Nope, i am running a 7B version on a pretty good CPU and it's so slow you want to kill yourself. But it's uncensored for sure. to run a 13B model you have to own at least a Nvidia RTX3060 GPU. Most people here don't know how to install Open Source repositories locally on their computers. So stop saying that it doesn't make sense to pay for it. Like anything else in the world we pay for convenience and it does make sense totally.doesnt matter, that just means people are not informed, with just few clicks you can run pretty good uncensored models based on llama with 30B parameters and more with just your CPU
It dont tell you anything illegal, whats the fun in it?The intention is not to make it reply with unethical approach, instead, if someone asks for something not really common. It will warn but respond, not censuring.
I also don't know and was wondering if there is any legal regulation on this topic.
Are really people buying that shit?
Look at polls result....
There are already plenty of those available, open source and free, there is no reason for anyone to buy it
Yeah but you need to run inference on them, costs and arm and a leg, GPUs aren't cheapThere are already a lot of uncensored models on Github
Yep that's right, my badYeah but you need to run inference on them, costs and arm and a leg, GPUs aren't cheap
yeah the 70B version needs 16 gb+ of ram just to utilize it . Running models is not cheapI've used llama2-uncensored and it's pretty good for some controversial topics, but its slow on my laptop i7 8750h, 32GB ddr4-2666mhz, 1060 Max Q 6GB.