Actually yes. Requested through Seerr, added automatically to Jellyfin, streamed to the Kodi box. Simpler interface than Netflix. Setting it up is though not simple, but if a toddler has learned to use the TV remote they’d be able to use it. And there’s about 30TB of content already so Seerr is rarely needed.
I am pretty sure my Debian box that runs Kodi has fewer trackers than my TV.
Definitely, I had also done a lot of investigation and assumed that when I isolated the problem to the RAM there was almost nothing more I could do expect cry and wait for the inevitable torturous moment I’d have to open my wallet amid sky high prices.
The server has now been running non-stop without an issue for 30 days, so I assume the problem was fixed by the suggestion!
BERT is a special case. The academic consensus is rightly considering it an LLM since it operates on text but the industry doesn’t consider it an LLM since LLM in its colloquial meaning has drifted to mean a model generating text (which BERT doesn’t). Whisper is not an LLM since it’s working on audio, so it is an Automatic Speech Recognition model (ASR).
Text in = LLM in academic circles (BERT and GPT)
Text in, text out = LLM in all circles (GPT only)
Audio in, text out ≠ LLM (Whisper)
Now, we can argue whether we want to accept the colloquial meaning of LLM, but the fact is that Whisper is not an LLM. And neither ASR nor BERT causes even a fraction of the damage GPT is doing, but that’s another discussion.
LLM is a subset of NLP. All LLMs are NLP but not all NLP are LLMs. To be more precise, LLM is a method of achieving NLP while NLP is the goal. Whisper is just a specialised speech-to-text engine. It doesn’t have the text reasoning engine that is the core of LLMs. Yes, it has a dataset to inform it’s encoder and decoder, but that doesn’t make it an LLM. If you strip out the encoder, then you’re getting closer.
Ah, that explains it
Neocities? I thought that was dead
Update, kernel panic happened today, so I updated the BIOS settings:
All the initial values were Auto values. I’ll let it sit for at least a month with no reboots before calling it fixed. I’ll set a reminder
Which TV show if I may ask?
I’ll try to remember to update you in 1-2 months then!
All the documentation/readme is AI generated, so I guess the difference is AI. I can’t speak for the code itself since I’m not a TS expert but almost the entire code is dumped in an initial commit which I definitely wouldn’t do, even when building the first prototype. But that’s not a dead giveaway.
For a project I liked I saw that the translation of the UI was only about 1% complete for my native language. I decided to dedicate the weekend to get it above 95%. This was apparently the part that seemed too hard for most to make a dent in, but across several new versions my language is still 100%. I think my groundwork made it easier for some sporadic contributors to get a small bit of dopamine when updating the few new strings in a release to see it hit 100% again.
I’ll try that next time it dies, see if it helps
The server currently runs on 32gig but I was planning on upgrading to 64 or 128gig DDR5. Could go as low as 16 if need be, but the the price is still hurting more than having to reboot 2-4 times a month.
Only thing I wish for Christmas is to be able to replace the RAM and mobo in one of my servers. It is failing with a kernel panic about once a week and it is an old DDR3 platform I planned to upgrade back in January. I have isolated the issue to a hardware fault but left without a solution…
I got my setup because the experience on streaming services sucked but I get it if you still use those. In that case an Android box is the closest you’d get without compromising on ease of use (just don’t buy the no-name ones from AliExpress).