float16x2# class nvmath.device.float16x2(x, y)[source]# Create a Numba compliant vector object for float16 with vector length 2. Methods __init__(*args)[source]#
