They just shut down the demo and I don't have the hardware to run it. Any public instances?
Alpaca demo suspended
Falling into your wing while paragliding is called 'gift wrapping' and turns you into a dirt torpedo pic.twitter.com/oQFKsVISkI— Mental Videos (@MentalVids) March 15, 2023
I'm sad guys. Every time I go online people are mean to me. Insult me and stuff. And same in person.
Or ignored. Urgh. Nobody ever cared about me. Goodbye everyone
do you like anime
then it's deserved
You would have said that regardless, so I said the opposite just to prove to myself my point
no, i would have offered my friendship if you said you did, but that bridge is burned now
gth anime hater
My feelings are hurt and my hope is lost.
you shouldn't have used somebody's good intentions as means of self deprecation
BUT supposing that you were indeed saying the opposite, you surely could name your favorite 2hu right
Alpaca.cpp? It requires a computer with a CPU and RAM. Not particularly much of either.
Which cpu should I buy for it then? I don't need light speed responses but not too slow either?
with 8 core / 16 threads ayymd cpu and 16 gb ram you will get ~250 ms per token.
and that's for 13B model.
>~250 ms per token.
What's that in wpm?
Whats the time per tokem in a 3050?
You can train the Lora's with low end GPUs. Pro tip: use hate speech prompt dataset.
There's a colab.
You are probably thinking of alpaca-lora, which is not made by the stanford team. It also is probably worse in quality, from what I've seen.