---
dataset_info:
  features:
  - name: id
    dtype: string
  - name: conversations
    list:
    - name: content
      dtype: string
    - name: from
      dtype: string
  splits:
  - name: train
    num_bytes: 30583910
    num_examples: 2959
  download_size: 30538658
  dataset_size: 30583910
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---

Translating [Jackrong/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled](https://huggingface.co/Jackrong/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled)'s training dataset via TranslateGemma.

The dataset consists of high-quality, filtered reasoning distillation data:

| Dataset Name | Description / Purpose |
|--------------|-----------------------|
| [nohurry/Opus-4.6-Reasoning-3000x-filtered](https://huggingface.co/datasets/nohurry/Opus-4.6-Reasoning-3000x-filtered) | Provides comprehensive Claude 4.6 Opus reasoning trajectories. |
| [Jackrong/Qwen3.5-reasoning-700x](https://huggingface.co/datasets/Jackrong/Qwen3.5-reasoning-700x) | Additional curated reasoning samples designed to strengthen structured step-by-step problem solving and improve reasoning diversity. |