Sequence To Sequence Learning With Tensorflow & Keras Tutorial Series

The Seq2Seq Learning Tutorial Series aims to build an Encoder-Decoder Model with Attention. I would like to develop a solution by showing the shortcomings of other possible approaches as well. Therefore, in the first 2 parts, we will observe that initial models have their own weaknesses. We will also understand why the Encoder-Decoder paradigm is so successful. 

You can access all the parts in the below links.

Photo by Clay Banks on Unsplash

PARTS

Part A: AN INTRODUCTION TO SEQ2SEQ LEARNING AND A SAMPLE SOLUTION WITH MLP NETWORK

YouTube Videos in ENGLISH or TURKISH / Post / Colab Notebook

Part B: SEQ2SEQ LEARNING WITH RECURRENT NEURAL NETWORKS (LSTM)

YouTube Video in ENGLISH or TURKISH / Post / Colab Notebook

Part C: SEQ2SEQ LEARNING WITH A BASIC ENCODER DECODER MODEL

YouTube Video in ENGLISH or TURKISH  / Post / Colab Notebook

Part D: SEQ2SEQ LEARNING WITH AN ENCODER DECODER MODEL + TEACHER FORCING

YouTube Video in ENGLISH or TURKISH / Post / Colab Notebook

Part E: SEQ2SEQ LEARNING WITH AN ENCODER DECODER MODEL WITH TEACHER FORCING FOR VARIABLE INPUT AND OUTPUT SIZE: MASKING & PADDING

YouTube Video in ENGLISH or TURKISH / Post / Colab Notebook

Part F: SEQ2SEQ LEARNING WITH AN ENCODER DECODER MODEL + BAHDANAU ATTENTION + LUONG ATTENTION

YouTube Video in ENGLISH or TURKISH / Post / Colab Notebook

I hope you enjoy the Seq2Seq Learning Tutorial Series!

You can access Murat Karakaya Akademi via:

YouTube

Facebook

Instagram

LinkedIn

Github

Kaggle

muratkarakaya.net