This is a merge of pre-trained language models created using
mergekit.
This model was merged using the
Model Noiser merge method.
1merge_method: model_noiser
2models:
3 - model: Qwen/Qwen2.5-0.5B-Instruct
4parameters:
5 normalize: true
6 noise_stddev: 0.02 # adjust as needed
7
The YAML config used does NOT work in mergekit normally.
You need my fork for this.
