Now, this approach works
very well until you have very large bodies of text
with many many words. So for example, you could
try the complete works of Shakespeare and you'll
likely hit memory errors, as assigning the one-hot
encodings of the labels to matrices that have
over 31,477 elements, which is the number of unique
words in the collection, and there are over
15 million sequences generated using the algorithm
that we showed here. So the labels alone would require the storage of
many terabytes of RAM. So for your next task, you'll go through
a workbook by yourself that uses character-based
prediction. The full number of unique
characters in a corpus is far less than the full number of unique words, at
least in English. So the same principles
that you use to predict words can be
used to apply here. The workbook is at
this URL, so try it out, and once you've done,
that you'll be ready for this week's final exercise.