I'm trying the R interpret package, the sample size is 100k with 100 features.
But I found that the training process is very slow and the function ebm_classify has no n_jobs argument.