import numpy as np
import matplotlib.pylab as plt
from data_loader import dataset_loader
from common.util import shuffle_dataset
from common.multi_layer_net import MultiLayerNet
from common.trainer import Trainer
x_train, t_train, x_test, t_test = dataset_loader()
x_train = x_train[:500]
t_train = t_train[:500]
validation_rate = 0.2
validation_num = int(x_train.shape[0] * validation_rate)
x_train, t_train = shuffle_dataset(x_train, t_train)
x_valuation = x_train[:validation_num]
t_valuation = t_train[:validation_num]
x_train = x_train[validation_num:]
t_train = t_train[validation_num:]
def __train(lr, weight_decay, epoches=50):
net = MultiLayerNet(input_size=784,
hidden_size_list=[100, 100, 100, 100, 100, 100],
output_size=10,
weight_decay_lambda=weight_decay)
trainer = Trainer(net,
x_train,
t_train,
import numpy as np
import matplotlib.pylab as plt
from data_loader import dataset_loader
from simple_convnet import SimpleConvnet
from common.trainer import Trainer
x_train, t_train, x_test, t_test = dataset_loader(
"/home/zjy/nlp/Machine_Learning_and_Deep_Learning_Algorithms/datasets",
flatten=False)
# x_train = x_train[:500]
# t_train = t_train[:500]
max_epoches = 20
net = SimpleConvnet(input_dim=(1, 28, 28),
conv_param={
'filter_num': 30,
'filter_size': 5,
'pad': 0,
'stride': 1
},
hidden_size=100,
output_size=10,
weight_init_std=0.01)
trainer = Trainer(net,
x_train,
t_train,
x_test,
t_test,
epoches=max_epoches,
mini_batch_size=100,