use multiprocessing for making wordcloud.fit

[vaaaaanquish]

slowly: https://github.com/vaaaaanquish/cloudia/blob/master/cloudia/main.py#L29

        wordcloud_list = []
        wcsize = self._calc_wc_size(rate)
        for name, words in self.wd:
            wordcloud = WordCloud(font_path=japanize_matplotlib.get_font_ttf_path(),
                                  background_color=self._color(dark_theme),
                                  width=wcsize[0],
                                  height=wcsize[1])
            wordcloud.fit_words(words)
            wordcloud_list.append((name, wordcloud))
        return wordcloud_list