README.md

metadata

license: llama3.2
library_name: transformers
tags:
  - mergekit
  - merge
base_model:
  - meta-llama/Llama-3.2-3B
  - meta-llama/Llama-3.2-3B-Instruct
  - PJMixers-Dev/LLaMa-3.2-Instruct-JankMix-v0.1-SFT-3B
model-index:
  - name: LLaMa-3.2-Instruct-JankMixBread-v0.1-3B
    results:
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: IFEval (0-Shot)
          type: HuggingFaceH4/ifeval
          args:
            num_few_shot: 0
        metrics:
          - type: inst_level_strict_acc and prompt_level_strict_acc
            value: 50.41
            name: strict accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=PJMixers-Dev/LLaMa-3.2-Instruct-JankMixBread-v0.1-3B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: BBH (3-Shot)
          type: BBH
          args:
            num_few_shot: 3
        metrics:
          - type: acc_norm
            value: 22.76
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=PJMixers-Dev/LLaMa-3.2-Instruct-JankMixBread-v0.1-3B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MATH Lvl 5 (4-Shot)
          type: hendrycks/competition_math
          args:
            num_few_shot: 4
        metrics:
          - type: exact_match
            value: 10.73
            name: exact match
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=PJMixers-Dev/LLaMa-3.2-Instruct-JankMixBread-v0.1-3B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: GPQA (0-shot)
          type: Idavidrein/gpqa
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 4.36
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=PJMixers-Dev/LLaMa-3.2-Instruct-JankMixBread-v0.1-3B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MuSR (0-shot)
          type: TAUR-Lab/MuSR
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 4.68
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=PJMixers-Dev/LLaMa-3.2-Instruct-JankMixBread-v0.1-3B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MMLU-PRO (5-shot)
          type: TIGER-Lab/MMLU-Pro
          config: main
          split: test
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 23.15
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=PJMixers-Dev/LLaMa-3.2-Instruct-JankMixBread-v0.1-3B
          name: Open LLM Leaderboard

LLaMa-3.2-Instruct-JankMixBread-v0.1-3B

This is a merge of pre-trained language models created using mergekit.

Merge Details

Merge Method

This model was merged using the breadcrumbs_ties merge method using meta-llama/Llama-3.2-3B as a base.

Models Merged

The following models were included in the merge:

• meta-llama/Llama-3.2-3B-Instruct
• PJMixers-Dev/LLaMa-3.2-Instruct-JankMix-v0.1-SFT-3B

Configuration

The following YAML configuration was used to produce this model:

merge_method: breadcrumbs_ties
base_model: meta-llama/Llama-3.2-3B
tokenizer_source: PJMixers-Dev/LLaMa-3.2-Instruct-JankMix-v0.1-SFT-3B
dtype: bfloat16
parameters:
  normalize: true
models:
  - model: meta-llama/Llama-3.2-3B-Instruct
    parameters:
      weight: 1
      density: 0.9
      gamma: 0.01
  - model: PJMixers-Dev/LLaMa-3.2-Instruct-JankMix-v0.1-SFT-3B
    parameters:
      weight: 1
      density: 0.9
      gamma: 0.01

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	19.35
IFEval (0-Shot)	50.41
BBH (3-Shot)	22.76
MATH Lvl 5 (4-Shot)	10.73
GPQA (0-shot)	4.36
MuSR (0-shot)	4.68
MMLU-PRO (5-shot)	23.15