Nice nice nice. I have a playlist of good russian rap, now I think I should start a collection of general foreign rap because well yea, I'll edit it in a sec
"And here you find civilized man. Civilized man refused to adapt himself to his environment. Instead he adapted his environment to suit him. So he built cities, roads, vehicles, machinery. And he put up power lines to run his labour-saving devices. But he some how didn't know when to stop.
The more he improved his surroundings to make life easier the more complicated he made it.
So now his children are sentenced to 10 to 15 years of school, just to learn how to survive in this complex and hazardous habitat they were born into.
And civilized man, who refused to adapt to his surroundings now finds he has to adapt and re-adapt every hour of the day to his self-created environment." - The Gods Must Be Crazy
It has some perks, is a bit more expressive in some cases, but overall is trained on really noisy data, uses more memory, and isn't that fast - I'm talking about the (7b?) version that they released then removed quickly (vibevoice-community on github) - I still use chatterbox turbo and sometimes qwen TTS.
Agree on the framework, last week you could get a strix halo for $2700 shipped now it's over $3500, find a deal on a NVME and the framework with the noctua is probably going to be the quietest, some of them are pretty loud and hot.
I run qwen 122b with Claude code and nanoclaw, it's pretty decent but this stuff is nowhere prime time ready, but super fun to tinker with. I have to keep updating drivers and see speed increases and stability being worked on. I can even run much larger models with llama.cpp (--fit on) like qwen 397b and I suppose any larger model like GLM, it's slow but smart.
https://www.youtube.com/watch?v=_fgypD0ODA8 - This album is more chill, I like 17 По району (My District) and 9 Оп, мусорок (Look out, cop)