"Experimental Karcher-Mean Mix" using "Mergekit".

Don't ask me how i could merge it with only "16GBs of RAM", all models are mixed equaly and all their data have same importance. This is a mix of some known "RP Uncensored" models. I'm not responsible of the use of this.

The recommended max context size is "8K (8192)". Some of the models in the mix, used a "fake rope_theta hack" to support "1M" of context size, and this model coppied that configuration. This are not a real numbers, and would NOT give you a "meaningful long‑context behavior".

I couldn't find a way to have the "Tokenizer clean". I don't know if it's about a "mergekit bug" or something related to the models, but everytime i tried to merge the mix with a clean tokenizer, the result was HORRIBLE. That's why these days, i went on a "try/test period", and I ended doing "SO MANY MERGES" that i can't count now... Trying to understand the reason, but, at least, learning a lot...

The model will drift if any "Chat Template" based on "ChatML" or "Mistral" is used. "The Tokenizer" is a fuckin mess, so understand this. The recommended chat template to use is "Alpaca (with "RAW" inputs)". The configuration files are now "Tweaked", leading to don't use any "Chat Template" now.