license: apache-2.0 task_categories: - question-answering language: - en - zh tags: - agent - math - science - VQA size_categories: - 10K<n<100K
This dataset is used to train CodeV-RL, as described in the referenced paper.