Conversational Agents using LLM Models
github link
Large Language Model Playground: Interacting with conversational agents using the following LLM models: Llama2(70B, 13B, 7B), Falcon(7B), GPT, and Flan UL2 and tune chatbot performance (temperature and max token length)
Pro-Task
github link
Time Management REACT Application with Voice Command Control using AI Voice Recognition Model
Arabic Part Of Speech Tagging
github link
Arabic part of speech tagging using arabic PUD dataset using bidirectioanl LSTM for sequential labeling classification
AI Voice Assistant
github link
AI voice assistant web app built using SpeechRecognition,pyttsx3, and streamlit open-source libraries
Arabic Word Embedding
github link
Arabic Word Embedding models SkipGram, and GLoVE are trained over Arabic Wiki data Dump 2018 dataset from scratch using Gensim and GLoVE python libraries. Then the models are evaluated on three NLP tasks and its results are visualized in T-SNE
Genetic Algorithm Solving N Queens Problem
github link
The project analyzes the N-Queens problem and defines the cost function, solution represntations as(Phenotype, Genotype) where the Genetic Algorithm is applied and experimented with different chess board dimensions (N x N)
Question-Answering-System
github link
Using End to End memory model and bAbI dataset I developed a Recurrent Attention model that can understand the context of a sentence. the system generates an answer based on a story and a question
Path Finding Problem using A Star Algoritm
github link
Unity Project to simulate obstacle avoidance and path finding (using A star search algorithm) to help the agent (robot) solve any maze
Bakery Web App
github link
A REACT-based web app suitable for small/large bakery business that allows costumers to convenient buy freshly baked goods, reserve table in the actual restaurant ahead and see the daily updated menu
Public Transportation User Statisfaction Dashboard
website link
My BI analysis system utilizes the public transportation data provided by ASCC for U.A.E, Ajman 2018-2021. Features of drilling down where users investigate and narrow down analysis to a specified group of year/s in addition to seamless switching between languages Arabic and English. The dashboard is enhanced with interactive graphs & maps updating the visuals by area-based filters
