Deep-Learning-with-python

This repository contains all the work I've done with the Tensorflow and Keras frameworks.

MNIST:

This folder contains two deep learning networks, that are supposed to work with the dataset MNIST. It will tell you the digit its receiving as an input image of 28x28, from 0 to 9. The first one is a simple one, with just 3 DENSE layers, while the other one is using a convolutional neural network. If you'd like to try it, feel free to use the 'MNIST_pred.py' program to make a prediction with one of them. Now there is also an Autoencoder using this dataset.

PRE-TRAINED:

This folder contains some projects that make use of some pre-trainend Deep Learning models from the tensorflow.keras library.

TEXT GENERATOR:

A deep learning model trained to predict a headline, using training data from thousands of New York Times headlines. Not always the words predicted make sense within the sentences, but it's fun to use. Right now it predicts up to 5 more words added from the text seed you write as an input. Of course, it can be changed.

LIBRARIES AND DEPENDENCIES:

• Python 3.9 (thats the version I'm using)
• Tensorflow 2
• Pandas
• Numpy