Pinned
Quantization can make an LLM 4x smaller and 2x faster, with barely any quality loss. But what *is* it? @samwhoo crafted a beautiful interactive essay explaining it from first principles, aimed at coders, not mathematicians.
ngrok
2,013 posts
ngrok
@ngrokHQ
- This prompt caching post is still a banger and perhaps the perfect read for a Friday afternoon if you're light on token spend.
- In 1996, AOL went down for 19 hours. It pushed "NASA finds evidence of life on Mars" off the front page of the New York Times. We sponsored an SRE to write a human postmortem of that outage: the stuff five whys leave out.
- Gather round and come on a journey with us back to 1996 when AOL went down ngrok.com/blog/aol-was-d…
GIF - Day 2 of vibe coding
- Yup, @samwhoo is a wild one for this. But we want to teach the invisible layers of modern software in ways grounded in reality. Ways that'll last for a long, long time. Sometimes that means 100k lines of TypeScript just to earn the right to start telling the story at all.Sam has also talked at length with our very own @ryan_blunden on how the build has come together, the tools he's used, and all the k8s internals he *had no idea about*. youtube.com/watch?v=wRSyz0…
- ngrok reposted
How can u describe yourself as a backend engineer if you’ve never heard of ngrok - ngrok reposted
Made the demo app for my Kubernetes-in-the-browser library more fun. All runs in the browser, no server-side stuff (you have to create images using a JS API I’ve made, they’re not real Docker images). Requests you see travelling between pods are actual inter-pod traffic.
00:00 - Congrats to the OpenHands team! We're just chuffed seeing ngrok in the mix as the best way to make self-hosted Agent Canvas installs both secure and accessible from anywhere.PR review on a schedule. Issue triage every morning. Repo summary every Monday. Your agent can do all of it. It just needed somewhere to live. Introducing Agent Canvas 🙌 hubs.la/Q04lBvDf0
- Self hosting n8n with docker and ngrok✅
- ngrok reposted
I bought some Raspberry Pis and OLEDs and a switch and fun cable-y bits to build a working little internet on a piece of plywood in my office. Time to build everything I've half-understood at best, figure it all out, and chronicle everything I learn along the way.
00:00 - HERMES + ELEVENLABS TURNED A $3,000/MO ANSWERING SERVICE INTO A ~$145/MO AI THAT NEVER MISSES A CALL. A missed call after 6pm isn't a missed call. It's a customer booking with your competitor. Most businesses pay $2,000-$4,000/mo for a human to make sure that doesn't happen.
00:00
