caTUNABeagle / README.md
leaderboard-pr-bot's picture
Adding Evaluation Results
2e47f26 verified
|
raw
history blame
4.04 kB
metadata
license: apache-2.0
tags:
  - merge
  - mergekit
  - lazymergekit
  - fblgit/UNA-TheBeagle-7b-v1
  - rishiraj/CatPPT-base
model-index:
  - name: caTUNABeagle
    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: 70.05
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Eric111/caTUNABeagle
          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: 87.35
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Eric111/caTUNABeagle
          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: 65.02
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Eric111/caTUNABeagle
          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: 65.31
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Eric111/caTUNABeagle
          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: 81.22
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Eric111/caTUNABeagle
          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: 71.34
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Eric111/caTUNABeagle
          name: Open LLM Leaderboard

caTUNABeagle

caTUNABeagle is a merge of the following models using mergekit:

🧩 Configuration

slices:
  - sources:
      - model: fblgit/UNA-TheBeagle-7b-v1
        layer_range: [0, 32]
      - model: rishiraj/CatPPT-base
        layer_range: [0, 32]
merge_method: slerp
base_model: fblgit/UNA-TheBeagle-7b-v1
parameters:
  t:
    - filter: self_attn
      value: [0, 0.5, 0.3, 0.7, 1]
    - filter: mlp
      value: [1, 0.5, 0.7, 0.3, 0]
    - value: 0.5
dtype: bfloat16

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric Value
Avg. 73.38
AI2 Reasoning Challenge (25-Shot) 70.05
HellaSwag (10-Shot) 87.35
MMLU (5-Shot) 65.02
TruthfulQA (0-shot) 65.31
Winogrande (5-shot) 81.22
GSM8k (5-shot) 71.34