Fix type checks to not fail if first value in series is NaN/missing.

[henrifroese]

Currently, when using our InputSeries decorator, it always throws an error if the first Series value is NaN. This of course makes no sense for users.

Example:

>>> import pandas as pd
>>> import texthero as hero
>>> s = pd.Series([pd.NA, "test"])
>>> hero.clean(s)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/henri/Prog/SummerOfCode/texthero/texthero/texthero/_types.py", line 233, in wrapper
    allowed_hero_series_type.check_series(s)
  File "/home/henri/Prog/SummerOfCode/texthero/texthero/texthero/_types.py", line 124, in check_series
    raise TypeError(error_string)
TypeError: The input Series should consist only of strings in every cell. See help(hero.HeroSeries) for more information.

We fix this by finding the first non-NaN value now and checking this.

[henrifroese]

Accidentally pushed the nlp tutorial on here, so had to force-push one commit back. It's now back at the correct commit (0cca7d9) 🙃

[jbesomi]

Good catch, thanks! 🎉

When reviewing, I noticed a couple of other issues (see #180 ) that preferably need to be dealt with separately (one PR for each issue). Would you mind looking into that, please?