I do test a bunch of models, as of today I would say it is dolphin-2_2-yi-34b

I’ve used the gguf version of Xwin-MLewd-13b and it’s the smartest 13b I’ve found so far

I’ve played with it for a bit and I agree with most people here. It seems to be as smart as a 70b, which is a big deal IMO.

1 - Horny stuff

2 - Waiting for a smart enough model that can be fed a JSON in a fixed interval, providing to it information of the environment, as in a simplified version of our brain getting a constant stream of info. That object sometimes will contain user input or not, (user input as in the regular questions we asks these models). Some models can keep up with this for a bit but eventually lose track. If anyone has done anything like this or has any tips / suggestions , I’ll happily accept them.