How to Get Started with Natural Language Processing in Tensorflow 2

In this python tutorial we'll learn how to get started with natural language processing and word embeddings in tensorflow 2. Modern artificial intelligence frameworks make natural language processing a breeze, and we can get started in only a few dozen lines of code.

Original code for this video is here

Learn how to turn deep reinforcement learning papers into code:

Get instant access to all my courses, including the new Prioritized Experience Replay course, with my subscription service. $29 a month gives you instant access to 42 hours of instructional content plus access to future updates, added monthly.

Or, pickup my Udemy courses here:

Deep Q Learning:

Actor Critic Methods:

Curiosity Driven Deep Reinforcement Learning

Natural Language Processing from First Principles:
Reinforcement Learning Fundamentals

Here are some books / courses I recommend (affiliate links):

Come hang out on Discord here:

[MachineLearningwithPhil]

This content is sponsored by my Udemy courses. Level up your skills by learning to turn papers into code. See the links in the description.

[RedShipsofSpainAgain]

Thanks Phil. Also great explanation of word embeddings:

4:05 "the integer encoding of our dictionary forms a basis in some higher dimensional space. But all of those vectors are orthogonal. So if you take their dot product they're essentially at right angles to each other in a higher dimension space and so their dot product is zero. So there's no projection of one word, one vector in another. There's no overlap in the meaning between the words.

Now word embeddings fix this problem by keeping the integer encoding but doing a transformation to a totally different space. So we introduce a new space of a vector of some arbitrary length (it's a hyperparameter of our model, much like the number of layers of your model). The length of the embedding layer is a hyperparameter. We'll just say it's 8. So the word king has 8 floating point elements that describe its relationship to all the other vectors in that space. And so what that allows you to do is to take dot products between two arbitrary words in your dictionary and you get non-zero components. So what that means in practical terms is you get a semantic relationship between words that emerges as a consequence of training your model.

[DanielWeikert]

Thanks for your tutorial Phil. Could you do a video where you dive a little bit deeper into the new tensorflow 2.0 methods like .take, .batch, .from_tensor_slices, ... explaining them and showing examples?
Really like your videos. Thanks for the channel. Best regards and enjoy your weekend

[TheMasterfulcreator]

Hi Phil, in the future if you get the time, could you talk more about your transition from academia into industry after you graduated? I watched your video on your transition in machine learning but I'm specifically confused about the software engineering skills required. This summer I'm graduating with a PhD in applied math and I'm interested in working in industry as a machine learning engineer. As a math guy I've picked up the linear algebra and optimization techniques pretty quickly but I'm confused about how good my software engineering skills will need to be. I've been working on learning algorithms, coding, etc for about a year now but I'm concerned I'd end up laughably in over my head when it comes to getting products into production.

[SeanCotton]

Phil, do you have an earlier video that shows how to get set up my computer so that I can code along with you on this video, for example: what code editor are you using?

[delllaptop]

Hey could you do more NLP tutorials in tensorflow?

[geo]

Thabk you for doing NLP tutorials!

Do you have any documentation on how did you setup your enviroment? (Cuda drivers, and did you compile TF from source? )

[nynzlol]

I am wondering why you don't make use of jupyter notebook to avoid having to re-train the model everytime the code is changed

[MohammadAli-qpgx]

I did not understand why you didn't upload the .tsv files to the browser at the end.

[athreyaabhijith]

Thanks for this video. The embedding visualization which you showed at the end is not from the IMDB dataset. If I load the vecs.tsv, it works fine. However, the metadata.tsv is not in the right format. It is not working.

[mohithshivu]

Sir my name is mohith I am final year BE student can you help me out some doubt on nlp I am working on data generalization and data sanitization our task is identifying given text weather it is sanitized or not generalized or not how it work in python can you help out sir please.... it is helpfull to me

[JousefM]

Comment for the "Algorithm" :D