Local LLM. Chapter 1

Mon, 23 Mar 2026 23:54:03 +0100

Hello there! 🖖

Recap

Tool: llama.cpp
OS: Windows

1. [Hugging Face Models](https://huggingface.co/models)
2. App: llama.cpp
3. Model: SmolVLM-500M-Instruct-GGUF
4. Win + R -> winget install llama.cpp
5. CMD -> llama-cli -hf ggml-org/SmolVLM-500M-Instruct-GGUF:Q8_0
/// Cleanup
1. winget list llama.cpp
2. winget uninstall --id ggml.llamacpp

Step-by-Step Implementation

Quick Way to Run llama.cpp on Windows

Navigate to Hugging Face Models

App: llama.cpp
Model: SmolVLM-500M-Instruct-GGUF

Click “Use this model” -> “llama.cpp”

A modal window will appear with instructions on how to install llama.cpp and a command to run the selected model.

Use WinGet to install and run:

# Press Win + R and type powershell (or use Terminal/CMD)
# 1. Install llama.cpp via Windows Package Manager
winget install llama.cpp
# 2. Download and run the model directly from Hugging Face in the console
llama-cli -hf ggml-org/SmolVLM-500M-Instruct-GGUF:Q8_0

Hello World!

Wed, 11 Mar 2026 13:01:13 +0100

Hello there! 🖖

My name is Aleksandr, and I am a software engineer.

Why this blog? Well, LLMs changed everything. To be honest, they’ve blown me away.

For me, coding isn’t just a job — it’s a hobby. Over the years, I’ve felt that “wow” effect only a few times:

When my dad took my brother and me to his lab, showed us a room full of “Corvette” computers, and launched a game from an 8″ floppy disk 😅;
When I first saw Windows 95;
When I created a snowman in QuickBasic following my dad’s instructions;
When I wrote my first Delphi app and everything looked so “beautiful” with almost no effort;
When I discovered version control systems (and the nightmare of manually archiving folders by date became a thing of the past);
And last but not least, when I tried LLMs for the first time.

The world of LLMs and the infrastructure around them is changing so fast that I struggle to keep up… In this blog, I want to share my thoughts and findings. I’m just a regular coder, not some ‘Holy-Logic-Batman’ genius, so everything here is filtered through my perception and is by no means the ultimate truth.

Posts on The .NET + LM Sandbox

Local LLM. Chapter 1

Hello there! 🖖

Recap

Step-by-Step Implementation

Quick Way to Run llama.cpp on Windows

Hello World!

Hello there! 🖖