Torque * rps > = Mass * Acceleration * Velocity / (2 * pi)
(kg*m^2/s^2)* (round/s) = (kg)*(m/s^2)*(m/s)/(2*pi)
Do the units even match up? Given 1 round = 2 pi radians, and assuming the (2 * pi) is in radians, the radians are not on the same side (one is in the numerator, the other in the denominator).
To derive it you would probably start from F = m*a (force = mass * acceleration), T=F*d (torque = force *distance), and some rotational mechanics like Torque = I *alpha (torque = I * angular acceleration), where I= moment of inertia and alpha = a/r where r is radius.