I honestly prefer if those that have Claude fix their forks keep said forks private and go on with their lives.They overwhelm the main with their commits and forces the developer to wade through their AI slop that they couldn't care less to check for quality and security themselves. Then they get angry that their commits aren't accepted as the fix.
I got so many offline installers from gog and it feels so good. I don't have to worry about a game disappearing on me or getting a patch that ruins mod compatibility years after release. And even if my internet goes down it's available.
I usually end up doing it very simple with huge /24 ipv4 networks, f.e.10.100.10.0/24 = VLAN 10 = User devices and purely internal servers10.100.20.0/24 = VLAN 20 = IoT10.100.30.0/24 = VLAN 30 = Servers that are reachable from outside10.100.40.0/24 = VLAN 40 = Guests
The main thing for me is to ensure that traffic that wants to pass between VLANs go through my firewall/router and allow Suricata to do its IPS work.
If you want a webui for the debian server that gives you logs, services, ssh terminal and more then I can recommend checking out Cockpithttps://cockpit-project.org/
If you decide you want to you can install KVM/Qemu on the debian host to get into full virtualization that way. The webui can be used to configure and manage the VMs too with https://github.com/cockpit-project/cockpit-machines
Now I can start throwing more stuff on there once I figure out backup for the game world incase I bork it.
Step 1. Find out where the docker image you run saves the volumesF.e. https://github.com/mornedhels/icarus-server saves here:VolumesVolume Description/home/icarus/drive_c/icarus Server config files and saves/opt/icarus Game files (steam download path)
Problem with smaller footprint is cooling and how audible it becomes.One idea is to use fiber optic hdmi cables and a usb extender to hide the pc away in another room.
If you want smaller footprint then the keyword to use is "Unified memory", it can be reasonable fast for 30B models and a slow thinker mode for 70B ones.
When it comes to Nvidia GPUs the VRAM is the main thing to look for.For consumer cards it is:Entry level - RTX 5060 Ti 16GB RAM with a price point around 500-550 euroMid - Buying a used RTX 3090 24GB RAM with a price point around 830 euro when I look at swedish second hand marketsHigh - RTX 5090 32GB RAM with a price point around 3500 euro
After that you end up looking at the RTX Pro Blackwell cards:Entry - RTX PRO 5000 Blackwell 48GB RAM ~5300 euroMid - RTX PRO 6000 Blackwell 96GB RAM ~10100 euro
It all depends on which models you want to run, you can definitely start playing around with Llama 3 8B and similar models with a 5060 Ti 16GB.
If you're looking at 24B-30B models you need the 24GB VRAM that RTX 3090 offers and get a larger context window if you go for the RTX 5090.
If you're looking to run Llama 3 70B then you need to go into the RTX Pro level of vram.
All of this is based on running it with Nvidia cards, there's also other setups such as Mac Studios with huge amount of RAM. They're slower but allow for much larger models at the same price point.You could also run with AMD/Intel gpus but much software is built primarily for running CUDA (and Nvidia) gpus so it's more work and not always compatible.
I know you said no "monster rack" but I don't really know what you classify as a monster. :)An ordinary gaming pc is also a good starter AI pc, so something like this allows you to do both:https://pcpartpicker.com/list/sFp4qd
When I wanted my current server to be available for more experimental loads I decided I needed a new machine and at that time all hdds had gone up in price, but a package of ds225+ with 2x16TB had not yet been price increased so I got the NAS "for free".I would've repurposed an old gaming machine otherwise.Worth looking for those kind of "deals" where someone has forgotten to increase a bundle price.
I added 16GB RAM (Crucial 16GB (1x16GB) DDR4 3200MHz CL22 SODIMM) and the github.com/007revad/Transcode_for_x25 fix to enable proper transcoding for a docker install of Jellyfin and enough RAM to add other "production load" dockers in the future.Now I can play around all I want on my old server while Jellyfin and SMB shares remain available on redundant drives with automated backups. I've also blocked the NAS from the internet in my router now to ensure that Synology can't "fix" the transcode fix.
Depends on which Piefed instance you're trying to contact and for what reason.In the cases of feature requests a good place to start would be piefed.social/c/piefed_meta - !piefed_meta@piefed.social
What I meant with lifestyle was the fact that vegan used to mean not using any products that cause animal suffering, be it a hamburger, leather shoes or makeup tested on animals. None of it is vegan.
While I don't believe IaaS to be selfhosting I do believe self-managed services on IaaS should be allowed here. It's the same software stack and requires the same skills so both parties gain from having the discussion in the same place.Not because I think selfhosting is a badge but because I think it makes sense to call things for what they are.
But I'm an old grumpy who thinks ovo-lakto vegetarians shouldn't have been allowed to steal the meaning of vegetarian or vegetarians steal it from vegans (and now we no longer got a word to describe old school vegans that makes it a lifestyle not a diet.)
If they already to the work for esm-apps repo then they could at least send those fixes over to the universe repo until the release version is EoL one would think. On the other hand I have no idea what lives in universe and what lives in main.
I honestly prefer if those that have Claude fix their forks keep said forks private and go on with their lives.They overwhelm the main with their commits and forces the developer to wade through their AI slop that they couldn't care less to check for quality and security themselves. Then they get angry that their commits aren't accepted as the fix.