def test(status_var, t):
numEpoch = config.numEpoch
trainExamples = config.trainExamples
valExamples = config.valExamples
batchSize = config.batchSize
saveModel = config.saveModel
modelFile = config.modelFile
validate = config.validate
pretrain = config.pretrain
trainedModel = config.trainedModel
log = config.log
trainlog = config.trainlog
vallog = config.vallog
net = cnn()
model = scio.loadmat("./models/1_0.01_model.mat")
net.Weights = np.asarray(model['weights'][0])
for i in range(5):
net.Biases[i] = model['biases'][0][i][0]
mnist = scio.loadmat('./data/mnist_2D.mat')
trainList = [np.random.randint(0, 60000) for i in range(trainExamples)]
valList = [np.random.randint(0, 10000) for i in range(valExamples)]
valLabel = np.asarray([[0 for i in range(10)] for j in range(valExamples)])
valData = np.zeros((valExamples, mnist['X_train'][0].shape[0],
mnist['X_train'][0].shape[1]))
j = 0
for i in valList:
valLabel[j, mnist['Y_test'][i]] = 1
valData[j] = mnist['X_test'][i]
j += 1
### Validation
acc = 0
val_loss = 0
for i in range(valExamples):
[predict, loss] = net.validate(valData[i], valLabel[i])
if valLabel[i][predict] == 1:
acc += 1
val_loss += loss
if validate:
test_accuracy_info = '\n Validation Loss: {0} ,Accuracy: {1} '.format(
str(val_loss / valExamples), str(acc * 100.0 / valExamples))
print(test_accuracy_info)
t.insert(END, test_accuracy_info)
import config
import signal
import sys
numEpoch = config.numEpoch
trainExamples = config.trainExamples
valExamples = config.valExamples
batchSize = config.batchSize
modelFile = config.modelFile
validate = config.validate
net = cnn()
mnist = scio.loadmat('../data/mnist_2D.mat')
numIter = 1
for epoch in range(numEpoch):
trainList = [np.random.randint(0,60000) for i in range(trainExamples)]
valList = [np.random.randint(0,10000) for i in range(valExamples)]
trainLabel = np.asarray([[0 for i in range(10)] for j in range(trainExamples)])