v1olet/

v1olet_marcoroni-go-bruins-merge-7B

slices:
  - sources:
      - model: AIDC-ai-business/Marcoroni-7B-v3
        layer_range: [0, 32]
      - model: rwitz/go-bruins-v2
        layer_range: [0, 32]
merge_method: slerp
base_model: AIDC-ai-business/Marcoroni-7B-v3
parameters:
  t:
    - filter: self_attn
      value: [0, 0.5, 0.3, 0.7, 1]
    - filter: mlp
      value: [1, 0.5, 0.7, 0.3, 0]
    - value: 0.5 
dtype: float16

template_format = """{system}
### Instruction:
{prompt}

### Response:
"""