python-3.x - 配置 Fast-Rcnn.config 以使用 Adam 优化器和其他参数

Question

我有以下 fast_rcnn_resnet101_coco.config ( here )。在这个配置文件中，我用亚当优化器替换了momentum_optimizer，如下所示:

train_config: {
  batch_size: 1
  optimizer {
    #momentum_optimizer: {
    adam_optimizer: {
      learning_rate: {
        manual_step_learning_rate {
          initial_learning_rate: 0.00001
          schedule {
            step: 4500
            learning_rate: .00001
          }
          schedule {
            step: 10000
            learning_rate: .000001
          }
        }
      }
      #momentum_optimizer_value: 0.9
    }
    use_moving_average: false
  }
  gradient_clipping_by_norm: 10.0
  fine_tune_checkpoint: "faster_rcnn_resnet101_coco_2018_01_28/model.ckpt"
  from_detection_checkpoint: true
  data_augmentation_options {
    random_horizontal_flip {
    }
  }
}

我已转介 Tensorflow Object Detection: use Adam instead of RMSProp做这个改变。我的目标是配置我更快的 rcnnresnet101.config 文件( attached here )以匹配该文件的配置:

我的目标是我的 .config 文件应该包含 .yaml 文件中提到的所有参数。到目前为止，我仅针对一个参数(“学习率”)成功地做到了这一点。如何在我的配置文件中集成 rpn_batch 大小、步长等参数？

Answer 1

您需要了解的基本事实如下:

配置文件必须与消息 TrainEvalPipelineConfig 匹配.现在该消息由多个组件组成。因此，如果要修改组件中的某些内容，则应转到定义该组件消息的 proto 文件，查看其中可能的参数，然后根据该参数修改配置文件。这正是您最终为更改优化器所做的。

给你一个提示，如果你想改变RPN批量大小，你必须修改this parameter .因此，在 proto 文件中查找它并将其添加到您的最终配置文件中。

举个例子，如果我使用原始配置文件进行一个小改动，即 RPN 批量大小为 128，我的配置文件将如下所示:

# Faster R-CNN with Resnet-101 (v1), configuration for MSCOCO 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 {
  faster_rcnn {
    num_classes: 90
    image_resizer {
      keep_aspect_ratio_resizer {
        min_dimension: 600
        max_dimension: 1024
      }
    }
    feature_extractor {
      type: 'faster_rcnn_resnet101'
      first_stage_features_stride: 16
    }
    first_stage_anchor_generator {
      grid_anchor_generator {
        scales: [0.25, 0.5, 1.0, 2.0]
        aspect_ratios: [0.5, 1.0, 2.0]
        height_stride: 16
        width_stride: 16
      }
    }
    first_stage_box_predictor_conv_hyperparams {
      op: CONV
      regularizer {
        l2_regularizer {
          weight: 0.0
        }
      }
      initializer {
        truncated_normal_initializer {
          stddev: 0.01
        }
      }
    }
    first_stage_nms_score_threshold: 0.0
    first_stage_nms_iou_threshold: 0.7
    first_stage_max_proposals: 300
    first_stage_localization_loss_weight: 2.0
    first_stage_objectness_loss_weight: 1.0
    # below i modify the RPN batch size to 128
    first_stage_minibatch_size: 128 
    initial_crop_size: 14
    maxpool_kernel_size: 2
    maxpool_stride: 2
    second_stage_box_predictor {
      mask_rcnn_box_predictor {
        use_dropout: false
        dropout_keep_probability: 1.0
        fc_hyperparams {
          op: FC
          regularizer {
            l2_regularizer {
              weight: 0.0
            }
          }
          initializer {
            variance_scaling_initializer {
              factor: 1.0
              uniform: true
              mode: FAN_AVG
            }
          }
        }
      }
    }
    second_stage_post_processing {
      batch_non_max_suppression {
        score_threshold: 0.0
        iou_threshold: 0.6
        max_detections_per_class: 100
        max_total_detections: 300
      }
      score_converter: SOFTMAX
    }
    second_stage_localization_loss_weight: 2.0
    second_stage_classification_loss_weight: 1.0
  }
}

train_config: {
  batch_size: 1
  optimizer {
    momentum_optimizer: {
      learning_rate: {
        manual_step_learning_rate {
          initial_learning_rate: 0.0003
          schedule {
            step: 900000
            learning_rate: .00003
          }
          schedule {
            step: 1200000
            learning_rate: .000003
          }
        }
      }
      momentum_optimizer_value: 0.9
    }
    use_moving_average: false
  }
  gradient_clipping_by_norm: 10.0
  fine_tune_checkpoint: "PATH_TO_BE_CONFIGURED/model.ckpt"
  from_detection_checkpoint: true
  data_augmentation_options {
    random_horizontal_flip {
    }
  }
}

train_input_reader: {
  tf_record_input_reader {
    input_path: "PATH_TO_BE_CONFIGURED/mscoco_train.record-?????-of-00100"
  }
  label_map_path: "PATH_TO_BE_CONFIGURED/mscoco_label_map.pbtxt"
}

eval_config: {
  num_examples: 8000
  # Note: The below line limits the evaluation process to 10 evaluations.
  # Remove the below line to evaluate indefinitely.
  max_evals: 10
}

eval_input_reader: {
  tf_record_input_reader {
    input_path: "PATH_TO_BE_CONFIGURED/mscoco_val.record-?????-of-00010"
  }
  label_map_path: "PATH_TO_BE_CONFIGURED/mscoco_label_map.pbtxt"
  shuffle: false
  num_readers: 1
}