Patent-Instruct-Extended

Patent-Instruct-Extended is an extension of the following models using the passthrough method and mergekit:

🧩 Configuration

  slices:
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 0
          - 4
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 2
          - 4
        parameters:
          scale:
            - filter: o_proj
              value: 0
            - filter: down_proj
              value: 0
            - value: 1
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 4
          - 8
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 6
          - 8
        parameters:
          scale:
            - filter: o_proj
              value: 0
            - filter: down_proj
              value: 0
            - value: 1
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 8
          - 12
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 10
          - 12
        parameters:
          scale:
            - filter: o_proj
              value: 0
            - filter: down_proj
              value: 0
            - value: 1
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 12
          - 16
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 14
          - 16
        parameters:
          scale:
            - filter: o_proj
              value: 0
            - filter: down_proj
              value: 0
            - value: 1
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 16
          - 20
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 18
          - 20
        parameters:
          scale:
            - filter: o_proj
              value: 0
            - filter: down_proj
              value: 0
            - value: 1
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 20
          - 24
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 22
          - 24
        parameters:
          scale:
            - filter: o_proj
              value: 0
            - filter: down_proj
              value: 0
            - value: 1
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 24
          - 28
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 26
          - 28
        parameters:
          scale:
            - filter: o_proj
              value: 0
            - filter: down_proj
              value: 0
            - value: 1
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 28
          - 32
  - sources:
      - model: arcee-ai/Patent-Instruct-7b
        layer_range:
          - 30
          - 32
        parameters:
          scale:
            - filter: o_proj
              value: 0
            - filter: down_proj
              value: 0
            - value: 1
  merge_method: passthrough
  dtype: bfloat16


Downloads last month
4
Safetensors
Model size
9.98B params
Tensor type
BF16
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.