Eyal Zakkay
Deep Learning Projects and Tutorials
-
TeleGrad - Monitor & Control DL Experiments Using Your Phone
A Telegram bot to monitor and control deep learning experiments. DL training processes can run for many hours/days, and not always you are near your station to see how it’s progressing or to make adjustments. Using this Telegram bot I have built you can seamlessly get constant updates and even...
-
Keras Tutorial - Constructing Complex Custom Losses and Metrics
Second post in my series of advanced Keras tutorials: on constructing complex custom losses and metrics, published on TowardsDataScience.com. In this tutorial I cover a simple trick that will allow you to construct custom loss functions in Keras which can receive arguments other than y_true and y_pred.
-
Keras Tutorial - Accurately Resuming Training
A blog post I published on TowardsDataScience.com, presenting a use case of the Keras API in which resuming a training process from a loaded checkpoint needs to be handled differently than usual. Includes examples and a tutorial on how to solve this issue.
-
Keras Implementation of Sketch-RNN
A Keras Implementation of Sketch-RNN In this repo there’s a Kares implementation of the Sketch-RNN algorithm, as described in the paper A Neural Representation of Sketch Drawings by David Ha and Douglas Eck (Google AI). The implementation is ported from the official Tensorflow implementation that was released under project Magenta...
-
BCI - From Hackathon Project to Master's Thesis and National News
Brain Computer Interfaces (BCI) allow direct control over technology using brain activity alone. Neuroscience has always fascinated me, which is why on my 2nd year as an undergrad Electrical Engineering student I started working as a research assistant at the Computational Psychiatry Lab at Ben Gurion University. Finding this intersection...
-
Throne2Vec
A Jupyter notebook with a Tensorflow implementation of the word2vec algorithm. Trained on a data-set made out of the entire Game of Thrones book collection! Includes text pre-processing, word analogies, tSNE of embedded space and more.
-
Neural Style Transfer
Tensorflow implementation of the paper “A Neural Algorithm of Artistic Style”. The algorithm renders an image that keeps the content of one reference image while copying the style of another. Dependencies: Tensorflow (version > 1.0) tensorflow-vgg Usage: Most parameters can be configured when running using command line, though some parameters...
More Coming Soon...