Tensorflow 模型恢复时，设备指定与当前不匹配报错的解决方法

错误内容：nvalidArgumentError (see above for traceback): Cannot assign a device for operation ‘Reshape’: Operation was explicitly assigned to /job:worker/task:0 but available devices are [ /job:localhost/replica:0/task:0/device:CPU:0, /job:localhost/replica:0/task:0/device:GPU:0 ]. Make sure the device specification refers to a valid device.

错误原因：训练时指定的设备与当前环境下的设备不匹配

解决方法：在恢复.meta的时候，添加一个参数clear devices = True即可清除之前配置。
代码样本如下：

saver = tf.train.import_meta_graph('./model.ckpt.meta',clear_devices=True)

我也是查了很久，自己无意中看源码看到了这个参数，才解决了问题，所以分享下，帮助后来人！