Python Tensorflow标签\u图像在重新培训后未发现错误

我已经在自定义数据集上运行了tensoflow的

retain.py

脚本。培训过程结束后，我得到了预期的输出文件

output\u graph.pb

和

output\u labels.txt

。 但是，当我试图使用最新的tensorflow的

label_image.py

脚本预测一张看不见的图像上的标签时，我遇到了以下错误：

NotFoundError (see above for traceback): Error while reading resource variable module/InceptionV3/Mixed_7c/Branch_3/Conv2d_0b_1x1/BatchNorm/moving_mean from Container: localhost. This could mean that the variable was uninitialized. Not found: Container localhost does not exist. (Could not find resource: localhost/module/InceptionV3/Mixed_7c/Branch_3/Conv2d_0b_1x1/BatchNorm/moving_mean)
         [[Node: import/module_apply_default/InceptionV3/InceptionV3/Mixed_7c/Branch_3/Conv2d_0b_1x1/BatchNorm/FusedBatchNorm/ReadVariableOp = ReadVariableOp[dtype=DT_FLOAT, _device="/job:localhost/replica:0/task:0/device:CPU:0"](import/module/InceptionV3/Mixed_7c/Branch_3/Conv2d_0b_1x1/BatchNorm/moving_mean)]]

有人知道为什么某些变量可能被错误初始化吗