cnn_mnist example which trains a CNN network on MNIST data stays at random (10%) accuracy over epochs;cnn_from_keras example which loads a pre-trained CNN from Keras and achieves expected high accuracy (90.14%)
The above suggests that the forward passes of conv2d, maxpool2d, and flatten layers are implemented correctly.
The culprit may be in the implementation of backward methods for any of these layers, or in the backward flow of data.
This should be fixed before the release of v0.13.0.
cnn_mnistexample which trains a CNN network on MNIST data stays at random (10%) accuracy over epochs;cnn_from_kerasexample which loads a pre-trained CNN from Keras and achieves expected high accuracy (90.14%)The above suggests that the forward passes of
conv2d,maxpool2d, andflattenlayers are implemented correctly.The culprit may be in the implementation of
backwardmethods for any of these layers, or in the backward flow of data.This should be fixed before the release of v0.13.0.