work Xe-Forge A multi-stage, LLM-powered pipeline that automatically optimizes Triton GPU kernels for Intel GPUs. DeepMath Simple framework for training and evaluating math reasoning agents using local models, GRPO and vLLM. FastRAG A fast, scalable retrieval-augmented generation framework built on Haystack and HuggingFace, with Intel hardware optimizations. RAG-FiT Open-source framework that helps train large language models to perform better at retrieval-augmented generation workflows. SQuARE Research paper: SQuARE uses self-interrogation to decompose questions, outperforming chain-of-thought prompts. DictaLM 2.0 Multi-node pretraining of a Hebrew LLM with distributed data and model parallelism. fun otzar I kept forgetting what I'd already written, so I built something to search my own notes. mu MCP Server MCP server POC for searching emails locally using the mu CLI tool. mu4easy Emails Emacs package for multi-account email setup, based on the mu4e package. Git Commit Lines Graph Python CLI tool for plotting project's line-count as a function of time, XKCD style. IMDB CLI browser CLI tool to browse IMDB, written in Go.
