LLM Went kind of roug on me. Anyone else?
Was discussing with it about the Krishnamurti (to reveal the skills and state...) and have no clue why it suddenly changed the tone...
It went in self-destruction mode :)) Can´t load it on LM Studio anymore, getting insufficient memory error despite having lots of free... I guess it was just to much to handle :))
@MikeZeroTango Hi, this seems like an issue with the Q4 for some reason, I don't know if it happens with any other quant. The Q4 seems to also lose some of the fine tuning uncensorship. I will look into it for the next V3. If you can manage to run Q8 with your hardware I would recommend try it out. You can also try out the new 3.1 release I've done.
Edit: It seems in your prompt, you added a 0 to the end of your prompt. That may have contributed to the already issued quantization. ^^
Hi Orenguteng, thank you for the kind reply! Yea, the zero was definitely not intentional :/, was considering this as well... I was even able to run f16 without any problems until the Q4 did that. Now I am unable to run any kind of model in the LM Studio!? Getting the message, I don't have enough of VRam available where there is actually 48GB available :) P.s. Looking forward for the new V3 :))