In [1]:
import sys
sys.path.insert(0, '../')
import time
import numpy as np
np.set_printoptions(precision=3, linewidth=200, suppress=True)
from library.datasets.cifar10 import CIFAR10
from library.plot_tools import plot_tools
from library.utils import file_utils
import tensorflow as tf
from library.tf.LinearClassifier import TFLinearClassifier
None
In [2]:
from library.hog.hog import HOG
In [3]:
# from tensorflow.python.client import device_lib
# local_device_protos = device_lib.list_local_devices()
# cpu_devices = [x.name for x in local_device_protos if x.device_type == 'CPU']
# gpu_devices = [x.name for x in local_device_protos if x.device_type == 'GPU']
# print('Available CPU Devices: ', end='')
# print(cpu_devices)
# print('Available GPU Devices: ', end='')
# print(gpu_devices)
# if len(gpu_devices) == 0:
# if len(cpu_devices) > 0:
# device_name = '/cpu:0'
# print('Using CPU: %s' %device_name)
# else:
# print('No CPU present in the system!!!')
# else:
# device_name = '/gpu:0'
# print('Using GPU: %s' %device_name)
In [4]:
total_time = 0
In [5]:
exp_no = 105
data_source = 'Website'
num_images_required = 1.0
file_no = 10
In [6]:
device_name = '/gpu:0'
learning_rate = 0.01
training_epochs = 1000
display_step = 1
regularize=True
regularization_const = 0.01
train_validate_split_data = None
train_validate_split = 0.9
transform=True
transform_method='StandardScaler'
learning_rate_type='constant'
init_weights='zeros'
init_bias='ones'
dataset = 'cifar10'
In [7]:
block_size = (8,8)
cell_size = (2,2)
nbins = 9
In [8]:
log_dir = '../logs/' + dataset + '/' + str(file_no).zfill(2) + '_tf_linear_hog/exp_no_' + str(exp_no).zfill(3) + '/'
log_file = log_dir + 'linear_classifier.ckpt'
print('Writing tensorboard logs to %s' %log_file)
print('view logs by running tensorboard: ', end='')
print('\"tensorboard --logdir=\'./%s/10_tf_linear_hog/\' --port 61111\"' %dataset)
Writing tensorboard logs to ../logs/cifar10/10_tf_linear_hog/exp_no_105/linear_classifier.ckpt
view logs by running tensorboard: "tensorboard --logdir='./cifar10/10_tf_linear_hog/' --port 61111"
In [9]:
start = time.time()
one_hot=True
cifar10 = CIFAR10(one_hot_encode=one_hot, num_images=num_images_required, image_mode='grey',
train_validate_split=train_validate_split_data, endian='little')
cifar10.load_data(train=True, test=True, data_directory='./datasets/cifar10/')
end = time.time()
print('[ Step 0] Dataset loaded in %5.6f ms' %((end-start)*1000))
print('Dataset size: ' + str(cifar10.train.data.shape))
num_train_images = cifar10.train.data.shape[0]
total_time += (end-start)
Loading CIFAR 10 Dataset
Downloading and extracting CIFAR 10 file
MD5sum of the file: ./datasets/cifar10/cifar-10.tar.gz is verified
Loading 50000 train images
Loading CIFAR 10 Training Dataset
Reading unpicked data file: ./datasets/cifar10/cifar-10-batches/data_batch_1
Reading unpicked data file: ./datasets/cifar10/cifar-10-batches/data_batch_2
Reading unpicked data file: ./datasets/cifar10/cifar-10-batches/data_batch_3
Reading unpicked data file: ./datasets/cifar10/cifar-10-batches/data_batch_4
Reading unpicked data file: ./datasets/cifar10/cifar-10-batches/data_batch_5
Loading 10000 test images
Loading CIFAR 10 Test Dataset
Unpickling test file: ./datasets/cifar10/cifar-10-batches/test_batch
Reading unpicked test file: ./datasets/cifar10/cifar-10-batches/test_batch
Loaded CIFAR 10 Dataset in 2.2825 seconds
[ Step 0] Dataset loaded in 2283.517361 ms
Dataset size: (50000, 3072)
In [10]:
print('Train data shape:', cifar10.train.data.shape)
if one_hot is True:
print('Train labels shape:', cifar10.train.one_hot_labels.shape)
print('Train class labels shape:', cifar10.train.class_labels.shape)
if train_validate_split_data is not None:
print('Validate data shape:', cifar10.validate.data.shape)
if one_hot is True:
print('Validate labels shape:', cifar10.validate.one_hot_labels.shape)
print('Validate class labels shape:', cifar10.validate.class_labels.shape)
print('Test data shape:', cifar10.test.data.shape)
if one_hot is True:
print('Test labels shape:', cifar10.test.one_hot_labels.shape)
print('Test class labels shape:', cifar10.test.class_labels.shape)
Train data shape: (50000, 3072)
Train labels shape: (50000, 10)
Train class labels shape: (50000,)
Test data shape: (10000, 3072)
Test labels shape: (10000, 10)
Test class labels shape: (10000,)
In [11]:
print('Training images')
print(cifar10.train.data[:5])
if one_hot is True:
print('Training labels')
print(cifar10.train.one_hot_labels[:5])
print('Training classes')
print(cifar10.train.class_labels[:5])
print('Testing images')
print(cifar10.test.data[:5])
if one_hot is True:
print('Testing labels')
print(cifar10.test.one_hot_labels[:5])
print('Testing classes')
print(cifar10.test.class_labels[:5])
Training images
[[ 59 43 50 ..., 140 84 72]
[154 126 105 ..., 139 142 144]
[255 253 253 ..., 83 83 84]
[ 28 37 38 ..., 28 37 46]
[170 168 177 ..., 82 78 80]]
Training labels
[[ 0. 0. 0. 0. 0. 0. 1. 0. 0. 0.]
[ 0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]
[ 0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]
[ 0. 0. 0. 0. 1. 0. 0. 0. 0. 0.]
[ 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]]
Training classes
[6 9 9 4 1]
Testing images
[[158 159 165 ..., 124 129 110]
[235 231 232 ..., 178 191 199]
[158 158 139 ..., 8 3 7]
[155 167 176 ..., 50 52 50]
[ 65 70 48 ..., 136 146 117]]
Testing labels
[[ 0. 0. 0. 1. 0. 0. 0. 0. 0. 0.]
[ 0. 0. 0. 0. 0. 0. 0. 0. 1. 0.]
[ 0. 0. 0. 0. 0. 0. 0. 0. 1. 0.]
[ 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]
[ 0. 0. 0. 0. 0. 0. 1. 0. 0. 0.]]
Testing classes
[3 8 8 0 6]
In [12]:
cifar10.plot_sample(plot_data=True, plot_test=True, fig_size=(7, 7))
Plotting CIFAR 10 Train Dataset
Plotting CIFAR 10 Test Dataset
In [13]:
cifar10.plot_images(cifar10.train.images[:50, :], cifar10.train.class_names[:50],
nrows=5, ncols=10, fig_size=(20,50), fontsize=35, convert=False)
Out[13]:
True
In [14]:
start = time.time()
data_hog = []
feature_size = 0
print('Block size : ' + str(block_size))
print('Cell size : ' + str(cell_size))
print('Number of bins : ' + str(nbins))
hog = HOG(block_size=block_size, cell_size=cell_size, nbins=nbins)
print('Generating HOG features for %d data images' %cifar10.train.images.shape[0])
for fig_num in range(cifar10.train.images.shape[0]):
img = cifar10.train.images[fig_num, :]
gradients = hog.make_hog_gradients(img.astype('uint8'))
data_hog.append(gradients.flatten())
feature_size = gradients.size
data_hog = np.array(data_hog)
print('HOG Features for data: ' + str(data_hog.shape))
end = time.time()
print('Generated HOG for train images in %.6f ms' %((end-start)*1000))
Block size : (8, 8)
Cell size : (2, 2)
Number of bins : 9
Generating HOG features for 50000 data images
HOG Features for data: (50000, 2304)
Generated HOG for train images in 44533.255339 ms
In [15]:
start = time.time()
test_hog = []
feature_size = 0
print('Generating HOG features for %d test images' %cifar10.test.images.shape[0])
for fig_num in range(cifar10.test.images.shape[0]):
img = cifar10.test.images[fig_num, :]
gradients = hog.make_hog_gradients(img.astype('uint8'))
test_hog.append(gradients.flatten())
feature_size = gradients.size
test_hog = np.array(test_hog)
print('HOG Features for test: ' + str(test_hog.shape))
end = time.time()
print('Generated HOG for test images in %.6f ms' %((end-start)*1000))
Generating HOG features for 10000 test images
HOG Features for test: (10000, 2304)
Generated HOG for test images in 9177.517176 ms
In [16]:
num_features = data_hog.shape[1]
num_classes = 10
In [17]:
tf_lc = TFLinearClassifier(verbose=True, device=device_name, session_type='interactive', learning_rate=learning_rate,
num_iterations=training_epochs, display_step=display_step, save_model=True, restore=False,
regularize=regularize, reg_const=regularization_const, train_validate_split=train_validate_split,
init_weights=init_weights, init_bias=init_bias, model_name=log_file, log_dir=log_dir,
learning_rate_type=learning_rate_type, test_log=True, separate_writer=False,
transform_method=transform_method)
print(tf_lc)
<library.tf.LinearClassifier.TFLinearClassifier object at 0x7efbd7634eb8>
In [18]:
start = time.time()
tf_lc.create_graph(num_features=num_features, num_classes=num_classes)
end = time.time()
print('Generated the tensorflow graph in %.4f ms' %((end-start)*1000))
total_time = (end-start)
Tensorflow graph created in 0.1827 seconds
Generated the tensorflow graph in 182.9076 ms
In [19]:
print(tf_lc.print_parameters())
>> Parameters for Linear Classifier
Activation function : softmax
Gradient Descent Method : gradient
Learning rate type : constant
Learning rate : 0.01
Regularization constant : 0.01
Error Tolerance : 1e-07
Data Transformation method : StandardScaler
Weights initializer : zeros
Bias initializer : ones
>> Inputs for Tensorflow Graph
X : Tensor("Inputs/Data/X:0", shape=(?, 2304), dtype=float32, device=/device:GPU:0)
Y_true : Tensor("Inputs/Train_Labels/y_label:0", shape=(?, 10), dtype=float32, device=/device:GPU:0)
Y_true_cls : Tensor("Inputs/Train_Labels/y_class:0", shape=(?,), dtype=int64, device=/device:GPU:0)
Device to use : /gpu:0
>> Output parameters for Tensorflow Graph
Restore model : False
W : Tensor("Parameters/Weights/W_zeros/read:0", shape=(2304, 10), dtype=float32, device=/device:GPU:0)
b : Tensor("Parameters/Bias/b_ones/read:0", shape=(10,), dtype=float32, device=/device:GPU:0)
logits : Tensor("Predictions/add:0", shape=(?, 10), dtype=float32, device=/device:GPU:0)
Y_pred : Tensor("Predictions/Softmax:0", shape=(?, 10), dtype=float32, device=/device:GPU:0)
Y_pred_cls : Tensor("Predictions/ArgMax:0", shape=(?,), dtype=int64, device=/device:GPU:0)
cross_entropy : Tensor("Cross_Entropy/Reshape_2:0", shape=(?,), dtype=float32, device=/device:GPU:0)
train_loss : Tensor("Loss_Function/Add:0", shape=(), dtype=float32, device=/device:GPU:0)
optimizer : name: "Optimizer/GradientDescent"
op: "NoOp"
input: "^Optimizer/GradientDescent/update_Parameters/Weights/W_zeros/ApplyGradientDescent"
input: "^Optimizer/GradientDescent/update_Parameters/Bias/b_ones/ApplyGradientDescent"
device: "/device:GPU:0"
correct_prediction : Tensor("Equal:0", shape=(?,), dtype=bool, device=/device:GPU:0)
>> Accuracy parameters
Train Accuracy : Tensor("Accuracy/Mean:0", shape=(), dtype=float32, device=/device:GPU:0)
Validate Accuracy : Tensor("Accuracy/Mean_1:0", shape=(), dtype=float32, device=/device:GPU:0)
Test Accuracy : Tensor("Accuracy/Mean_2:0", shape=(), dtype=float32, device=/device:GPU:0)
None
In [20]:
tf_lc.fit(data_hog, cifar10.train.one_hot_labels, cifar10.train.class_labels,
test_data=test_hog, test_labels=cifar10.test.one_hot_labels, test_classes=cifar10.test.class_labels)
Using GPU: /gpu:0
Session: <tensorflow.python.client.session.InteractiveSession object at 0x7efbf3f16908>
Data shape : (50000, 2304)
Labels shape : (50000, 10)
Classes shape : (50000,)
Train Data shape : (45000, 2304)
Train Labels shape : (45000, 10)
Train Classes shape : (45000,)
Validate Data shape : (5000, 2304)
Validate Labels shape : (5000, 10)
Validate Classes shape : (5000,)
Length of train loss : 0
Length of train accuracy : 0
Length of validate loss : 0
Length of validate accuracy : 0
Length of test accuracy : 0
Restoring training from epoch : 0
>>> Epoch [ 0/1000]
train_loss: 2.3026 | train_acc: 0.0997 | val_loss: 2.1684 | val_acc: 0.3596 | test_acc: 0.3528 | Time: 1.5429 s
>>> Epoch [ 1/1000]
train_loss: 2.1712 | train_acc: 0.3546 | val_loss: 2.0765 | val_acc: 0.3660 | test_acc: 0.3622 | Time: 1.2977 s
>>> Epoch [ 2/1000]
train_loss: 2.0813 | train_acc: 0.3647 | val_loss: 2.0100 | val_acc: 0.3734 | test_acc: 0.3705 | Time: 1.2051 s
>>> Epoch [ 3/1000]
train_loss: 2.0161 | train_acc: 0.3737 | val_loss: 1.9591 | val_acc: 0.3802 | test_acc: 0.3772 | Time: 1.3037 s
>>> Epoch [ 4/1000]
train_loss: 1.9660 | train_acc: 0.3827 | val_loss: 1.9183 | val_acc: 0.3888 | test_acc: 0.3835 | Time: 1.2872 s
>>> Epoch [ 5/1000]
train_loss: 1.9258 | train_acc: 0.3897 | val_loss: 1.8845 | val_acc: 0.3964 | test_acc: 0.3889 | Time: 1.1883 s
>>> Epoch [ 6/1000]
