TensorFlow Tutorial #20 Natural Language Processing

How to process human language in a Recurrent Neural Network (LSTM / GRU) in TensorFlow and Keras. Demonstrated on Sentiment Analysis of the IMDB dataset.

This tutorial has been updated to work with TensorFlow 2.1 and possibly later versions.

[kalpitdamahe]

Probably the best explanation on the text classification using RNN

[ribzi-kela-bru]

Finally someone who takes teaching seriously. Thanks for the clarity and effort you put into your video and github resources. +1 sub

[victorsung]

Holy SHIT! I have been waiting a long time for something like this!

More sequence models in Tensorflow would be invaluable additions to this already amazing series.

Keep up the great work!

[djcoolio]

You've got the most brief and precise explanations ever. Thanks and keep em' comin' :)

[sachavanweeren]

Thanks Magnus for the wonderful tutorial. Very well explained and very complete.

[probirsil]

Excellent tutorial on NLP. I am new to this field and was struggling with the idea of Embedding. This made it clear. Thanks!

[stevecoxiscool]

Great explanation. Embeddings started clicking in my mind after watching this video.

[VijaySharma-jqnz]

I work on AI & ML and i found this as a great help to me. Thanks for sharing on YouTube.

[muhammedbuyukknac]

This is a great video and you are a great guy. Great explanation of RNN's.

[fygarOnTheRun]

Det är en mycket fantastisk video, thanks you so much!

[sinaasadi]

Tak skal du ha! I think I have met you in one of the Pydata meetups in Copenhagen. Keep it up ! Great moves.

[mertsefatarhan]

It's crazy, but maybe those numbers (7&8) imply the reviewers' score for the movie. You know, 'I give 7/10 to this movie'. So, 7 & 8 are positive (1) and 2 & 3 are negative (0). That might be the reason why the word 'great' was correlated to number '8'.
Anyway, bahut badhiya dhanyawad again :D

[saurabhkanekar]

Great work dude! I was wondering the padded inputs would have effect on learning (embedding as well as on model prediction)

[eff]

Gained urself a like and a subscriber… thnx bro

[snarcraft]

OMG! you are amazing, thank you so much

[raghavsaxena]

Can you explain how to analyze video and audio data

[arfasobirhanu]

wooow Fantastic video. continue other on cnn based machine translation.

[BoneCrushGaming]

I don't know what I am doing wrong but I have checked and have included EVERYTHING and it is having a hard time loading the data.


After I do imbd.loda_data() it says ValueError: Object arrays cannot be loaded when allow_pickle=False.


Help!

[JoseCastillo-fljn]

I was like 'wut??' when I saw the embeddings dimensions.. :v
Anyway, your explanation is really easy to understand, keep it up 👆

[Plumatino]

The nlp libraries are also able for Java? By the way congrats for your amazing job!