kaggle 训练自己的cycleGan

1：Kaggle上面命令Proceed (y/n)的处理

• 直接在后面输入yes
• https://blog.csdn.net/gdufsTFknight/article/details/107957433

2：查看tensorflow版本

import tensorflow as td
tf.__version__

3：keras-contrib

• pip install git+https://www.github.com/keras-team/keras-contrib.git

  4：kaggle约束

  tensorflow-cloud 0.1.13 requires tensorboard>=2.3.0, but you have tensorboard 1.13.1 which is incompatible.
  tensorflow-cloud 0.1.13 requires tensorflow<3.0,>=1.15.0, but you have tensorflow 1.13.1 which is incompatible.
  

  4：kaggle 的 TF要求

• kaggle要求tensorboard > 2.3.0， 1.15.0 < tensorflow < 3.0

5：No module named 'keras.backend.tensorflow_backen

• ModuleNotFoundError: No module named 'keras.backend.tensorflow_backend'; 'keras.backend' is not a package
  

• pip install tensorflow==2.2.0
  then 
  pip install Keras==2.2.0
  This worked for me with Python 3.7.
  

• https://stackoverflow.com/questions/60970093/modulenotfounderror-no-module-named-keras-backend-tensorflow-backend

6：module ‘tensorflow’ has no attribute ‘Session’

import tensorflow as tf
tf.Session()

# 在tf2之后，使用下面这个
tf.compat.v1.Session()
# 或者是重新安装会tf1 的版本