---
pretty_name: VisuLogic-Train (OmniTool)
language: [en]
license: apache-2.0
size_categories: [1K<n<10K]
task_categories: [image-text-to-text]
tags: [multimodal, visual-reasoning, sft]
# Make the Viewer know which files are in which split/config:
configs:
  - config_name: solution
    data_files: visulogic_train_solution.jsonl
  - config_name: qwen
    data_files: visulogic_train_qwen.jsonl
  - config_name: internvl
    data_files: visulogic_train_internvl.jsonl
---

# VisuLogic-Train (OmniTool)

Short description of the dataset…

## Usage

```python
from datasets import load_dataset
ds = load_dataset("OmniTool/VisuLogic-Train", "solution", split="train")"internvl"
print(ds[0])