train_loss: 1.8925 | train_acc: 0.3962 | val_loss: 1.8559 | val_acc: 0.4016 | test_acc: 0.3936 | Time: 1.2375 s
>>> Epoch [ 7/1000]
train_loss: 1.8642 | train_acc: 0.4019 | val_loss: 1.8313 | val_acc: 0.4054 | test_acc: 0.3985 | Time: 1.2292 s
>>> Epoch [ 8/1000]
train_loss: 1.8398 | train_acc: 0.4071 | val_loss: 1.8099 | val_acc: 0.4084 | test_acc: 0.4028 | Time: 1.2125 s
>>> Epoch [ 9/1000]
train_loss: 1.8185 | train_acc: 0.4107 | val_loss: 1.7910 | val_acc: 0.4110 | test_acc: 0.4054 | Time: 1.2485 s
>>> Epoch [ 10/1000]
train_loss: 1.7996 | train_acc: 0.4135 | val_loss: 1.7741 | val_acc: 0.4136 | test_acc: 0.4095 | Time: 1.2083 s
>>> Epoch [ 11/1000]
train_loss: 1.7828 | train_acc: 0.4163 | val_loss: 1.7591 | val_acc: 0.4160 | test_acc: 0.4135 | Time: 1.1819 s
>>> Epoch [ 12/1000]
train_loss: 1.7677 | train_acc: 0.4187 | val_loss: 1.7456 | val_acc: 0.4174 | test_acc: 0.4164 | Time: 1.2756 s
>>> Epoch [ 13/1000]
train_loss: 1.7541 | train_acc: 0.4215 | val_loss: 1.7333 | val_acc: 0.4200 | test_acc: 0.4192 | Time: 1.2286 s
>>> Epoch [ 14/1000]
train_loss: 1.7418 | train_acc: 0.4235 | val_loss: 1.7222 | val_acc: 0.4234 | test_acc: 0.4206 | Time: 1.3314 s
>>> Epoch [ 15/1000]
train_loss: 1.7305 | train_acc: 0.4258 | val_loss: 1.7120 | val_acc: 0.4244 | test_acc: 0.4237 | Time: 1.3533 s
>>> Epoch [ 16/1000]
train_loss: 1.7202 | train_acc: 0.4279 | val_loss: 1.7027 | val_acc: 0.4252 | test_acc: 0.4257 | Time: 1.2649 s
>>> Epoch [ 17/1000]
train_loss: 1.7107 | train_acc: 0.4299 | val_loss: 1.6940 | val_acc: 0.4262 | test_acc: 0.4269 | Time: 1.3258 s
>>> Epoch [ 18/1000]
train_loss: 1.7020 | train_acc: 0.4318 | val_loss: 1.6861 | val_acc: 0.4288 | test_acc: 0.4295 | Time: 1.3568 s
>>> Epoch [ 19/1000]
train_loss: 1.6938 | train_acc: 0.4332 | val_loss: 1.6787 | val_acc: 0.4300 | test_acc: 0.4310 | Time: 1.3138 s
>>> Epoch [ 20/1000]
train_loss: 1.6863 | train_acc: 0.4348 | val_loss: 1.6719 | val_acc: 0.4310 | test_acc: 0.4328 | Time: 1.3349 s
>>> Epoch [ 21/1000]
train_loss: 1.6793 | train_acc: 0.4368 | val_loss: 1.6655 | val_acc: 0.4332 | test_acc: 0.4341 | Time: 1.2693 s
>>> Epoch [ 22/1000]
train_loss: 1.6727 | train_acc: 0.4382 | val_loss: 1.6595 | val_acc: 0.4352 | test_acc: 0.4357 | Time: 1.2559 s
>>> Epoch [ 23/1000]
train_loss: 1.6666 | train_acc: 0.4398 | val_loss: 1.6539 | val_acc: 0.4374 | test_acc: 0.4369 | Time: 1.3627 s
>>> Epoch [ 24/1000]
train_loss: 1.6608 | train_acc: 0.4412 | val_loss: 1.6487 | val_acc: 0.4396 | test_acc: 0.4383 | Time: 1.2053 s
>>> Epoch [ 25/1000]
train_loss: 1.6554 | train_acc: 0.4426 | val_loss: 1.6437 | val_acc: 0.4414 | test_acc: 0.4393 | Time: 1.3351 s
>>> Epoch [ 26/1000]
train_loss: 1.6502 | train_acc: 0.4439 | val_loss: 1.6390 | val_acc: 0.4420 | test_acc: 0.4397 | Time: 1.2821 s
>>> Epoch [ 27/1000]
train_loss: 1.6454 | train_acc: 0.4450 | val_loss: 1.6346 | val_acc: 0.4432 | test_acc: 0.4410 | Time: 1.2479 s
>>> Epoch [ 28/1000]
train_loss: 1.6408 | train_acc: 0.4457 | val_loss: 1.6304 | val_acc: 0.4456 | test_acc: 0.4419 | Time: 1.1673 s
>>> Epoch [ 29/1000]
train_loss: 1.6364 | train_acc: 0.4466 | val_loss: 1.6265 | val_acc: 0.4466 | test_acc: 0.4429 | Time: 1.2406 s
>>> Epoch [ 30/1000]
train_loss: 1.6323 | train_acc: 0.4476 | val_loss: 1.6227 | val_acc: 0.4486 | test_acc: 0.4442 | Time: 1.2902 s
>>> Epoch [ 31/1000]
train_loss: 1.6283 | train_acc: 0.4486 | val_loss: 1.6191 | val_acc: 0.4500 | test_acc: 0.4454 | Time: 1.4164 s
>>> Epoch [ 32/1000]
train_loss: 1.6246 | train_acc: 0.4499 | val_loss: 1.6157 | val_acc: 0.4516 | test_acc: 0.4458 | Time: 1.2744 s
>>> Epoch [ 33/1000]
train_loss: 1.6210 | train_acc: 0.4506 | val_loss: 1.6124 | val_acc: 0.4516 | test_acc: 0.4469 | Time: 1.3389 s
>>> Epoch [ 34/1000]
train_loss: 1.6175 | train_acc: 0.4518 | val_loss: 1.6093 | val_acc: 0.4530 | test_acc: 0.4477 | Time: 1.3186 s
>>> Epoch [ 35/1000]
train_loss: 1.6142 | train_acc: 0.4527 | val_loss: 1.6063 | val_acc: 0.4542 | test_acc: 0.4483 | Time: 1.3064 s
>>> Epoch [ 36/1000]
train_loss: 1.6110 | train_acc: 0.4536 | val_loss: 1.6034 | val_acc: 0.4552 | test_acc: 0.4493 | Time: 1.3472 s
>>> Epoch [ 37/1000]
train_loss: 1.6080 | train_acc: 0.4543 | val_loss: 1.6006 | val_acc: 0.4556 | test_acc: 0.4495 | Time: 1.1745 s
>>> Epoch [ 38/1000]
train_loss: 1.6051 | train_acc: 0.4551 | val_loss: 1.5980 | val_acc: 0.4558 | test_acc: 0.4505 | Time: 1.3540 s
>>> Epoch [ 39/1000]
train_loss: 1.6023 | train_acc: 0.4560 | val_loss: 1.5954 | val_acc: 0.4572 | test_acc: 0.4515 | Time: 1.4130 s
>>> Epoch [ 40/1000]
train_loss: 1.5995 | train_acc: 0.4565 | val_loss: 1.5930 | val_acc: 0.4566 | test_acc: 0.4513 | Time: 1.1707 s
>>> Epoch [ 41/1000]
train_loss: 1.5969 | train_acc: 0.4569 | val_loss: 1.5906 | val_acc: 0.4576 | test_acc: 0.4516 | Time: 1.3683 s
>>> Epoch [ 42/1000]
train_loss: 1.5944 | train_acc: 0.4574 | val_loss: 1.5883 | val_acc: 0.4580 | test_acc: 0.4525 | Time: 1.2703 s
>>> Epoch [ 43/1000]
train_loss: 1.5920 | train_acc: 0.4582 | val_loss: 1.5861 | val_acc: 0.4580 | test_acc: 0.4532 | Time: 1.3028 s
>>> Epoch [ 44/1000]
train_loss: 1.5896 | train_acc: 0.4590 | val_loss: 1.5840 | val_acc: 0.4586 | test_acc: 0.4533 | Time: 1.3947 s
>>> Epoch [ 45/1000]
train_loss: 1.5873 | train_acc: 0.4595 | val_loss: 1.5819 | val_acc: 0.4596 | test_acc: 0.4544 | Time: 1.2192 s
>>> Epoch [ 46/1000]
train_loss: 1.5851 | train_acc: 0.4601 | val_loss: 1.5799 | val_acc: 0.4602 | test_acc: 0.4547 | Time: 1.3576 s
>>> Epoch [ 47/1000]
train_loss: 1.5830 | train_acc: 0.4610 | val_loss: 1.5780 | val_acc: 0.4604 | test_acc: 0.4551 | Time: 1.3837 s
>>> Epoch [ 48/1000]
train_loss: 1.5809 | train_acc: 0.4616 | val_loss: 1.5761 | val_acc: 0.4602 | test_acc: 0.4556 | Time: 1.3996 s
>>> Epoch [ 49/1000]
train_loss: 1.5789 | train_acc: 0.4623 | val_loss: 1.5743 | val_acc: 0.4604 | test_acc: 0.4560 | Time: 1.4082 s
>>> Epoch [ 50/1000]
train_loss: 1.5769 | train_acc: 0.4628 | val_loss: 1.5726 | val_acc: 0.4604 | test_acc: 0.4565 | Time: 1.3408 s
>>> Epoch [ 51/1000]
train_loss: 1.5750 | train_acc: 0.4632 | val_loss: 1.5709 | val_acc: 0.4616 | test_acc: 0.4569 | Time: 1.3452 s
>>> Epoch [ 52/1000]
train_loss: 1.5732 | train_acc: 0.4638 | val_loss: 1.5692 | val_acc: 0.4622 | test_acc: 0.4573 | Time: 1.2115 s
>>> Epoch [ 53/1000]
train_loss: 1.5714 | train_acc: 0.4643 | val_loss: 1.5676 | val_acc: 0.4620 | test_acc: 0.4574 | Time: 1.3670 s
>>> Epoch [ 54/1000]
train_loss: 1.5696 | train_acc: 0.4647 | val_loss: 1.5660 | val_acc: 0.4632 | test_acc: 0.4583 | Time: 1.4449 s
>>> Epoch [ 55/1000]
train_loss: 1.5679 | train_acc: 0.4651 | val_loss: 1.5645 | val_acc: 0.4634 | test_acc: 0.4585 | Time: 1.1936 s
>>> Epoch [ 56/1000]
train_loss: 1.5662 | train_acc: 0.4657 | val_loss: 1.5630 | val_acc: 0.4636 | test_acc: 0.4584 | Time: 1.2374 s
>>> Epoch [ 57/1000]
train_loss: 1.5646 | train_acc: 0.4660 | val_loss: 1.5615 | val_acc: 0.4636 | test_acc: 0.4584 | Time: 1.3022 s
>>> Epoch [ 58/1000]
train_loss: 1.5630 | train_acc: 0.4664 | val_loss: 1.5601 | val_acc: 0.4644 | test_acc: 0.4590 | Time: 1.2372 s
>>> Epoch [ 59/1000]
train_loss: 1.5615 | train_acc: 0.4666 | val_loss: 1.5587 | val_acc: 0.4644 | test_acc: 0.4598 | Time: 1.3372 s
>>> Epoch [ 60/1000]
train_loss: 1.5600 | train_acc: 0.4670 | val_loss: 1.5574 | val_acc: 0.4648 | test_acc: 0.4603 | Time: 1.2097 s
>>> Epoch [ 61/1000]
train_loss: 1.5585 | train_acc: 0.4672 | val_loss: 1.5561 | val_acc: 0.4658 | test_acc: 0.4609 | Time: 1.2952 s
>>> Epoch [ 62/1000]
train_loss: 1.5570 | train_acc: 0.4679 | val_loss: 1.5548 | val_acc: 0.4668 | test_acc: 0.4615 | Time: 1.2762 s
>>> Epoch [ 63/1000]
train_loss: 1.5556 | train_acc: 0.4684 | val_loss: 1.5535 | val_acc: 0.4670 | test_acc: 0.4614 | Time: 1.3002 s
>>> Epoch [ 64/1000]
train_loss: 1.5543 | train_acc: 0.4688 | val_loss: 1.5523 | val_acc: 0.4678 | test_acc: 0.4621 | Time: 1.3035 s
>>> Epoch [ 65/1000]
train_loss: 1.5529 | train_acc: 0.4693 | val_loss: 1.5511 | val_acc: 0.4682 | test_acc: 0.4626 | Time: 1.4536 s
>>> Epoch [ 66/1000]
train_loss: 1.5516 | train_acc: 0.4695 | val_loss: 1.5499 | val_acc: 0.4684 | test_acc: 0.4630 | Time: 1.2296 s
>>> Epoch [ 67/1000]
train_loss: 1.5503 | train_acc: 0.4699 | val_loss: 1.5488 | val_acc: 0.4686 | test_acc: 0.4634 | Time: 1.3629 s
>>> Epoch [ 68/1000]
train_loss: 1.5490 | train_acc: 0.4705 | val_loss: 1.5477 | val_acc: 0.4694 | test_acc: 0.4634 | Time: 1.2453 s
>>> Epoch [ 69/1000]
train_loss: 1.5478 | train_acc: 0.4712 | val_loss: 1.5466 | val_acc: 0.4702 | test_acc: 0.4642 | Time: 1.2296 s
>>> Epoch [ 70/1000]
train_loss: 1.5466 | train_acc: 0.4715 | val_loss: 1.5455 | val_acc: 0.4708 | test_acc: 0.4648 | Time: 1.2596 s
>>> Epoch [ 71/1000]
train_loss: 1.5454 | train_acc: 0.4720 | val_loss: 1.5444 | val_acc: 0.4708 | test_acc: 0.4647 | Time: 1.1904 s
>>> Epoch [ 72/1000]
train_loss: 1.5442 | train_acc: 0.4724 | val_loss: 1.5434 | val_acc: 0.4704 | test_acc: 0.4648 | Time: 1.2101 s
>>> Epoch [ 73/1000]
train_loss: 1.5431 | train_acc: 0.4726 | val_loss: 1.5424 | val_acc: 0.4704 | test_acc: 0.4650 | Time: 1.3650 s
>>> Epoch [ 74/1000]
train_loss: 1.5419 | train_acc: 0.4730 | val_loss: 1.5414 | val_acc: 0.4708 | test_acc: 0.4654 | Time: 1.4172 s
>>> Epoch [ 75/1000]
train_loss: 1.5408 | train_acc: 0.4731 | val_loss: 1.5404 | val_acc: 0.4716 | test_acc: 0.4658 | Time: 1.4331 s
>>> Epoch [ 76/1000]
train_loss: 1.5397 | train_acc: 0.4736 | val_loss: 1.5395 | val_acc: 0.4722 | test_acc: 0.4660 | Time: 1.3134 s
>>> Epoch [ 77/1000]
train_loss: 1.5387 | train_acc: 0.4739 | val_loss: 1.5385 | val_acc: 0.4726 | test_acc: 0.4664 | Time: 1.3741 s
>>> Epoch [ 78/1000]
train_loss: 1.5376 | train_acc: 0.4746 | val_loss: 1.5376 | val_acc: 0.4726 | test_acc: 0.4667 | Time: 1.4860 s
>>> Epoch [ 79/1000]
train_loss: 1.5366 | train_acc: 0.4748 | val_loss: 1.5367 | val_acc: 0.4720 | test_acc: 0.4677 | Time: 1.3630 s
>>> Epoch [ 80/1000]
train_loss: 1.5356 | train_acc: 0.4750 | val_loss: 1.5358 | val_acc: 0.4724 | test_acc: 0.4684 | Time: 1.4566 s
>>> Epoch [ 81/1000]
train_loss: 1.5346 | train_acc: 0.4754 | val_loss: 1.5349 | val_acc: 0.4728 | test_acc: 0.4689 | Time: 1.1743 s
>>> Epoch [ 82/1000]
train_loss: 1.5336 | train_acc: 0.4756 | val_loss: 1.5341 | val_acc: 0.4734 | test_acc: 0.4685 | Time: 1.3488 s
>>> Epoch [ 83/1000]
train_loss: 1.5326 | train_acc: 0.4759 | val_loss: 1.5332 | val_acc: 0.4740 | test_acc: 0.4684 | Time: 1.4324 s
>>> Epoch [ 84/1000]
train_loss: 1.5317 | train_acc: 0.4764 | val_loss: 1.5324 | val_acc: 0.4744 | test_acc: 0.4685 | Time: 1.3897 s
>>> Epoch [ 85/1000]
train_loss: 1.5307 | train_acc: 0.4770 | val_loss: 1.5316 | val_acc: 0.4744 | test_acc: 0.4687 | Time: 1.3979 s
>>> Epoch [ 86/1000]
train_loss: 1.5298 | train_acc: 0.4771 | val_loss: 1.5308 | val_acc: 0.4750 | test_acc: 0.4688 | Time: 1.2100 s
>>> Epoch [ 87/1000]
train_loss: 1.5289 | train_acc: 0.4775 | val_loss: 1.5300 | val_acc: 0.4758 | test_acc: 0.4691 | Time: 1.2562 s
>>> Epoch [ 88/1000]
train_loss: 1.5280 | train_acc: 0.4779 | val_loss: 1.5292 | val_acc: 0.4760 | test_acc: 0.4693 | Time: 1.4501 s
>>> Epoch [ 89/1000]
train_loss: 1.5272 | train_acc: 0.4783 | val_loss: 1.5285 | val_acc: 0.4768 | test_acc: 0.4697 | Time: 1.3122 s
>>> Epoch [ 90/1000]
train_loss: 1.5263 | train_acc: 0.4786 | val_loss: 1.5277 | val_acc: 0.4770 | test_acc: 0.4703 | Time: 1.3871 s
>>> Epoch [ 91/1000]
train_loss: 1.5254 | train_acc: 0.4789 | val_loss: 1.5270 | val_acc: 0.4768 | test_acc: 0.4710 | Time: 1.2335 s
>>> Epoch [ 92/1000]
train_loss: 1.5246 | train_acc: 0.4793 | val_loss: 1.5262 | val_acc: 0.4774 | test_acc: 0.4715 | Time: 1.3809 s
>>> Epoch [ 93/1000]
train_loss: 1.5238 | train_acc: 0.4793 | val_loss: 1.5255 | val_acc: 0.4770 | test_acc: 0.4719 | Time: 1.4255 s
>>> Epoch [ 94/1000]
train_loss: 1.5230 | train_acc: 0.4798 | val_loss: 1.5248 | val_acc: 0.4770 | test_acc: 0.4723 | Time: 1.3873 s
>>> Epoch [ 95/1000]
train_loss: 1.5222 | train_acc: 0.4803 | val_loss: 1.5241 | val_acc: 0.4768 | test_acc: 0.4722 | Time: 1.4920 s
>>> Epoch [ 96/1000]
train_loss: 1.5214 | train_acc: 0.4806 | val_loss: 1.5234 | val_acc: 0.4774 | test_acc: 0.4726 | Time: 1.3138 s
