README.md

metadata

license: apache-2.0
library_name: transformers
base_model:
  - Qwen/Qwen2.5-14B-Instruct
model-index:
  - name: Qwen2.5-Gutenberg-Doppel-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: 80.91
            name: strict accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=nbeerbower/Qwen2.5-Gutenberg-Doppel-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: 48.24
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=nbeerbower/Qwen2.5-Gutenberg-Doppel-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: 0
            name: exact match
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=nbeerbower/Qwen2.5-Gutenberg-Doppel-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: 11.07
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=nbeerbower/Qwen2.5-Gutenberg-Doppel-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: 10.02
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=nbeerbower/Qwen2.5-Gutenberg-Doppel-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: 43.57
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=nbeerbower/Qwen2.5-Gutenberg-Doppel-14B
          name: Open LLM Leaderboard

Qwen2.5-Gutenberg-Doppel-14B

Qwen/Qwen2.5-14B-Instruct finetuned on jondurbin/gutenberg-dpo-v0.1 and nbeerbower/gutenberg2-dpo.

Method

ORPO tuned with 4x A40 for 3 epochs.

Thank you @ParasiticRogue for sponsoring.

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	32.30
IFEval (0-Shot)	80.91
BBH (3-Shot)	48.24
MATH Lvl 5 (4-Shot)	0.00
GPQA (0-shot)	11.07
MuSR (0-shot)	10.02
MMLU-PRO (5-shot)	43.57