[gem]gem5-smaug deeplearning workload on gem5-Aladdin

[samialabed]

Would be nice to be able to benchmark deeplearning workloads on the gem5-Aladdin environment.

Either:

• Map the names to smaug
  https://github.com/yaoyuannnn/experiments/blob/435747732815c24e2bff4718597f31d8e6aa2cb1/sweeps/params.py

• Or properly resolve the issue by addressing the concerns in this issue Use of xenon generator yaoyuannnn/experiments#2