Half the data was geared towards better reasoning (EvolKit-20k and reasoning-base-20k), the other half will help to de-censor the model (WizardLM data set).
Merge
This is a merge of pre-trained language models created using mergekit.
Merge Details
Merge Method
This model was merged using the SLERP merge method.
Models Merged
The following models were included in the merge:
Configuration
The following YAML configuration was used to produce this model:
slices:
- sources:
- model: theprint/ReWiz-7B
layer_range: [0, 32]
- model: theprint/WorldBuilder-7B
layer_range: [0, 32]
merge_method: slerp
base_model: theprint/ReWiz-7B
parameters:
t:
- filter: self_attn
value: [0.1, 0.5, 0.3, 0.7, 0.9]
- filter: mlp
value: [0.9, 0.5, 0.7, 0.3, 0.1]
- value: 0.5
dtype: bfloat16
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 15.66 |
| IFEval (0-Shot) | 25.10 |
| BBH (3-Shot) | 25.08 |
| MATH Lvl 5 (4-Shot) | 2.95 |
| GPQA (0-shot) | 2.57 |
| MuSR (0-shot) | 16.39 |
| MMLU-PRO (5-shot) | 21.90 |