>>> Epoch [ 97/1000]
train_loss: 1.5206 | train_acc: 0.4810 | val_loss: 1.5228 | val_acc: 0.4774 | test_acc: 0.4726 | Time: 1.4158 s
>>> Epoch [ 98/1000]
train_loss: 1.5198 | train_acc: 0.4814 | val_loss: 1.5221 | val_acc: 0.4770 | test_acc: 0.4726 | Time: 1.3066 s
>>> Epoch [ 99/1000]
train_loss: 1.5191 | train_acc: 0.4816 | val_loss: 1.5214 | val_acc: 0.4772 | test_acc: 0.4728 | Time: 1.2786 s
>>> Epoch [ 100/1000]
train_loss: 1.5183 | train_acc: 0.4821 | val_loss: 1.5208 | val_acc: 0.4772 | test_acc: 0.4731 | Time: 1.4809 s
>>> Epoch [ 101/1000]
train_loss: 1.5176 | train_acc: 0.4824 | val_loss: 1.5201 | val_acc: 0.4776 | test_acc: 0.4732 | Time: 1.3653 s
>>> Epoch [ 102/1000]
train_loss: 1.5168 | train_acc: 0.4826 | val_loss: 1.5195 | val_acc: 0.4780 | test_acc: 0.4739 | Time: 1.4701 s
>>> Epoch [ 103/1000]
train_loss: 1.5161 | train_acc: 0.4830 | val_loss: 1.5189 | val_acc: 0.4786 | test_acc: 0.4745 | Time: 1.3525 s
>>> Epoch [ 104/1000]
train_loss: 1.5154 | train_acc: 0.4832 | val_loss: 1.5183 | val_acc: 0.4784 | test_acc: 0.4744 | Time: 1.3649 s
>>> Epoch [ 105/1000]
train_loss: 1.5147 | train_acc: 0.4835 | val_loss: 1.5177 | val_acc: 0.4784 | test_acc: 0.4747 | Time: 1.4726 s
>>> Epoch [ 106/1000]
train_loss: 1.5140 | train_acc: 0.4836 | val_loss: 1.5171 | val_acc: 0.4786 | test_acc: 0.4748 | Time: 1.5270 s
>>> Epoch [ 107/1000]
train_loss: 1.5133 | train_acc: 0.4837 | val_loss: 1.5165 | val_acc: 0.4790 | test_acc: 0.4751 | Time: 1.4202 s
>>> Epoch [ 108/1000]
train_loss: 1.5126 | train_acc: 0.4839 | val_loss: 1.5159 | val_acc: 0.4790 | test_acc: 0.4753 | Time: 1.4171 s
>>> Epoch [ 109/1000]
train_loss: 1.5120 | train_acc: 0.4843 | val_loss: 1.5153 | val_acc: 0.4796 | test_acc: 0.4757 | Time: 1.5647 s
>>> Epoch [ 110/1000]
train_loss: 1.5113 | train_acc: 0.4845 | val_loss: 1.5148 | val_acc: 0.4796 | test_acc: 0.4759 | Time: 1.2232 s
>>> Epoch [ 111/1000]
train_loss: 1.5106 | train_acc: 0.4847 | val_loss: 1.5142 | val_acc: 0.4802 | test_acc: 0.4763 | Time: 1.3195 s
>>> Epoch [ 112/1000]
train_loss: 1.5100 | train_acc: 0.4850 | val_loss: 1.5137 | val_acc: 0.4802 | test_acc: 0.4768 | Time: 1.3075 s
>>> Epoch [ 113/1000]
train_loss: 1.5094 | train_acc: 0.4853 | val_loss: 1.5131 | val_acc: 0.4800 | test_acc: 0.4769 | Time: 1.2425 s
>>> Epoch [ 114/1000]
train_loss: 1.5087 | train_acc: 0.4857 | val_loss: 1.5126 | val_acc: 0.4804 | test_acc: 0.4770 | Time: 1.4259 s
>>> Epoch [ 115/1000]
train_loss: 1.5081 | train_acc: 0.4861 | val_loss: 1.5120 | val_acc: 0.4804 | test_acc: 0.4772 | Time: 1.2993 s
>>> Epoch [ 116/1000]
train_loss: 1.5075 | train_acc: 0.4864 | val_loss: 1.5115 | val_acc: 0.4808 | test_acc: 0.4773 | Time: 1.4285 s
>>> Epoch [ 117/1000]
train_loss: 1.5069 | train_acc: 0.4868 | val_loss: 1.5110 | val_acc: 0.4812 | test_acc: 0.4774 | Time: 1.3714 s
>>> Epoch [ 118/1000]
train_loss: 1.5063 | train_acc: 0.4872 | val_loss: 1.5105 | val_acc: 0.4814 | test_acc: 0.4777 | Time: 1.4593 s
>>> Epoch [ 119/1000]
train_loss: 1.5057 | train_acc: 0.4875 | val_loss: 1.5100 | val_acc: 0.4812 | test_acc: 0.4775 | Time: 1.3441 s
>>> Epoch [ 120/1000]
train_loss: 1.5051 | train_acc: 0.4877 | val_loss: 1.5095 | val_acc: 0.4818 | test_acc: 0.4773 | Time: 1.3887 s
>>> Epoch [ 121/1000]
train_loss: 1.5045 | train_acc: 0.4881 | val_loss: 1.5090 | val_acc: 0.4822 | test_acc: 0.4776 | Time: 1.5740 s
>>> Epoch [ 122/1000]
train_loss: 1.5039 | train_acc: 0.4883 | val_loss: 1.5085 | val_acc: 0.4822 | test_acc: 0.4779 | Time: 1.1849 s
>>> Epoch [ 123/1000]
train_loss: 1.5033 | train_acc: 0.4885 | val_loss: 1.5080 | val_acc: 0.4820 | test_acc: 0.4779 | Time: 1.4435 s
>>> Epoch [ 124/1000]
train_loss: 1.5028 | train_acc: 0.4888 | val_loss: 1.5075 | val_acc: 0.4824 | test_acc: 0.4784 | Time: 1.4245 s
>>> Epoch [ 125/1000]
train_loss: 1.5022 | train_acc: 0.4888 | val_loss: 1.5070 | val_acc: 0.4822 | test_acc: 0.4791 | Time: 1.4460 s
>>> Epoch [ 126/1000]
train_loss: 1.5016 | train_acc: 0.4891 | val_loss: 1.5066 | val_acc: 0.4828 | test_acc: 0.4795 | Time: 1.5322 s
>>> Epoch [ 127/1000]
train_loss: 1.5011 | train_acc: 0.4894 | val_loss: 1.5061 | val_acc: 0.4834 | test_acc: 0.4797 | Time: 1.4488 s
>>> Epoch [ 128/1000]
train_loss: 1.5006 | train_acc: 0.4895 | val_loss: 1.5056 | val_acc: 0.4830 | test_acc: 0.4798 | Time: 1.5489 s
>>> Epoch [ 129/1000]
train_loss: 1.5000 | train_acc: 0.4897 | val_loss: 1.5052 | val_acc: 0.4832 | test_acc: 0.4799 | Time: 1.3455 s
>>> Epoch [ 130/1000]
train_loss: 1.4995 | train_acc: 0.4898 | val_loss: 1.5047 | val_acc: 0.4828 | test_acc: 0.4800 | Time: 1.4097 s
>>> Epoch [ 131/1000]
train_loss: 1.4989 | train_acc: 0.4899 | val_loss: 1.5043 | val_acc: 0.4834 | test_acc: 0.4800 | Time: 1.2872 s
>>> Epoch [ 132/1000]
train_loss: 1.4984 | train_acc: 0.4899 | val_loss: 1.5038 | val_acc: 0.4836 | test_acc: 0.4801 | Time: 1.4485 s
>>> Epoch [ 133/1000]
train_loss: 1.4979 | train_acc: 0.4901 | val_loss: 1.5034 | val_acc: 0.4838 | test_acc: 0.4798 | Time: 1.4110 s
>>> Epoch [ 134/1000]
train_loss: 1.4974 | train_acc: 0.4901 | val_loss: 1.5030 | val_acc: 0.4840 | test_acc: 0.4802 | Time: 1.4586 s
>>> Epoch [ 135/1000]
train_loss: 1.4969 | train_acc: 0.4904 | val_loss: 1.5025 | val_acc: 0.4840 | test_acc: 0.4804 | Time: 1.6081 s
>>> Epoch [ 136/1000]
train_loss: 1.4964 | train_acc: 0.4906 | val_loss: 1.5021 | val_acc: 0.4838 | test_acc: 0.4805 | Time: 1.3646 s
>>> Epoch [ 137/1000]
train_loss: 1.4959 | train_acc: 0.4906 | val_loss: 1.5017 | val_acc: 0.4842 | test_acc: 0.4806 | Time: 1.4992 s
>>> Epoch [ 138/1000]
train_loss: 1.4954 | train_acc: 0.4907 | val_loss: 1.5013 | val_acc: 0.4842 | test_acc: 0.4810 | Time: 1.3498 s
>>> Epoch [ 139/1000]
train_loss: 1.4949 | train_acc: 0.4911 | val_loss: 1.5009 | val_acc: 0.4848 | test_acc: 0.4813 | Time: 1.3937 s
>>> Epoch [ 140/1000]
train_loss: 1.4944 | train_acc: 0.4915 | val_loss: 1.5005 | val_acc: 0.4848 | test_acc: 0.4815 | Time: 1.4858 s
>>> Epoch [ 141/1000]
train_loss: 1.4939 | train_acc: 0.4916 | val_loss: 1.5001 | val_acc: 0.4856 | test_acc: 0.4817 | Time: 1.4752 s
>>> Epoch [ 142/1000]
train_loss: 1.4934 | train_acc: 0.4917 | val_loss: 1.4997 | val_acc: 0.4854 | test_acc: 0.4820 | Time: 1.3913 s
>>> Epoch [ 143/1000]
train_loss: 1.4930 | train_acc: 0.4918 | val_loss: 1.4993 | val_acc: 0.4856 | test_acc: 0.4820 | Time: 1.4557 s
>>> Epoch [ 144/1000]
train_loss: 1.4925 | train_acc: 0.4920 | val_loss: 1.4989 | val_acc: 0.4858 | test_acc: 0.4821 | Time: 1.6062 s
>>> Epoch [ 145/1000]
train_loss: 1.4920 | train_acc: 0.4922 | val_loss: 1.4985 | val_acc: 0.4860 | test_acc: 0.4820 | Time: 1.4205 s
>>> Epoch [ 146/1000]
train_loss: 1.4916 | train_acc: 0.4925 | val_loss: 1.4981 | val_acc: 0.4860 | test_acc: 0.4821 | Time: 1.7116 s
>>> Epoch [ 147/1000]
train_loss: 1.4911 | train_acc: 0.4925 | val_loss: 1.4977 | val_acc: 0.4862 | test_acc: 0.4826 | Time: 1.4162 s
>>> Epoch [ 148/1000]
train_loss: 1.4907 | train_acc: 0.4927 | val_loss: 1.4973 | val_acc: 0.4864 | test_acc: 0.4829 | Time: 1.6455 s
>>> Epoch [ 149/1000]
train_loss: 1.4902 | train_acc: 0.4929 | val_loss: 1.4970 | val_acc: 0.4866 | test_acc: 0.4829 | Time: 1.3175 s
>>> Epoch [ 150/1000]
train_loss: 1.4898 | train_acc: 0.4931 | val_loss: 1.4966 | val_acc: 0.4870 | test_acc: 0.4831 | Time: 1.4218 s
>>> Epoch [ 151/1000]
train_loss: 1.4893 | train_acc: 0.4935 | val_loss: 1.4962 | val_acc: 0.4874 | test_acc: 0.4834 | Time: 1.4311 s
>>> Epoch [ 152/1000]
train_loss: 1.4889 | train_acc: 0.4938 | val_loss: 1.4959 | val_acc: 0.4876 | test_acc: 0.4835 | Time: 1.5022 s
>>> Epoch [ 153/1000]
train_loss: 1.4884 | train_acc: 0.4941 | val_loss: 1.4955 | val_acc: 0.4874 | test_acc: 0.4835 | Time: 1.3078 s
>>> Epoch [ 154/1000]
train_loss: 1.4880 | train_acc: 0.4943 | val_loss: 1.4951 | val_acc: 0.4874 | test_acc: 0.4835 | Time: 1.3669 s
>>> Epoch [ 155/1000]
train_loss: 1.4876 | train_acc: 0.4943 | val_loss: 1.4948 | val_acc: 0.4872 | test_acc: 0.4835 | Time: 1.5830 s
>>> Epoch [ 156/1000]
train_loss: 1.4871 | train_acc: 0.4943 | val_loss: 1.4944 | val_acc: 0.4870 | test_acc: 0.4836 | Time: 1.2981 s
>>> Epoch [ 157/1000]
train_loss: 1.4867 | train_acc: 0.4944 | val_loss: 1.4941 | val_acc: 0.4870 | test_acc: 0.4838 | Time: 1.4654 s
>>> Epoch [ 158/1000]
train_loss: 1.4863 | train_acc: 0.4945 | val_loss: 1.4937 | val_acc: 0.4870 | test_acc: 0.4838 | Time: 1.4417 s
>>> Epoch [ 159/1000]
train_loss: 1.4859 | train_acc: 0.4948 | val_loss: 1.4934 | val_acc: 0.4872 | test_acc: 0.4843 | Time: 1.4797 s
>>> Epoch [ 160/1000]
train_loss: 1.4855 | train_acc: 0.4949 | val_loss: 1.4931 | val_acc: 0.4874 | test_acc: 0.4843 | Time: 1.3373 s
>>> Epoch [ 161/1000]
train_loss: 1.4851 | train_acc: 0.4950 | val_loss: 1.4927 | val_acc: 0.4880 | test_acc: 0.4843 | Time: 1.2976 s
>>> Epoch [ 162/1000]
train_loss: 1.4847 | train_acc: 0.4951 | val_loss: 1.4924 | val_acc: 0.4878 | test_acc: 0.4846 | Time: 1.5298 s
>>> Epoch [ 163/1000]
train_loss: 1.4843 | train_acc: 0.4953 | val_loss: 1.4921 | val_acc: 0.4878 | test_acc: 0.4846 | Time: 1.4553 s
>>> Epoch [ 164/1000]
train_loss: 1.4839 | train_acc: 0.4954 | val_loss: 1.4917 | val_acc: 0.4874 | test_acc: 0.4847 | Time: 1.6157 s
>>> Epoch [ 165/1000]
train_loss: 1.4835 | train_acc: 0.4954 | val_loss: 1.4914 | val_acc: 0.4874 | test_acc: 0.4848 | Time: 1.4037 s
>>> Epoch [ 166/1000]
train_loss: 1.4831 | train_acc: 0.4956 | val_loss: 1.4911 | val_acc: 0.4876 | test_acc: 0.4849 | Time: 1.4706 s
>>> Epoch [ 167/1000]
train_loss: 1.4827 | train_acc: 0.4956 | val_loss: 1.4908 | val_acc: 0.4876 | test_acc: 0.4851 | Time: 1.4048 s
>>> Epoch [ 168/1000]
train_loss: 1.4823 | train_acc: 0.4958 | val_loss: 1.4904 | val_acc: 0.4874 | test_acc: 0.4851 | Time: 1.4250 s
>>> Epoch [ 169/1000]
train_loss: 1.4819 | train_acc: 0.4960 | val_loss: 1.4901 | val_acc: 0.4878 | test_acc: 0.4851 | Time: 1.4143 s
>>> Epoch [ 170/1000]
train_loss: 1.4815 | train_acc: 0.4960 | val_loss: 1.4898 | val_acc: 0.4878 | test_acc: 0.4851 | Time: 1.4296 s
>>> Epoch [ 171/1000]
train_loss: 1.4811 | train_acc: 0.4963 | val_loss: 1.4895 | val_acc: 0.4878 | test_acc: 0.4851 | Time: 1.5686 s
>>> Epoch [ 172/1000]
train_loss: 1.4807 | train_acc: 0.4965 | val_loss: 1.4892 | val_acc: 0.4878 | test_acc: 0.4852 | Time: 1.3928 s
>>> Epoch [ 173/1000]
train_loss: 1.4804 | train_acc: 0.4965 | val_loss: 1.4889 | val_acc: 0.4880 | test_acc: 0.4856 | Time: 1.6564 s
>>> Epoch [ 174/1000]
train_loss: 1.4800 | train_acc: 0.4966 | val_loss: 1.4886 | val_acc: 0.4880 | test_acc: 0.4856 | Time: 1.3088 s
>>> Epoch [ 175/1000]
train_loss: 1.4796 | train_acc: 0.4966 | val_loss: 1.4883 | val_acc: 0.4880 | test_acc: 0.4859 | Time: 1.4980 s
>>> Epoch [ 176/1000]
train_loss: 1.4793 | train_acc: 0.4968 | val_loss: 1.4880 | val_acc: 0.4884 | test_acc: 0.4862 | Time: 1.5183 s
>>> Epoch [ 177/1000]
train_loss: 1.4789 | train_acc: 0.4968 | val_loss: 1.4877 | val_acc: 0.4884 | test_acc: 0.4864 | Time: 1.5660 s
>>> Epoch [ 178/1000]
train_loss: 1.4785 | train_acc: 0.4970 | val_loss: 1.4874 | val_acc: 0.4886 | test_acc: 0.4866 | Time: 1.4620 s
>>> Epoch [ 179/1000]
train_loss: 1.4782 | train_acc: 0.4971 | val_loss: 1.4871 | val_acc: 0.4886 | test_acc: 0.4865 | Time: 1.5543 s
>>> Epoch [ 180/1000]
train_loss: 1.4778 | train_acc: 0.4972 | val_loss: 1.4868 | val_acc: 0.4890 | test_acc: 0.4866 | Time: 1.4256 s
>>> Epoch [ 181/1000]
train_loss: 1.4775 | train_acc: 0.4973 | val_loss: 1.4865 | val_acc: 0.4896 | test_acc: 0.4862 | Time: 1.4899 s
>>> Epoch [ 182/1000]
train_loss: 1.4771 | train_acc: 0.4973 | val_loss: 1.4862 | val_acc: 0.4894 | test_acc: 0.4864 | Time: 1.4894 s
>>> Epoch [ 183/1000]
train_loss: 1.4768 | train_acc: 0.4974 | val_loss: 1.4859 | val_acc: 0.4896 | test_acc: 0.4863 | Time: 1.5257 s
>>> Epoch [ 184/1000]
train_loss: 1.4764 | train_acc: 0.4977 | val_loss: 1.4857 | val_acc: 0.4896 | test_acc: 0.4868 | Time: 1.4413 s
>>> Epoch [ 185/1000]
train_loss: 1.4761 | train_acc: 0.4978 | val_loss: 1.4854 | val_acc: 0.4900 | test_acc: 0.4869 | Time: 1.4216 s
>>> Epoch [ 186/1000]
train_loss: 1.4757 | train_acc: 0.4979 | val_loss: 1.4851 | val_acc: 0.4904 | test_acc: 0.4874 | Time: 1.4945 s
>>> Epoch [ 187/1000]
