Is it possible to do Text Classification with pretained model and the existing vocab + new vocal?

I am following the video tutorial 8 - NLP
here is the notebook

Once we trained the model with the imdb texts and saved it as a model.

Then we can load the pre-trained model again with the same vocabulary like this:

dls_clas = DataBlock(
    blocks=(TextBlock.from_folder(path, vocab=dls_lm.vocab),CategoryBlock),
    get_y = parent_label,
    get_items=partial(get_text_files, folders=['train', 'test']),
    splitter=GrandparentSplitter(valid_name='test')
).dataloaders(path, path=path, bs=128, seq_len=72)

Now, I have a totally different NLP task which is to do classification(10 categories) for comments.
I would like to add the vocabulary from my new task into the existing vocabulary, so it recognizes something like: COVID-19

How can I do it?