Chapter 15. Processing Sequences Using RNNs and CNNs

🎼

Chapter 15. Processing Sequences Using RNNs and CNNs

This is a short companion page to our internal reading group of the book “Hands-on Machine Learning with Scikit-Learn, Keras, and TensorFlow, 2nd Edition”. However I unashamedly used a lot of PyTorch examples.

Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow, 2nd Edition

Chapter 15. Processing Sequences Using RNNs and CNNs The batter hits the ball. The outfielder immediately starts running, anticipating the ball's trajectory. He tracks it, adapts his movements, and finally ... - Selection from Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow, 2nd Edition [Book]

https://learning.oreilly.com/library/view/hands-on-machine-learning/9781492032632/ch15.html

Understanding LSTM

Understanding LSTM Networks

Posted on August 27, 2015 Humans don't start their thinking from scratch every second. As you read this essay, you understand each word based on your understanding of previous words. You don't throw everything away and start thinking from scratch again. Your thoughts have persistence.

https://colah.github.io/posts/2015-08-Understanding-LSTMs/

https://atcold.github.io/pytorch-Deep-Learning/en/week06/06-3/

Input and Output

https://karpathy.github.io/2015/05/21/rnn-effectiveness/

one to many: image -> caption sentence

many to one: sentence -> sentiment (positive / negative label)

many to many: a sentence in English -> a sentence in Turkish

the other many to many: frames of video -> coordinates of bounding boxes around an object

NMT and Seq2Seq

Visualizing A Neural Machine Translation Model (Mechanics of Seq2seq Models With Attention)

Translations: Chinese (Simplified), Japanese, Korean, Russian, Turkish Watch: MIT's Deep Learning State of the Art lecture referencing this post May 25th update: New graphics (RNN animation, word embedding graph), color coding, elaborated on the final attention example. Note: The animations below are videos.

https://jalammar.github.io/visualizing-neural-machine-translation-mechanics-of-seq2seq-models-with-attention/

LSTM Cell

LSTM

Stacked LSTM

WaveNet

WaveNet: A Generative Model for Raw Audio

This post presents WaveNet, a deep generative model of raw audio waveforms. We show that WaveNets are able to generate speech which mimics any human voice and which sounds more natural than the best existing Text-to-Speech systems, reducing the gap with human performance by over 50%.

https://deepmind.com/blog/article/wavenet-generative-model-raw-audio