README.md

metadata

license: llama2
tags:
  - llama2
  - merge
  - mergekit
model-index:
  - name: llama-polyglot-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: 59.81
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=chargoddard/llama-polyglot-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: 81.27
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=chargoddard/llama-polyglot-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: 55.04
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=chargoddard/llama-polyglot-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: 48.71
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=chargoddard/llama-polyglot-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: 76.72
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=chargoddard/llama-polyglot-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: 22.59
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=chargoddard/llama-polyglot-13b
          name: Open LLM Leaderboard

Experimental multi-lingual model using a new merge technique.

Mergekit configuration (experimental branch):

models:
  - model: clibrain/Llama-2-13b-ft-instruct-es
  - model: LeoLM/leo-hessianai-13b
  - model: daekeun-ml/Llama-2-ko-DPO-13B
  - model: pleisto/yuren-13b-chatml
  - model: bofenghuang/vigogne-2-13b-instruct
  - model: OpenBuddy/openbuddy-llama2-13b-v8.1-fp16
merge_method: dare_ties
base_model: TheBloke/Llama-2-13B-fp16
dtype: float16
parameters:
  density: 0.3
  weight: 1.0
  normalize: true
  int8_mask: true
tokenizer_source: base

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	57.36
AI2 Reasoning Challenge (25-Shot)	59.81
HellaSwag (10-Shot)	81.27
MMLU (5-Shot)	55.04
TruthfulQA (0-shot)	48.71
Winogrande (5-shot)	76.72
GSM8k (5-shot)	22.59