Pycharm is slow at importing during debugging - what can I do apart from Attach to subprocess automatically while debugging?

After installing Pycharm 2018.3.2, I realize that during debugging the import of other packages is getting really slow.

Taking the imports below resulted (with Pycharms default settings) to a runtime of 10 seconds.

It seems to be related to the setting Attach to subprocess automatically while debugging (see: File | Settings | Build, Execution, Deployment | Python Debugger). When I disable this, the debugging gets (much) faster:

/home/user/anaconda3/bin/python3 /home/user/4me/Technologie/0010-install/0220-pycharm-installation/2018-3/pycharm-2018.3/helpers/pydev/pydevd.py --cmd-line --client 127.0.0.1 --port 38085 --file /home/user/PycharmProjects/0540_sqlalchemy_and_pooling/attic.py
pydev debugger: starting

Connected to pydev debugger (build 183.4886.43)
pydev debugger: New process is launching (breakpoints won't work in the new process).
pydev debugger: To debug that process please enable 'Attach to subprocess automatically while debugging?' option in the debugger settings.

time used ----: 2.4612913131713867
fin

Process finished with exit code 1

I opened already a bug at Pycharm ( https://youtrack.jetbrains.com/issue/PY-33464 ) for this.

Why is this so that this flag makes debugging much faster and are there other workarounds available, too?

import time
my_time = time.time()
from sklearn import datasets, linear_model
from sklearn.metrics import mean_squared_error, r2_score
import numba as nb
import time
import sys
import matplotlib
matplotlib.use('TkAgg')
import matplotlib.pyplot as plt
from matplotlib.ticker import FuncFormatter
import numpy as np
import pandas as pd
import datetime

from typing import Dict, List, Optional
import datetime
import pickle
from os import listdir
import os
from os.path import isfile, join
import logging
import re
import traceback
from dateutil.easter import *
import copy

import multiprocessing
from multiprocessing import Pool
from multiprocessing import cpu_count
from fbprophet import Prophet
print("time used ----:", time.time() - my_time)
print("fin")
import sys
sys.exit(1)