| Main site: | https://loysharosen.com |
| Blog: | https://blog.loysharosen.com |
| Main site: | https://loysharosen.com |
| Blog: | https://blog.loysharosen.com |
Where's the hype around Local LLMs?
Local inference seems like the most reasonable way forward, in order to avoid security and privacy concerns.
I'm currently trying out Ollama and LMStudio, primarily using Qwen3:8B. It's not fast with my hardware (CPU only), but it's exciting to see it works.
Every month the United States provides even more reason to boycott its products and companies — the strikes on Iran being just the latest.
In light of that, it’s a great time to ditch the products and services of US tech companies. I made a comprehensive guide to help you that I’ve been regularly updating for months.
This is brilliant advice. I've been using an initial comma for all my custom commands since reading this. Especially useful in combination with tab completion.
I used to be an avid runner. PB in a race is a half marathon at pace 4:30 min/km.
Various commitments got in my way, stopped running.
Last year, my wife gave me a beautiful son.
I'd like for him to grow up seeing his father run. To see that with methodical practice, you can literally go far. So I signed up for a #marathon this November.
Here we go. 💪
I started learning #Python three years ago and, surprisingly, loved it.
As this year closes, I've switched over to Debian 13 for my main laptop and have a Proxmox homelab running 24/7, with a plethora of useful services.
In retrospect, tech used to feel patronizing and intrusive. I love how #FLOSS has made my digital world decluttered, deliberate.