You'll want to install a reverse proxy of your choice on the VPS. Have clients access it over ipv4 and configure it to proxy pass it to your ipv6 address. Nginx at least is capable of doing ipv4 & ipv6 -> ipv4, I think the inverse should also be possible.
exu
joined 1 year ago
You can use OpenCL instead of ROCm for GPU offloading. In my tests with llama.cpp that improved performance massively.
Definitely do benchmarks for how many layers you can offload to the GPU. You'll see when it's too many, as performance will crater.
By launching llama.cpp as a server you'll actually be able to continue to use openwebui as you currently have.
It would probably take someone to sue them, but they would have to implement it.
It's probably still more efficient to keep a 192k opus and a 320k mp3 around than one flac.
Cron sadly does not offer precision in the seconds range.
I don't know how good it is, but for Turing and later GPUs there's a new official open source driver being developed. You'd likely have to use a more bleeding edge distro to get that.
VR works depending on your headset. Index is fine, Oculus doesn't work.
You could make a usb stick of your desired distro and test everything without permanent changes before commiting.
With Mozilla Location Services going away soon, I'm wondering at the legality of using the Apple data to seed a replacement.
Yeah, they're more power hungry, but they're also way more performant than a pi 4.
VAAPI is the "standard" interface for hardware en-/decoding on Linux. It should work with any GPU using the open source drivers and mesa.
I don't know how QSV can be installed; AMF, the AMD equivalent, is limited to their proprietary driver.
Besides maybe confusing the codecs, hardware encoders, especially the AMD ones, are always less space efficient than software encoders.
If you want to convert video for long-term storage, please use a software encoder.
AirVPN, but only for its port forwarding to sail the high seas.
Creative Commons is likely more appropriate for FOSR.