python train.py 时报错 libselector.so

>python train.py --run_mode=local --config=config.train.json
运行以上命令报如下错误，这个应该如何解决呢

Start by reproducing `python train.py --run_mode=local --config=config.train.json` and inspect the reported `libselector.so` error in the attached image. Read `train.py` and `config.train.json` to identify how local training loads the selector library. Done means the command runs without the reported shared-library error.

Written by the indexing model from the issue text.