RicardoEstep/RPBizkitRemiX-v1-12B
"Experimental Own Mix" using... "Model Stock merge using Mergekit".
This is a "RemiX" of all my models in one using RicardoEstep/RPBizkit-v6-12B as base.
This one was merged using "32GBs of RAM", i tried to mix all models equaly, and giving all their data have same importance.

Models Merged
The following models were included in the merge:
- RicardoEstep/AngelSlayerKrix-12B
- RicardoEstep/RPBizkit-v2-12B
- RicardoEstep/RPBizkit-v4-12B
- RicardoEstep/RPBizkit-v4-12B_Lorablated
- RicardoEstep/RPBizkit-v5-12B-Lorablated
- RicardoEstep/RPBizkit-v6-12B
It's actually pretty stable as "V6 version", with a "Clean Tokenizer & Embedding Sizes (131072)" based on "Mistral". But, the model will drift if any "Chat Template" is used. The recommended chat template to use is "Alpaca (with "RAW" inputs)". The configuration files are already "Tweaked", to don't use any "Chat Template".
While it's supposed to support a "128K (131072)" context size. The recommended max context size is "8K (8192)", as is a "LLaMA 3 limitation".
- Quantized Model, Here RicardoEstep/RPBizkitRemiX-v1-12B-GGUF.