Patent-Instruct-LLaMA-Pro

Patent-Instruct-LLaMA-Pro is a merge of the following models using mergekit:

• arcee-ai/Patent-Instruct-7b
• TencentARC/LLaMA-Pro-8B-Instruct

🧩 Configuration

  merge_method: passthrough
  dtype: bfloat16
  slices:
  - sources:
      - model: arcee-ai/Patent-Instruct-7b 
        layer_range:
          - 0
          - 4
  - sources:
      - model: TencentARC/LLaMA-Pro-8B-Instruct
        layer_range:
          - 4
          - 5
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 4
          - 8
  - sources:
      - model: TencentARC/LLaMA-Pro-8B-Instruct
        layer_range:
          - 9
          - 10
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 8
          - 12
  - sources:
      - model: TencentARC/LLaMA-Pro-8B-Instruct
        layer_range:
          - 14
          - 15
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 12
          - 16
  - sources:
      - model: TencentARC/LLaMA-Pro-8B-Instruct
        layer_range:
          - 19
          - 20
  - sources:
       - model: arcee-ai/Patent-Instruct-7b
         layer_range:
          - 16
          - 20
  - sources:
      - model: TencentARC/LLaMA-Pro-8B-Instruct
        layer_range:
          - 24
          - 25
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 20
          - 24
  - sources:
      - model: TencentARC/LLaMA-Pro-8B-Instruct
        layer_range:
          - 29
          - 30
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 24
          - 28
  - sources:
      - model: TencentARC/LLaMA-Pro-8B-Instruct
        layer_range:
          - 34
          - 35
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 28
          - 32
  - sources:
      - model: TencentARC/LLaMA-Pro-8B-Instruct
        layer_range:
          - 39
          - 40