<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>bilawal.net</title><description>Thoughts on infrastructure, platforms &amp; building things</description><link>https://bilawal.net/</link><language>en-gb</language><item><title>LLM fundamentals: what the papers actually say</title><link>https://bilawal.net/blog/llm-fundamentals-18-papers/</link><guid isPermaLink="true">https://bilawal.net/blog/llm-fundamentals-18-papers/</guid><description>A technical introduction to how LLMs work, how they are trained, and what fine-tuning really does. Built from 18 papers read together, organised around one idea: post-training surfaces knowledge, it does not add it.</description><pubDate>Thu, 30 Jul 2026 00:00:00 GMT</pubDate><category>ML-AI</category><category>Software-Engineering</category></item><item><title>RAG: a diagnostic reference for real systems</title><link>https://bilawal.net/blog/rag-everything-you-need-to-know/</link><guid isPermaLink="true">https://bilawal.net/blog/rag-everything-you-need-to-know/</guid><description>A working reference for retrieval-augmented generation: how to find the stage that is failing, what to measure, and when RAG is the wrong tool. Retrieval, chunking, metadata, hybrid search, reranking, evaluation, observability, permissions, freshness, cost, and the alternatives.</description><pubDate>Thu, 30 Jul 2026 00:00:00 GMT</pubDate><category>ML-AI</category><category>Software-Engineering</category></item><item><title>I trained a 1.5B model to deobfuscate Python. It scored 93% and survived the obfuscated Python Olympics.</title><link>https://bilawal.net/blog/semrec-python-deobfuscation/</link><guid isPermaLink="true">https://bilawal.net/blog/semrec-python-deobfuscation/</guid><description>SemRec: a 940MB fine-tuned Qwen2.5-Coder model that recovers clean semantics from obfuscated Python. 93.36% overall, 11/11 on competition-grade obfuscation.</description><pubDate>Sat, 20 Jun 2026 00:00:00 GMT</pubDate><category>ML-AI</category></item><item><title>PetLM: teaching a tiny language model to be a virtual pet</title><link>https://bilawal.net/blog/petlm-tiny-virtual-pet-language-model/</link><guid isPermaLink="true">https://bilawal.net/blog/petlm-tiny-virtual-pet-language-model/</guid><description>A 25M parameter model, a strict little DSL, and a browser toy that runs locally with WebGPU or WASM.</description><pubDate>Wed, 13 May 2026 00:00:00 GMT</pubDate><category>Projects</category></item><item><title>Self-hosted SMS gateway with an old Android phone</title><link>https://bilawal.net/blog/self-hosted-sms-gateway/</link><guid isPermaLink="true">https://bilawal.net/blog/self-hosted-sms-gateway/</guid><description>How I built a self-hosted SMS gateway using a spare Android phone, a custom FastAPI relay, and Tailscale for secure, zero-cost text messaging from my AI agent.</description><pubDate>Sat, 28 Mar 2026 00:00:00 GMT</pubDate><category>Projects</category></item><item><title>educated.space — AI-Powered Educational Resource Generation</title><link>https://bilawal.net/blog/eduforge-ai-education-platform/</link><guid isPermaLink="true">https://bilawal.net/blog/eduforge-ai-education-platform/</guid><description>AI-powered K-12 educational platform. Generates curriculum-aligned worksheets, quizzes, and info sheets. Live at https://educated.space</description><pubDate>Sat, 14 Mar 2026 00:00:00 GMT</pubDate><category>Projects</category></item><item><title>Fixing Quran Audio Segment Timings with WhisperX</title><link>https://bilawal.net/blog/fixing-quran-audio-segments-with-whisperx/</link><guid isPermaLink="true">https://bilawal.net/blog/fixing-quran-audio-segments-with-whisperx/</guid><description>How I used WhisperX forced alignment to fix inconsistent word timings in a Quran reading app.</description><pubDate>Sat, 21 Feb 2026 00:00:00 GMT</pubDate><category>Projects</category></item><item><title>QuranZen: A Hybrid Architecture Quran App</title><link>https://bilawal.net/blog/quranzen/</link><guid isPermaLink="true">https://bilawal.net/blog/quranzen/</guid><description>Building a full-stack Quran exploration app with React 19, FastAPI, and a hybrid database design.</description><pubDate>Wed, 28 Jan 2026 00:00:00 GMT</pubDate><category>Projects</category></item><item><title>Running LLMs on a Steam Deck in 5 minutes</title><link>https://bilawal.net/blog/running-llms-on-steam-deck/</link><guid isPermaLink="true">https://bilawal.net/blog/running-llms-on-steam-deck/</guid><description>A step-by-step guide to setting up your Steam Deck to run local Large Language Models (LLMs) using llama.cpp and a Distrobox container.</description><pubDate>Tue, 09 Sep 2025 00:00:00 GMT</pubDate><category>ML-AI</category></item><item><title>Fine-tuning a 3B Model for Transcript Analysis: Better Than 70B Models?</title><link>https://bilawal.net/blog/finetuning-llama32-3b-for-transcripts/</link><guid isPermaLink="true">https://bilawal.net/blog/finetuning-llama32-3b-for-transcripts/</guid><description>How I fine-tuned Llama 3.2 3B to process transcripts locally, achieving better results than models 5-8x larger through specialized training.</description><pubDate>Mon, 01 Sep 2025 00:00:00 GMT</pubDate><category>ML-AI</category></item><item><title>ToddlerTones: offline, toddler-proof soundboard</title><link>https://bilawal.net/blog/toddler-tones/</link><guid isPermaLink="true">https://bilawal.net/blog/toddler-tones/</guid><description>Designing and building a distraction-free Android app in Kotlin/Compose with low-latency audio and a gentle &apos;Find it&apos; game.</description><pubDate>Tue, 26 Aug 2025 00:00:00 GMT</pubDate><category>Side-Projects</category></item><item><title>Reluctant Buddha: Fine-Tuned LLM for shitty insights</title><link>https://bilawal.net/blog/reluctant-buddha/</link><guid isPermaLink="true">https://bilawal.net/blog/reluctant-buddha/</guid><description>Turning a 1B LLM into a deranged cosmic entity running on CPU only.</description><pubDate>Sat, 08 Mar 2025 00:00:00 GMT</pubDate><category>Projects</category></item></channel></rss>