About

Tiny Weights is a blog focused on small language models — the Gemmas, SmolLMs, Phis, and Mistrals of the world.

The goal is simple: cut through the hype and focus on what actually runs well on modest hardware. Every post is written with the assumption that you have a laptop, a cheap VPS, or a Raspberry Pi — not a data centre.

Topics covered:

• Model releases and benchmarks
• Running models locally with Ollama, llama.cpp, and LlamaEdge
• Quantization and efficiency
• Fine-tuning at small scale
• Edge and on-device inference