train_loss: 1.4754 | train_acc: 0.4978 | val_loss: 1.4848 | val_acc: 0.4904 | test_acc: 0.4877 | Time: 1.4356 s
>>> Epoch [ 188/1000]
train_loss: 1.4750 | train_acc: 0.4979 | val_loss: 1.4846 | val_acc: 0.4902 | test_acc: 0.4878 | Time: 1.6463 s
>>> Epoch [ 189/1000]
train_loss: 1.4747 | train_acc: 0.4979 | val_loss: 1.4843 | val_acc: 0.4900 | test_acc: 0.4879 | Time: 1.5168 s
>>> Epoch [ 190/1000]
train_loss: 1.4744 | train_acc: 0.4979 | val_loss: 1.4840 | val_acc: 0.4900 | test_acc: 0.4881 | Time: 1.5241 s
>>> Epoch [ 191/1000]
train_loss: 1.4740 | train_acc: 0.4980 | val_loss: 1.4837 | val_acc: 0.4898 | test_acc: 0.4881 | Time: 1.5415 s
>>> Epoch [ 192/1000]
train_loss: 1.4737 | train_acc: 0.4980 | val_loss: 1.4835 | val_acc: 0.4898 | test_acc: 0.4882 | Time: 1.6142 s
>>> Epoch [ 193/1000]
train_loss: 1.4734 | train_acc: 0.4982 | val_loss: 1.4832 | val_acc: 0.4902 | test_acc: 0.4880 | Time: 1.5526 s
>>> Epoch [ 194/1000]
train_loss: 1.4730 | train_acc: 0.4983 | val_loss: 1.4830 | val_acc: 0.4902 | test_acc: 0.4880 | Time: 1.6319 s
>>> Epoch [ 195/1000]
train_loss: 1.4727 | train_acc: 0.4984 | val_loss: 1.4827 | val_acc: 0.4902 | test_acc: 0.4884 | Time: 1.4541 s
>>> Epoch [ 196/1000]
train_loss: 1.4724 | train_acc: 0.4986 | val_loss: 1.4824 | val_acc: 0.4902 | test_acc: 0.4884 | Time: 1.6771 s
>>> Epoch [ 197/1000]
train_loss: 1.4721 | train_acc: 0.4987 | val_loss: 1.4822 | val_acc: 0.4906 | test_acc: 0.4886 | Time: 1.4284 s
>>> Epoch [ 198/1000]
train_loss: 1.4717 | train_acc: 0.4989 | val_loss: 1.4819 | val_acc: 0.4910 | test_acc: 0.4886 | Time: 1.7002 s
>>> Epoch [ 199/1000]
train_loss: 1.4714 | train_acc: 0.4989 | val_loss: 1.4817 | val_acc: 0.4912 | test_acc: 0.4888 | Time: 1.5344 s
>>> Epoch [ 200/1000]
train_loss: 1.4711 | train_acc: 0.4993 | val_loss: 1.4814 | val_acc: 0.4912 | test_acc: 0.4890 | Time: 1.6664 s
>>> Epoch [ 201/1000]
train_loss: 1.4708 | train_acc: 0.4995 | val_loss: 1.4812 | val_acc: 0.4912 | test_acc: 0.4890 | Time: 1.5285 s
>>> Epoch [ 202/1000]
train_loss: 1.4705 | train_acc: 0.4996 | val_loss: 1.4809 | val_acc: 0.4910 | test_acc: 0.4890 | Time: 1.6902 s
>>> Epoch [ 203/1000]
train_loss: 1.4702 | train_acc: 0.4999 | val_loss: 1.4807 | val_acc: 0.4912 | test_acc: 0.4889 | Time: 1.6038 s
>>> Epoch [ 204/1000]
train_loss: 1.4699 | train_acc: 0.5000 | val_loss: 1.4804 | val_acc: 0.4912 | test_acc: 0.4889 | Time: 1.6418 s
>>> Epoch [ 205/1000]
train_loss: 1.4696 | train_acc: 0.5001 | val_loss: 1.4802 | val_acc: 0.4912 | test_acc: 0.4890 | Time: 1.5067 s
>>> Epoch [ 206/1000]
train_loss: 1.4692 | train_acc: 0.5001 | val_loss: 1.4799 | val_acc: 0.4912 | test_acc: 0.4891 | Time: 1.6281 s
>>> Epoch [ 207/1000]
train_loss: 1.4689 | train_acc: 0.5001 | val_loss: 1.4797 | val_acc: 0.4914 | test_acc: 0.4895 | Time: 1.4907 s
>>> Epoch [ 208/1000]
train_loss: 1.4686 | train_acc: 0.5002 | val_loss: 1.4794 | val_acc: 0.4916 | test_acc: 0.4896 | Time: 1.7260 s
>>> Epoch [ 209/1000]
train_loss: 1.4683 | train_acc: 0.5003 | val_loss: 1.4792 | val_acc: 0.4916 | test_acc: 0.4897 | Time: 1.4960 s
>>> Epoch [ 210/1000]
train_loss: 1.4680 | train_acc: 0.5004 | val_loss: 1.4790 | val_acc: 0.4920 | test_acc: 0.4895 | Time: 1.4407 s
>>> Epoch [ 211/1000]
train_loss: 1.4677 | train_acc: 0.5004 | val_loss: 1.4787 | val_acc: 0.4922 | test_acc: 0.4894 | Time: 1.4885 s
>>> Epoch [ 212/1000]
train_loss: 1.4674 | train_acc: 0.5006 | val_loss: 1.4785 | val_acc: 0.4922 | test_acc: 0.4894 | Time: 1.4884 s
>>> Epoch [ 213/1000]
train_loss: 1.4672 | train_acc: 0.5007 | val_loss: 1.4783 | val_acc: 0.4924 | test_acc: 0.4895 | Time: 1.5337 s
>>> Epoch [ 214/1000]
train_loss: 1.4669 | train_acc: 0.5007 | val_loss: 1.4780 | val_acc: 0.4920 | test_acc: 0.4895 | Time: 1.5283 s
>>> Epoch [ 215/1000]
train_loss: 1.4666 | train_acc: 0.5009 | val_loss: 1.4778 | val_acc: 0.4924 | test_acc: 0.4897 | Time: 1.5694 s
>>> Epoch [ 216/1000]
train_loss: 1.4663 | train_acc: 0.5011 | val_loss: 1.4776 | val_acc: 0.4922 | test_acc: 0.4902 | Time: 1.5659 s
>>> Epoch [ 217/1000]
train_loss: 1.4660 | train_acc: 0.5012 | val_loss: 1.4774 | val_acc: 0.4924 | test_acc: 0.4903 | Time: 1.4667 s
>>> Epoch [ 218/1000]
train_loss: 1.4657 | train_acc: 0.5013 | val_loss: 1.4771 | val_acc: 0.4926 | test_acc: 0.4902 | Time: 1.4868 s
>>> Epoch [ 219/1000]
train_loss: 1.4654 | train_acc: 0.5013 | val_loss: 1.4769 | val_acc: 0.4930 | test_acc: 0.4902 | Time: 1.6002 s
>>> Epoch [ 220/1000]
train_loss: 1.4651 | train_acc: 0.5013 | val_loss: 1.4767 | val_acc: 0.4932 | test_acc: 0.4901 | Time: 1.5548 s
>>> Epoch [ 221/1000]
train_loss: 1.4649 | train_acc: 0.5014 | val_loss: 1.4765 | val_acc: 0.4940 | test_acc: 0.4903 | Time: 1.5170 s
>>> Epoch [ 222/1000]
train_loss: 1.4646 | train_acc: 0.5015 | val_loss: 1.4762 | val_acc: 0.4942 | test_acc: 0.4903 | Time: 1.5291 s
>>> Epoch [ 223/1000]
train_loss: 1.4643 | train_acc: 0.5015 | val_loss: 1.4760 | val_acc: 0.4942 | test_acc: 0.4902 | Time: 1.3677 s
>>> Epoch [ 224/1000]
train_loss: 1.4640 | train_acc: 0.5015 | val_loss: 1.4758 | val_acc: 0.4944 | test_acc: 0.4900 | Time: 1.4310 s
>>> Epoch [ 225/1000]
train_loss: 1.4637 | train_acc: 0.5017 | val_loss: 1.4756 | val_acc: 0.4946 | test_acc: 0.4901 | Time: 1.6180 s
>>> Epoch [ 226/1000]
train_loss: 1.4635 | train_acc: 0.5018 | val_loss: 1.4754 | val_acc: 0.4946 | test_acc: 0.4902 | Time: 1.5258 s
>>> Epoch [ 227/1000]
train_loss: 1.4632 | train_acc: 0.5020 | val_loss: 1.4752 | val_acc: 0.4946 | test_acc: 0.4901 | Time: 1.6915 s
>>> Epoch [ 228/1000]
train_loss: 1.4629 | train_acc: 0.5021 | val_loss: 1.4749 | val_acc: 0.4946 | test_acc: 0.4903 | Time: 1.5379 s
>>> Epoch [ 229/1000]
train_loss: 1.4627 | train_acc: 0.5023 | val_loss: 1.4747 | val_acc: 0.4946 | test_acc: 0.4903 | Time: 1.3557 s
>>> Epoch [ 230/1000]
train_loss: 1.4624 | train_acc: 0.5024 | val_loss: 1.4745 | val_acc: 0.4948 | test_acc: 0.4903 | Time: 1.4776 s
>>> Epoch [ 231/1000]
train_loss: 1.4621 | train_acc: 0.5027 | val_loss: 1.4743 | val_acc: 0.4954 | test_acc: 0.4902 | Time: 1.6785 s
>>> Epoch [ 232/1000]
train_loss: 1.4619 | train_acc: 0.5026 | val_loss: 1.4741 | val_acc: 0.4956 | test_acc: 0.4901 | Time: 1.4481 s
>>> Epoch [ 233/1000]
train_loss: 1.4616 | train_acc: 0.5026 | val_loss: 1.4739 | val_acc: 0.4956 | test_acc: 0.4907 | Time: 1.4709 s
>>> Epoch [ 234/1000]
train_loss: 1.4613 | train_acc: 0.5026 | val_loss: 1.4737 | val_acc: 0.4956 | test_acc: 0.4911 | Time: 1.5752 s
>>> Epoch [ 235/1000]
train_loss: 1.4611 | train_acc: 0.5026 | val_loss: 1.4735 | val_acc: 0.4956 | test_acc: 0.4912 | Time: 1.7500 s
>>> Epoch [ 236/1000]
train_loss: 1.4608 | train_acc: 0.5026 | val_loss: 1.4733 | val_acc: 0.4958 | test_acc: 0.4912 | Time: 1.5528 s
>>> Epoch [ 237/1000]
train_loss: 1.4605 | train_acc: 0.5028 | val_loss: 1.4731 | val_acc: 0.4958 | test_acc: 0.4913 | Time: 1.7687 s
>>> Epoch [ 238/1000]
train_loss: 1.4603 | train_acc: 0.5029 | val_loss: 1.4729 | val_acc: 0.4960 | test_acc: 0.4915 | Time: 1.5029 s
>>> Epoch [ 239/1000]
train_loss: 1.4600 | train_acc: 0.5029 | val_loss: 1.4727 | val_acc: 0.4958 | test_acc: 0.4917 | Time: 1.6761 s
>>> Epoch [ 240/1000]
train_loss: 1.4598 | train_acc: 0.5030 | val_loss: 1.4725 | val_acc: 0.4958 | test_acc: 0.4917 | Time: 1.6695 s
>>> Epoch [ 241/1000]
train_loss: 1.4595 | train_acc: 0.5030 | val_loss: 1.4723 | val_acc: 0.4958 | test_acc: 0.4919 | Time: 1.5981 s
>>> Epoch [ 242/1000]
train_loss: 1.4593 | train_acc: 0.5031 | val_loss: 1.4721 | val_acc: 0.4960 | test_acc: 0.4921 | Time: 1.6191 s
>>> Epoch [ 243/1000]
train_loss: 1.4590 | train_acc: 0.5032 | val_loss: 1.4719 | val_acc: 0.4964 | test_acc: 0.4921 | Time: 1.5344 s
>>> Epoch [ 244/1000]
train_loss: 1.4588 | train_acc: 0.5033 | val_loss: 1.4717 | val_acc: 0.4960 | test_acc: 0.4922 | Time: 1.8015 s
>>> Epoch [ 245/1000]
train_loss: 1.4585 | train_acc: 0.5033 | val_loss: 1.4715 | val_acc: 0.4962 | test_acc: 0.4922 | Time: 1.5728 s
>>> Epoch [ 246/1000]
train_loss: 1.4583 | train_acc: 0.5032 | val_loss: 1.4713 | val_acc: 0.4960 | test_acc: 0.4921 | Time: 1.5932 s
>>> Epoch [ 247/1000]
train_loss: 1.4580 | train_acc: 0.5034 | val_loss: 1.4711 | val_acc: 0.4960 | test_acc: 0.4924 | Time: 1.5171 s
>>> Epoch [ 248/1000]
train_loss: 1.4578 | train_acc: 0.5034 | val_loss: 1.4709 | val_acc: 0.4962 | test_acc: 0.4924 | Time: 1.5427 s
>>> Epoch [ 249/1000]
train_loss: 1.4575 | train_acc: 0.5034 | val_loss: 1.4707 | val_acc: 0.4964 | test_acc: 0.4926 | Time: 1.3856 s
>>> Epoch [ 250/1000]
train_loss: 1.4573 | train_acc: 0.5035 | val_loss: 1.4706 | val_acc: 0.4962 | test_acc: 0.4924 | Time: 1.5189 s
>>> Epoch [ 251/1000]
train_loss: 1.4570 | train_acc: 0.5035 | val_loss: 1.4704 | val_acc: 0.4964 | test_acc: 0.4927 | Time: 1.4406 s
>>> Epoch [ 252/1000]
train_loss: 1.4568 | train_acc: 0.5036 | val_loss: 1.4702 | val_acc: 0.4966 | test_acc: 0.4926 | Time: 1.5857 s
>>> Epoch [ 253/1000]
train_loss: 1.4565 | train_acc: 0.5038 | val_loss: 1.4700 | val_acc: 0.4970 | test_acc: 0.4928 | Time: 1.5351 s
>>> Epoch [ 254/1000]
train_loss: 1.4563 | train_acc: 0.5040 | val_loss: 1.4698 | val_acc: 0.4972 | test_acc: 0.4928 | Time: 1.6841 s
>>> Epoch [ 255/1000]
train_loss: 1.4561 | train_acc: 0.5042 | val_loss: 1.4696 | val_acc: 0.4972 | test_acc: 0.4929 | Time: 1.7761 s
>>> Epoch [ 256/1000]
train_loss: 1.4558 | train_acc: 0.5043 | val_loss: 1.4694 | val_acc: 0.4974 | test_acc: 0.4928 | Time: 1.6542 s
>>> Epoch [ 257/1000]
train_loss: 1.4556 | train_acc: 0.5045 | val_loss: 1.4693 | val_acc: 0.4978 | test_acc: 0.4932 | Time: 1.8887 s
>>> Epoch [ 258/1000]
train_loss: 1.4553 | train_acc: 0.5046 | val_loss: 1.4691 | val_acc: 0.4978 | test_acc: 0.4933 | Time: 1.4913 s
>>> Epoch [ 259/1000]
train_loss: 1.4551 | train_acc: 0.5047 | val_loss: 1.4689 | val_acc: 0.4980 | test_acc: 0.4936 | Time: 1.6402 s
>>> Epoch [ 260/1000]
train_loss: 1.4549 | train_acc: 0.5048 | val_loss: 1.4687 | val_acc: 0.4980 | test_acc: 0.4936 | Time: 1.4229 s
>>> Epoch [ 261/1000]
train_loss: 1.4546 | train_acc: 0.5048 | val_loss: 1.4685 | val_acc: 0.4982 | test_acc: 0.4934 | Time: 1.4902 s
>>> Epoch [ 262/1000]
train_loss: 1.4544 | train_acc: 0.5051 | val_loss: 1.4684 | val_acc: 0.4980 | test_acc: 0.4933 | Time: 1.4241 s
>>> Epoch [ 263/1000]
train_loss: 1.4542 | train_acc: 0.5052 | val_loss: 1.4682 | val_acc: 0.4980 | test_acc: 0.4932 | Time: 1.6857 s
>>> Epoch [ 264/1000]
train_loss: 1.4540 | train_acc: 0.5053 | val_loss: 1.4680 | val_acc: 0.4982 | test_acc: 0.4932 | Time: 1.6885 s
>>> Epoch [ 265/1000]
train_loss: 1.4537 | train_acc: 0.5054 | val_loss: 1.4678 | val_acc: 0.4986 | test_acc: 0.4932 | Time: 1.3996 s
>>> Epoch [ 266/1000]
train_loss: 1.4535 | train_acc: 0.5055 | val_loss: 1.4677 | val_acc: 0.4982 | test_acc: 0.4933 | Time: 1.7117 s
>>> Epoch [ 267/1000]
train_loss: 1.4533 | train_acc: 0.5057 | val_loss: 1.4675 | val_acc: 0.4984 | test_acc: 0.4936 | Time: 1.5344 s
>>> Epoch [ 268/1000]
train_loss: 1.4530 | train_acc: 0.5058 | val_loss: 1.4673 | val_acc: 0.4984 | test_acc: 0.4935 | Time: 1.8806 s
>>> Epoch [ 269/1000]
train_loss: 1.4528 | train_acc: 0.5058 | val_loss: 1.4671 | val_acc: 0.4984 | test_acc: 0.4935 | Time: 1.6606 s
>>> Epoch [ 270/1000]
train_loss: 1.4526 | train_acc: 0.5060 | val_loss: 1.4670 | val_acc: 0.4984 | test_acc: 0.4937 | Time: 1.6581 s
>>> Epoch [ 271/1000]
train_loss: 1.4524 | train_acc: 0.5062 | val_loss: 1.4668 | val_acc: 0.4984 | test_acc: 0.4938 | Time: 1.5349 s
>>> Epoch [ 272/1000]
train_loss: 1.4521 | train_acc: 0.5062 | val_loss: 1.4666 | val_acc: 0.4986 | test_acc: 0.4939 | Time: 1.5856 s
>>> Epoch [ 273/1000]
train_loss: 1.4519 | train_acc: 0.5064 | val_loss: 1.4665 | val_acc: 0.4990 | test_acc: 0.4936 | Time: 1.6792 s
>>> Epoch [ 274/1000]
train_loss: 1.4517 | train_acc: 0.5063 | val_loss: 1.4663 | val_acc: 0.4990 | test_acc: 0.4937 | Time: 1.6822 s
>>> Epoch [ 275/1000]
train_loss: 1.4515 | train_acc: 0.5064 | val_loss: 1.4661 | val_acc: 0.4990 | test_acc: 0.4935 | Time: 1.7808 s
>>> Epoch [ 276/1000]
train_loss: 1.4513 | train_acc: 0.5065 | val_loss: 1.4660 | val_acc: 0.4990 | test_acc: 0.4938 | Time: 1.6274 s
>>> Epoch [ 277/1000]
