How to use custom sigma value in gaussian kernel function for fitcsvm function

Hi,

Can anyone please explain how to input custom sigma value for gaussian kernel while using fitcsvm function for SVM classifier. I am using rbf as the kernel function, but i want to manipulate the sigma value for the gaussian kernel instead of using the default sigma value. I tried using kernel scale with different inputs, but i think it doesn't change the sigma value. Any help would be greatly appreciated. Thanks!