A laptop screen showing Python code and a flowchart of a tokenizer

Hands‑On BPE Tokenizer: Pure‑Python Incremental Merge Learning with Trie Lookup

This post walks you through building a from‑scratch BPE tokenizer in pure Python, complete with online incremental merge learning and a trie for fast tokenization. You’ll end up with a runnable library you can showcase on LinkedIn or a résumé.

October 21, 2026 · 9 min · 1824 words · martinuke0
Python code editor with a neural network diagram

Hands‑On Build Guide: Pure‑Python GGUF Weight Loader & Minimal Transformer Inference Engine

A step‑by‑step guide to loading GGUF weights from safetensors and running a minimal transformer in pure Python, ready to showcase on your CV.

October 19, 2026 · 9 min · 1917 words · martinuke0
A developer's workstation with code visualizing vector embeddings

Build a From-Scratch IVF-Flat Vector Index with Exact Reconstruction and Greedy Search in Pure Python

Learn how to construct a minimal yet functional IVF-Flat index, implement greedy search, and verify exact reconstruction — all in pure Python, ready to showcase on your CV.

October 19, 2026 · 7 min · 1454 words · martinuke0
A laptop screen showing Python code for an attention engine

From‑Scratch Paged Attention Engine: LRU KV‑Cache Eviction and GPU‑Free Sliding‑Window Simulation in Python

Build a from‑scratch paged attention engine in pure Python, complete with LRU‑based KV‑cache eviction and GPU‑free sliding‑window simulation. Great for demonstrating systems‑level ML engineering skill.

October 16, 2026 · 8 min · 1611 words · martinuke0
A sleek Python logo over a compressed text ribbon.

Dynamic Prompt Compression System with Learned Importance Scores

A hands‑on guide to building a from‑scratch prompt compression system that learns token importance and truncates context for fixed‑size LLM windows.

October 15, 2026 · 10 min · 1923 words · martinuke0
Feedback