train_loss: 1.4510 | train_acc: 0.5066 | val_loss: 1.4658 | val_acc: 0.4990 | test_acc: 0.4939 | Time: 1.9589 s
>>> Epoch [ 278/1000]
train_loss: 1.4508 | train_acc: 0.5067 | val_loss: 1.4656 | val_acc: 0.4990 | test_acc: 0.4941 | Time: 1.5645 s
>>> Epoch [ 279/1000]
train_loss: 1.4506 | train_acc: 0.5068 | val_loss: 1.4655 | val_acc: 0.4992 | test_acc: 0.4941 | Time: 1.7701 s
>>> Epoch [ 280/1000]
train_loss: 1.4504 | train_acc: 0.5071 | val_loss: 1.4653 | val_acc: 0.4992 | test_acc: 0.4942 | Time: 1.6579 s
>>> Epoch [ 281/1000]
train_loss: 1.4502 | train_acc: 0.5071 | val_loss: 1.4651 | val_acc: 0.4990 | test_acc: 0.4943 | Time: 1.6490 s
>>> Epoch [ 282/1000]
train_loss: 1.4500 | train_acc: 0.5071 | val_loss: 1.4650 | val_acc: 0.4988 | test_acc: 0.4945 | Time: 1.7920 s
>>> Epoch [ 283/1000]
train_loss: 1.4498 | train_acc: 0.5073 | val_loss: 1.4648 | val_acc: 0.4988 | test_acc: 0.4945 | Time: 1.5243 s
>>> Epoch [ 284/1000]
train_loss: 1.4495 | train_acc: 0.5072 | val_loss: 1.4647 | val_acc: 0.4988 | test_acc: 0.4946 | Time: 1.9566 s
>>> Epoch [ 285/1000]
train_loss: 1.4493 | train_acc: 0.5073 | val_loss: 1.4645 | val_acc: 0.4988 | test_acc: 0.4949 | Time: 1.4995 s
>>> Epoch [ 286/1000]
train_loss: 1.4491 | train_acc: 0.5074 | val_loss: 1.4643 | val_acc: 0.4986 | test_acc: 0.4949 | Time: 1.8208 s
>>> Epoch [ 287/1000]
train_loss: 1.4489 | train_acc: 0.5074 | val_loss: 1.4642 | val_acc: 0.4986 | test_acc: 0.4951 | Time: 1.4854 s
>>> Epoch [ 288/1000]
train_loss: 1.4487 | train_acc: 0.5075 | val_loss: 1.4640 | val_acc: 0.4986 | test_acc: 0.4953 | Time: 1.6948 s
>>> Epoch [ 289/1000]
train_loss: 1.4485 | train_acc: 0.5077 | val_loss: 1.4639 | val_acc: 0.4988 | test_acc: 0.4952 | Time: 1.6568 s
>>> Epoch [ 290/1000]
train_loss: 1.4483 | train_acc: 0.5077 | val_loss: 1.4637 | val_acc: 0.4990 | test_acc: 0.4952 | Time: 1.7031 s
>>> Epoch [ 291/1000]
train_loss: 1.4481 | train_acc: 0.5079 | val_loss: 1.4636 | val_acc: 0.4988 | test_acc: 0.4952 | Time: 1.8176 s
>>> Epoch [ 292/1000]
train_loss: 1.4479 | train_acc: 0.5080 | val_loss: 1.4634 | val_acc: 0.4988 | test_acc: 0.4952 | Time: 1.8028 s
>>> Epoch [ 293/1000]
train_loss: 1.4477 | train_acc: 0.5080 | val_loss: 1.4633 | val_acc: 0.4988 | test_acc: 0.4952 | Time: 1.9054 s
>>> Epoch [ 294/1000]
train_loss: 1.4475 | train_acc: 0.5081 | val_loss: 1.4631 | val_acc: 0.4990 | test_acc: 0.4953 | Time: 1.8318 s
>>> Epoch [ 295/1000]
train_loss: 1.4473 | train_acc: 0.5081 | val_loss: 1.4629 | val_acc: 0.4990 | test_acc: 0.4954 | Time: 1.7888 s
>>> Epoch [ 296/1000]
train_loss: 1.4471 | train_acc: 0.5082 | val_loss: 1.4628 | val_acc: 0.4994 | test_acc: 0.4953 | Time: 2.0648 s
>>> Epoch [ 297/1000]
train_loss: 1.4469 | train_acc: 0.5082 | val_loss: 1.4626 | val_acc: 0.4994 | test_acc: 0.4951 | Time: 1.7616 s
>>> Epoch [ 298/1000]
train_loss: 1.4467 | train_acc: 0.5083 | val_loss: 1.4625 | val_acc: 0.4996 | test_acc: 0.4953 | Time: 1.7362 s
>>> Epoch [ 299/1000]
train_loss: 1.4465 | train_acc: 0.5083 | val_loss: 1.4623 | val_acc: 0.5000 | test_acc: 0.4953 | Time: 1.6757 s
>>> Epoch [ 300/1000]
train_loss: 1.4463 | train_acc: 0.5083 | val_loss: 1.4622 | val_acc: 0.5000 | test_acc: 0.4954 | Time: 1.5121 s
>>> Epoch [ 301/1000]
train_loss: 1.4461 | train_acc: 0.5085 | val_loss: 1.4620 | val_acc: 0.5004 | test_acc: 0.4953 | Time: 2.0594 s
>>> Epoch [ 302/1000]
train_loss: 1.4459 | train_acc: 0.5085 | val_loss: 1.4619 | val_acc: 0.5004 | test_acc: 0.4953 | Time: 1.9399 s
>>> Epoch [ 303/1000]
train_loss: 1.4457 | train_acc: 0.5086 | val_loss: 1.4618 | val_acc: 0.5000 | test_acc: 0.4954 | Time: 1.7856 s
>>> Epoch [ 304/1000]
train_loss: 1.4455 | train_acc: 0.5085 | val_loss: 1.4616 | val_acc: 0.5000 | test_acc: 0.4956 | Time: 2.0360 s
>>> Epoch [ 305/1000]
train_loss: 1.4453 | train_acc: 0.5086 | val_loss: 1.4615 | val_acc: 0.4998 | test_acc: 0.4960 | Time: 1.6670 s
>>> Epoch [ 306/1000]
train_loss: 1.4451 | train_acc: 0.5086 | val_loss: 1.4613 | val_acc: 0.5000 | test_acc: 0.4960 | Time: 1.7551 s
>>> Epoch [ 307/1000]
train_loss: 1.4449 | train_acc: 0.5088 | val_loss: 1.4612 | val_acc: 0.5002 | test_acc: 0.4961 | Time: 1.9349 s
>>> Epoch [ 308/1000]
train_loss: 1.4447 | train_acc: 0.5088 | val_loss: 1.4610 | val_acc: 0.5002 | test_acc: 0.4961 | Time: 1.7651 s
>>> Epoch [ 309/1000]
train_loss: 1.4445 | train_acc: 0.5090 | val_loss: 1.4609 | val_acc: 0.5002 | test_acc: 0.4958 | Time: 1.9929 s
>>> Epoch [ 310/1000]
train_loss: 1.4443 | train_acc: 0.5091 | val_loss: 1.4607 | val_acc: 0.5004 | test_acc: 0.4959 | Time: 1.8170 s
>>> Epoch [ 311/1000]
train_loss: 1.4441 | train_acc: 0.5091 | val_loss: 1.4606 | val_acc: 0.5004 | test_acc: 0.4957 | Time: 1.7703 s
>>> Epoch [ 312/1000]
train_loss: 1.4439 | train_acc: 0.5092 | val_loss: 1.4605 | val_acc: 0.5006 | test_acc: 0.4956 | Time: 2.1930 s
>>> Epoch [ 313/1000]
train_loss: 1.4437 | train_acc: 0.5093 | val_loss: 1.4603 | val_acc: 0.5004 | test_acc: 0.4956 | Time: 1.8423 s
>>> Epoch [ 314/1000]
train_loss: 1.4435 | train_acc: 0.5094 | val_loss: 1.4602 | val_acc: 0.5004 | test_acc: 0.4954 | Time: 1.7874 s
>>> Epoch [ 315/1000]
train_loss: 1.4433 | train_acc: 0.5095 | val_loss: 1.4600 | val_acc: 0.5004 | test_acc: 0.4954 | Time: 1.7336 s
>>> Epoch [ 316/1000]
train_loss: 1.4432 | train_acc: 0.5096 | val_loss: 1.4599 | val_acc: 0.5008 | test_acc: 0.4954 | Time: 2.0048 s
>>> Epoch [ 317/1000]
train_loss: 1.4430 | train_acc: 0.5096 | val_loss: 1.4598 | val_acc: 0.5008 | test_acc: 0.4954 | Time: 1.9429 s
>>> Epoch [ 318/1000]
train_loss: 1.4428 | train_acc: 0.5096 | val_loss: 1.4596 | val_acc: 0.5008 | test_acc: 0.4954 | Time: 2.0196 s
>>> Epoch [ 319/1000]
train_loss: 1.4426 | train_acc: 0.5098 | val_loss: 1.4595 | val_acc: 0.5006 | test_acc: 0.4957 | Time: 1.8618 s
>>> Epoch [ 320/1000]
train_loss: 1.4424 | train_acc: 0.5098 | val_loss: 1.4593 | val_acc: 0.5008 | test_acc: 0.4956 | Time: 1.9147 s
>>> Epoch [ 321/1000]
train_loss: 1.4422 | train_acc: 0.5099 | val_loss: 1.4592 | val_acc: 0.5008 | test_acc: 0.4956 | Time: 1.8770 s
>>> Epoch [ 322/1000]
train_loss: 1.4420 | train_acc: 0.5100 | val_loss: 1.4591 | val_acc: 0.5008 | test_acc: 0.4955 | Time: 1.7584 s
>>> Epoch [ 323/1000]
train_loss: 1.4419 | train_acc: 0.5100 | val_loss: 1.4589 | val_acc: 0.5010 | test_acc: 0.4955 | Time: 2.1271 s
>>> Epoch [ 324/1000]
train_loss: 1.4417 | train_acc: 0.5102 | val_loss: 1.4588 | val_acc: 0.5010 | test_acc: 0.4956 | Time: 1.7565 s
>>> Epoch [ 325/1000]
train_loss: 1.4415 | train_acc: 0.5102 | val_loss: 1.4587 | val_acc: 0.5008 | test_acc: 0.4955 | Time: 1.7959 s
>>> Epoch [ 326/1000]
train_loss: 1.4413 | train_acc: 0.5103 | val_loss: 1.4585 | val_acc: 0.5008 | test_acc: 0.4954 | Time: 1.9800 s
>>> Epoch [ 327/1000]
train_loss: 1.4411 | train_acc: 0.5104 | val_loss: 1.4584 | val_acc: 0.5008 | test_acc: 0.4957 | Time: 1.8733 s
>>> Epoch [ 328/1000]
train_loss: 1.4409 | train_acc: 0.5106 | val_loss: 1.4583 | val_acc: 0.5010 | test_acc: 0.4958 | Time: 1.9025 s
>>> Epoch [ 329/1000]
train_loss: 1.4408 | train_acc: 0.5107 | val_loss: 1.4581 | val_acc: 0.5012 | test_acc: 0.4958 | Time: 1.9788 s
>>> Epoch [ 330/1000]
train_loss: 1.4406 | train_acc: 0.5108 | val_loss: 1.4580 | val_acc: 0.5012 | test_acc: 0.4961 | Time: 1.9592 s
>>> Epoch [ 331/1000]
train_loss: 1.4404 | train_acc: 0.5109 | val_loss: 1.4579 | val_acc: 0.5014 | test_acc: 0.4962 | Time: 1.9274 s
>>> Epoch [ 332/1000]
train_loss: 1.4402 | train_acc: 0.5108 | val_loss: 1.4577 | val_acc: 0.5014 | test_acc: 0.4962 | Time: 2.2137 s
>>> Epoch [ 333/1000]
train_loss: 1.4401 | train_acc: 0.5110 | val_loss: 1.4576 | val_acc: 0.5016 | test_acc: 0.4963 | Time: 1.9041 s
>>> Epoch [ 334/1000]
train_loss: 1.4399 | train_acc: 0.5110 | val_loss: 1.4575 | val_acc: 0.5016 | test_acc: 0.4963 | Time: 2.0802 s
>>> Epoch [ 335/1000]
train_loss: 1.4397 | train_acc: 0.5111 | val_loss: 1.4573 | val_acc: 0.5016 | test_acc: 0.4962 | Time: 2.1874 s
>>> Epoch [ 336/1000]
train_loss: 1.4395 | train_acc: 0.5112 | val_loss: 1.4572 | val_acc: 0.5020 | test_acc: 0.4962 | Time: 2.0163 s
>>> Epoch [ 337/1000]
train_loss: 1.4393 | train_acc: 0.5112 | val_loss: 1.4571 | val_acc: 0.5020 | test_acc: 0.4963 | Time: 2.0270 s
>>> Epoch [ 338/1000]
train_loss: 1.4392 | train_acc: 0.5113 | val_loss: 1.4570 | val_acc: 0.5018 | test_acc: 0.4962 | Time: 2.1632 s
>>> Epoch [ 339/1000]
train_loss: 1.4390 | train_acc: 0.5113 | val_loss: 1.4568 | val_acc: 0.5018 | test_acc: 0.4962 | Time: 1.8292 s
>>> Epoch [ 340/1000]
train_loss: 1.4388 | train_acc: 0.5113 | val_loss: 1.4567 | val_acc: 0.5020 | test_acc: 0.4963 | Time: 2.1024 s
>>> Epoch [ 341/1000]
train_loss: 1.4386 | train_acc: 0.5114 | val_loss: 1.4566 | val_acc: 0.5022 | test_acc: 0.4963 | Time: 2.2360 s
>>> Epoch [ 342/1000]
train_loss: 1.4385 | train_acc: 0.5114 | val_loss: 1.4564 | val_acc: 0.5022 | test_acc: 0.4963 | Time: 2.1973 s
>>> Epoch [ 343/1000]
train_loss: 1.4383 | train_acc: 0.5114 | val_loss: 1.4563 | val_acc: 0.5022 | test_acc: 0.4965 | Time: 1.9319 s
>>> Epoch [ 344/1000]
train_loss: 1.4381 | train_acc: 0.5115 | val_loss: 1.4562 | val_acc: 0.5030 | test_acc: 0.4964 | Time: 2.2023 s
>>> Epoch [ 345/1000]
train_loss: 1.4380 | train_acc: 0.5115 | val_loss: 1.4561 | val_acc: 0.5030 | test_acc: 0.4965 | Time: 2.2798 s
>>> Epoch [ 346/1000]
train_loss: 1.4378 | train_acc: 0.5116 | val_loss: 1.4559 | val_acc: 0.5030 | test_acc: 0.4966 | Time: 2.1453 s
>>> Epoch [ 347/1000]
train_loss: 1.4376 | train_acc: 0.5117 | val_loss: 1.4558 | val_acc: 0.5034 | test_acc: 0.4966 | Time: 1.9655 s
>>> Epoch [ 348/1000]
train_loss: 1.4374 | train_acc: 0.5117 | val_loss: 1.4557 | val_acc: 0.5034 | test_acc: 0.4967 | Time: 2.1002 s
>>> Epoch [ 349/1000]
train_loss: 1.4373 | train_acc: 0.5118 | val_loss: 1.4556 | val_acc: 0.5034 | test_acc: 0.4970 | Time: 2.0831 s
>>> Epoch [ 350/1000]
train_loss: 1.4371 | train_acc: 0.5118 | val_loss: 1.4555 | val_acc: 0.5034 | test_acc: 0.4971 | Time: 1.9397 s
>>> Epoch [ 351/1000]
train_loss: 1.4369 | train_acc: 0.5120 | val_loss: 1.4553 | val_acc: 0.5036 | test_acc: 0.4972 | Time: 2.1840 s
>>> Epoch [ 352/1000]
train_loss: 1.4368 | train_acc: 0.5120 | val_loss: 1.4552 | val_acc: 0.5036 | test_acc: 0.4972 | Time: 2.2252 s
>>> Epoch [ 353/1000]
train_loss: 1.4366 | train_acc: 0.5122 | val_loss: 1.4551 | val_acc: 0.5036 | test_acc: 0.4970 | Time: 2.0260 s
>>> Epoch [ 354/1000]
train_loss: 1.4364 | train_acc: 0.5122 | val_loss: 1.4550 | val_acc: 0.5036 | test_acc: 0.4971 | Time: 2.1190 s
>>> Epoch [ 355/1000]
train_loss: 1.4363 | train_acc: 0.5123 | val_loss: 1.4548 | val_acc: 0.5038 | test_acc: 0.4971 | Time: 2.1758 s
>>> Epoch [ 356/1000]
train_loss: 1.4361 | train_acc: 0.5124 | val_loss: 1.4547 | val_acc: 0.5036 | test_acc: 0.4974 | Time: 2.0449 s
>>> Epoch [ 357/1000]
train_loss: 1.4359 | train_acc: 0.5125 | val_loss: 1.4546 | val_acc: 0.5034 | test_acc: 0.4974 | Time: 2.0076 s
>>> Epoch [ 358/1000]
train_loss: 1.4358 | train_acc: 0.5126 | val_loss: 1.4545 | val_acc: 0.5032 | test_acc: 0.4975 | Time: 2.0214 s
>>> Epoch [ 359/1000]
train_loss: 1.4356 | train_acc: 0.5126 | val_loss: 1.4544 | val_acc: 0.5032 | test_acc: 0.4975 | Time: 2.1634 s
>>> Epoch [ 360/1000]
train_loss: 1.4355 | train_acc: 0.5127 | val_loss: 1.4543 | val_acc: 0.5032 | test_acc: 0.4975 | Time: 1.8317 s
>>> Epoch [ 361/1000]
train_loss: 1.4353 | train_acc: 0.5127 | val_loss: 1.4541 | val_acc: 0.5032 | test_acc: 0.4975 | Time: 2.2350 s
>>> Epoch [ 362/1000]
train_loss: 1.4351 | train_acc: 0.5128 | val_loss: 1.4540 | val_acc: 0.5032 | test_acc: 0.4979 | Time: 2.2577 s
>>> Epoch [ 363/1000]
train_loss: 1.4350 | train_acc: 0.5129 | val_loss: 1.4539 | val_acc: 0.5034 | test_acc: 0.4980 | Time: 2.2732 s
>>> Epoch [ 364/1000]
train_loss: 1.4348 | train_acc: 0.5129 | val_loss: 1.4538 | val_acc: 0.5036 | test_acc: 0.4981 | Time: 1.8570 s
>>> Epoch [ 365/1000]
train_loss: 1.4346 | train_acc: 0.5130 | val_loss: 1.4537 | val_acc: 0.5036 | test_acc: 0.4981 | Time: 2.1319 s
>>> Epoch [ 366/1000]
train_loss: 1.4345 | train_acc: 0.5130 | val_loss: 1.4536 | val_acc: 0.5036 | test_acc: 0.4980 | Time: 2.1596 s
>>> Epoch [ 367/1000]
