all you have to do is load a session from the session set, and in the 'expert' tab you see some options- there's one option for the length of the sweep, which is usually set to 45 or 55 seconds (not sure), you can change that to be smaller (which i believe will degrade the definition of the frequency response if you lower it much). you can't really cut so much time here, so instead-
there is another option for the number of dynamic steps you want to sample, in other words the number of total sine sweeps (one for each dynamic level). you could set that to one, which means your audio file will only be about a minute long. you won't have any dynamic behavior in whatever you sample or process it with though.
so if you change that dynamic steps settings to 1, then hit generate, the wav it generates will only have one dynamic step. much smaller. IF you choose to save this session, you should remember to name it something else instead of overwriting the original.