LM transformer (FP32)
│   ├── flow_lm_main_int8.onnx     #  76 MB - Flow LM transformer (INT8)
│   ├── flow_lm_flow.onnx          #  39 MB - Flow network (FP32)
│   ├── flow_lm_flow_int8.onnx     #  10 MB - Flow network (INT8)
│   ├── mimi_decoder.onnx          #  42 MB - Audio decoder (FP32)
│   ├── mimi_decoder_int8.onnx     #  23 MB - Audio decoder (INT8)
│   ├── mimi_encoder.onnx          #  73 MB - Voice encoder
│   └── text_conditioner.onnx      #  16 MB - Text embeddings
├── reference_sample.wav           # Example voice reference
├── tokenizer.model                # SentencePiece tokenizer
├── pocket_tts_onnx.py             # Inference wrapper
├── generate.py                    # CLI script
├── requirements.txt               # Python dependencies
└── README.md

Requirements

onnxruntime>=1.16.0
numpy
soundfile
sentencepiece
scipy  # Only needed if resampling from non-24kHz audio

Install with:

pip install -r requirements.txt

License

• Models: CC BY 4.0 (inherited from kyutai/pocket-tts)
• Code: Apache 2.0

Prohibited Use

Use of our model must comply with all applicable laws and regulations and must not result in, involve, or facilitate any illegal, harmful, deceptive, fraudulent, or unauthorized activity. Prohibited uses include, without limitation, voice impersonation or cloning without explicit and lawful consent; misinformation, disinformation, or deception (including fake news, fraudulent calls, or presenting generated content as genuine recordings of real people or events); and the generation of unlawful, harmful, libelous, abusive, harassing, discriminatory, hateful, or privacy-invasive content. We disclaim all liability for any non-compliant use.

Acknowledgments

• Kyutai for the original Pocket TTS model