train_loss: 1.4343 | train_acc: 0.5132 | val_loss: 1.4534 | val_acc: 0.5036 | test_acc: 0.4980 | Time: 2.0453 s
>>> Epoch [ 368/1000]
train_loss: 1.4342 | train_acc: 0.5132 | val_loss: 1.4533 | val_acc: 0.5038 | test_acc: 0.4981 | Time: 1.9934 s
>>> Epoch [ 369/1000]
train_loss: 1.4340 | train_acc: 0.5134 | val_loss: 1.4532 | val_acc: 0.5038 | test_acc: 0.4984 | Time: 2.2246 s
>>> Epoch [ 370/1000]
train_loss: 1.4338 | train_acc: 0.5134 | val_loss: 1.4531 | val_acc: 0.5038 | test_acc: 0.4984 | Time: 2.2702 s
>>> Epoch [ 371/1000]
train_loss: 1.4337 | train_acc: 0.5134 | val_loss: 1.4530 | val_acc: 0.5042 | test_acc: 0.4985 | Time: 2.0342 s
>>> Epoch [ 372/1000]
train_loss: 1.4335 | train_acc: 0.5135 | val_loss: 1.4529 | val_acc: 0.5042 | test_acc: 0.4985 | Time: 2.1460 s
>>> Epoch [ 373/1000]
train_loss: 1.4334 | train_acc: 0.5136 | val_loss: 1.4528 | val_acc: 0.5046 | test_acc: 0.4985 | Time: 2.1351 s
>>> Epoch [ 374/1000]
train_loss: 1.4332 | train_acc: 0.5136 | val_loss: 1.4526 | val_acc: 0.5046 | test_acc: 0.4985 | Time: 1.8651 s
>>> Epoch [ 375/1000]
train_loss: 1.4331 | train_acc: 0.5136 | val_loss: 1.4525 | val_acc: 0.5046 | test_acc: 0.4985 | Time: 2.0842 s
>>> Epoch [ 376/1000]
train_loss: 1.4329 | train_acc: 0.5137 | val_loss: 1.4524 | val_acc: 0.5050 | test_acc: 0.4985 | Time: 2.2247 s
>>> Epoch [ 377/1000]
train_loss: 1.4327 | train_acc: 0.5137 | val_loss: 1.4523 | val_acc: 0.5050 | test_acc: 0.4986 | Time: 2.3327 s
>>> Epoch [ 378/1000]
train_loss: 1.4326 | train_acc: 0.5137 | val_loss: 1.4522 | val_acc: 0.5048 | test_acc: 0.4986 | Time: 2.1885 s
>>> Epoch [ 379/1000]
train_loss: 1.4324 | train_acc: 0.5137 | val_loss: 1.4521 | val_acc: 0.5048 | test_acc: 0.4986 | Time: 1.9148 s
>>> Epoch [ 380/1000]
train_loss: 1.4323 | train_acc: 0.5137 | val_loss: 1.4520 | val_acc: 0.5050 | test_acc: 0.4985 | Time: 2.2494 s
>>> Epoch [ 381/1000]
train_loss: 1.4321 | train_acc: 0.5139 | val_loss: 1.4519 | val_acc: 0.5052 | test_acc: 0.4984 | Time: 2.1793 s
>>> Epoch [ 382/1000]
train_loss: 1.4320 | train_acc: 0.5140 | val_loss: 1.4518 | val_acc: 0.5052 | test_acc: 0.4985 | Time: 2.1238 s
>>> Epoch [ 383/1000]
train_loss: 1.4318 | train_acc: 0.5140 | val_loss: 1.4516 | val_acc: 0.5056 | test_acc: 0.4986 | Time: 1.9361 s
>>> Epoch [ 384/1000]
train_loss: 1.4317 | train_acc: 0.5140 | val_loss: 1.4515 | val_acc: 0.5056 | test_acc: 0.4990 | Time: 2.2292 s
>>> Epoch [ 385/1000]
train_loss: 1.4315 | train_acc: 0.5141 | val_loss: 1.4514 | val_acc: 0.5056 | test_acc: 0.4990 | Time: 2.2310 s
>>> Epoch [ 386/1000]
train_loss: 1.4314 | train_acc: 0.5141 | val_loss: 1.4513 | val_acc: 0.5058 | test_acc: 0.4992 | Time: 1.9363 s
>>> Epoch [ 387/1000]
train_loss: 1.4312 | train_acc: 0.5142 | val_loss: 1.4512 | val_acc: 0.5058 | test_acc: 0.4993 | Time: 2.0650 s
>>> Epoch [ 388/1000]
train_loss: 1.4311 | train_acc: 0.5142 | val_loss: 1.4511 | val_acc: 0.5060 | test_acc: 0.4994 | Time: 2.1304 s
>>> Epoch [ 389/1000]
train_loss: 1.4309 | train_acc: 0.5142 | val_loss: 1.4510 | val_acc: 0.5060 | test_acc: 0.4995 | Time: 2.3269 s
>>> Epoch [ 390/1000]
train_loss: 1.4308 | train_acc: 0.5143 | val_loss: 1.4509 | val_acc: 0.5062 | test_acc: 0.4995 | Time: 2.5395 s
>>> Epoch [ 391/1000]
train_loss: 1.4306 | train_acc: 0.5145 | val_loss: 1.4508 | val_acc: 0.5062 | test_acc: 0.4996 | Time: 2.0278 s
>>> Epoch [ 392/1000]
train_loss: 1.4305 | train_acc: 0.5146 | val_loss: 1.4507 | val_acc: 0.5062 | test_acc: 0.4998 | Time: 2.0820 s
>>> Epoch [ 393/1000]
train_loss: 1.4303 | train_acc: 0.5148 | val_loss: 1.4506 | val_acc: 0.5060 | test_acc: 0.4998 | Time: 2.3067 s
>>> Epoch [ 394/1000]
train_loss: 1.4302 | train_acc: 0.5149 | val_loss: 1.4505 | val_acc: 0.5060 | test_acc: 0.4998 | Time: 2.2585 s
>>> Epoch [ 395/1000]
train_loss: 1.4300 | train_acc: 0.5149 | val_loss: 1.4504 | val_acc: 0.5060 | test_acc: 0.4999 | Time: 2.2442 s
>>> Epoch [ 396/1000]
train_loss: 1.4299 | train_acc: 0.5149 | val_loss: 1.4503 | val_acc: 0.5060 | test_acc: 0.4997 | Time: 2.1472 s
>>> Epoch [ 397/1000]
train_loss: 1.4297 | train_acc: 0.5150 | val_loss: 1.4502 | val_acc: 0.5060 | test_acc: 0.4997 | Time: 2.3172 s
>>> Epoch [ 398/1000]
train_loss: 1.4296 | train_acc: 0.5152 | val_loss: 1.4500 | val_acc: 0.5058 | test_acc: 0.4998 | Time: 2.4523 s
>>> Epoch [ 399/1000]
train_loss: 1.4294 | train_acc: 0.5152 | val_loss: 1.4499 | val_acc: 0.5058 | test_acc: 0.4998 | Time: 2.4962 s
>>> Epoch [ 400/1000]
train_loss: 1.4293 | train_acc: 0.5153 | val_loss: 1.4498 | val_acc: 0.5056 | test_acc: 0.4997 | Time: 2.4037 s
>>> Epoch [ 401/1000]
train_loss: 1.4291 | train_acc: 0.5153 | val_loss: 1.4497 | val_acc: 0.5058 | test_acc: 0.4997 | Time: 1.9040 s
>>> Epoch [ 402/1000]
train_loss: 1.4290 | train_acc: 0.5153 | val_loss: 1.4496 | val_acc: 0.5060 | test_acc: 0.4996 | Time: 2.1006 s
>>> Epoch [ 403/1000]
train_loss: 1.4288 | train_acc: 0.5154 | val_loss: 1.4495 | val_acc: 0.5062 | test_acc: 0.4998 | Time: 2.1397 s
>>> Epoch [ 404/1000]
train_loss: 1.4287 | train_acc: 0.5154 | val_loss: 1.4494 | val_acc: 0.5062 | test_acc: 0.4999 | Time: 2.2186 s
>>> Epoch [ 405/1000]
train_loss: 1.4286 | train_acc: 0.5154 | val_loss: 1.4493 | val_acc: 0.5062 | test_acc: 0.5000 | Time: 2.3215 s
>>> Epoch [ 406/1000]
train_loss: 1.4284 | train_acc: 0.5154 | val_loss: 1.4492 | val_acc: 0.5062 | test_acc: 0.4999 | Time: 2.0668 s
>>> Epoch [ 407/1000]
train_loss: 1.4283 | train_acc: 0.5154 | val_loss: 1.4491 | val_acc: 0.5062 | test_acc: 0.4999 | Time: 2.3220 s
>>> Epoch [ 408/1000]
train_loss: 1.4281 | train_acc: 0.5154 | val_loss: 1.4490 | val_acc: 0.5062 | test_acc: 0.4999 | Time: 2.2948 s
>>> Epoch [ 409/1000]
train_loss: 1.4280 | train_acc: 0.5154 | val_loss: 1.4489 | val_acc: 0.5064 | test_acc: 0.5001 | Time: 2.3540 s
>>> Epoch [ 410/1000]
train_loss: 1.4278 | train_acc: 0.5155 | val_loss: 1.4488 | val_acc: 0.5064 | test_acc: 0.5000 | Time: 2.2148 s
>>> Epoch [ 411/1000]
train_loss: 1.4277 | train_acc: 0.5156 | val_loss: 1.4487 | val_acc: 0.5066 | test_acc: 0.5000 | Time: 2.0927 s
>>> Epoch [ 412/1000]
train_loss: 1.4276 | train_acc: 0.5156 | val_loss: 1.4486 | val_acc: 0.5066 | test_acc: 0.5000 | Time: 2.4040 s
>>> Epoch [ 413/1000]
train_loss: 1.4274 | train_acc: 0.5157 | val_loss: 1.4485 | val_acc: 0.5068 | test_acc: 0.4999 | Time: 2.7192 s
>>> Epoch [ 414/1000]
train_loss: 1.4273 | train_acc: 0.5159 | val_loss: 1.4484 | val_acc: 0.5068 | test_acc: 0.4999 | Time: 2.0740 s
>>> Epoch [ 415/1000]
train_loss: 1.4271 | train_acc: 0.5160 | val_loss: 1.4483 | val_acc: 0.5070 | test_acc: 0.5000 | Time: 2.1891 s
>>> Epoch [ 416/1000]
train_loss: 1.4270 | train_acc: 0.5161 | val_loss: 1.4482 | val_acc: 0.5070 | test_acc: 0.5002 | Time: 2.3146 s
>>> Epoch [ 417/1000]
train_loss: 1.4268 | train_acc: 0.5162 | val_loss: 1.4481 | val_acc: 0.5070 | test_acc: 0.5005 | Time: 2.3167 s
>>> Epoch [ 418/1000]
train_loss: 1.4267 | train_acc: 0.5164 | val_loss: 1.4480 | val_acc: 0.5068 | test_acc: 0.5005 | Time: 2.5273 s
>>> Epoch [ 419/1000]
train_loss: 1.4266 | train_acc: 0.5164 | val_loss: 1.4479 | val_acc: 0.5068 | test_acc: 0.5004 | Time: 2.4963 s
>>> Epoch [ 420/1000]
train_loss: 1.4264 | train_acc: 0.5165 | val_loss: 1.4478 | val_acc: 0.5066 | test_acc: 0.5004 | Time: 2.5684 s
>>> Epoch [ 421/1000]
train_loss: 1.4263 | train_acc: 0.5165 | val_loss: 1.4477 | val_acc: 0.5070 | test_acc: 0.5003 | Time: 2.4748 s
>>> Epoch [ 422/1000]
train_loss: 1.4262 | train_acc: 0.5167 | val_loss: 1.4476 | val_acc: 0.5070 | test_acc: 0.5005 | Time: 2.2590 s
>>> Epoch [ 423/1000]
train_loss: 1.4260 | train_acc: 0.5168 | val_loss: 1.4476 | val_acc: 0.5072 | test_acc: 0.5007 | Time: 2.2925 s
>>> Epoch [ 424/1000]
train_loss: 1.4259 | train_acc: 0.5168 | val_loss: 1.4475 | val_acc: 0.5074 | test_acc: 0.5007 | Time: 2.5101 s
>>> Epoch [ 425/1000]
train_loss: 1.4257 | train_acc: 0.5168 | val_loss: 1.4474 | val_acc: 0.5074 | test_acc: 0.5007 | Time: 2.6650 s
>>> Epoch [ 426/1000]
train_loss: 1.4256 | train_acc: 0.5168 | val_loss: 1.4473 | val_acc: 0.5076 | test_acc: 0.5007 | Time: 2.4693 s
>>> Epoch [ 427/1000]
train_loss: 1.4255 | train_acc: 0.5170 | val_loss: 1.4472 | val_acc: 0.5078 | test_acc: 0.5006 | Time: 2.4774 s
>>> Epoch [ 428/1000]
train_loss: 1.4253 | train_acc: 0.5170 | val_loss: 1.4471 | val_acc: 0.5082 | test_acc: 0.5006 | Time: 2.0563 s
>>> Epoch [ 429/1000]
train_loss: 1.4252 | train_acc: 0.5170 | val_loss: 1.4470 | val_acc: 0.5080 | test_acc: 0.5007 | Time: 2.3388 s
>>> Epoch [ 430/1000]
train_loss: 1.4251 | train_acc: 0.5170 | val_loss: 1.4469 | val_acc: 0.5080 | test_acc: 0.5008 | Time: 2.3602 s
>>> Epoch [ 431/1000]
train_loss: 1.4249 | train_acc: 0.5171 | val_loss: 1.4468 | val_acc: 0.5082 | test_acc: 0.5008 | Time: 2.7072 s
>>> Epoch [ 432/1000]
train_loss: 1.4248 | train_acc: 0.5172 | val_loss: 1.4467 | val_acc: 0.5082 | test_acc: 0.5008 | Time: 2.1819 s
>>> Epoch [ 433/1000]
train_loss: 1.4247 | train_acc: 0.5173 | val_loss: 1.4466 | val_acc: 0.5084 | test_acc: 0.5008 | Time: 2.2265 s
>>> Epoch [ 434/1000]
train_loss: 1.4245 | train_acc: 0.5173 | val_loss: 1.4465 | val_acc: 0.5086 | test_acc: 0.5009 | Time: 2.3385 s
>>> Epoch [ 435/1000]
train_loss: 1.4244 | train_acc: 0.5174 | val_loss: 1.4464 | val_acc: 0.5086 | test_acc: 0.5010 | Time: 2.4624 s
>>> Epoch [ 436/1000]
train_loss: 1.4243 | train_acc: 0.5175 | val_loss: 1.4463 | val_acc: 0.5084 | test_acc: 0.5010 | Time: 2.5262 s
>>> Epoch [ 437/1000]
train_loss: 1.4241 | train_acc: 0.5176 | val_loss: 1.4462 | val_acc: 0.5086 | test_acc: 0.5010 | Time: 2.3852 s
>>> Epoch [ 438/1000]
train_loss: 1.4240 | train_acc: 0.5176 | val_loss: 1.4461 | val_acc: 0.5088 | test_acc: 0.5009 | Time: 2.5899 s
>>> Epoch [ 439/1000]
train_loss: 1.4239 | train_acc: 0.5177 | val_loss: 1.4461 | val_acc: 0.5088 | test_acc: 0.5008 | Time: 2.3577 s
>>> Epoch [ 440/1000]
train_loss: 1.4237 | train_acc: 0.5178 | val_loss: 1.4460 | val_acc: 0.5086 | test_acc: 0.5008 | Time: 2.4363 s
>>> Epoch [ 441/1000]
train_loss: 1.4236 | train_acc: 0.5179 | val_loss: 1.4459 | val_acc: 0.5086 | test_acc: 0.5006 | Time: 2.7985 s
>>> Epoch [ 442/1000]
train_loss: 1.4235 | train_acc: 0.5180 | val_loss: 1.4458 | val_acc: 0.5086 | test_acc: 0.5008 | Time: 2.5316 s
>>> Epoch [ 443/1000]
train_loss: 1.4233 | train_acc: 0.5180 | val_loss: 1.4457 | val_acc: 0.5084 | test_acc: 0.5010 | Time: 2.7125 s
>>> Epoch [ 444/1000]
train_loss: 1.4232 | train_acc: 0.5181 | val_loss: 1.4456 | val_acc: 0.5088 | test_acc: 0.5009 | Time: 2.6721 s
>>> Epoch [ 445/1000]
train_loss: 1.4231 | train_acc: 0.5182 | val_loss: 1.4455 | val_acc: 0.5088 | test_acc: 0.5009 | Time: 2.3054 s
>>> Epoch [ 446/1000]
train_loss: 1.4229 | train_acc: 0.5182 | val_loss: 1.4454 | val_acc: 0.5092 | test_acc: 0.5010 | Time: 2.7229 s
>>> Epoch [ 447/1000]
train_loss: 1.4228 | train_acc: 0.5184 | val_loss: 1.4453 | val_acc: 0.5088 | test_acc: 0.5012 | Time: 2.6407 s
>>> Epoch [ 448/1000]
train_loss: 1.4227 | train_acc: 0.5185 | val_loss: 1.4452 | val_acc: 0.5088 | test_acc: 0.5011 | Time: 2.5801 s
>>> Epoch [ 449/1000]
train_loss: 1.4226 | train_acc: 0.5186 | val_loss: 1.4451 | val_acc: 0.5088 | test_acc: 0.5011 | Time: 2.9053 s
>>> Epoch [ 450/1000]
train_loss: 1.4224 | train_acc: 0.5185 | val_loss: 1.4451 | val_acc: 0.5088 | test_acc: 0.5014 | Time: 3.0750 s
>>> Epoch [ 451/1000]
train_loss: 1.4223 | train_acc: 0.5185 | val_loss: 1.4450 | val_acc: 0.5088 | test_acc: 0.5014 | Time: 2.8781 s
>>> Epoch [ 452/1000]
train_loss: 1.4222 | train_acc: 0.5185 | val_loss: 1.4449 | val_acc: 0.5088 | test_acc: 0.5014 | Time: 2.5647 s
>>> Epoch [ 453/1000]
train_loss: 1.4220 | train_acc: 0.5185 | val_loss: 1.4448 | val_acc: 0.5086 | test_acc: 0.5014 | Time: 2.2560 s
>>> Epoch [ 454/1000]
train_loss: 1.4219 | train_acc: 0.5186 | val_loss: 1.4447 | val_acc: 0.5086 | test_acc: 0.5015 | Time: 2.6401 s
>>> Epoch [ 455/1000]
train_loss: 1.4218 | train_acc: 0.5186 | val_loss: 1.4446 | val_acc: 0.5084 | test_acc: 0.5017 | Time: 2.5785 s
>>> Epoch [ 456/1000]
train_loss: 1.4217 | train_acc: 0.5187 | val_loss: 1.4445 | val_acc: 0.5082 | test_acc: 0.5017 | Time: 2.7240 s
>>> Epoch [ 457/1000]
