Hi!
My cheif area of interest is in NLP. If I’m mainly looking at this area of deep learning, am I better off starting with the 2020 course Part 1 or the specific NLP curse, linked here:
Because this is such a changing field (NLP specifically) I’m wondering if i’d be better off going through the deep learning course, rather than the NLP course given how much the field may have changed over the last 2 years.
The NLP couse is quite good because it gives a wide view of the subject. The videos are on youtube.
You might also find these interesting. The latter I believe was used on NLP.
Also Open AI have created GTP which is very powerful.
Conwyn,
thanks for sharing these resources. I also appreciate your respone about the course.
it was exactly because of things like GTP that I was wondering if a course taught in 2019 still has the same utility. It’s changing so rapidly.
Would you start with the NLP course or go through the deep learning course first, give your experience?
Best,
The deep learning course does not provide you with a lot of stuff on NLP. It will just cover the basics of deep learning in a bunch of subdomains such as Computer Vision, Tabular Data, Recommender Systems, and yes even NLP. But then again it does not give you a wide overview of NLP.
I have recently started working on the NLP course, so far it has been mostly working on Naive Bayes, Logistic Regression and Regex in the initial 5-6 videos or so. But the course seems to give a really good deep dive into the topic of NLP.
One of the other pathways you can also take up is going through the NLP chapters of the fast.ai Part 1 course and then go through the NLP course from 2019. This would be a good solution as it will help you cover all the aspects of NLP that are there in the fast.ai courses.
Personal recommendation: start with Deep Learning Course. The state of the art for NLP is all about deep learning, so this will help you a lot. As a next step, consider either fastai NLP course or the HuggingFace course. Or ideally both