Datasetpublic

Shadow0482/breeze-tts-2-audiocpp-q8-dataset

BreezeTTS 2 synthetic speech dataset Generated with the prebuilt 0xShug0/audio.cpp v0.8.1 CUDA 12.8 bundle. Samples: 100 Audio: 0.11 hours Format: 24 kHz mono WAV Engine: 0xShug0/audio.cpp Version: v0.8.1 prebuilt CUDA 12.8 Family: breeze_tts Model: BreezeTTS 2 Precision: Q8_0 GGUF Backend: CUDA Model package audio-cpp/audio.cpp-gguf Breeze-TTS-2-GGUF/breeze-tts-2-q8_0.gguf Metadata metadata.jsonl includes the original manifest fields plus… See the full description on the dataset page: https://huggingface.co/datasets/Shadow0482/breeze-tts-2-audiocpp-q8-dataset.

sourceHugging Faceotherupdated 5d agoView on Hugging Face

0likes39downloads

Dataset Card

BreezeTTS 2 synthetic speech dataset

Generated with the prebuilt 0xShug0/audio.cpp v0.8.1 CUDA 12.8 bundle.

—Samples: 100
—Audio: 0.11 hours
—Format: 24 kHz mono WAV
—Engine: 0xShug0/audio.cpp
—Version: v0.8.1 prebuilt CUDA 12.8
—Family: breeze_tts
—Model: BreezeTTS 2
—Precision: Q8_0 GGUF
—Backend: CUDA

Model package

audio-cpp/audio.cpp-gguf

Breeze-TTS-2-GGUF/breeze-tts-2-q8_0.gguf

Metadata

metadata.jsonl includes the original manifest fields plus generated duration, seed, guidance scale, sampling parameters, server, GPU and QC information.

QC

qc_ok=false means the generated duration was outside the configured duration ratio range after retries.

Check the BreezeTTS model license and the audio.cpp/model package terms before redistributing or using the generated dataset commercially.