Docs
A collection of personal notes and technical guides accumulated over the years — some written from scratch, others distilled from documentation and kept as bookmarks. They are shared here in the hope that they might be useful to others.
- LLM internals: tokens, embeddings, attention How transformer-based LLMs convert text to vectors, resolve meaning through attention, and how the architecture is structured. Reference for developers.
- A practical guide to LLM variants on Hugging Face Architecture, post-training, quantization, derivation — what the labels mean, how they combine, and how to pick.
- MLOps concepts for data engineers Feature stores, drift, experiment tracking, model registry, and inference patterns — the mental model before touching any platform tooling.
- MLOps on Databricks Feature Engineering Client, MLflow, Unity Catalog model registry, Lakehouse Monitoring, and Model Serving — the Databricks ML stack for data engineers.
- Apache Spark internals: Reading notes Detailed notes on Learning Spark (O'Reilly). Architecture, Catalyst optimizer, Tungsten, Structured APIs, and performance tuning.
- SQL window functions Syntax reference for window functions — OVER, PARTITION BY, ranking, offsets, and frame clauses, with working examples.
- systemd timers Cheat sheet for systemd timers — unit file structure, OnCalendar syntax, and debugging commands.