How to change number CPU core used to train deep learning?

How to change the Execution Environment, number CPU core used to train deep learning model?

Answer 1

[Ameer Hamza]

Are you training using trainNetwork() function? If yes, then the function accepts an input "options". You can define the option using trainingoptions(): https://www.mathworks.com/help/releases/R2020b/deeplearning/ref/trainingoptions.html function. In the trainingoptions, there is a Name-Value pair called 'ExecutionEnvironment', you can set it to 'parallel' and it will use the number of cores according to your cluster profile. I guess you will need a parallel Computing toolbox for this. Read here on how to configure the cluster profile: https://www.mathworks.com/help/parallel-computing/discover-clusters-and-use-cluster-profiles.html