README.md

---
license: apache-2.0
library_name: transformers
tags:
- mergekit
- merge
base_model:
- lodrick-the-lafted/Olethros-8B
- lodrick-the-lafted/Limon-8B
- lodrick-the-lafted/Rummage-8B
- cgato/L3-TheSpice-8b-v0.8.3
- unsloth/llama-3-8b-Instruct
- Edgerunners/meta-llama-3-8b-instruct-hf-ortho-baukit-10fail-1000total
model-index:
- name: Kudzu-8B
  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: 62.46
      name: normalized accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=lodrick-the-lafted/Kudzu-8B
      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: 80.28
      name: normalized accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=lodrick-the-lafted/Kudzu-8B
      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: 68.14
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=lodrick-the-lafted/Kudzu-8B
      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: 52.77
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=lodrick-the-lafted/Kudzu-8B
      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: 76.8
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=lodrick-the-lafted/Kudzu-8B
      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: 63.08
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=lodrick-the-lafted/Kudzu-8B
      name: Open LLM Leaderboard
---

<img src=https://huggingface.co/lodrick-the-lafted/Kudzu-8B/resolve/main/kudzu.png>

Kudzu-8B

Fresh out of the mergekit-evolve kitchen, this is a merge model between:
* [lodrick-the-lafted/Olethros-8B](https://huggingface.co/lodrick-the-lafted/Olethros-8B)
* [lodrick-the-lafted/Limon-8B](https://huggingface.co/lodrick-the-lafted/Limon-8B)
* [lodrick-the-lafted/Rummage-8B](https://huggingface.co/lodrick-the-lafted/Rummage-8B)
* [Edgerunners/meta-llama-3-8b-instruct-hf-ortho-baukit-10fail-1000total](https://huggingface.co/Edgerunners/meta-llama-3-8b-instruct-hf-ortho-baukit-10fail-1000total)
* [cgato/L3-TheSpice-8b-v0.8.3](https://huggingface.co/cgato/L3-TheSpice-8b-v0.8.3)


Used wmdp as the scoring method for evolve.  In my limited testing, it has not done the usual Llama-3 "Ahaha!" interjections while retaining a good portion of the intelligence. There are several ablated models in the mix so don't be surprised if it gives you what you ask for.
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_lodrick-the-lafted__Kudzu-8B)

|             Metric              |Value|
|---------------------------------|----:|
|Avg.                             |67.25|
|AI2 Reasoning Challenge (25-Shot)|62.46|
|HellaSwag (10-Shot)              |80.28|
|MMLU (5-Shot)                    |68.14|
|TruthfulQA (0-shot)              |52.77|
|Winogrande (5-shot)              |76.80|
|GSM8k (5-shot)                   |63.08|