Ml on martinuke0's Blog https://martinuke0.github.io/tags/ml/ Recent content in Ml on martinuke0's Blog martinuke0's Blog https://martinuke0.github.io/%3Clink%20or%20path%20of%20image%20for%20opengraph,%20twitter-cards%3E https://martinuke0.github.io/%3Clink%20or%20path%20of%20image%20for%20opengraph,%20twitter-cards%3E Hugo -- 0.152.2 en Wed, 16 Sep 2026 16:01:49 +0000 Hands‑On Build Guide: Page‑Transformer Inference with Greedy KV‑Cache Eviction and Paged Attention in Numpy https://martinuke0.github.io/posts/2026-09-16-handson-build-guide-pagetransformer-inference-with-greedy-kvcache-eviction-and-paged-attention-in-numpy/ Wed, 16 Sep 2026 16:01:49 +0000 https://martinuke0.github.io/posts/2026-09-16-handson-build-guide-pagetransformer-inference-with-greedy-kvcache-eviction-and-paged-attention-in-numpy/ Build a minimal page‑transformer inference engine from scratch using NumPy, demonstrating greedy KV‑cache eviction and paged attention for a portfolio project that signals systems skills to hiring managers. Build a Minimal Mixture-of-Experts Inference Engine in Pure Python https://martinuke0.github.io/posts/2026-09-14-build-a-minimal-mixture-of-experts-inference-engine-in-pure-python/ Mon, 14 Sep 2026 16:01:34 +0000 https://martinuke0.github.io/posts/2026-09-14-build-a-minimal-mixture-of-experts-inference-engine-in-pure-python/ Learn to build a minimal mixture-of-experts inference engine with a learned top-k router, auxiliary load-balancing loss, and feed-forward experts in pure Python.