README.md

metadata

license: other
tags:
  - merge
license_name: yi-34b
license_link: https://huggingface.co/01-ai/Yi-34B/blob/main/LICENSE
model-index:
  - name: Nous-Hermes-2-SUS-Chat-34B-Linear
    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: 66.38
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Weyaxi/Nous-Hermes-2-SUS-Chat-34B-Linear
          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: 84.94
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Weyaxi/Nous-Hermes-2-SUS-Chat-34B-Linear
          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: 76.82
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Weyaxi/Nous-Hermes-2-SUS-Chat-34B-Linear
          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: 59.19
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Weyaxi/Nous-Hermes-2-SUS-Chat-34B-Linear
          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: 82.79
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Weyaxi/Nous-Hermes-2-SUS-Chat-34B-Linear
          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: 72.02
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Weyaxi/Nous-Hermes-2-SUS-Chat-34B-Linear
          name: Open LLM Leaderboard

Nous-Hermes-2-SUS-Chat-34B-Linear

This is the model for Nous-Hermes-2-SUS-Chat-34B-Linear. I used mergekit to merge models.

Yaml Config

models:
  - model: Nous-Hermes-2-Yi-34B
    parameters:
      weight: 0.5
  - model: SUS-Chat-34B
    parameters:
      weight: 0.3
merge_method: linear
dtype: bfloat16

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	73.69
AI2 Reasoning Challenge (25-Shot)	66.38
HellaSwag (10-Shot)	84.94
MMLU (5-Shot)	76.82
TruthfulQA (0-shot)	59.19
Winogrande (5-shot)	82.79
GSM8k (5-shot)	72.02