Improve the efficiency of your code by using command-line and configuration files in Python — If you have worked a lot with machine learning and deep learning models, you probably tried to change hyperparameters a million times on your Jupyter notebook or python script to improve the performance on your dataset. This can be time-consuming if you go through your code every time you edit…

A gentle overview of this fast open-source with examples from the Pokemon dataset with Pytorch — This article is in continuation with the story A Comprehensive Guide to Image Augmentation using Pytorch. In the previous post, different image augmentation techniques were shown to increase the size and the diversity of the images present in the dataset. The goal of image augmentation is to improve the generalization…

Fast way to scrape data without coding using Octoparse — There are always two suggested image datasets in Computer vision projects, MNIST and fashion MNIST. They respectively provide grayscale images of handwritten digits and Zalando’s article images. But unfortunately, the images are more complex than these popular toy datasets and not always black and white in the real world. For…

A way to increase the amount of data and make the model more robust — Lately, while working on my research project, I began to understand the importance of image augmentation techniques. The aim of this project is to train a robust generative model able to reconstruct the original images. The problem addressed is anomaly detection, which is quite challenging since there is a small…

A simple and funny way to convert a video to GIF using MoviePy and Streamlit — A few months ago I experimented with a lot of websites to convert the videos into gifs, but I didn’t find any satisfactory solution to edit and convert the video without becoming crazy. Suddenly, I discovered MoviePy, a python library specialized in video editing. It allows performing many basic editing…

A demonstration of this platform can help you to track your model results efficiently — This article is in continuation with the story An intuitive Guide to Track your ML Experiments using MLflow. In the previous post, I introduced MLflow, an online source platform that allows saving the hyperparameters and metrics of your model. The only drawback I found is that it doesn’t provide an…

An overview to clone a repository, push changes, and make pull requests in VS code — For a lot of time, I updated my GitHub repository by simply inserting the files. But doing this can waste a lot of time when you have to repeat the same operation more than one time and when you are working with other people…

A brief overview of an interactive package to manipulate Covid-19 Vaccine Data — If you have worked with real data in Python, you surely have noticed that data manipulation and data cleaning takes most of your time, even up to 70% of the time. …

Helping you on improving the reproducibility, management and tracking of experiments for your Sklearn model — I have lost a lot of time running my Jupiter notebook to build a Machine Learning model with the best performances. …