Im new to tensorflow,and when I check the output, I find that the Lenet isn't converged....
So I think it may be caused by the failure of backward propagation....
I think the main problem is in the set of session, but .I can't find it...
My code is set in github
tf_mnist.py is the main python,and then the tf_network.py
It can be directly run by python3 tf_mnist.py
So can anyone help me check these two files?there is no need to check the other files .
It can be run, but evidently it isn't converged....
The result is as following:
EPOCH 1: Loss = 2.30815, Validation Accuracy = 0.11578
EPOCH 2: Loss = 2.30507, Validation Accuracy = 0.11578
EPOCH 3: Loss = 2.30479, Validation Accuracy = 0.11578
EPOCH 4: Loss = 2.30477, Validation Accuracy = 0.11578
EPOCH 5: Loss = 2.30477, Validation Accuracy = 0.11578
EPOCH 6: Loss = 2.30476, Validation Accuracy = 0.11578
EPOCH 7: Loss = 2.30476, Validation Accuracy = 0.11578
EPOCH 8: Loss = 2.30476, Validation Accuracy = 0.11578
EPOCH 9: Loss = 2.30476, Validation Accuracy = 0.11578
EPOCH 10: Loss = 2.30476, Validation Accuracy = 0.11578
Thank you for your help
