Terrasque

I still use http a lot for internal stuff running in my own network. There’s no spying there… I hope … And ssl for local network only services is a total pita.

So I really hope browsers won’t adapt https only

But even if you use GoMommy extra super duper triple snake oil security checked ssl cert, if I trick LetsEncrypt to sign a key for that domain I still have a valid cert for your site.

I doubt the disk will bottleneck at 40mb/s when doing sequential write. Torrent downloads are usually heavy random writes, which is the worst you can do to a HDD.

I was trying to find an article I read about a year ago, about an experiment where AI was assisting a doctor. Where it suggested questions and possible diagnosis for the doctor to look into.

IIRC the result was both faster and more accurate diagnosis. Too bad I can’t find it again now :(

You’re not great taking medical advice from a doctor either, seeing how often they’re wrong.

I remember back in the day this automated downloader program… the links had a limit of one download at a time and you had to solve a captcha to start each download.

So the downloader had built in “solve other’s captcha” system, where you could build up credit.

So when you had say 20 links to download you spent some minutes solving other’s captchas and get some credit, then the program would use that crowdsourcing to solve yours as they popped up.

Sell them to zoomers as 3d save button coasters. $19.95 each

Llama3 8b can be run at 6gb vram, and it’s fairly competent. Gemma has a 9b I think, which would also be worth looking into.

And just to top it off, make this pythonscript a dialect of rust

Better background backups

Rework background backups to be more reliable

Hilarious for a system which main point / feature is photo backup

🫰🤙🫵👌✊🫳🫸🤲🤌

I worked on one where the columns were datanasename_tablename_column

They said it makes things “less confusing”

I mean, I totally agree with you. But that also kinda ignores all the useful things a dog can be trained to do.

It’s less the calculations and more about memory bandwidth. To generate a token you need to go through all the model data, and that’s usually many many gigabytes. So the time it takes to read through in memory is usually longer than the compute time. GPUs have gb’s of RAM that’s many times faster than the CPU’s ram, which is the main reason it’s faster for llm’s.

Most tpu’s don’t have much ram, and especially cheap ones.

Reasonable smart… that works preferably be a 70b model, but maybe phi3-14b or llama3 8b could work. They’re rather impressive for their size.

For just the model, if one of the small ones work, you probably need 6+ gb VRAM. If 70b you need roughly 40gb.

And then for the context. Most models are optimized for around 4k to 8k tokens. One word is roughly 3-4 tokens. The VRAM needed for the context varies a bit, but is not trivial. For 4k I’d say right half a gig to a gig of VRAM.

As you go higher context size the VRAM requirement for that start to eclipse the model VRAM cost, and you will need specialized models to handle that big context without going off the rails.

So no, you’re not loading all the notes directly, and you won’t have a smart model.

For your hardware and use case… try phi3-mini with a RAG system as a start.

So you’re saying it’s already feature complete with most json libraries out there?

https://y.yarn.co/00b19fc5-cc98-4af3-8298-93092ffc4ff6_text.gif

Who are you?

What do you want?

Also, I think good and bad is a bit fluid there. It’s just people with different agendas. Well, except emperor Cartagia. And perhaps Bester.

Yep, I usually make docker environments for cuda workloads because of these things. Much more reliable

Yeah, it also has the effect that when starting up say a new postgres or web server is one simple command, a few seconds and a few mb of disk and ram, you do it more for smaller stuff.

Instead of setting up one nginx for multiple sites you run one nginx per site and have the settings for that as part of the site repository. Or when a service needs a DB, just start a new one just for that. And if that file analyzer ran in it’s own image instead of being part of the web service, you could scale that separately… oh, and it needs a redis instance and a rabbitmq server, that’s two more containers, that serves just that web service. And so on…

Things that were a huge hassle before, like separate mini vm’s for each sub-service, and unique sub-services for each service doesn’t just become practical but easy. You can define all the services and their relations in one file and docker will recreate the whole stack with all services with one command.

And then it also gets super easy to start more than one of them, for example for testing or if you have a different client. … which is how you easily reach a hundred instances running.

So instead of a service you have a service blueprint, which can be used in service stack blueprints, which allows you to set up complex systems relatively easily. With a granularity that would traditionally be insanity for anything other than huge, serious big-company deployments.