Mark-Arcee's picture
Update README.md
d795489 verified
metadata
license: apache-2.0
tags:
  - merge
  - mergekit
  - CorticalStack/pastiche-crown-clown-7b-dare-dpo
  - Equall/Saul-Instruct-v1

image/webp

Saul-Instruct-Clown-7b

Saul-Instruct-Clown-7b is a merge of the following models using mergekit:

πŸ† Evaluation

OpenLLM

Saul-Instruct-Clown-7b OpenLLM benchmark suite

Model Average arc HellaSwag mmlu TruthfulQA gsm8k
arcee-ai/Saul-Instruct-Clown-7b 72.79 68.26 86.28 63.12 64.68 83.43

🧩 Configuration

  slices:
    - sources:
        - model: CorticalStack/pastiche-crown-clown-7b-dare-dpo
          layer_range: [0, 32]
        - model: Equall/Saul-Instruct-v1
          layer_range: [0, 32]
  merge_method: slerp
  base_model: CorticalStack/pastiche-crown-clown-7b-dare-dpo
  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: bfloat16