BIG-Bench但不需要官方版本的恶心依赖项（tensorflow，pypi-bigbench，protobuf）。

dataset = load_dataset("tasksource/bigbench",'movie_recommendation')

重现代码： https://colab.research.google.com/drive/1MKdLdF7oqrSQCeavAcsEnPdI85kD0LzU?usp=sharing

将数据集限制为50k个示例，以保持轻巧。我还删除了默认拆分，当训练可用时，默认=train+val，以节省空间。

@article{srivastava2022beyond,
  title={Beyond the imitation game: Quantifying and extrapolating the capabilities of language models},
  author={Srivastava, Aarohi and Rastogi, Abhinav and Rao, Abhishek and Shoeb, Abu Awal Md and Abid, Abubakar and Fisch, Adam and Brown, Adam R and Santoro, Adam and Gupta, Aditya and Garriga-Alonso, Adri{\`a} and others},
  journal={arXiv preprint arXiv:2206.04615},
  year={2022}
}