Language Breakdown
Lines of code distribution across 25 owned repositories
I-Shaped Developer
I-shapedSpecialist — deep expertise in Jupyter Notebook
Collaboration Network
Global Impact visualization
Repos
25
PRs
0
Growth
+18%
Top Collaborators
No collaborator data yet.
Coding Streak
Contribution activity over the past year
zhao tuo
@zhaotuo0815
Amieur Zineb Ichrake
@IchraqAmieur
Mohammed BEKKOUCHE
@mdbekkouche
Delba de Oliveira
@delbaoliveira
tharwat samy
@tharwatsamy
Top Repositories
A Deep Reinforcement Learning-based agent for intelligent traffic light control at a four-way intersection using SUMO. Trained with DQN variants to reduce congestion and adapt to real-time traf
A local AI setup using Continue.dev with Ollama and a Playwright-based MCP server. This project integrates Devstral models and browser automation into an AI coding workflow.
A local Retrieval-Augmented Generation (RAG) chatbot built with Python, LangChain, and Ollama for both embeddings and LLM inference. Uses Bright Data to scrape Wikipedia articles by keywords, stores them in a Chroma vector database, and provides intelligent, context-aware answers to user queries — all running locally.
A Gradio-based AI agent powered by smolagents and connected to external tools via Model Context Protocol (MCP). The agent uses a Hugging Face Inference API model to process user queries and dynamically execute code or call remote tools.
Simple sentiment analysis server using Gradio and the MCP protocol (Gradio SDK)
dermaScan is a deep learning-powered web application that detects types of skin cancer from medical images. Built with transfer learning using ConvNeXt-Tiny, trained on the HAM10000 dataset, and deployed via a FastAPI backend and a responsive Next.js frontend.
A collection of hands-on NLP and LLM projects built during my learning journey. This repository includes practical experiments with transformers, tokenizers, embeddings, semantic search, fine-tuning, and more — using Hugging Face, PyTorch, and other modern NLP tools
A semantic search demo using Hugging Face Transformers and FAISS. This project indexes GitHub issue threads using embeddings and retrieves the most relevant comments for a user query based on vector similarity.
Fine-tuning bert-base-uncased on the GLUE MRPC dataset using Hugging Face's Trainer API and a custom PyTorch loop. Part of my learning journey in understanding transformers and LLM fine-tuning.
Fine-tuning unsloth/Meta-Llama-3.1-8B-Instruct-bnb-4bit using QLoRA on the pookie3000/pg_chat dataset as part of my LLM fine-tuning journey
Open Source Impact
Contributions to external projects
No external contributions found.