merged

🛑 Premise

Alright so. Here we are again with my second merge, this time i used Model Stock as the merging method and SicariusSicariiStuff/Impish_Nemo_12B as a base since I tried it and I really liked it

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

Data

well... i didn't add any new data so nothing, I just merged the models

🧪 Intended Use

This model is designed mainly for:

• Roleplay and creative writing

• Storytelling & world-building

😷 Ethical Containment

This model can:

• ⚠️ Generating unfiltered creative content
• ⚠️ Producing potentially disturbing narratives
• ⚠️ Creating NSFW content

⚖️ License

Follow the licensing terms of each merged model:

Each source model’s license applies

Merge Details

Models Merged

The following models were included in the merge:

• DreadPoor/Irix-12B-Model_Stock
• MrRikyz/Kitsune-Symphony-V0.0-12B
• SicariusSicariiStuff/Impish_Nemo_12B

Configuration

The following YAML configuration was used to produce this model:

base_model: SicariusSicariiStuff/Impish_Nemo_12B
dtype: float16
merge_method: model_stock
modules:
  default:
    slices:
    - sources:
      - layer_range: [0, 40]
        model: MrRikyz/Kitsune-Symphony-V0.0-12B
        parameters:
          weight: 0.35
      - layer_range: [0, 40]
        model: SicariusSicariiStuff/Impish_Nemo_12B
        parameters:
          weight: 0.55
      - layer_range: [0, 40]
        model: DreadPoor/Irix-12B-Model_Stock
        parameters:
          weight: 0.3
tokenizer:
  source: base

✨ Acknowledgements

Thanks to the authors of the original models for their incredible work:

• SicariusSicariiStuff for Impish Nemo 12B (Great model I really liked it)

• DreadPoor for Irix 12B

• MrRikyz for Kitsune-Symphony (Myself)