Block’s Dessa is a full-duplex speech model trained only on human speech, using 1 NVIDIA GB200 NVL72 rack.