README.md

metadata

license: apache-2.0
tags:
  - merge
datasets:
  - jondurbin/truthy-dpo-v0.1
model-index:
  - name: NeuralBeagle-11B-truthy
    results:
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: AI2 Reasoning Challenge (25-Shot)
          type: ai2_arc
          config: ARC-Challenge
          split: test
          args:
            num_few_shot: 25
        metrics:
          - type: acc_norm
            value: 73.63
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=vicgalle/NeuralBeagle-11B-truthy
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: HellaSwag (10-Shot)
          type: hellaswag
          split: validation
          args:
            num_few_shot: 10
        metrics:
          - type: acc_norm
            value: 87.86
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=vicgalle/NeuralBeagle-11B-truthy
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MMLU (5-Shot)
          type: cais/mmlu
          config: all
          split: test
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 63.11
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=vicgalle/NeuralBeagle-11B-truthy
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: TruthfulQA (0-shot)
          type: truthful_qa
          config: multiple_choice
          split: validation
          args:
            num_few_shot: 0
        metrics:
          - type: mc2
            value: 75.92
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=vicgalle/NeuralBeagle-11B-truthy
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: Winogrande (5-shot)
          type: winogrande
          config: winogrande_xl
          split: validation
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 82.08
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=vicgalle/NeuralBeagle-11B-truthy
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: GSM8k (5-shot)
          type: gsm8k
          config: main
          split: test
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 49.73
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=vicgalle/NeuralBeagle-11B-truthy
          name: Open LLM Leaderboard

NeuralBeagle-11B

DPO'd from vicgalle/franken-Beagle-11B, a Beagle-like model upscaled to 11B. It is a frankenmerge model created using mergekit. Then, we applied DPO over a high-quality preference dataset.

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	72.06
AI2 Reasoning Challenge (25-Shot)	73.63
HellaSwag (10-Shot)	87.86
MMLU (5-Shot)	63.11
TruthfulQA (0-shot)	75.92
Winogrande (5-shot)	82.08
GSM8k (5-shot)	49.73