200K Training Dataset

该数据集包含了大约20万个示例，旨在评估模型的性能和扩展能力。此外，该数据集还用于评估模型在大数据量下的可扩展性。具体来说，它包含了大约20万个训练示例，涉及的任务是机器人操作与接触合成。

This dataset contains approximately 200,000 examples, which is designed to evaluate the performance and scalability of models. Furthermore, this dataset is also utilized to assess the scalability of models when dealing with large-scale data volumes. Specifically, it comprises approximately 200,000 training examples, with the tasks focusing on robot manipulation and contact synthesis.