evaluate.py 对 tensorflow poets 中其他架构的修改

Question

在 tensorflow-for-poets-2/scripts/evaluate.py. That script was originally written for mobilenet models but can be modified for other architectures accordingly. What should I write in line 41

有一个用于测量模型性能的脚本

logits = graph.get_tensor_by_name("final_training_ops/Wx_plus_b/add:0"). 

我收到 `KeyError:

The name final_training_ops/Wx_plus_b/add:0' refers to a Tensor which does not exist. The operation, 'final_training_ops/Wx_plus_b/add', does not exist in the graph.

Answer 1

这是我自己想出来的。给定一个 .pb 文件，我们可以根据我们的 .pb 文件使用这个 code. So, we need to replace the line 41 和最后一个操作来查看冻结的 .pb TensorFlow 图中的操作。