Modelpublic

OccultAI/Soulblighter-24B-v1

sourceHugging Faceapache-2.0updated 8mo agoView on Hugging Face

5likes30downloads

Model Card

Soulblighter 24B v1

Soulblighter

This is a creative merge made using the karcher method. It has some refusals but should be easy to ablate or jailbreak if needed.

Similar to the StormSeeker merge, except this also blends Magistral finetunes. You can probably enable "reasoning mode" (untested).

Impish Magic had to be excluded due to various issues (tokenizers, vocab_size, etc.) so Allura's Angel was added in place.

architecture: MistralForCausalLM
merge_method: karcher
models:
  - model: B:\hub\!BeaverAI_Fallen-Mistral-Small-3.1-24B-v1e_textonly # Q0F Pass
  - model: B:\hub\!models--aixonlab--Eurydice-24b-v3.5 # Q0F Pass
  - model: B:\hub\!models--allura-forge--ms32-final-TEXTONLY # Q0F Pass
  - model: B:\hub\!models--CrucibleLab--M3.2-24B-Loki-V2 # Q0F Pass
  - model: B:\hub\!models--LatitudeGames--Hearthfire-24B # Q0F Pass
  - model: B:\hub\!models--PocketDoc--Dans-PersonalityEngine-V1.3.0-24b # Q0F Fail
  - model: B:\hub\!models--ReadyArt--Dark-Nexus-24B-v2.0 # Q0F Pass
  # - model: B:\hub\!models--SicariusSicariiStuff--Impish_Magic_24B\fixed # Q0F Pass
  - model: B:\hub\!models--TheDrummer--Precog-24B-v1 # Q0F Pass
  - model: B:\hub\!models--trashpanda-org--MS3.2-24B-Mullein-v2 # Q0F Fail (but still impressive)
  - model: B:\hub\!models--zerofata--MS3.2-PaintedFantasy-v3-24B # Q0F Fail # v2 pass # v1 fail
dtype: float32
out_dtype: bfloat16
parameters:
tokenizer:
  source: union
chat_template: auto
name: Soulblighter-24B-v1