README.md

metadata

license: cc-by-sa-3.0
model-index:
  - name: Palworld-SME-13b
    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: 55.55
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ericpolewski/Palworld-SME-13b
          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: 80.81
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ericpolewski/Palworld-SME-13b
          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: 53.64
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ericpolewski/Palworld-SME-13b
          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: 46.67
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ericpolewski/Palworld-SME-13b
          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: 74.82
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ericpolewski/Palworld-SME-13b
          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: 2.2
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=ericpolewski/Palworld-SME-13b
          name: Open LLM Leaderboard

This is a Subject Matter Expert (SME) bot trained on the Palworld Fandom Wiki as a test of a new SME model pipeline. There's no RAG. All information is embedded in the model. It uses the OpenOrca-Platypus-13b fine-tune as a base.

Should work in any applicable loader/app, though I only tested it in EricLLM and TGWUI.

All SME bots are generally useful, but focus on a topic. Contact me if you're interested in having one built.

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	52.28
AI2 Reasoning Challenge (25-Shot)	55.55
HellaSwag (10-Shot)	80.81
MMLU (5-Shot)	53.64
TruthfulQA (0-shot)	46.67
Winogrande (5-shot)	74.82
GSM8k (5-shot)	2.20