This model is a merged pre-trained language model created using MergeKit with the TIES merge method. It uses Qwen/Qwen2.5-32B-Instruct as the base and combines Qwen/QwQ-32B-Preview and deepseek-ai/DeepSeek-R1-Distill-Qwen-32B with equal weight and density. The merge configuration includes normalization, int8 masking, and bfloat16 precision for optimized performance.
This is a merge of pre-trained language models created using
mergekit.
This model was merged using the
TIES merge method using
Qwen/Qwen2.5-32B-Instruct as a base.
1models:
2 - model: deepseek-ai/DeepSeek-R1-Distill-Qwen-32B
3 - model: Qwen/QwQ-32B-Preview
4 parameters:
5 weight: 1
6 density: 1
7merge_method: ties
8base_model: Qwen/Qwen2.5-32B-Instruct
9parameters:
10 weight: 1
11 density: 1
12 normalize: true
13 int8_mask: true
14dtype: bfloat16
