PyTorch

The ImagesAnnotator application project` annotations format data exporters library for NN training software

if u're 555, i'm 666

Practical tasks on Deep Learning (DL) and Neural Networks (NN). 🤖

Self-hosted forecasting + prediction service. Five zero-shot time-series foundation models (Chronos-2, TimesFM 2.5, Moirai-2, Toto-1, Sundial) across six forecast types, plus nine supervised tabular ML backends (LightGBM, XGBoost, sklearn family) with calibrated / stacking / diversified meta-learners. Unified REST API + MCP server.

A Contextual-Enhanced LightGCN for Movie Recommendation Systems

貪吃蛇 AI 訓練

Fundamental theory and practice in Data Science (DS). 🧮

Instance-level 3D shape completion on ScanNet: an IF-Net-style implicit occupancy network trained on ShapeNet, extended with human-annotated occupancy supervision (red/black mesh annotations) and per-category IoU benchmarking.

IF-Net fork + complete preprocessed ShapeNet archive (tar.xz bundles via git LFS). ShapeNet derivatives: non-commercial research use only.

PyTorch reimplementation of CheXNet: multi-label classification and CAM/Grad-CAM localization of 14 thoracic diseases on ChestX-ray14, with full-resolution (1024px) ResNet50 training.

A set of data machine learning/deep learning projects exploring various datasets — a place to test ideas, models, and modern stacks.

Modular AI toolkit designed to assist animation studio workflows through computer vision.

Echo is a computer vision app for brain tumor detection and classification on MRI scans: end-to-end computer vision pipeline, interactive web interface, and production-grade infrastructure.

Automated LLM Benchmarking on GPU - tokens/sec, latency percentiles, VRAM profiling, multi-format support (HuggingFace, GGUF, GPTQ)

https://ayinedjimi-consultants.fr

GPU-Accelerated Hash Cracking Engine - MD5/SHA1/SHA256/SHA512/NTLM/bcrypt with CUDA PyTorch acceleration, dictionary/bruteforce/mask/combination attacks, 50+ mutation rules. For authorized pentesting only.

https://ayinedjimi-consultants.fr

Extreme KV Cache Compression for LLM Inference — C++17/CUDA implementation of TurboQuant (arXiv 2504.19874). 7.5x compression, <2% quality loss.

https://ayinedjimi-consultants.fr

Detect and classify objects in Panoramax pictures

txt/pdf -> mp3

In real time inference simulator using the MVTec dataset, aproach via DINOv2 and Claude Sonnet 3.5 (Amazon Bedrock's API)