Deep Learning with TensorFlow

In this course, we will talk about the basics of neural networks and deep learning, how to use deep learning ConvNets, transfer learning, transformers - a deep learning architecture that has revolutionized the traditional NLP.

Neural Network Foundations with TF

What TensorFlow and Keras are?

An introduction to neural networks

The perceptron and multi-layer perceptron

Recognizing handwritten digits

Activation functions

Backpropagation

Regularization methods

Regression and Classification

Regression with TF

Classification using the TensorFlow

Logistic regression to identify handwritten digits

Linear regression to estimate the price of a house

Convolutional Neural Networks

Deep convolutional neural networks

ConvNets in TensorFlow

DNNs for large-scale image recognition

Deep Inception V3 networks for transfer learning

Other CNN architectures

Word Embeddings

Word embeddings – origins and fundamentals

Distributed representations

Creating Azerbaijani embedding with Gensim

Exploring the embedding space with Gensim

Character and subword embedding

Word2Vec models: CBOW and Skip-gram arthitectures

Language model-based embeddings: BERT, ULMFiT, GPT

Recurrent Neural Networks

The basic RNN cell

RNN cell variants

Encoder-decoder architectures – seq2seq

Attention mechanism

Text cleaning

Novel generation using Recurrent Neural Networks

Building seq2seq network with/without attention mechanism

Transformers

Transformers: Understanding architecture

Language models: BERT, Reformer, GPTs and etc.

Using Hugging Face models

Applying BERT models to variety of tasks

Generative Models

What is a GAN?

Deep convolutional GANs

Applications of GANs

Flow-based generative models

Diffusion models for data generation

Appendix

Introduction to Docker, Flask, FastAPI

Deploying Deep Learning Model in Heroku using flask