162 combine loras 146 15 etc with these for truly unique outputs
12B-Suite
community
AI & ML interests
None defined yet.
Recent Activity
Test of ablated Gemma 12B for finetune base. Step 1) Test the uncensored model, Step 2) Finetune the model and test again for refusals
-
MuXodious/gemma-4-12B-it-QAT-SOMPOA-heresy
Any-to-Any • 12B • Updated • 128 • 5 -
huihui-ai/Huihui-gemma-4-12B-it-abliterated
Any-to-Any • 12B • Updated • 15k • 203 -
igorls/gemma-4-12B-it-heretic-v1
Text Generation • 12B • Updated • 432 • 16 -
RumiaChannel/gemma-4-12B-it-uncensored-ara-Refusals2
Any-to-Any • 12B • Updated • 39 • 2
162 combine loras 146 15 etc with these for truly unique outputs
This specific set of models, when merged with a 4th finetune via the della method, produces very interesting "boosted" output (using chatML)
Test of ablated Gemma 12B for finetune base. Step 1) Test the uncensored model, Step 2) Finetune the model and test again for refusals
-
MuXodious/gemma-4-12B-it-QAT-SOMPOA-heresy
Any-to-Any • 12B • Updated • 128 • 5 -
huihui-ai/Huihui-gemma-4-12B-it-abliterated
Any-to-Any • 12B • Updated • 15k • 203 -
igorls/gemma-4-12B-it-heretic-v1
Text Generation • 12B • Updated • 432 • 16 -
RumiaChannel/gemma-4-12B-it-uncensored-ara-Refusals2
Any-to-Any • 12B • Updated • 39 • 2