Error in Tensorflow Object Detection running on Cloud ML: no module object_detection.train

Question

Not sure if this is a bug (file report), or I did something wrong.

System information:

Linux 17.04 TensorFlow version: 1.9.0 Python version: 2.7.13

Command I used:

gcloud ml-engine jobs submit training object_detection_$(date +%Y%m%d_%H%M%S)  \
    --job-dir="gs://mybucket/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 /home/me/Desktop/die_detection/config.yml \
    -- \
    --train_dir="gs://mybucket/train" \
    --pipeline_config_path="gs://mybucket/data/pipeline_cloud.config"

Tried following this sample, but with my own data: https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/running_pets.md

Works fine locally. Running on CloudML Engine, I get a non-0 exit status. From the logs it seems object_detection.train cannot be found.

Source code / logs

E  The replica ps 0 exited with a non-zero status of 1. Termination reason: Error. To find out more about why your job exited please check the logs: https://console.cloud.google.com/logs/viewer?project=730275006403&resource=ml_job%2Fjob_id%2Fobject_detection_20180725_090524&advancedFilter=resource.type%3D%22ml_job%22%0Aresource.labels.job_id%3D%22object_detection_20180725_090524%22 
  undefined
E  ps-replica-0 Command '['python', '-m', u'object_detection.train', u'--train_dir=gs://mybucket/train', u'--pipeline_config_path=gs://mybucket/data/pipeline_cloud.config', '--job-dir', u'gs://mybucket/train']' returned non-zero exit status 1 ps-replica-0 
  undefined
E  ps-replica-0 /usr/bin/python: No module named object_detection.train ps-replica-0 
  undefined

My pipeline.config:

# SSD with Mobilenet v1, configured for Oxford-IIIT Pets Dataset.
# Users should configure the fine_tune_checkpoint field in the train config as
# well as the label_map_path and input_path fields in the train_input_reader and
# eval_input_reader. Search for "PATH_TO_BE_CONFIGURED" to find the fields that
# should be configured.

model {
  ssd {
    num_classes: 1
    box_coder {
      faster_rcnn_box_coder {
        y_scale: 10.0
        x_scale: 10.0
        height_scale: 5.0
        width_scale: 5.0
      }
    }
    matcher {
      argmax_matcher {
        matched_threshold: 0.5
        unmatched_threshold: 0.5
        ignore_thresholds: false
        negatives_lower_than_unmatched: true
        force_match_for_each_row: true
      }
    }
    similarity_calculator {
      iou_similarity {
      }
    }
    anchor_generator {
      ssd_anchor_generator {
        num_layers: 6
        min_scale: 0.2
        max_scale: 0.95
        aspect_ratios: 1.0
        aspect_ratios: 2.0
        aspect_ratios: 0.5
        aspect_ratios: 3.0
        aspect_ratios: 0.3333
      }
    }
    image_resizer {
      fixed_shape_resizer {
        height: 300
        width: 300
      }
    }
    box_predictor {
      convolutional_box_predictor {
        min_depth: 0
        max_depth: 0
        num_layers_before_predictor: 0
        use_dropout: false
        dropout_keep_probability: 0.8
        kernel_size: 1
        box_code_size: 4
        apply_sigmoid_to_scores: false
        conv_hyperparams {
          activation: RELU_6,
          regularizer {
            l2_regularizer {
              weight: 0.00004
            }
          }
          initializer {
            truncated_normal_initializer {
              stddev: 0.03
              mean: 0.0
            }
          }
          batch_norm {
            train: true,
            scale: true,
            center: true,
            decay: 0.9997,
            epsilon: 0.001,
          }
        }
      }
    }
    feature_extractor {
      type: 'ssd_mobilenet_v1'
      min_depth: 16
      depth_multiplier: 1.0
      conv_hyperparams {
        activation: RELU_6,
        regularizer {
          l2_regularizer {
            weight: 0.00004
          }
        }
        initializer {
          truncated_normal_initializer {
            stddev: 0.03
            mean: 0.0
          }
        }
        batch_norm {
          train: true,
          scale: true,
          center: true,
          decay: 0.9997,
          epsilon: 0.001,
        }
      }
    }
    loss {
      classification_loss {
        weighted_sigmoid {
        }
      }
      localization_loss {
        weighted_smooth_l1 {
        }
      }
      hard_example_miner {
        num_hard_examples: 3000
        iou_threshold: 0.99
        loss_type: CLASSIFICATION
        max_negatives_per_positive: 3
        min_negatives_per_image: 0
      }
      classification_weight: 1.0
      localization_weight: 1.0
    }
    normalize_loss_by_num_matches: true
    post_processing {
      batch_non_max_suppression {
        score_threshold: 1e-8
        iou_threshold: 0.6
        max_detections_per_class: 100
        max_total_detections: 100
      }
      score_converter: SIGMOID
    }
  }
}

train_config: {
  batch_size: 24
  optimizer {
    rms_prop_optimizer: {
      learning_rate: {
        exponential_decay_learning_rate {
          initial_learning_rate: 0.0004
          decay_steps: 800720
          decay_factor: 0.95
        }
      }
      momentum_optimizer_value: 0.9
      decay: 0.9
      epsilon: 1.0
    }
  }

  num_steps: 20000
  data_augmentation_options {
    random_horizontal_flip {
    }
  }
  data_augmentation_options { 
    ssd_random_crop {
    }
  }
}

train_input_reader: {
  tf_record_input_reader {
    input_path: "gs://mybucket/data/train.record"
  }
  label_map_path: "gs://mybucket/data/object-detection.pbtxt"
}

eval_config: {
  metrics_set: "coco_detection_metrics"
  num_examples: 32
}

eval_input_reader: {
  tf_record_input_reader {
    input_path: "gs://mybucket/data/val.record""
  }
  label_map_path: "gs://mybucket/data/object-detection.pbtxt"
  shuffle: false
  num_readers: 1
}

My config.yml

trainingInput:
  runtimeVersion: "1.0"
  scaleTier: CUSTOM
  masterType: standard_gpu
  workerCount: 1
  workerType: standard_gpu
  parameterServerCount: 1
  parameterServerType: standard

lwz1992 · Accepted Answer · 2018-07-27 20:57:39Z

5

I assumed you're using the unmodified object detection sample. According to https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/running_pets.md#starting-training-and-evaluation-jobs-on-google-cloud-ml-engine, the --module-name should be object_detection.model_main instead of object_detection.train. Could you please double check in your dist/object_detection-0.1.tar.gz file?

answered Jul 27, 2018 at 20:57

lwz1992

3041 silver badge4 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Farzad Vertigo · Accepted Answer · 2019-02-01 04:53:10Z

1

Copy the train.py from your models\research\object_detection\legacy dir and paste to the models\research\object_detection and cd to models\research and run the following cmd : python setup.py sdist. This will create a new object_detection-0.1.tar.gz in your models-master\research\dist, your can then run your commands again:

gcloud ml-engine jobs submit training object_detection_$(date +%Y%m%d_%H%M%S)  \
--job-dir="gs://mybucket/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 /home/me/Desktop/die_detection/config.yml \
-- \
--train_dir="gs://mybucket/train" \
--pipeline_config_path="gs://mybucket/data/pipeline_cloud.config"

edited Feb 1, 2019 at 4:53

Farzad Vertigo

2,8481 gold badge33 silver badges35 bronze badges

answered Feb 1, 2019 at 2:35

Leo

111 bronze badge

Collectives™ on Stack Overflow

Error in Tensorflow Object Detection running on Cloud ML: no module object_detection.train

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related