models:
  - model: x0000001/mergekit-task_arithmetic-vlehhex+grimjim/Llama-3-Instruct-abliteration-LoRA-8B
    parameters:
      weight: 1.0
merge_method: linear
dtype: float32