README.md

metadata

license: apache-2.0
library_name: transformers
base_model:
  - mistralai/Mistral-Nemo-Instruct-2407
datasets:
  - tasksource/ScienceQA_text_only
model-index:
  - name: mistral-nemo-wissenschaft-12B
    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: 65.2
            name: strict accuracy
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=nbeerbower/mistral-nemo-wissenschaft-12B
          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: 29.57
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=nbeerbower/mistral-nemo-wissenschaft-12B
          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: 6.57
            name: exact match
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=nbeerbower/mistral-nemo-wissenschaft-12B
          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: 5.7
            name: acc_norm
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=nbeerbower/mistral-nemo-wissenschaft-12B
          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: 12.29
            name: acc_norm
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=nbeerbower/mistral-nemo-wissenschaft-12B
          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: 28.14
            name: accuracy
        source:
          url: >-
            https://huggingface.co./spaces/open-llm-leaderboard/open_llm_leaderboard?query=nbeerbower/mistral-nemo-wissenschaft-12B
          name: Open LLM Leaderboard

mistral-nemo-wissenschaft-12B

mistralai/Mistral-Nemo-Instruct-2407 finetuned on tasksource/ScienceQA_text_only.

Method

Finetuned using an A100 on Google Colab for 1 epoch. Correct answers were selected as the chosen answer, a random wrong answer was selected as "rejected."

Fine-tune Llama 3 with ORPO

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	24.58
IFEval (0-Shot)	65.20
BBH (3-Shot)	29.57
MATH Lvl 5 (4-Shot)	6.57
GPQA (0-shot)	5.70
MuSR (0-shot)	12.29
MMLU-PRO (5-shot)	28.14