Research

Edge0: 35B Models in 3 GB of Memory

Audio8-ASR: From On-Device Speech Recognition to High-Precision Large Models

Audio8-TTS: High-Fidelity Zero-Shot Speech Synthesis