PermissionDeniedError

I followed the Docker installation from https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/tf1.md. To run a local training job from https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/tf1_training_and_evaluation.md:

tensorflow@_:~/models/research$ python object_detection/model_main.py
--pipeline_config_path=./models/detection_model/pipeline.config
--model_dir=./models/detection_model/
--num_train_steps=50000
--sample_1_of_n_eval_examples=1
--alsologtostderr

Also tried:

tensorflow@_:~/models/research$ python object_detection/model_main.py
--pipeline_config_path=models/detection_model/pipeline.config
--model_dir=models/detection_model/
--num_train_steps=50000
--sample_1_of_n_eval_examples=1
--alsologtostderr

Paths in the pipeline.config file were changed in line with the changes above. For either command, the training job raises:

tensorflow.python.framework.errors_impl.PermissionDeniedError: models/detection_model/graph.pbtxt.tmp9f_; Permission denied.

Is there a workaround?