fufuchiu / speechloom

git clone https://github.com/fufuchiu/speechloom.git

End-to-end speech model research: waveform encoding, causal decoding, codec tokens, joint losses and streaming.