FREE 11 Hour NLP Transformers Course (Next 3 Days Only)
00:00:00.000 | I want to introduce you to this course I've been working on.
00:00:03.360 | I've just released it and I wanted to give a lot of you guys who subscribed and follow
00:00:11.840 | me on Medium or Twitter, I wanted to give you guys a chance to get this course for free.
00:00:20.240 | So for the next three days it is completely free, you just use this code.
00:00:24.160 | But I just want to talk very quickly about what it actually covers.
00:00:28.940 | Now obviously you can see from the title it's NLP and it's with Transformers and Python.
00:00:35.440 | Now if we scroll down a little bit we come to this course overview video and I'll just
00:00:40.880 | quickly go through this because it's quite long and I don't want to take too much of
00:00:47.440 | your time and we cover a lot of things.
00:00:50.360 | So first thing is NLP and Transformers where I give a quick summary of NLP in general,
00:00:56.840 | the history of NLP leading up to Transformers.
00:01:02.240 | Then we move into a bit of pre-processing for NLP.
00:01:04.880 | Now this is just your basic stuff, I think the most relevant one here for us in Transformers
00:01:11.880 | is Unicode normalization and tokenization special tokens.
00:01:17.660 | Then I move through a few lectures on attention, how attention works and describing the logic
00:01:25.080 | behind it.
00:01:28.720 | I always see this as like the hello world of NLP which is sentiment analysis.
00:01:34.120 | I think it's a great introduction and we introduce Transformers in this section here.
00:01:44.120 | And it's worth pointing out as well that I use a lot of different frameworks throughout
00:01:48.440 | this course.
00:01:49.440 | So Flare is the very first one, we also use Hockey Face Transformers, that's obviously
00:01:55.440 | the primary one that we'll be using throughout the course, TensorFlow, PyTorch, NLTK, Spacey
00:02:04.280 | and many others as well.
00:02:06.680 | So there's a lot in there, of course using a lot of BERT.
00:02:13.400 | So there's two projects in the course as well, the first of those is sentiment analysis,
00:02:21.040 | the second one is question answering.
00:02:23.320 | Both of them I think are great because they take you all the way through from the very
00:02:27.000 | start of your project, so getting data, all the way through to actually building your
00:02:31.080 | model and applying it to your data.
00:02:36.360 | So moving on to named entity recognition, question answering, how we measure the performance
00:02:46.400 | of our models which is of course very important, a full question answering stack using another
00:02:55.760 | library called Haystack which I think this is one of the coolest things in the course
00:02:59.440 | in my opinion and in NLP in general, this sort of stuff is incredibly cool.
00:03:07.880 | Then like I said, there's that second project, the Q&A project.
00:03:12.480 | Before we move on to similarity, now similarity is super important in NLP and I think probably
00:03:19.400 | one of the most promising areas in the future for further research and just impact that
00:03:26.480 | it could have on industry.
00:03:29.520 | I think this is really a super cool place to be.
00:03:34.080 | Then finally we move on to fine tuning.
00:03:38.160 | So that's the course in a nutshell, all together there's 11 hours of content so it's I think
00:03:48.000 | comparatively long when you look at other NLP courses, so we see this 11, 10, 10, 3
00:03:57.600 | and 6 and as far as I'm aware it's the first course that focuses on Transformers, on Udemy.
00:04:06.640 | So if you're into NLP, obviously Transformers are really the models that you want to be
00:04:14.280 | using, check out the course in the next few days, it's completely free using this code.
00:04:21.040 | So thank you for watching and I hope you enjoy the course.