Run the AADL computations on cpu by default for now

ORNL / AADL

Anderson Acceleration for Deep Learning

12 stars 6 forks source link

Closed henrymai closed 2 years ago

henrymai commented 2 years ago

This is temporary until we can figure out why the results are worse on gpu compared to cpu.