README.md

metadata

license: apache-2.0
library_name: transformers
tags:
  - mergekit
  - merge
base_model:
  - jpacifico/Chocolatine-14B-Instruct-4k-DPO
  - jpacifico/Chocolatine-14B-Instruct-DPO-v1.2
  - failspy/Phi-3-medium-4k-instruct-abliterated-v3
model-index:
  - name: Ph3task1-14B
    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: 46.95
            name: strict accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=allknowingroger/Ph3task1-14B
          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: 47.93
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=allknowingroger/Ph3task1-14B
          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: 13.9
            name: exact match
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=allknowingroger/Ph3task1-14B
          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: 13.42
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=allknowingroger/Ph3task1-14B
          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: 16.81
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=allknowingroger/Ph3task1-14B
          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: 41.49
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=allknowingroger/Ph3task1-14B
          name: Open LLM Leaderboard

merge

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

Merge Details

Merge Method

This model was merged using the task arithmetic merge method using jpacifico/Chocolatine-14B-Instruct-DPO-v1.2 as a base.

Models Merged

The following models were included in the merge:

• jpacifico/Chocolatine-14B-Instruct-4k-DPO
• failspy/Phi-3-medium-4k-instruct-abliterated-v3

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: jpacifico/Chocolatine-14B-Instruct-DPO-v1.2
    parameters:
      weight: 1.0
  - model: failspy/Phi-3-medium-4k-instruct-abliterated-v3
    parameters:
      weight: 1.0
  - model: jpacifico/Chocolatine-14B-Instruct-4k-DPO
    parameters:
      weight: 1.0
merge_method: task_arithmetic
base_model: jpacifico/Chocolatine-14B-Instruct-DPO-v1.2
parameters:
    normalize: true
dtype: float16

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	30.08
IFEval (0-Shot)	46.95
BBH (3-Shot)	47.93
MATH Lvl 5 (4-Shot)	13.90
GPQA (0-shot)	13.42
MuSR (0-shot)	16.81
MMLU-PRO (5-shot)	41.49