Skip to content

Dataset · OpenDataArena

Multimodal reasoning on 1.8M examples

OpenDataArena/MMFineReason-1.8M-Qwen3-VL-235B-Thinking

1.8M multimodal examples with detailed reasoning distilled from Qwen3-VL-235B-Thinking.

What's inside

The dataset contains 1.8 million multimodal examples and 5.1 billion solution tokens in English. Detailed reasoning annotations are distilled from the Qwen3-VL-235B-A22B-Thinking model. The format is not specified in the card.

What you can do with it

  • SFT of a multimodal model for image reasoning
  • Fine-tuning VLMs for visual question answering
  • Evaluating reasoning quality on math and comprehension benchmarks
Tasks
Visual question answering, Question answering, Text generation, Image text → text
Languages
English
Size
1M–10M rows
Downloads this month
2.7K
Likes
127

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("OpenDataArena/MMFineReason-1.8M-Qwen3-VL-235B-Thinking")