---
configs:
- config_name: default
  data_files:
  - split: clozet_train
    path: clozet/train*
  - split: clozet_val
    path: clozet/val*
  - split: clozet_test
    path: clozet/test*
  - split: senpos_train
    path: senpos/train*
  - split: senpos_val
    path: senpos/val*
  - split: senpos_test
    path: senpos/test*
  - split: plotcom_train
    path: plotcom/train*
  - split: plotcom_val
    path: plotcom/val*
  - split: plotcom_test
    path: plotcom/test*
  - split: outgen_train
    path: outgen/train*
  - split: outgen_val
    path: outgen/val*
  - split: outgen_test
    path: outgen/test*

---
LOT: A Benchmark for Evaluating Chinese Long Text Understanding and Generation

Details are shown in our [TACL 2022 Paper](https://aclanthology.org/2022.tacl-1.25/) and [Github Repo](https://github.com/thu-coai/LOT-LongLM).