djuna/L3.1-Noraian

merge

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

Merge Details

Merge Method

This model was merged using the della merge method using vicgalle/Configurable-Hermes-3-Llama-3.1-8B as a base.

Models Merged

The following models were included in the merge:

• djuna/L3.1-Romes-Ninomos
• DreadPoor/Aurora_faustus-8B-LINEAR + grimjim/Llama-3-Instruct-abliteration-LoRA-8B
• ValiantLabs/Llama3.1-8B-ShiningValiant2
• v000000/L3.1-Storniitova-8B

Configuration

The following YAML configuration was used to produce this model:

merge_method: della
dtype: bfloat16
tokenizer_source: union
parameters:
  lambda: 0.78
  epsilon: 0.1
  normalize: false
base_model: vicgalle/Configurable-Hermes-3-Llama-3.1-8B
models:
  - model: ValiantLabs/Llama3.1-8B-ShiningValiant2
    parameters: 
      weight: 0.2
      density: 0.5
  - model: djuna/L3.1-Romes-Ninomos
    parameters:
      weight: 0.15
      density: 0.55
  - model: DreadPoor/Aurora_faustus-8B-LINEAR+grimjim/Llama-3-Instruct-abliteration-LoRA-8B
    parameters: 
      weight: 0.14
      density: 0.56
  - model: v000000/L3.1-Storniitova-8B
    parameters:
      weight: 0.2
      density: 0.5