L3.1-Sallowv1-12B

sallow

This is a merge of pre-trained language models created using mergekit.

Merge Details

Merge Method

This model was merged using the Passthrough merge method.

Models Merged

The following models were included in the merge:

• merge/polar

Configuration

The following YAML configuration was used to produce this model:

dtype: float32
merge_method: passthrough
modules:
  default:
    slices:
    - sources:
      - layer_range: [0, 4]
        model: merge/polar
    - sources:
      - layer_range: [2, 4]
        model: merge/polar
        parameters:
          scale:
          - filter: o_proj
            value: 0.0
          - filter: down_proj
            value: 0.0
          - value: 1.0
    - sources:
      - layer_range: [4, 6]
        model: merge/polar
    - sources:
      - layer_range: [4, 6]
        model: merge/polar
        parameters:
          scale:
          - filter: o_proj
            value: 0.0
          - filter: down_proj
            value: 0.0
          - value: 1.0
    - sources:
      - layer_range: [6, 8]
        model: merge/polar
    - sources:
      - layer_range: [6, 8]
        model: merge/polar
        parameters:
          scale:
          - filter: o_proj
            value: 0.0
          - filter: down_proj
            value: 0.0
          - value: 1.0
    - sources:
      - layer_range: [8, 10]
        model: merge/polar
    - sources:
      - layer_range: [8, 10]
        model: merge/polar
        parameters:
          scale:
          - filter: o_proj
            value: 0.0
          - filter: down_proj
            value: 0.0
          - value: 1.0
    - sources:
      - layer_range: [10, 14]
        model: merge/polar
    - sources:
      - layer_range: [12, 14]
        model: merge/polar
        parameters:
          scale:
          - filter: o_proj
            value: 0.0
          - filter: down_proj
            value: 0.0
          - value: 1.0
    - sources:
      - layer_range: [14, 18]
        model: merge/polar
    - sources:
      - layer_range: [16, 18]
        model: merge/polar
        parameters:
          scale:
          - filter: o_proj
            value: 0.0
          - filter: down_proj
            value: 0.0
          - value: 1.0
    - sources:
      - layer_range: [18, 28]
        model: merge/polar
    - sources:
      - layer_range: [26, 28]
        model: merge/polar
        parameters:
          scale:
          - filter: o_proj
            value: 0.0
          - filter: down_proj
            value: 0.0
          - value: 1.0
    - sources:
      - layer_range: [28, 30]
        model: merge/polar
    - sources:
      - layer_range: [28, 30]
        model: merge/polar
        parameters:
          scale:
          - filter: o_proj
            value: 0.0
          - filter: down_proj
            value: 0.0
          - value: 1.0
    - sources:
      - layer_range: [30, 32]
        model: merge/polar
    - sources:
      - layer_range: [30, 32]
        model: merge/polar
        parameters:
          scale:
          - filter: o_proj
            value: 0.0
          - filter: down_proj
            value: 0.0
          - value: 1.0