File size: 1,507 Bytes
8bf0e31 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 |
---
base_model:
- CultriX/Qwen2.5-14B-Wernickev5
- CultriX/Qwen2.5-14B-Wernickev3
library_name: transformers
tags:
- mergekit
- merge
---
# merge
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
## Merge Details
### Merge Method
This model was merged using the SLERP merge method.
### Models Merged
The following models were included in the merge:
* [CultriX/Qwen2.5-14B-Wernickev5](https://huggingface.co/CultriX/Qwen2.5-14B-Wernickev5)
* [CultriX/Qwen2.5-14B-Wernickev3](https://huggingface.co/CultriX/Qwen2.5-14B-Wernickev3)
### Configuration
The following YAML configuration was used to produce this model:
```yaml
models:
- model: CultriX/Qwen2.5-14B-Wernickev3
- model: CultriX/Qwen2.5-14B-Wernickev5
merge_method: slerp
base_model: CultriX/Qwen2.5-14B-Wernickev3
dtype: bfloat16
parameters:
t: [0, 0.5, 1, 0.5, 0]
dtype: bfloat16
adaptive_merge_parameters:
task_weights:
tinyArc: 1.2 # Emphasize Arc's logical and reasoning tasks
tinyHellaswag: 1.1 # Maintain strong performance in contextual predictions
tinyMMLU: 1.2 # Ensure domain knowledge is well-preserved
tinyTruthfulQA: 1.3 # Leverage v3's top TruthfulQA score
tinyTruthfulQA_mc1: 1.1 # Balance for both models' performance
tinyWinogrande: 1.2 # Enhance contextual understanding
smoothing_factor: 0.2 # Moderate blending for stable integration
gradient_clipping: 1.0 # Prevent over-contribution from any single model
```
|