train_loss: 1.4215 | train_acc: 0.5187 | val_loss: 1.4444 | val_acc: 0.5086 | test_acc: 0.5017 | Time: 2.7432 s
>>> Epoch [ 458/1000]
train_loss: 1.4214 | train_acc: 0.5187 | val_loss: 1.4444 | val_acc: 0.5086 | test_acc: 0.5016 | Time: 2.7401 s
>>> Epoch [ 459/1000]
train_loss: 1.4213 | train_acc: 0.5188 | val_loss: 1.4443 | val_acc: 0.5086 | test_acc: 0.5016 | Time: 2.9117 s
>>> Epoch [ 460/1000]
train_loss: 1.4212 | train_acc: 0.5189 | val_loss: 1.4442 | val_acc: 0.5086 | test_acc: 0.5015 | Time: 2.9967 s
>>> Epoch [ 461/1000]
train_loss: 1.4210 | train_acc: 0.5188 | val_loss: 1.4441 | val_acc: 0.5084 | test_acc: 0.5018 | Time: 2.9914 s
>>> Epoch [ 462/1000]
train_loss: 1.4209 | train_acc: 0.5189 | val_loss: 1.4440 | val_acc: 0.5084 | test_acc: 0.5018 | Time: 2.9901 s
>>> Epoch [ 463/1000]
train_loss: 1.4208 | train_acc: 0.5190 | val_loss: 1.4439 | val_acc: 0.5086 | test_acc: 0.5019 | Time: 2.9751 s
>>> Epoch [ 464/1000]
train_loss: 1.4207 | train_acc: 0.5190 | val_loss: 1.4438 | val_acc: 0.5084 | test_acc: 0.5018 | Time: 3.1901 s
>>> Epoch [ 465/1000]
train_loss: 1.4205 | train_acc: 0.5190 | val_loss: 1.4438 | val_acc: 0.5078 | test_acc: 0.5017 | Time: 2.8342 s
>>> Epoch [ 466/1000]
train_loss: 1.4204 | train_acc: 0.5191 | val_loss: 1.4437 | val_acc: 0.5076 | test_acc: 0.5016 | Time: 2.8632 s
>>> Epoch [ 467/1000]
train_loss: 1.4203 | train_acc: 0.5192 | val_loss: 1.4436 | val_acc: 0.5074 | test_acc: 0.5017 | Time: 2.7004 s
>>> Epoch [ 468/1000]
train_loss: 1.4202 | train_acc: 0.5192 | val_loss: 1.4435 | val_acc: 0.5072 | test_acc: 0.5016 | Time: 2.9775 s
>>> Epoch [ 469/1000]
train_loss: 1.4200 | train_acc: 0.5193 | val_loss: 1.4434 | val_acc: 0.5072 | test_acc: 0.5016 | Time: 2.5248 s
>>> Epoch [ 470/1000]
train_loss: 1.4199 | train_acc: 0.5194 | val_loss: 1.4433 | val_acc: 0.5072 | test_acc: 0.5016 | Time: 3.3201 s
>>> Epoch [ 471/1000]
train_loss: 1.4198 | train_acc: 0.5194 | val_loss: 1.4432 | val_acc: 0.5072 | test_acc: 0.5017 | Time: 2.9511 s
>>> Epoch [ 472/1000]
train_loss: 1.4197 | train_acc: 0.5194 | val_loss: 1.4432 | val_acc: 0.5070 | test_acc: 0.5017 | Time: 2.3092 s
>>> Epoch [ 473/1000]
train_loss: 1.4195 | train_acc: 0.5195 | val_loss: 1.4431 | val_acc: 0.5074 | test_acc: 0.5017 | Time: 2.3089 s
>>> Epoch [ 474/1000]
train_loss: 1.4194 | train_acc: 0.5196 | val_loss: 1.4430 | val_acc: 0.5074 | test_acc: 0.5017 | Time: 2.2324 s
>>> Epoch [ 475/1000]
train_loss: 1.4193 | train_acc: 0.5196 | val_loss: 1.4429 | val_acc: 0.5074 | test_acc: 0.5017 | Time: 2.2693 s
>>> Epoch [ 476/1000]
train_loss: 1.4192 | train_acc: 0.5196 | val_loss: 1.4428 | val_acc: 0.5070 | test_acc: 0.5017 | Time: 2.2461 s
>>> Epoch [ 477/1000]
train_loss: 1.4191 | train_acc: 0.5196 | val_loss: 1.4427 | val_acc: 0.5070 | test_acc: 0.5016 | Time: 2.1336 s
>>> Epoch [ 478/1000]
train_loss: 1.4189 | train_acc: 0.5196 | val_loss: 1.4427 | val_acc: 0.5070 | test_acc: 0.5016 | Time: 2.2309 s
>>> Epoch [ 479/1000]
train_loss: 1.4188 | train_acc: 0.5197 | val_loss: 1.4426 | val_acc: 0.5070 | test_acc: 0.5018 | Time: 2.5630 s
>>> Epoch [ 480/1000]
train_loss: 1.4187 | train_acc: 0.5197 | val_loss: 1.4425 | val_acc: 0.5070 | test_acc: 0.5019 | Time: 2.0120 s
>>> Epoch [ 481/1000]
train_loss: 1.4186 | train_acc: 0.5198 | val_loss: 1.4424 | val_acc: 0.5072 | test_acc: 0.5020 | Time: 2.3140 s
>>> Epoch [ 482/1000]
train_loss: 1.4184 | train_acc: 0.5199 | val_loss: 1.4423 | val_acc: 0.5074 | test_acc: 0.5019 | Time: 2.1985 s
>>> Epoch [ 483/1000]
train_loss: 1.4183 | train_acc: 0.5199 | val_loss: 1.4423 | val_acc: 0.5076 | test_acc: 0.5022 | Time: 2.3513 s
>>> Epoch [ 484/1000]
train_loss: 1.4182 | train_acc: 0.5199 | val_loss: 1.4422 | val_acc: 0.5074 | test_acc: 0.5021 | Time: 2.4302 s
>>> Epoch [ 485/1000]
train_loss: 1.4181 | train_acc: 0.5200 | val_loss: 1.4421 | val_acc: 0.5074 | test_acc: 0.5021 | Time: 2.3032 s
>>> Epoch [ 486/1000]
train_loss: 1.4180 | train_acc: 0.5200 | val_loss: 1.4420 | val_acc: 0.5076 | test_acc: 0.5022 | Time: 2.2478 s
>>> Epoch [ 487/1000]
train_loss: 1.4179 | train_acc: 0.5201 | val_loss: 1.4419 | val_acc: 0.5076 | test_acc: 0.5022 | Time: 2.2324 s
>>> Epoch [ 488/1000]
train_loss: 1.4177 | train_acc: 0.5202 | val_loss: 1.4419 | val_acc: 0.5076 | test_acc: 0.5022 | Time: 2.1207 s
>>> Epoch [ 489/1000]
train_loss: 1.4176 | train_acc: 0.5201 | val_loss: 1.4418 | val_acc: 0.5076 | test_acc: 0.5021 | Time: 2.3893 s
>>> Epoch [ 490/1000]
train_loss: 1.4175 | train_acc: 0.5201 | val_loss: 1.4417 | val_acc: 0.5076 | test_acc: 0.5020 | Time: 2.2985 s
>>> Epoch [ 491/1000]
train_loss: 1.4174 | train_acc: 0.5201 | val_loss: 1.4416 | val_acc: 0.5078 | test_acc: 0.5021 | Time: 2.0845 s
>>> Epoch [ 492/1000]
train_loss: 1.4173 | train_acc: 0.5201 | val_loss: 1.4415 | val_acc: 0.5078 | test_acc: 0.5022 | Time: 2.1455 s
>>> Epoch [ 493/1000]
train_loss: 1.4171 | train_acc: 0.5202 | val_loss: 1.4415 | val_acc: 0.5078 | test_acc: 0.5022 | Time: 2.2382 s
>>> Epoch [ 494/1000]
train_loss: 1.4170 | train_acc: 0.5202 | val_loss: 1.4414 | val_acc: 0.5078 | test_acc: 0.5024 | Time: 2.3904 s
>>> Epoch [ 495/1000]
train_loss: 1.4169 | train_acc: 0.5202 | val_loss: 1.4413 | val_acc: 0.5078 | test_acc: 0.5023 | Time: 2.5266 s
>>> Epoch [ 496/1000]
train_loss: 1.4168 | train_acc: 0.5203 | val_loss: 1.4412 | val_acc: 0.5082 | test_acc: 0.5024 | Time: 2.4218 s
>>> Epoch [ 497/1000]
train_loss: 1.4167 | train_acc: 0.5204 | val_loss: 1.4411 | val_acc: 0.5080 | test_acc: 0.5024 | Time: 2.0491 s
>>> Epoch [ 498/1000]
train_loss: 1.4166 | train_acc: 0.5204 | val_loss: 1.4411 | val_acc: 0.5080 | test_acc: 0.5023 | Time: 2.3162 s
>>> Epoch [ 499/1000]
train_loss: 1.4164 | train_acc: 0.5205 | val_loss: 1.4410 | val_acc: 0.5080 | test_acc: 0.5024 | Time: 2.1184 s
>>> Epoch [ 500/1000]
train_loss: 1.4163 | train_acc: 0.5205 | val_loss: 1.4409 | val_acc: 0.5082 | test_acc: 0.5023 | Time: 2.3917 s
>>> Epoch [ 501/1000]
train_loss: 1.4162 | train_acc: 0.5206 | val_loss: 1.4408 | val_acc: 0.5082 | test_acc: 0.5025 | Time: 2.0601 s
>>> Epoch [ 502/1000]
train_loss: 1.4161 | train_acc: 0.5206 | val_loss: 1.4408 | val_acc: 0.5082 | test_acc: 0.5025 | Time: 1.9873 s
>>> Epoch [ 503/1000]
train_loss: 1.4160 | train_acc: 0.5206 | val_loss: 1.4407 | val_acc: 0.5082 | test_acc: 0.5024 | Time: 2.2377 s
>>> Epoch [ 504/1000]
train_loss: 1.4159 | train_acc: 0.5206 | val_loss: 1.4406 | val_acc: 0.5084 | test_acc: 0.5025 | Time: 2.3231 s
>>> Epoch [ 505/1000]
train_loss: 1.4158 | train_acc: 0.5208 | val_loss: 1.4405 | val_acc: 0.5084 | test_acc: 0.5025 | Time: 2.5011 s
>>> Epoch [ 506/1000]
train_loss: 1.4156 | train_acc: 0.5209 | val_loss: 1.4404 | val_acc: 0.5084 | test_acc: 0.5025 | Time: 2.0055 s
>>> Epoch [ 507/1000]
train_loss: 1.4155 | train_acc: 0.5208 | val_loss: 1.4404 | val_acc: 0.5084 | test_acc: 0.5028 | Time: 1.9749 s
>>> Epoch [ 508/1000]
train_loss: 1.4154 | train_acc: 0.5208 | val_loss: 1.4403 | val_acc: 0.5088 | test_acc: 0.5028 | Time: 2.2216 s
>>> Epoch [ 509/1000]
train_loss: 1.4153 | train_acc: 0.5209 | val_loss: 1.4402 | val_acc: 0.5088 | test_acc: 0.5029 | Time: 2.4169 s
>>> Epoch [ 510/1000]
train_loss: 1.4152 | train_acc: 0.5209 | val_loss: 1.4401 | val_acc: 0.5088 | test_acc: 0.5030 | Time: 2.6139 s
>>> Epoch [ 511/1000]
train_loss: 1.4151 | train_acc: 0.5209 | val_loss: 1.4401 | val_acc: 0.5088 | test_acc: 0.5031 | Time: 2.0957 s
>>> Epoch [ 512/1000]
train_loss: 1.4150 | train_acc: 0.5209 | val_loss: 1.4400 | val_acc: 0.5088 | test_acc: 0.5032 | Time: 2.3267 s
>>> Epoch [ 513/1000]
train_loss: 1.4149 | train_acc: 0.5209 | val_loss: 1.4399 | val_acc: 0.5088 | test_acc: 0.5034 | Time: 2.3907 s
>>> Epoch [ 514/1000]
train_loss: 1.4147 | train_acc: 0.5210 | val_loss: 1.4398 | val_acc: 0.5090 | test_acc: 0.5032 | Time: 2.3210 s
>>> Epoch [ 515/1000]
train_loss: 1.4146 | train_acc: 0.5210 | val_loss: 1.4398 | val_acc: 0.5090 | test_acc: 0.5032 | Time: 2.5032 s
>>> Epoch [ 516/1000]
train_loss: 1.4145 | train_acc: 0.5211 | val_loss: 1.4397 | val_acc: 0.5090 | test_acc: 0.5033 | Time: 2.6041 s
>>> Epoch [ 517/1000]
train_loss: 1.4144 | train_acc: 0.5212 | val_loss: 1.4396 | val_acc: 0.5092 | test_acc: 0.5034 | Time: 2.6656 s
>>> Epoch [ 518/1000]
train_loss: 1.4143 | train_acc: 0.5212 | val_loss: 1.4395 | val_acc: 0.5092 | test_acc: 0.5030 | Time: 2.3160 s
>>> Epoch [ 519/1000]
train_loss: 1.4142 | train_acc: 0.5213 | val_loss: 1.4395 | val_acc: 0.5090 | test_acc: 0.5031 | Time: 2.4049 s
>>> Epoch [ 520/1000]
train_loss: 1.4141 | train_acc: 0.5212 | val_loss: 1.4394 | val_acc: 0.5090 | test_acc: 0.5031 | Time: 2.2249 s
>>> Epoch [ 521/1000]
train_loss: 1.4140 | train_acc: 0.5213 | val_loss: 1.4393 | val_acc: 0.5092 | test_acc: 0.5031 | Time: 2.1987 s
>>> Epoch [ 522/1000]
train_loss: 1.4139 | train_acc: 0.5213 | val_loss: 1.4392 | val_acc: 0.5092 | test_acc: 0.5032 | Time: 2.2983 s
>>> Epoch [ 523/1000]
train_loss: 1.4137 | train_acc: 0.5213 | val_loss: 1.4392 | val_acc: 0.5090 | test_acc: 0.5033 | Time: 2.6865 s
>>> Epoch [ 524/1000]
train_loss: 1.4136 | train_acc: 0.5214 | val_loss: 1.4391 | val_acc: 0.5090 | test_acc: 0.5032 | Time: 2.7935 s
>>> Epoch [ 525/1000]
train_loss: 1.4135 | train_acc: 0.5215 | val_loss: 1.4390 | val_acc: 0.5090 | test_acc: 0.5032 | Time: 2.2825 s
>>> Epoch [ 526/1000]
train_loss: 1.4134 | train_acc: 0.5215 | val_loss: 1.4390 | val_acc: 0.5090 | test_acc: 0.5034 | Time: 2.7921 s
>>> Epoch [ 527/1000]
train_loss: 1.4133 | train_acc: 0.5215 | val_loss: 1.4389 | val_acc: 0.5088 | test_acc: 0.5034 | Time: 2.7258 s
>>> Epoch [ 528/1000]
train_loss: 1.4132 | train_acc: 0.5215 | val_loss: 1.4388 | val_acc: 0.5088 | test_acc: 0.5034 | Time: 2.5906 s
>>> Epoch [ 529/1000]
train_loss: 1.4131 | train_acc: 0.5216 | val_loss: 1.4387 | val_acc: 0.5090 | test_acc: 0.5035 | Time: 2.7623 s
>>> Epoch [ 530/1000]
train_loss: 1.4130 | train_acc: 0.5215 | val_loss: 1.4387 | val_acc: 0.5092 | test_acc: 0.5035 | Time: 2.4279 s
>>> Epoch [ 531/1000]
train_loss: 1.4129 | train_acc: 0.5216 | val_loss: 1.4386 | val_acc: 0.5094 | test_acc: 0.5036 | Time: 2.6482 s
>>> Epoch [ 532/1000]
train_loss: 1.4128 | train_acc: 0.5216 | val_loss: 1.4385 | val_acc: 0.5092 | test_acc: 0.5038 | Time: 2.7302 s
>>> Epoch [ 533/1000]
train_loss: 1.4127 | train_acc: 0.5216 | val_loss: 1.4384 | val_acc: 0.5092 | test_acc: 0.5038 | Time: 2.6515 s
>>> Epoch [ 534/1000]
train_loss: 1.4125 | train_acc: 0.5216 | val_loss: 1.4384 | val_acc: 0.5092 | test_acc: 0.5039 | Time: 2.6239 s
>>> Epoch [ 535/1000]
train_loss: 1.4124 | train_acc: 0.5217 | val_loss: 1.4383 | val_acc: 0.5094 | test_acc: 0.5041 | Time: 2.1660 s
>>> Epoch [ 536/1000]
train_loss: 1.4123 | train_acc: 0.5217 | val_loss: 1.4382 | val_acc: 0.5096 | test_acc: 0.5042 | Time: 2.4851 s
>>> Epoch [ 537/1000]
train_loss: 1.4122 | train_acc: 0.5218 | val_loss: 1.4382 | val_acc: 0.5096 | test_acc: 0.5044 | Time: 2.5679 s
>>> Epoch [ 538/1000]
train_loss: 1.4121 | train_acc: 0.5218 | val_loss: 1.4381 | val_acc: 0.5096 | test_acc: 0.5044 | Time: 2.4615 s
>>> Epoch [ 539/1000]
train_loss: 1.4120 | train_acc: 0.5219 | val_loss: 1.4380 | val_acc: 0.5098 | test_acc: 0.5044 | Time: 2.7669 s
>>> Epoch [ 540/1000]
train_loss: 1.4119 | train_acc: 0.5219 | val_loss: 1.4379 | val_acc: 0.5100 | test_acc: 0.5046 | Time: 2.6108 s
>>> Epoch [ 541/1000]
train_loss: 1.4118 | train_acc: 0.5218 | val_loss: 1.4379 | val_acc: 0.5100 | test_acc: 0.5046 | Time: 2.3530 s
>>> Epoch [ 542/1000]
train_loss: 1.4117 | train_acc: 0.5219 | val_loss: 1.4378 | val_acc: 0.5100 | test_acc: 0.5047 | Time: 2.4329 s
>>> Epoch [ 543/1000]
train_loss: 1.4116 | train_acc: 0.5219 | val_loss: 1.4377 | val_acc: 0.5102 | test_acc: 0.5047 | Time: 2.7281 s
>>> Epoch [ 544/1000]
train_loss: 1.4115 | train_acc: 0.5219 | val_loss: 1.4377 | val_acc: 0.5102 | test_acc: 0.5048 | Time: 2.5787 s
>>> Epoch [ 545/1000]
train_loss: 1.4114 | train_acc: 0.5219 | val_loss: 1.4376 | val_acc: 0.5102 | test_acc: 0.5049 | Time: 2.8540 s
>>> Epoch [ 546/1000]
train_loss: 1.4113 | train_acc: 0.5220 | val_loss: 1.4375 | val_acc: 0.5100 | test_acc: 0.5050 | Time: 2.7892 s
>>> Epoch [ 547/1000]
