Dataset Card (Qwen3.5-reasoning-700x)

Dataset Summary

Qwen3.5-reasoning-700x is a high-quality distilled dataset.

This dataset uses the high-quality instructions constructed by Alibaba-Superior-Reasoning-Stage2 as the seed question set. By calling the latest Qwen3.5-27B full-parameter model on the Alibaba Cloud DashScope platform as the teacher model, it generates high-quality responses featuring long-text reasoning processes (Chain-of-Thought). It covers several major domains: Mathematics | Code | Logic | Science | Instructions.

Generation Parameters

• Seed Data: Alibaba-Superior-Reasoning-Stage2
• Teacher Model: qwen3.5-27b (Full-parameter model)
• API Provider: Alibaba Cloud DashScope
• Context Window: 16384 tokens.
• Temperature: 0.6.

Dataset Structure

{
  "id": "a1b2c3d4e5f6g7h8i9j0k1l2m3n4o5p6",
  "conversation": [
    {
      "from": "human",
      "value": "Seed question..."
    },
    {
      "from": "gpt",
      "value": "<think>\nComplete chain of thought process...\n</think>\n\nFinal answer..."
    }
  ],
  "input": "Original seed question...",
  "output": "<think>\nComplete chain of thought process...\n</think>\n\nFinal answer...",
  "domain": "Math | Code | Logic | Science | ...",
  "meta": {
    "training_stage": "stage2",
    "teacher_model": "Qwen3.5-27B"
  }
}

Disclaimers

• Language Model Hallucinations: Although Qwen3.5-27B is an exceptionally powerful model and a low Temperature was set to maintain rigor, there may still be a very small number of calculation errors or logical fallacies in the generated data. It is highly recommended to conduct sample quality inspections before proceeding with fine-tuning.
• License Compliance: The open-sourcing and usage of this dataset must strictly comply with the open-source license agreement of the Qwen models, as well as the terms of service of Alibaba Cloud DashScope.