README.md

metadata

license: apache-2.0
tags:
  - merge
  - mergekit
  - lazymergekit
  - MaziyarPanahi/calme-2.1-phi3.5-4b
  - jpacifico/Chocolatine-3B-Instruct-DPO-Revised
model-index:
  - name: merge-test
    results:
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: IFEval (0-Shot)
          type: HuggingFaceH4/ifeval
          args:
            num_few_shot: 0
        metrics:
          - type: inst_level_strict_acc and prompt_level_strict_acc
            value: 53.83
            name: strict accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=lesubra/merge-test
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: BBH (3-Shot)
          type: BBH
          args:
            num_few_shot: 3
        metrics:
          - type: acc_norm
            value: 33.35
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=lesubra/merge-test
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MATH Lvl 5 (4-Shot)
          type: hendrycks/competition_math
          args:
            num_few_shot: 4
        metrics:
          - type: exact_match
            value: 9.14
            name: exact match
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=lesubra/merge-test
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: GPQA (0-shot)
          type: Idavidrein/gpqa
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 9.62
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=lesubra/merge-test
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MuSR (0-shot)
          type: TAUR-Lab/MuSR
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 15.64
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=lesubra/merge-test
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MMLU-PRO (5-shot)
          type: TIGER-Lab/MMLU-Pro
          config: main
          split: test
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 31.93
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=lesubra/merge-test
          name: Open LLM Leaderboard

merge-test

merge-test is a merge of the following models using mergekit:

• MaziyarPanahi/calme-2.1-phi3.5-4b
• jpacifico/Chocolatine-3B-Instruct-DPO-Revised

🧩 Configuration

models:
  - model: microsoft/Phi-3.5-mini-instruct
    # no parameters necessary for base model
  - model: MaziyarPanahi/calme-2.1-phi3.5-4b
    parameters:
      density: 0.8
      weight: 0.5
  - model: jpacifico/Chocolatine-3B-Instruct-DPO-Revised 
    parameters:
      density: 0.8
      weight: 0.3
merge_method: ties
base_model: microsoft/Phi-3.5-mini-instruct
parameters:
  normalize: true
dtype: float16

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	25.58
IFEval (0-Shot)	53.83
BBH (3-Shot)	33.35
MATH Lvl 5 (4-Shot)	9.14
GPQA (0-shot)	9.62
MuSR (0-shot)	15.64
MMLU-PRO (5-shot)	31.93