models:
  - model: google/gemma-7b-it
    parameters:
      density: 0.5
      weight: 0.5
  - model: mlabonne/Gemmalpaca-7B
    parameters:
      density: 0.5
      weight: 0.5 # weight gradient
merge_method: ties
base_model: google/gemma-7b
parameters:
  normalize: true
  int8_mask: true
dtype: float16

arcee-ai/gemma-7b-alpaca-it-ties

gemma-7b-alpaca-it-ties

🧩 Configuration

arcee-ai/
gemma-7b-alpaca-it-ties