环境版本

TensorFlow 2.1.0
Python 3.7
keras 2.3.1

1. 数据集下载与参数配置

下载链接
- https://github.com/matterport/Mask_RCNN/releases
保存路径

balloon.py运行参数配置

train
	--dataset=..\\..\\balloon
	--wieghts=coco

在这里插入图片描述

2. 运行报错开始

报错1

AttributeError: module 'tensorflow' has no attribute 'random_shuffle'
将random_shuffle改为random.shuffle

报错2

AttributeError: module 'tensorflow' has no attribute 'log''
将log改为math.log

报错3

AttributeError:‘str’ object has no sttribute decode
h5py版本太高，降低版本
```
pip install h5py==0.10.2
```

报错4

tensorflow.python.framework.errors_impl.OperatorNotAllowedInGraphError: using a tf.Tensor as a Python bool is not allowed in Graph execution. Use Eager execution or decorate this function with @tf.function.
修改

报错5

AttributeError: ‘Model’ object has no attribute ‘metrics_tensors’
修改，添加代码self.keras_model.metrics_tensors = []

报错6

AttributeError: ‘Model’ object has no attribute ‘_get_distribution_strategy’
修改文件..\Lib\site-packages\tensorflow_core\python\keras\callbacks.py【自己找自己环境位置】

修改代码

    # In case this callback is used via native Keras, _get_distribution_strategy does not exist.
if hasattr(self.model, '_get_distribution_strategy'):
  # TensorBoard callback involves writing a summary file in a
  # possibly distributed settings.
  self._log_write_dir = distributed_file_utils.write_dirpath(
      self.log_dir, self.model._get_distribution_strategy())  # pylint: disable=protected-access
else:
  self._log_write_dir = self.log_dir

修改前文件
修改后文件

修改代码

   # In case this callback is used via native Keras, _get_distribution_strategy does not exist.
if hasattr(self.model, '_get_distribution_strategy'):
  # Safely remove the unneeded temp files.
  distributed_file_utils.remove_temp_dirpath(
      self.log_dir, self.model._get_distribution_strategy())  # pylint: disable=protected-access