Scikit-learn with Jupyter notebooks is a great workflow. Scikit-learn with parallel jobs is also wonderful. Verbose messages provide valuable status messages when running through the Search.

Unfortunately, the messages that appear when RandomizedSearchCV or GridSearchCV run in Jupyter notebooks with n_jobs = 1 do not appear in the notebook in the same way when n_jobs specifies running jobs in parallel (e.g. n_jobs = -1 or n_jobs = 4). Can the same information please be displayed in the notebook regardless of parallel or serial processing?

Thanks,
Aaron

PS I discussed this issue with Andreas briefly at PyDataNYC and he suggested adding it here.