merge

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

Merge Details

Merge Method

This model was merged using the DARE TIES merge method using unsloth/Mistral-Small-Instruct-2409 as a base.

Models Merged

The following models were included in the merge:

Configuration

The following YAML configuration was used to produce this model:

models:
  # Core Fiction and Character Detail Models (Increased Precision)
  - model: Kaoeiri/MS_Moingooistral-2409-22B  # Monster fiction core
    parameters:
      weight: 0.40  # Increased for better monster/character detail
      density: 1.30 # Increased for richer character descriptions
  
  - model: Kaoeiri/MS-Magpantheonsel-lark-v4x1.6.2-Cydonia-vXXX-22B-5  # Main writing engine
    parameters:
      weight: 1.0   # Maximized for core writing
      density: 0.85 # Slightly increased for deeper character development
  
  - model: anthracite-org/magnum-v4-22b  # Added for writing recap and precision
    parameters:
      weight: 0.95
      density: 0.84
  
  # World Building & Character Interaction
  - model: Kaoeiri/MS-Inky-2409-22B  # Descriptive world dynamics
    parameters:
      weight: 0.45  # Increased for richer character environments
      density: 0.82 # Increased for better world-character integration
  
  - model: Gryphe/Pantheon-RP-Pure-1.6.2-22b-Small  # Rich interaction
    parameters:
      weight: 0.42  # Increased for deeper character interactions
      density: 0.78
  
  # Character Development Core
  - model: DigitalSouls/BlackSheep-DigitalSoul-22B  # Combat and conflict
    parameters:
      weight: 0.30  # Increased for better character conflict
      density: 0.75
  
  # Magical Elements (Rebalanced NovusKyver Sister)
  - model: InferenceIllusionist/SorcererLM-22B
    parameters:
      weight: 0.15 # Increased for magical character traits
      density: 0.78 # Increased for better integration
  
  # Secondary Character Enhancement
  - model: TheDrummer/Cydonia-22B-v1.1
    parameters:
      weight: 0.15  # Slight increase for character depth
      density: 0.68
  
  - model: crestf411/MS-sunfall-v0.7.0
    parameters:
      weight: 0.18  # Increased for writing precision
      density: 0.72
  
  - model: Kaoeiri/MS_a-coolyte-2409-22B
    parameters:
      weight: 0.22  # Increased for fictional character detail
      density: 0.73
  
  # Character Personality Development  
  - model: Kaoeiri/MS-Quadrosiac-2409-22B
    parameters:
      weight: 0.18
      density: 0.73
  
  # Enhanced Story and Character Building
  - model: hf-100/Mistral-Small-Spellbound-StoryWriter-22B-instruct-0.2-chkpt-200-16-bit
    parameters:
      weight: 0.25  # Increased for better character narrative
      density: 0.72
  
  - model: ArliAI/Mistral-Small-22B-ArliAI-RPMax-v1.1
    parameters:
      weight: 0.15  # Increased for character roleplay
      density: 0.65
  
  - model: Darkknight535/MS-Moonlight-22B-v3
    parameters:
      weight: 0.25  # Increased for character detail
      density: 0.65
  
  # Reduced weight but maintained for diversity
  - model: Jellywibble/MistralSmall1500CTXDummy
    parameters:
      weight: 0.12
      density: 0.62
  
  # Cultural and Character Depth
  - model: Saxo/Linkbricks-Horizon-AI-Japanese-Superb-V1-22B
    parameters:
      weight: 0.22  # Increased for news and cultural character traits
      density: 0.62
  
  # Rebalanced NovusKyver (Modified to maintain benefits while reducing instruction avoidance)
  - model: Envoid/Mistral-Small-NovusKyver
    parameters:
      weight: 0.20  # Reduced to minimize instruction avoidance
      density: 0.74 # Increased for better integration
  
  - model: TroyDoesAI/BlackSheep-MermaidMistral-22B
    parameters:
      weight: 0.25  # Increased for character personality
      density: 0.73

merge_method: dare_ties
base_model: unsloth/Mistral-Small-Instruct-2409
parameters:
  density: 0.95    # Increased for maximum character detail
  epsilon: 0.035   # Reduced for more consistent character behavior
  lambda: 1.50     # Increased to enhance character creativity
dtype: bfloat16
tokenizer_source: union
Downloads last month
8
Safetensors
Model size
22.2B params
Tensor type
BF16
·
Inference Providers NEW
This model is not currently available via any of the supported third-party Inference Providers, and the model is not deployed on the HF Inference API.

Model tree for Kaoeiri/MS-MagpantheonselRP-22B-12.95-Fictional