Blog
Machine Learning & AI
Discover our articles on IT performance, cloud, data science and industry best practices.
All articles
Running a private Qwen3.6 Coding Agent on Scaleway
This is the Scaleway sequel to two previous posts on self hosted coding agents: the AWS run where we deployed a 27B model on a g5.xlarge, and the consumer GPU experiment where we…
![Running a Local LLM on a Consumer GPU [8 GB VRAM]](/_next/image?url=%2Fblog%2F20260418_running-a-local-llm-on-a-consumer-gpu-8-gb-vram%2Fimg%2F04_gpu_poor.jpeg&w=3840&q=75)
Running a Local LLM on a Consumer GPU [8 GB VRAM]
Any spare VRAM? In a previous post, we ran Qwen3.5 27B on an AWS g5.xlarge with an NVIDIA A10G (24 GB VRAM). This follow up covers the same workflow on a consumer laptop GPU with…
Running a Local LLM Coding Agent on AWS
We deploy Qwen3.5 27B on a single GPU instance, serve it with llama.cpp, wire it to OpenCode as a coding agent frontend, and run ToolCall 15 to measure tool calling quality. The…
Zipf's Law on "La Comédie humaine"
In this Python notebook, we're going to explore Zipf's law as applied to words. Zipf's law is an empirical law stating that when a list of measured values is sorted in decreasing…
Analyzing tramontane wind frequency
The tramontane wind, a defining feature of Pyrénées Orientales region, is known for sweeping through the south of France with its characteristic force and direction from the…
Calculating walking isochrones with Python
In this blog post, we'll explore how to calculate walking isochrones using Python, taking into account the slope of the terrain. An isochrone is a line connecting all points that…
Calculating daily mean temperatures with scikit-learn
The goal is of this post is to predict the daily mean air temperature TAVG from the following climate data variables: maximum and minimum daily temperatures and daily…
Built-in Expectations in Great Expectations
Great expectation is a Python tool for data testing, documentation, and profiling. Here is a figure from the documentation describing its purpose: Great Expectations makes it easy…
Saving a tf.keras model with data normalization
Training a DL model might take some time, and make use of some special hardware. So you may want to save the model for using it later, or on another computer. In this short Python…