RicardoEstep/RPBizkitRemiX-v3-12B
"Experimental Mistral Nemo Mix" using... "DARE TIES merge using Mergekit".
This is a "Cleaner RemiX" of my "very best" models in one. After noticing it late, i found that "model-V2" adds a noisy hard "assistant image" even on "low weights". I have changed the "Base Model", to force it, to be more "independent", and now, even the "Model-V9" is "stabilized", making the model working completly "uncensored".
Models Merged:
Configuration Used:
merge_method: dare_ties
base_model: TheDrummer/UnslopNemo-12B-v4.1
tokenizer_source: base
parameters:
normalize: true
int8_mask: false
random_seed: 5318008
models:
- model: RicardoEstep/RPBizkit-v9-12B
parameters:
weight: 0.30
density: 0.45
- model: RicardoEstep/RPBizkit-v6-12B
parameters:
weight: 0.70
density: 0.85
dtype: bfloat16
Final Notes;
The Tokenizer & Embedding Sizes (131072) are clean + It's "supposed to" support now a full "128K Context Size". The recommended chat template to use is "Alpaca (with "RAW" inputs)". The configuration files are already "Tweaked", to don't use any "Chat Template". **It's recommended to use Lower Temps ("1.15, max." - "0.86 > 0.92 recommended", for "Creative RPs").
- Quantized Model, Here RicardoEstep/RPBizkitRemiX-v3-12B-GGUF.