merge_method: task_anysize
base_model: princeton-nlp/Sheared-LLaMA-2.7B-ShareGPT
models:
  - model: SanjiWatsuki/Silicon-Maid-7B
    parameters:
      weight: 0.1
dtype: float16