SYLIN

Audio8

Maker · 3 entries

Everything from Audio8

NameWhat to knowPrice or memoryRanks
Audio8 TTSModel0.6B TTS with zero-shot voice cloning in 23 languages, 2.6 GB, Apache 2.0~2.6 GB
ARK-ASRModelOpen multilingual speech-to-text model in 3B and 0.6B sizes; Apache-2.0 weights, with training and inference code on GitHub~7.6 GB
Open Audio OPDTechniqueOnline policy distillation: student transcribes first, teacher grades that attempt live; 0.6B model trained on about 100,000 hours of audio