SaisExperiments's picture
Adding Evaluation Results (#1)
870c4d1 verified
metadata
language:
  - en
license: apache-2.0
base_model:
  - mistralai/Mistral-Small-24B-Base-2501
datasets:
  - anthracite-org/kalo-opus-instruct-22k-no-refusal
model-index:
  - name: Not-So-Small-Alpaca-24B
    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: 62.44
            name: strict accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=SaisExperiments/Not-So-Small-Alpaca-24B
          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: 33.02
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=SaisExperiments/Not-So-Small-Alpaca-24B
          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: 18.05
            name: exact match
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=SaisExperiments/Not-So-Small-Alpaca-24B
          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: 14.54
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=SaisExperiments/Not-So-Small-Alpaca-24B
          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: 12.09
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=SaisExperiments/Not-So-Small-Alpaca-24B
          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: 29.94
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=SaisExperiments/Not-So-Small-Alpaca-24B
          name: Open LLM Leaderboard

This model uses Alpaca

This is a lora finetune of mistralai/Mistral-Small-24B-Base-2501 using anthracite-org/kalo-opus-instruct-22k-no-refusal for ~6M tokens

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric Value
Avg. 28.35
IFEval (0-Shot) 62.44
BBH (3-Shot) 33.02
MATH Lvl 5 (4-Shot) 18.05
GPQA (0-shot) 14.54
MuSR (0-shot) 12.09
MMLU-PRO (5-shot) 29.94