| --- |
| license: mit |
| tags: |
| - tts |
| - text-to-speech |
| - piper |
| - vits |
| - gguf |
| - crispasr |
| base_model: rhasspy/piper-voices |
| pipeline_tag: text-to-speech |
| language: |
| - en |
| --- |
| |
| # Piper TTS (en_US lessac medium) — GGUF |
| |
| Native C++ GGUF conversion of the [Piper](https://github.com/rhasspy/piper) VITS voice `en_US-lessac-medium` for use with [CrispASR](https://github.com/CrispStrobe/CrispASR). |
|
|
| ## Files |
|
|
| | File | Size | Description | |
| |------|------|-------------| |
| | `piper-en_US-lessac-medium-f16.gguf` | 30 MB | F16 weights (full model) | |
|
|
| Piper models are small enough that quantization provides no meaningful savings. F16 is the only format. |
|
|
| ## Usage with CrispASR |
|
|
| ```bash |
| ./build/bin/crispasr --backend piper \ |
| -m piper-en_US-lessac-medium-f16.gguf \ |
| --tts "Hello, how are you today?" \ |
| --tts-output hello.wav |
| ``` |
|
|
| Phonemization uses espeak-ng (must be installed: `apt install espeak-ng`). |
|
|
| ## Architecture |
|
|
| - **VITS** (Conditional Variational Autoencoder with Adversarial Learning) |
| - **Text encoder**: 6-layer relative-position transformer (192-d, 2 heads) |
| - **Duration predictor**: Stochastic Duration Predictor with rational-quadratic spline flows |
| - **Flow**: 4 affine coupling blocks with WaveNet conditioning |
| - **Decoder**: HiFi-GAN (3 upsample stages, 9 MRF resblocks) |
| - **Output**: 22.05 kHz mono PCM |
| - **License**: MIT |
|
|
| ## Conversion |
|
|
| ```bash |
| python models/convert-piper-to-gguf.py \ |
| --onnx en_US-lessac-medium.onnx \ |
| --output piper-en_US-lessac-medium-f16.gguf |
| ``` |
|
|
| ## Provenance and EU AI Act Art. 53 note |
|
|
| - **Upstream model:** [rhasspy/piper-voices](https://huggingface.co/rhasspy/piper-voices) — published by `rhasspy`. |
| - **Upstream licence:** `mit`. This repository redistributes under the same terms; it grants no rights the upstream licence does not. |
| - **What was done here:** format conversion and/or quantisation only (GGUF). No training, no fine-tuning, no merging, no distillation, no change to architecture, vocabulary or capability. Only the numeric representation of the upstream weights differs. |
| - **Training data:** documented — where it is documented at all — by the upstream provider; see the upstream model card. No training data was used, added or selected by this repository. No training-content summary was found on the upstream model card at the time of writing; that documentation gap is upstream's and is not filled here. |
| - **Provider status:** under Regulation (EU) 2024/1689 the upstream authors remain the provider of this model. Converting the serialisation format does not make this repository the provider of a new general-purpose AI model, and no such claim is made. Questions about training content, copyright policy or model capability belong upstream. |
|
|
