metadata

language:
  - en
license: apache-2.0
library_name: transformers
tags:
  - mistral
  - merge
  - openchat
  - 7b
  - zephyr
model-index:
  - name: MadMix-v0.2
    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: 64.85
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Fredithefish/MadMix-v0.2
          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: 83.54
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Fredithefish/MadMix-v0.2
          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: 64.02
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Fredithefish/MadMix-v0.2
          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: 55.79
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Fredithefish/MadMix-v0.2
          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: 77.35
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Fredithefish/MadMix-v0.2
          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: 30.78
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Fredithefish/MadMix-v0.2
          name: Open LLM Leaderboard

🔥 MadMix: Merging the top Seven Billion-Parameter Mistral-Based Open Source Models 🔥

This Model is a ties merge between various high-quality open-source models, it was done by using the https://github.com/cg123/mergekit repository.

Details

Due to the apache-2 license, the model is fully commercially usable

Models included in the merge:
- OpenChat_3.5
- MetaMath 7B
- Zephyr 7B beta
- OpenHermes-2.5
- Neural-chat 7b v3.1
- Mistral-7B v0.1 (serving as the base for the merge)

Evals

(soon...)

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	62.72
AI2 Reasoning Challenge (25-Shot)	64.85
HellaSwag (10-Shot)	83.54
MMLU (5-Shot)	64.02
TruthfulQA (0-shot)	55.79
Winogrande (5-shot)	77.35
GSM8k (5-shot)	30.78