In September, machine learning engineers at Mozilla filed a bug report indicating that Firefox was consuming excessive memory and CPU resources while running Microsoft’s ONNX Runtime (a machine learning library) compiled to WebAssembly. This post describes how we addressed this … Read more
Introduction When modeling a Postgres database, you probably don’t give much thought to the order of columns in your tables. After all, it seems like the kind of thing that wouldn’t affect storage or performance. But what if I told … Read more
Guile is a rather niche language that I love dearly. Guile is a Scheme dialect that features an advanced optimizing bytecode compiler, a JIT compiler, and a modest set of developer tools for inspecting and debugging. Through my time spent … Read more
Optimizing a bignum library for fun – Austin Z. Henley Austin Z. Henley Associate Teaching ProfessorCarnegie Mellon University 7/9/2024 I’m down a rabbit hole of learning how bignums work. In the first post, making a bignum library for fun, I … Read more
Recently our our 7 piece Syzygy tablebase server was struggling to complete its periodic RAID integrity check while being hammered with tablebase requests. We decided to try a new approach, using dm-integrity on LVM. Now, instead of periodically checking every … Read more
Over the past year or two, I’ve worked on a project to convert a massive dataset into an SQLite database. The original data was in a compressed binary format known as OSMPBF, which stands for OpenStreetMap Protocol Buffer Format. This … Read more
For one of my programming classes this semester, my task was writing a program in Python. It was only an introductory course, so it didn’t have to be anything fancy. Whenever I get assignments like this one, my choice is … Read more
This article focuses on improving the modeling performance of LLMs by finetuning them using carefully curated datasets. Specifically, this article highlights strategies that involve modifying, utilizing, or manipulating the datasets for instruction-based finetuning rather than altering the model architecture or … Read more
In this article I will take a look at the classic concurrent ring buffer and how it can be optimized to increase throughput. I will show you how to significantly increase throughput from 5.5M items/s to 112M items/s, beating the … Read more
Your default hash table should be open-addressed, using Robin Hood linear probing with backward-shift deletion. When prioritizing deterministic performance over memory efficiency, two-way chaining is also a good choice. Code for this article may be found on GitHub. Most people … Read more