LKYDeepNN

LKYDeepNN copied to clipboard

How to save a trained weight matrix?

1

Just as the title. I want to save a trained weight matrix for a direct use in future.

STEAVETM

Header file 裡的 using namespace 需要限制範圍

2

mosdeo

原因：以免父類別指標解構子類別物件時，遺漏應該釋放的記憶體。

mosdeo

在 macOS 上發生「NSInternalInconsistencyException」

2

``` LKYs-MacBook-Air:LKYDeepNN lky$ ./LKYDeepNN Completed BackPropagation Layer InitializeWeights() Completed BackPropagation Layer InitializeWeights() Completed BackPropagation Layer InitializeWeights() Completed BackPropagation Layer InitializeWeights() Activation is SeLU. Activation is Softmax. LossFunction is CrossEntropy. ======...

mosdeo

這樣才能把訓練後的成果重複利用，可以參考 @ONNX 或 [libDNN](https://github.com/botonchou/libdnn/wiki/XML-model) @WiiWilson

mosdeo

mosdeo