System information
Describe the problem I could train the model locally. The accuracy was decent but I wanted to try and use GCP to get similar results. I followed the google documentation mentioned here:
https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/running_on_cloud.md
https://medium.com/google-cloud/object-detection-tensorflow-and-google-cloud-platform-72e0a3f3bdd6
When I try to run the training session on the cloud using the following command:
gcloud ml-engine jobs submit training object_detection_188001 \ --job-dir=gs://weeddetection/train \ --packages dist/object_detection-0.1.tar.gz,slim/dist/slim-0.1.tar.gz \ --module-name object_detection.train \ --region us-central1 \ --config object_detection/samples/cloud/cloud.yml \ --train_dir=gs://weeddetection/train \ --pipeline_config_path=gs://weeddetection/data/ssd_mobilenet_v1_pets.config
I get the following error:
ERROR: (gcloud.ml-engine.jobs.submit.training) unrecognized arguments:
\
\
\
\
\
\
--train_dir=gs://weeddetection/train
\
--pipeline_config_path=gs://weeddetection/data/ssd_mobilenet_v1_pets.config
weeddetection is my GC bucket name and 188001 is my shell ID.
When you include custom options like --train-dir
and --pipeline_config_path
, you need to put them at the end of the command. You also need to separate the command's options and your options with --
. Try this:
gcloud ml-engine jobs submit training object_detection_188001 --job-dir=gs://weeddetection/train --packages dist/object_detection-0.1.tar.gz,slim/dist/slim-0.1.tar.gz --module-name object_detection.train --region us-central1 --config object_detection/samples/cloud/cloud.yml -- --train_dir=gs://weeddetection/train --pipeline_config_path=gs://weeddetection/data/ssd_mobilenet_v1_pets.config