PlumMath

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

Merge Details

Merge Method

This model was merged using the della merge method using meta-llama/Llama-3.1-8B-Instruct as a base.

Models Merged

The following models were included in the merge:

Configuration

The following YAML configuration was used to produce this model:

merge_method: della
dtype: bfloat16
parameters:
  normalize: true
models:
  - model: ValiantLabs/Llama3.1-8B-ShiningValiant2
    parameters:
      density: 0.5
      weight: 0.3
  - model: ValiantLabs/Llama3.1-8B-Cobalt
    parameters:
      density: 0.5
      weight: 0.2
base_model: meta-llama/Llama-3.1-8B-Instruct

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	13.80
IFEval (0-Shot)	22.42
BBH (3-Shot)	16.45
MATH Lvl 5 (4-Shot)	3.93
GPQA (0-shot)	9.06
MuSR (0-shot)	8.98
MMLU-PRO (5-shot)	21.95

Model tree for sequelbox/Llama3.1-8B-PlumMath

Evaluation results

acc on Winogrande (5-Shot)
self-reported

72.380
acc on MathQA (5-Shot)
self-reported

40.270
strict accuracy on IFEval (0-Shot)
Open LLM Leaderboard

22.420
normalized accuracy on BBH (3-Shot)
Open LLM Leaderboard

16.450
exact match on MATH Lvl 5 (4-Shot)
Open LLM Leaderboard

3.930
acc_norm on GPQA (0-shot)
Open LLM Leaderboard

9.060
acc_norm on MuSR (0-shot)
Open LLM Leaderboard

8.980
accuracy on MMLU-PRO (5-shot)
test set Open LLM Leaderboard

21.950

View on Papers With Code