Zum Inhalt springen
ModelAufgelistetExtern
Browse files

Pocket Tts Mnn

MNN conversion of IgnitiveLabs/PocketTTS-ONNX, itself an ONNX export of Kyutai's Pocket TTS (english2026-04).

Quellehuggingface.co/developerabu/pocket-tts-mnn/tree/4e75a38d20a9c532ed55591186584a244c381095
ArtOtherBasismodellIgnitiveLabs/PocketTTS-ONNXVersionFlow lm main int8CBewertung

Modellquelle

Art
Other
Dateigröße
38,0–72,8 MB
Basismodell
IgnitiveLabs/PocketTTS-ONNX
Version
vFlow lm main int8
Aktualisiert
26. Okt. 2026
Quelle
Hugging Face

Quellenbeschreibung

Pocket TTS MNN (English)

MNN conversion of IgnitiveLabs/PocketTTS-ONNX, itself an ONNX export of Kyutai's Pocket TTS (english_2026-04).

Weights use INT8 weight quantization via MNNConvert --weightQuantBits 8 (text conditioner kept FP32). Unsupported ONNX::IsNaN ops were rewritten to Not(Equal(x, x)) before conversion.

Files

mnn/
  flow_lm_main_w8.mnn          # ~73 MB
  flow_lm_flow_w8.mnn          # ~9.5 MB
  mimi_decoder_w8.mnn          # ~10 MB
  mimi_encoder_w8.mnn          # ~9.5 MB (converted; runtime uses ONNX fallback)
  text_conditioner_fp32.mnn    # ~16 MB (preferred)
  text_conditioner_w8.mnn      # ~16 MB
flow_lm_main_int8.onnx         # empty-seq priming fallback
mimi_encoder.onnx              # voice encode fallback
bundle.json
tokenizer.model
bos_before_voice.npy
pocket_tts_mnn.py
pocket_tts_onnx.py
reference_sample.wav

Hybrid runtime

MNN runs the generation hot path (frame loop, flow net, decoder, text conditioner).

ONNX is still required for:

  1. Empty-sequence priming (seq_len=0) — not supported by MNN on flow_lm_main
  2. mimi_encoder — MNN fails streaming Concat shape inference

Quick start

pip install MNN onnxruntime numpy soundfile sentencepiece scipy huggingface_hub safetensors

huggingface-cli download developerabu/pocket-tts-mnn --local-dir pocket-tts-mnn
cd pocket-tts-mnn
python generate_mnn.py "Hello, this is a test." reference_sample.wav out.wav

Or in Python:

from pocket_tts_mnn import PocketTTSMnn

tts = PocketTTSMnn(models_dir=".")
audio = tts.generate("Hello, this is a test.", voice="reference_sample.wav")
tts.save_audio(audio, "out.wav")

Benchmark (Apple M1, CPU, 5 runs)

BackendMedium RTFxStream TTFB
ONNX INT8~4.1×~100 ms
MNN INT8 (this repo)~2.4×~123 ms

MNN is mainly targeted at mobile; desktop Python numbers above are not the expected mobile ranking.

Attribution

License

Original model weights: CC-BY-4.0 (provide attribution to Kyutai Labs).
Conversion scripts and packaging in this repo: MIT.

Acceptable use

Same restrictions as Kyutai / IgnitiveLabs releases: no unauthorized voice cloning, no deceptive synthetic audio presented as real, no harmful/unlawful content, and disclose AI-generated speech to end users.

Quellen

1 Quelle
Hugging FacePrimärVersion 4e75a38d20a9c532ed55591186584a244c381095
huggingface.co/developerabu/pocket-tts-mnn/tree/4e75a38d20a9c532ed55591186584a244c381095

Verifiziert 26. Okt.

Download files

Modellartefakte

2 Artefakte

flow_lm_main_int8.onnx

onnx · 72,8 MB · SHA-256 f9bd8106b79a…0116 · Hugging Face

Download file

mimi_encoder.onnx

onnx · 38,0 MB · SHA-256 9e230b04c85a…f708 · Hugging Face

Download file

Quellenauszüge

2 Auszüge
Readme7. Aug.huggingface.co/developerabu/pocket-tts-mnn

--- license: cc-by-4.0 language: - en library_name: mnn base_model: - kyutai/pocket-tts - IgnitiveLabs/PocketTTS-ONNX pipeline_tag: text-to-speech tags: - tts - voice-cloning - mnn - onnx - int8 - pocket-tts --- # Pocket TTS MNN (English) MNN conversion of [IgnitiveLabs/PocketTTS-ONNX](https://huggingface.co/IgnitiveLabs/PocketTTS-ONNX), itself an ONNX export of Kyutai's [Pocket TTS](https://huggingface.co/kyutai/pocket-tts) (`english_2026-04`). Weights use **INT8 weight quantization** via `MNNConvert --weightQuantBits 8` (text conditioner kept FP32). Unsupported `ONNX::IsNaN` ops were rewritten to `Not(Equal(x, x))` before conversion. ## Files ``` mnn/ flow_lm_main_w8.mnn # ~73 MB flow_lm_flow_w8.mnn # ~9.5 MB mimi_decoder_w8.mnn # ~10 MB mimi_encoder_w8.mnn # ~9.5 MB (converted; runtime uses ONNX fallback) text_conditioner_fp32.mnn # ~16 MB (preferred) text_conditioner_w8.mnn # ~16 MB flow_lm_main_int8.onnx # empty-seq priming fallback mimi_encoder.onnx # voice encode fallback bundle.json tokenizer.model bos_before_voice.npy pocket_tts_mnn.py pocket_tts_onnx.py reference_sample.wav ``` ## Hybrid runtime MNN runs the generation hot path (frame loop, flow net, decoder, text conditioner). ONNX is still required for: 1. **Empty-sequence priming** (`seq_len=0`) — not supported by MNN on `flow_lm_main` 2. **`mimi_encoder`** — MNN fails streaming Concat shape inference ## Quick start ```bash pip install MNN onnxruntime numpy soundfile sentencepiece scipy huggingface_hub safetensors huggingface-cli download developerabu/pocket-tts-mnn --local-dir pocket-tts-mnn cd pocket-tts-mnn python generate_mnn.py "Hello, this is a test." reference_sample.wav out.wav ``` Or in Python: ```python from pocket_tts_mnn import PocketTTSMnn tts = PocketTTSMnn(models_dir=".") audio = tts.generate("Hello, this is a test.", voice="reference_sample.wav") tts.save_audio(audio, "out.wav") ``` ## Benchmark (Apple M1, CPU, 5 runs) | Backend | Medium RTFx | Stream TTFB | |---------|-------------|-------------| | ONNX INT8 | ~4.1× | ~100 ms | | MNN INT8 (this repo) | ~2.4× | ~123 ms | MNN is mainly targeted at mobile; desktop Python numbers above are not the expected mobile ranking. ## Attribution - Original model: [Kyutai Labs](https://kyutai.org/) — [kyutai/pocket-tts](https://huggingface.co/kyutai/pocket-tts) (CC-BY-4.0) - ONNX export used as source: [IgnitiveLa...

Note7. Aug.huggingface.co/developerabu/pocket-tts-mnn

Source context: 0 downloads · 0 likes · Pipeline text-to-speech · Library mnn · Repo developerabu/pocket-tts-mnn