[DL021]tensorflow.Keras报错解决：UnknownError: Fail to find the dnn implementation.

环境：

• tensorflow2.3.0
• cuda10.1
• python3.8.3
• 未安装Keras，用的from tensorflow.keras import XXX

出错场景：
运行《python深度学习》（《Deep Learing with Python》）书中代码清单6-39时报错：

UnknownError: Fail to find the dnn implementation.
	 [[{{node CudnnRNN}}]] [[sequential/gru/PartitionedCall]] [Op:__inference_train_function_2948]


Function call stack: train_function -> train_function -> train_function

解决步骤;

1. 重启内核
2. 在第一个cell里面/程序首行，加入下列内容

from tensorflow.compat.v1 import ConfigProto
from tensorflow.compat.v1 import InteractiveSession

config = ConfigProto()
config.gpu_options.allow_growth = True
session = InteractiveSession(config=config)

就可以了
Reference

https://stackoverflow.com/questions/63315046/function-call-stack-train-function