Mahder Tesfaye Abebe's repositories
advanced-react-portfolio
Final project for Advanced react course on coursera
Autonomous-Driving---Car-Detection-using-YOLO
Implement object detection for autonomous driving using the YOLO model. Includes image pre-processing, loading a pre-trained YOLO model, and drawing bounding boxes around detected objects. Utilizes drive.ai dataset for training/testing. Implemented in Python with TensorFlow and Keras.
awesome-react
A collection of awesome things regarding React ecosystem
CBOW-Word-Embedding
A 2-layer neural network from scratch to train word embeddings, enhancing natural language understanding.
Cryptography-Playground
This app is a Django and React web app for experimenting with encryption and decryption using OTP, AES, and 3DES algorithms. Featuring an intuitive UI and RESTful APIs, it’s perfect for learning cryptography basics. Note: For educational use only—not production-ready.
Emoji-Prediction
LSTM-based classifier with custom embeddings to predict emojis for input sentences, enhancing user interaction in messaging apps.
Face-Recognition-System
A face recognition system using a pre-trained FaceNet model to encode face images into 128-dimensional vectors. This project includes face verification and recognition functionalities, leveraging triplet loss for training. The system can verify if a person is who they claim to be and recognize authorized individuals without the need for ID cards.
Jazz-Music-Generation
Built an LSTM-based system for jazz melody prediction, demonstrating creativity in AI-generated music.
Little-Lemon-reservation-web-app
Meta backend developer capstone project
little_lemon_restaurant_API
API project for the Little Lemon restaurant
LLM-Fine-tunning-Data-Format
This repo shows how you should format your input data to fine tune Gemma2 models on MahderAI app
Meta_django_web_framework_final_project
Final project for meta django web framework course.
Neural-Machine-Translation
An attention-based model for converting human-readable dates to machine-readable formats, improving data processing accuracy.
Neural_network_from_scratch
Designed and trained a deep neural network from scratch using numpy only for binary classification tasks (e.g., cat vs. dog).
Question-duplicate-identifier
A model to identify similar questions from quora using Siamese networks
restaurant-front-end
Front end web app for a small restaurant
SentencePiece-and-Byte-Pair-Encoding-BPE-Implementation
This repository provides a hands-on exploration of SentencePiece tokenization and Byte-Pair Encoding (BPE) .The code demonstrates data preprocessing steps like NFKC normalization and lossless tokenization, followed by a practical implementation of the BPE algorithm from scratch.
Sign-Language-Recognition
A multi-class classifier using TensorFlow on a sign language dataset, achieving 99% accuracy in training and over 95% accuracy in the validation dataset within 15 epochs, enhancing accessibility for the hearing impaired. used data from https://www.kaggle.com/datasets/datamunge/sign-language-mnist
tensorflow-Course-lab
All the lab questions of tensorflow developer professional certificate by deeplearning.ai
Transfer-Learning-InceptionV3-
Fine-tuned a pre-trained InceptionV3 model for binary classification (human vs. horse), achieving 99.9% training accuracy and 98.66% validation accuracy in just 3 epochs, improving model efficiency.
transformer-network
Implementation of the Transformer architecture including positional encodings, scaled dot-product self-attention, masked multi-head attention, and model training.
trigger-word-detection-
A deep learning project for detecting trigger words in audio clips using convolutional and recurrent neural networks.
Verzantlias-School-for-Future-Website
6 pages website for a school called Verzantlias School for Future using react and tailwindcss
VirtualR
A Simple landing page for a company called, "VirtualR"