README.md

metadata

language:
  - ja
  - en
license: apache-2.0
datasets:
  - augmxnt/ultra-orca-boros-en-ja-v1
model-index:
  - name: shisa-gamma-7b-v1
    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: 53.16
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=augmxnt/shisa-gamma-7b-v1
          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: 77.3
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=augmxnt/shisa-gamma-7b-v1
          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: 55.23
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=augmxnt/shisa-gamma-7b-v1
          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: 50.73
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=augmxnt/shisa-gamma-7b-v1
          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: 73.88
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=augmxnt/shisa-gamma-7b-v1
          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: 22.74
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=augmxnt/shisa-gamma-7b-v1
          name: Open LLM Leaderboard

shisa-gamma-7b-v1

For more information see our main Shisa 7B model

We applied a version of our fine-tune data set onto Japanese Stable LM Base Gamma 7B and it performed pretty well, just sharing since it might be of interest.

Check out our JA MT-Bench results.

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	55.50
AI2 Reasoning Challenge (25-Shot)	53.16
HellaSwag (10-Shot)	77.30
MMLU (5-Shot)	55.23
TruthfulQA (0-shot)	50.73
Winogrande (5-shot)	73.88
GSM8k (5-shot)	22.74