mail

Local LLM spam classifier — model shootout

by

We tested 9 local LLM models as spam classifiers on an AMD EPYC 4545P running Ollama. Here’s what we found. After setting up the SpamAssassin + Ollama integration (see the previous post), the obvious next question was: is qwen2.5:7b actually the best choice, or did we just get lucky picking it first? So we ran … Read more

Local LLM for mail spam/ham classification Part 1

by

In which I spend a Sunday morning asking a local AI to tell me whether emails about cheap Viagra are, in fact, about cheap Viagra. Spoiler: the 0.5b model cannot. The Problem SpamAssassin is great. Bayes is great. RBLs are great. But spam has gotten weird. Greek-language product spam from hacked domains with randomized subfolders … Read more

DNSBL: how does it work, and how to create and build an own

by

A DNSBL is a DNS based [DNS = Domain Name System] blackhole list, which can be used as countermeasure against unwanted mail spam. One of the most efficient ways to block mail spam is to do it on mail server level by denying incoming connects from spam sources, where the contacting remote machine is identified by its IP address … Read more

Building a Poor Man’s Barracuda – cPanel edition

by

This article will show you how to replicate most of the functionality of a Barracuda Spam Firewall on you cPanel server. The BSF has 12 layers that messages would be process through as illustrated in the graphic below. I’ll show you how to replicate most of these steps using customizations to the Exim and Spam Assassin. This … Read more