(Tensorflow)加载并重新保存模型,丢失一些参数

时间:2019-04-28 03:16:36

标签: python-3.x tensorflow

我还原了模型,然后再次保存。 当我使用mmconvert将模型从tensorflow转换为caffe时, 它提供了有关变量的差异信息。

原始模型转换信息

Parse file [model-lenet-17000.meta] with binary format successfully.
Tensorflow model file [model-lenet-17000.meta] loaded successfully.
Tensorflow checkpoint file [model-lenet-17000] loaded successfully. [36] variables loaded.

重新保存的模型转换信息。

Parse file [model-lenet-17000.meta] with binary format successfully.
Tensorflow model file [model-lenet-17000.meta] loaded successfully.
Tensorflow checkpoint file [model-lenet-17000] loaded successfully. [14] variables loaded

重新保存恢复的模型后会发生什么。 哪些变量被删除?

0 个答案:

没有答案