keras调用tensorborad时报错：
AttributeError: 'Model' object has no attribute '_get_distribution_strategy'
软件版本：
- python: 3.7.4
- keras: 2.3.1
- tensorboard: 2.1.0
- tensorflow: 2.1.0
解决办法A：
- 参考tensorflow的 pull #34870 : "Use _get_distribution_strategy only when it is available. " https://github.com/tensorflow/tensorflow/pull/34870
The changes introduced in 06d8f77 are not compatible with standalone
Keras(they are compatible withtf.Keras). akeras.Modeldoes not have a_get_distribution_strategymethod, which is now assumed for theTensorboardcallback.
- 直接修改
tensorflow/python/keras/callbacks.py文件，如下图。
相关代码如下。
# 1529行左右 : # distributed_file_utils.write_dirpath()
# 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
# 1732行左右也要改一下: # distributed_file_utils.remove_temp_dirpath()
# 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
解决办法B：
修改引用，改为从tensorflow.keras中引用。
import keras
from tensorflow.keras.callbacks import TensorBoard
from tensorflow.keras.layers import *
from tensorflow.keras.models import Sequential
# xxxxx
