ABX-AI/Silver-Sun-v2-11B-GGUF-IQ-Imatrix

Updated Apr 9 • 77 • 8

Note This is an updated version of Silver-Sun-11B, using Fimbulvetr-11B-v2 instead of v1. As expected, this is a HIGHLY uncensored model. It scores ever-so-slightly lower than Silver-Sun-11B (v1).

ABX-AI/Silver-Sun-11B-GGUF-IQ-Imatrix

Updated Apr 9 • 130 • 13

Note A highly uncensored RP model, powered by Solar. It likes to act and will go with any kind of RP imaginable without a problem. It has become my daily RP driver.

ABX-AI/Quantum-Citrus-9B-GGUF-IQ-Imatrix

Updated Apr 11 • 136 • 1

Note The concept behind this merge is to combine the power of Cerebrum, TheSpice, LemonadeRP, InfinityRP, and some of my previous merges containing Infinitely-Laydiculous, while returning to a layla v0.1 base for more stable results. I think the merge worked, as I have no GPT-isms, or alignment issues, and the output is rather creative and refined. It's my new favorite :)

ABX-AI/Infinite-Laymons-9B-GGUF-IQ-Imatrix

Updated Apr 11 • 189 • 7

Note Been having good results with this one, laying on the already good foundations of Infinity RP and Nitral-AI's Infinitely-Laydiculous. A generally good RP model, standing on the shoulders of other great models, and also the first of my merges which I was happy with.

ABX-AI/Cerebral-Lemonade-9B-GGUF-IQ-Imatrix

Updated Mar 29 • 58 • 4

Note Very happy with the results on this one. The responses are a more refined, and somewhat more intelligent than I expected. The "alignment reminders" still happen on this one, although predominantly on SFW cards. It still follows NSFW cards quite well and without refusal. The adherence to character card here is a bit stronger.

ABX-AI/Cosmic-Citrus-9B-GGUF-IQ-Imatrix

Updated Apr 2 • 51 • 5

Note An interesting merge, sometimes with very impressive responses. However, at times, it may spit out some GPT-ism or weird string of data. A rare case, but it can happen. Still, due to the original responses, I think it's worth trying it out!

ABX-AI/Starbral-Infinimons-9B-GGUF-IQ-Imatrix

Updated Mar 30 • 56 • 3

Note So far, I'm happy with the results from this one, one of my favorite merge experiments at the moment. The responses are a mix of smart and unhinged, at least with my cards, and there are virtually no refusals or "alignment reminders", even on more SFW cards, which is impressive.