This is a merge of pre-trained language models created using mergekit.
This model was merged using the Model Stock merge method using bamec66557/MNRP_0.5 as a base.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
models:
- model: bamec66557/MNRP_0.5
- model: Infermatic/MN-12B-Inferor-v0.1
- model: redrix/AngelSlayer-12B-Unslop-Mell-RPMax-DARKNESS
- model: crestf411/MN-Slush
- model: DavidAU/MN-GRAND-Gutenberg-Lyra4-Lyra-12B-DARKNESS
- model: nbeerbower/mistral-nemo-wissenschaft-12B
- model: ArliAI/Mistral-Nemo-12B-ArliAI-RPMax-v1.2
merge_method: model_stock
base_model: bamec66557/MNRP_0.5
parameters:
weight: 1
density: 1
normalize: true
int8_mask: true
tokenizer_source: union
parameters:
normalize: false
int8_mask: true
epsilon: 0.05
lambda: 1
dtype: bfloat16
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 22.57 |
| IFEval (0-Shot) | 38.52 |
| BBH (3-Shot) | 34.07 |
| MATH Lvl 5 (4-Shot) | 12.46 |
| GPQA (0-shot) | 9.40 |
| MuSR (0-shot) | 11.28 |
| MMLU-PRO (5-shot) | 29.69 |