DarkHazard-v2.0-24b

This is a merge of pre-trained language models created using mergekit.

Merge Details

Inspiration

This merge was inspired by

• Yoesph/Haphazard-v1.1-24b
• yvvki/Erotophobia-24B-v1.1

Changelog

v2.0

• Major version bump because of base model change: cognitivecomputations/Dolphin-Mistral-24B-Venice-Edition
• swapped TheDrummer/Cydonia-24B-v2.1 with ReadyArt/Forgotten-Safeword-24B-v4.0
• (I've been doing some tests with LatitudeGames/Harbinger-24B but it just seemed to introduce positivity bias to my test scenarios, so it stays out for now)

v1.3

• updated Eurydice to v3

v1.2

• replaced Yoesph/Haphazard-v1.1-24b with model: TheDrummer/Cydonia-24B-v2.1
• replaced ReadyArt/Safeword-Abomination-of-Omega-Darker-Gaslight_The-Final-Forgotten-Transgression-24B with ReadyArt/Broken-Tutu-24B

Merge Method

This model was merged using the Model Stock merge method using cognitivecomputations/Dolphin-Mistral-24B-Venice-Edition as a base.

Models Merged

The following models were included in the merge:

• ReadyArt/Forgotten-Safeword-24B-v4.0
• aixonlab/Eurydice-24b-v3
• ReadyArt/Broken-Tutu-24B
• PocketDoc/Dans-PersonalityEngine-V1.2.0-24b

Configuration

The following YAML configuration was used to produce this model:

base_model: cognitivecomputations/Dolphin-Mistral-24B-Venice-Edition
merge_method: model_stock
dtype: bfloat16
models:
  - model: aixonlab/Eurydice-24b-v3 # storytelling / RP
  - model: ReadyArt/Forgotten-Safeword-24B-v4.0 # uncensor + Cydonia
  - model: ReadyArt/Broken-Tutu-24B # uncensor + nsfw + Cydonia
  - model: PocketDoc/Dans-PersonalityEngine-V1.2.0-24b # Prompt Adherence