This is my lab, where I am messing with config tweaks, weights, parameters, model stack orders etc. These are not meant to be full release models but experiments and parts of larger merges. Many of these are pitted against each other for comparisons on how different settings affect the final merge. They are of course free to use, but don't expect models to be of the same quality as those on my main repo: https://huggingface.co/Tarek07