quantum-keek-v2-7b

quantum-keek-v2-7b is a merge of the following models using mergekit:

• Dongwei/DeepSeek-R1-Distill-Qwen-7B-GRPO
• lightblue/DeepSeek-R1-Distill-Qwen-7B-Multilingual
• huihui-ai/DeepSeek-R1-Distill-Qwen-7B-abliterated
• deepseek-ai/DeepSeek-R1-Distill-Qwen-7B

🧩 Configuration

base_model: deepseek-ai/DeepSeek-R1-Distill-Qwen-7B
tokenizer_source: base
dtype: bfloat16

models:
  - model: Dongwei/DeepSeek-R1-Distill-Qwen-7B-GRPO
    parameters:
      density: 0.40
      weight: 0.20

  - model: lightblue/DeepSeek-R1-Distill-Qwen-7B-Multilingual
    parameters:
      density: 0.35
      weight: 0.15

  - model: huihui-ai/DeepSeek-R1-Distill-Qwen-7B-abliterated
    parameters:
      density: 0.12
      weight: 0.05

  - model: deepseek-ai/DeepSeek-R1-Distill-Qwen-7B
    parameters:
      density: 0.30
      weight: 0.10

parameters:
  normalize: true```