如何在googlecolab上设置多个目录的PYTHONPATH

Question

我在谷歌Colab与GPU的工作。我想在colab上进行一些训练，因为它提供了GPU，这有助于减少训练时间。我已经上传了包含google驱动器中所有文件的文件夹。然后我使用以下代码安装了带有colab的google drive：

from google.colab import drive
drive.mount('/content/gdrive')

每当我运行命令进行训练时，都会一次又一次地出现同样的错误。命令和错误如下所示：

!python train.py --logtostderr --train_dir=training/ --pipeline_config_path=training/faster_rcnn_inception_v2_coco.config

错误：

Traceback (most recent call last):
  File "train.py", line 49, in <module>
    from object_detection.builders import dataset_builder
ModuleNotFoundError: No module named 'object_detection'

每当这种类型的问题出现在我的本地机器上时，问题总是因为PYTHONPATH的错误设置而发生。你知道吗

我已经尝试在colab上设置PYTHONPATH，但是同样的错误一再出现。设置PYTHONPATH的常用命令有：

1. 第一次尝试：

   !export PYTHONPATH="${PYTHONPATH}:/content/gdrive/'My Drive'/tensorflow1/models:/content/gdrive/'My Drive'/tensorflow1/models/research:/content/gdrive/'My Drive'/tensorflow1/models/research/slim:/content/gdrive/'My Drive'/tensorflow1/models/research/object_detection"
   

2. 第二次尝试：

   !export PYTHONPATH=$PYTHONPATH:/content/gdrive/'My Drive'/tensorflow1/models:/content/gdrive/'My Drive'/tensorflow1/models/research:/content/gdrive/'My Drive'/tensorflow1/models/research/slim:/content/gdrive/'My Drive'/tensorflow1/models/research/object_detection
   

3. 第三次尝试：

   os.environ['PYTHONPATH'] = "/content/gdrive/'My Drive'/tensorflow1/models:/content/gdrive/'My Drive'/tensorflow1/models/research:/content/gdrive/'My Drive'/tensorflow1/models/research/slim:/content/gdrive/My Drive/tensorflow1/models/research/object_detection"
   

4. 第四次尝试：

   %set_env PYTHONPATH=":/content/gdrive/My\ Drive/tensorflow1/models:/content/gdrive/My\ Drive/tensorflow1/models/research:/content/gdrive/My\ Drive/tensorflow1/models/research/slim:/content/gdrive/My\ Drive/tensorflow1/models/research/object_detection"
   

以上命令均未解决此错误。所以，我需要知道哪个是正确的方法来设置PYTHONPATH在colab上，如果设置PYTHONPATH不是正确的方法来克服这个错误，那么我应该尝试什么呢？你知道吗