train_loss: 1.4112 | train_acc: 0.5220 | val_loss: 1.4375 | val_acc: 0.5100 | test_acc: 0.5051 | Time: 2.8044 s
>>> Epoch [ 548/1000]
train_loss: 1.4111 | train_acc: 0.5220 | val_loss: 1.4374 | val_acc: 0.5098 | test_acc: 0.5051 | Time: 3.0962 s
>>> Epoch [ 549/1000]
train_loss: 1.4110 | train_acc: 0.5220 | val_loss: 1.4373 | val_acc: 0.5098 | test_acc: 0.5053 | Time: 2.7445 s
>>> Epoch [ 550/1000]
train_loss: 1.4109 | train_acc: 0.5222 | val_loss: 1.4373 | val_acc: 0.5098 | test_acc: 0.5052 | Time: 2.8909 s
>>> Epoch [ 551/1000]
train_loss: 1.4107 | train_acc: 0.5221 | val_loss: 1.4372 | val_acc: 0.5098 | test_acc: 0.5052 | Time: 2.7631 s
>>> Epoch [ 552/1000]
train_loss: 1.4106 | train_acc: 0.5221 | val_loss: 1.4371 | val_acc: 0.5100 | test_acc: 0.5053 | Time: 2.8779 s
>>> Epoch [ 553/1000]
train_loss: 1.4105 | train_acc: 0.5222 | val_loss: 1.4370 | val_acc: 0.5100 | test_acc: 0.5054 | Time: 2.2849 s
>>> Epoch [ 554/1000]
train_loss: 1.4104 | train_acc: 0.5223 | val_loss: 1.4370 | val_acc: 0.5100 | test_acc: 0.5055 | Time: 2.8017 s
>>> Epoch [ 555/1000]
train_loss: 1.4103 | train_acc: 0.5223 | val_loss: 1.4369 | val_acc: 0.5098 | test_acc: 0.5055 | Time: 2.5550 s
>>> Epoch [ 556/1000]
train_loss: 1.4102 | train_acc: 0.5223 | val_loss: 1.4368 | val_acc: 0.5098 | test_acc: 0.5053 | Time: 2.7361 s
>>> Epoch [ 557/1000]
train_loss: 1.4101 | train_acc: 0.5224 | val_loss: 1.4368 | val_acc: 0.5100 | test_acc: 0.5053 | Time: 2.7378 s
>>> Epoch [ 558/1000]
train_loss: 1.4100 | train_acc: 0.5223 | val_loss: 1.4367 | val_acc: 0.5100 | test_acc: 0.5052 | Time: 2.5589 s
>>> Epoch [ 559/1000]
train_loss: 1.4099 | train_acc: 0.5222 | val_loss: 1.4366 | val_acc: 0.5102 | test_acc: 0.5052 | Time: 3.2786 s
>>> Epoch [ 560/1000]
train_loss: 1.4098 | train_acc: 0.5222 | val_loss: 1.4366 | val_acc: 0.5100 | test_acc: 0.5053 | Time: 2.8517 s
>>> Epoch [ 561/1000]
train_loss: 1.4097 | train_acc: 0.5223 | val_loss: 1.4365 | val_acc: 0.5098 | test_acc: 0.5053 | Time: 2.6616 s
>>> Epoch [ 562/1000]
train_loss: 1.4096 | train_acc: 0.5224 | val_loss: 1.4364 | val_acc: 0.5098 | test_acc: 0.5053 | Time: 2.9371 s
>>> Epoch [ 563/1000]
train_loss: 1.4095 | train_acc: 0.5224 | val_loss: 1.4364 | val_acc: 0.5098 | test_acc: 0.5052 | Time: 2.7757 s
>>> Epoch [ 564/1000]
train_loss: 1.4094 | train_acc: 0.5225 | val_loss: 1.4363 | val_acc: 0.5096 | test_acc: 0.5052 | Time: 2.9294 s
>>> Epoch [ 565/1000]
train_loss: 1.4093 | train_acc: 0.5225 | val_loss: 1.4362 | val_acc: 0.5096 | test_acc: 0.5052 | Time: 2.6828 s
>>> Epoch [ 566/1000]
train_loss: 1.4092 | train_acc: 0.5225 | val_loss: 1.4362 | val_acc: 0.5094 | test_acc: 0.5052 | Time: 2.9098 s
>>> Epoch [ 567/1000]
train_loss: 1.4091 | train_acc: 0.5225 | val_loss: 1.4361 | val_acc: 0.5094 | test_acc: 0.5052 | Time: 2.5274 s
>>> Epoch [ 568/1000]
train_loss: 1.4090 | train_acc: 0.5225 | val_loss: 1.4360 | val_acc: 0.5096 | test_acc: 0.5052 | Time: 2.5336 s
>>> Epoch [ 569/1000]
train_loss: 1.4089 | train_acc: 0.5226 | val_loss: 1.4360 | val_acc: 0.5094 | test_acc: 0.5052 | Time: 2.5794 s
>>> Epoch [ 570/1000]
train_loss: 1.4088 | train_acc: 0.5226 | val_loss: 1.4359 | val_acc: 0.5094 | test_acc: 0.5052 | Time: 3.1068 s
>>> Epoch [ 571/1000]
train_loss: 1.4087 | train_acc: 0.5228 | val_loss: 1.4358 | val_acc: 0.5094 | test_acc: 0.5051 | Time: 3.1605 s
>>> Epoch [ 572/1000]
train_loss: 1.4086 | train_acc: 0.5229 | val_loss: 1.4358 | val_acc: 0.5098 | test_acc: 0.5050 | Time: 2.8162 s
>>> Epoch [ 573/1000]
train_loss: 1.4085 | train_acc: 0.5229 | val_loss: 1.4357 | val_acc: 0.5098 | test_acc: 0.5050 | Time: 3.3145 s
>>> Epoch [ 574/1000]
train_loss: 1.4084 | train_acc: 0.5229 | val_loss: 1.4356 | val_acc: 0.5098 | test_acc: 0.5051 | Time: 2.9955 s
>>> Epoch [ 575/1000]
train_loss: 1.4083 | train_acc: 0.5230 | val_loss: 1.4356 | val_acc: 0.5098 | test_acc: 0.5052 | Time: 2.9034 s
>>> Epoch [ 576/1000]
train_loss: 1.4082 | train_acc: 0.5231 | val_loss: 1.4355 | val_acc: 0.5098 | test_acc: 0.5052 | Time: 2.8500 s
>>> Epoch [ 577/1000]
train_loss: 1.4081 | train_acc: 0.5232 | val_loss: 1.4355 | val_acc: 0.5098 | test_acc: 0.5053 | Time: 2.8985 s
>>> Epoch [ 578/1000]
train_loss: 1.4080 | train_acc: 0.5232 | val_loss: 1.4354 | val_acc: 0.5098 | test_acc: 0.5054 | Time: 2.9132 s
>>> Epoch [ 579/1000]
train_loss: 1.4079 | train_acc: 0.5233 | val_loss: 1.4353 | val_acc: 0.5098 | test_acc: 0.5054 | Time: 2.9951 s
>>> Epoch [ 580/1000]
train_loss: 1.4078 | train_acc: 0.5233 | val_loss: 1.4353 | val_acc: 0.5100 | test_acc: 0.5053 | Time: 2.8489 s
>>> Epoch [ 581/1000]
train_loss: 1.4077 | train_acc: 0.5232 | val_loss: 1.4352 | val_acc: 0.5100 | test_acc: 0.5053 | Time: 3.0441 s
>>> Epoch [ 582/1000]
train_loss: 1.4076 | train_acc: 0.5233 | val_loss: 1.4351 | val_acc: 0.5100 | test_acc: 0.5052 | Time: 2.9000 s
>>> Epoch [ 583/1000]
train_loss: 1.4075 | train_acc: 0.5234 | val_loss: 1.4351 | val_acc: 0.5102 | test_acc: 0.5052 | Time: 2.8608 s
>>> Epoch [ 584/1000]
train_loss: 1.4074 | train_acc: 0.5233 | val_loss: 1.4350 | val_acc: 0.5102 | test_acc: 0.5051 | Time: 2.8404 s
>>> Epoch [ 585/1000]
train_loss: 1.4073 | train_acc: 0.5233 | val_loss: 1.4349 | val_acc: 0.5102 | test_acc: 0.5050 | Time: 2.9734 s
>>> Epoch [ 586/1000]
train_loss: 1.4072 | train_acc: 0.5233 | val_loss: 1.4349 | val_acc: 0.5100 | test_acc: 0.5051 | Time: 2.8385 s
>>> Epoch [ 587/1000]
train_loss: 1.4071 | train_acc: 0.5233 | val_loss: 1.4348 | val_acc: 0.5104 | test_acc: 0.5051 | Time: 2.7085 s
>>> Epoch [ 588/1000]
train_loss: 1.4070 | train_acc: 0.5233 | val_loss: 1.4348 | val_acc: 0.5104 | test_acc: 0.5051 | Time: 2.8569 s
>>> Epoch [ 589/1000]
train_loss: 1.4069 | train_acc: 0.5233 | val_loss: 1.4347 | val_acc: 0.5104 | test_acc: 0.5051 | Time: 3.0017 s
>>> Epoch [ 590/1000]
train_loss: 1.4068 | train_acc: 0.5233 | val_loss: 1.4346 | val_acc: 0.5104 | test_acc: 0.5052 | Time: 2.9191 s
>>> Epoch [ 591/1000]
train_loss: 1.4067 | train_acc: 0.5233 | val_loss: 1.4346 | val_acc: 0.5104 | test_acc: 0.5052 | Time: 2.8317 s
>>> Epoch [ 592/1000]
train_loss: 1.4066 | train_acc: 0.5234 | val_loss: 1.4345 | val_acc: 0.5104 | test_acc: 0.5052 | Time: 2.8127 s
>>> Epoch [ 593/1000]
train_loss: 1.4065 | train_acc: 0.5233 | val_loss: 1.4344 | val_acc: 0.5108 | test_acc: 0.5052 | Time: 2.9213 s
>>> Epoch [ 594/1000]
train_loss: 1.4064 | train_acc: 0.5233 | val_loss: 1.4344 | val_acc: 0.5108 | test_acc: 0.5053 | Time: 2.6941 s
>>> Epoch [ 595/1000]
train_loss: 1.4063 | train_acc: 0.5234 | val_loss: 1.4343 | val_acc: 0.5106 | test_acc: 0.5055 | Time: 2.9000 s
>>> Epoch [ 596/1000]
train_loss: 1.4062 | train_acc: 0.5235 | val_loss: 1.4343 | val_acc: 0.5106 | test_acc: 0.5055 | Time: 2.8344 s
>>> Epoch [ 597/1000]
train_loss: 1.4062 | train_acc: 0.5236 | val_loss: 1.4342 | val_acc: 0.5106 | test_acc: 0.5056 | Time: 2.8001 s
>>> Epoch [ 598/1000]
train_loss: 1.4061 | train_acc: 0.5236 | val_loss: 1.4341 | val_acc: 0.5108 | test_acc: 0.5056 | Time: 2.9969 s
>>> Epoch [ 599/1000]
train_loss: 1.4060 | train_acc: 0.5236 | val_loss: 1.4341 | val_acc: 0.5108 | test_acc: 0.5055 | Time: 2.8093 s
>>> Epoch [ 600/1000]
train_loss: 1.4059 | train_acc: 0.5236 | val_loss: 1.4340 | val_acc: 0.5108 | test_acc: 0.5056 | Time: 2.6545 s
>>> Epoch [ 601/1000]
train_loss: 1.4058 | train_acc: 0.5237 | val_loss: 1.4339 | val_acc: 0.5108 | test_acc: 0.5055 | Time: 2.6281 s
>>> Epoch [ 602/1000]
train_loss: 1.4057 | train_acc: 0.5237 | val_loss: 1.4339 | val_acc: 0.5108 | test_acc: 0.5056 | Time: 2.6199 s
>>> Epoch [ 603/1000]
train_loss: 1.4056 | train_acc: 0.5238 | val_loss: 1.4338 | val_acc: 0.5108 | test_acc: 0.5056 | Time: 2.9032 s
>>> Epoch [ 604/1000]
train_loss: 1.4055 | train_acc: 0.5238 | val_loss: 1.4338 | val_acc: 0.5108 | test_acc: 0.5057 | Time: 3.0356 s
>>> Epoch [ 605/1000]
train_loss: 1.4054 | train_acc: 0.5238 | val_loss: 1.4337 | val_acc: 0.5108 | test_acc: 0.5057 | Time: 2.8421 s
>>> Epoch [ 606/1000]
train_loss: 1.4053 | train_acc: 0.5238 | val_loss: 1.4336 | val_acc: 0.5108 | test_acc: 0.5056 | Time: 2.8678 s
>>> Epoch [ 607/1000]
train_loss: 1.4052 | train_acc: 0.5238 | val_loss: 1.4336 | val_acc: 0.5108 | test_acc: 0.5057 | Time: 2.7741 s
>>> Epoch [ 608/1000]
train_loss: 1.4051 | train_acc: 0.5238 | val_loss: 1.4335 | val_acc: 0.5106 | test_acc: 0.5057 | Time: 2.7779 s
>>> Epoch [ 609/1000]
train_loss: 1.4050 | train_acc: 0.5238 | val_loss: 1.4335 | val_acc: 0.5108 | test_acc: 0.5057 | Time: 3.1018 s
>>> Epoch [ 610/1000]
train_loss: 1.4049 | train_acc: 0.5237 | val_loss: 1.4334 | val_acc: 0.5106 | test_acc: 0.5058 | Time: 2.6544 s
>>> Epoch [ 611/1000]
train_loss: 1.4048 | train_acc: 0.5237 | val_loss: 1.4333 | val_acc: 0.5106 | test_acc: 0.5057 | Time: 2.6535 s
>>> Epoch [ 612/1000]
train_loss: 1.4047 | train_acc: 0.5237 | val_loss: 1.4333 | val_acc: 0.5108 | test_acc: 0.5057 | Time: 2.8235 s
>>> Epoch [ 613/1000]
train_loss: 1.4046 | train_acc: 0.5237 | val_loss: 1.4332 | val_acc: 0.5108 | test_acc: 0.5058 | Time: 2.9093 s
>>> Epoch [ 614/1000]
train_loss: 1.4045 | train_acc: 0.5238 | val_loss: 1.4332 | val_acc: 0.5106 | test_acc: 0.5058 | Time: 2.8975 s
>>> Epoch [ 615/1000]
train_loss: 1.4045 | train_acc: 0.5238 | val_loss: 1.4331 | val_acc: 0.5106 | test_acc: 0.5058 | Time: 3.0269 s
>>> Epoch [ 616/1000]
train_loss: 1.4044 | train_acc: 0.5239 | val_loss: 1.4330 | val_acc: 0.5106 | test_acc: 0.5056 | Time: 2.7935 s
>>> Epoch [ 617/1000]
train_loss: 1.4043 | train_acc: 0.5239 | val_loss: 1.4330 | val_acc: 0.5106 | test_acc: 0.5056 | Time: 3.0135 s
>>> Epoch [ 618/1000]
train_loss: 1.4042 | train_acc: 0.5240 | val_loss: 1.4329 | val_acc: 0.5108 | test_acc: 0.5058 | Time: 2.9932 s
>>> Epoch [ 619/1000]
train_loss: 1.4041 | train_acc: 0.5241 | val_loss: 1.4329 | val_acc: 0.5108 | test_acc: 0.5059 | Time: 2.8523 s
>>> Epoch [ 620/1000]
train_loss: 1.4040 | train_acc: 0.5242 | val_loss: 1.4328 | val_acc: 0.5108 | test_acc: 0.5059 | Time: 2.8109 s
>>> Epoch [ 621/1000]
train_loss: 1.4039 | train_acc: 0.5242 | val_loss: 1.4327 | val_acc: 0.5110 | test_acc: 0.5059 | Time: 3.0304 s
>>> Epoch [ 622/1000]
train_loss: 1.4038 | train_acc: 0.5241 | val_loss: 1.4327 | val_acc: 0.5110 | test_acc: 0.5058 | Time: 2.8822 s
>>> Epoch [ 623/1000]
train_loss: 1.4037 | train_acc: 0.5242 | val_loss: 1.4326 | val_acc: 0.5110 | test_acc: 0.5058 | Time: 2.9880 s
>>> Epoch [ 624/1000]
train_loss: 1.4036 | train_acc: 0.5242 | val_loss: 1.4326 | val_acc: 0.5112 | test_acc: 0.5057 | Time: 2.7051 s
>>> Epoch [ 625/1000]
train_loss: 1.4035 | train_acc: 0.5243 | val_loss: 1.4325 | val_acc: 0.5112 | test_acc: 0.5058 | Time: 2.7533 s
>>> Epoch [ 626/1000]
train_loss: 1.4034 | train_acc: 0.5243 | val_loss: 1.4325 | val_acc: 0.5112 | test_acc: 0.5059 | Time: 2.7766 s
>>> Epoch [ 627/1000]
train_loss: 1.4033 | train_acc: 0.5242 | val_loss: 1.4324 | val_acc: 0.5112 | test_acc: 0.5059 | Time: 2.8882 s
>>> Epoch [ 628/1000]
train_loss: 1.4033 | train_acc: 0.5243 | val_loss: 1.4323 | val_acc: 0.5112 | test_acc: 0.5059 | Time: 2.8752 s
>>> Epoch [ 629/1000]
train_loss: 1.4032 | train_acc: 0.5243 | val_loss: 1.4323 | val_acc: 0.5112 | test_acc: 0.5059 | Time: 2.6107 s
>>> Epoch [ 630/1000]
train_loss: 1.4031 | train_acc: 0.5244 | val_loss: 1.4322 | val_acc: 0.5112 | test_acc: 0.5057 | Time: 2.7618 s
>>> Epoch [ 631/1000]
train_loss: 1.4030 | train_acc: 0.5244 | val_loss: 1.4322 | val_acc: 0.5112 | test_acc: 0.5058 | Time: 2.6630 s
>>> Epoch [ 632/1000]
train_loss: 1.4029 | train_acc: 0.5244 | val_loss: 1.4321 | val_acc: 0.5112 | test_acc: 0.5058 | Time: 2.5934 s
>>> Epoch [ 633/1000]
train_loss: 1.4028 | train_acc: 0.5245 | val_loss: 1.4320 | val_acc: 0.5112 | test_acc: 0.5058 | Time: 2.3866 s
>>> Epoch [ 634/1000]
train_loss: 1.4027 | train_acc: 0.5245 | val_loss: 1.4320 | val_acc: 0.5112 | test_acc: 0.5058 | Time: 2.3134 s
>>> Epoch [ 635/1000]
train_loss: 1.4026 | train_acc: 0.5246 | val_loss: 1.4319 | val_acc: 0.5112 | test_acc: 0.5058 | Time: 2.4745 s
>>> Epoch [ 636/1000]
train_loss: 1.4025 | train_acc: 0.5246 | val_loss: 1.4319 | val_acc: 0.5112 | test_acc: 0.5058 | Time: 2.4582 s
>>> Epoch [ 637/1000]
train_loss: 1.4024 | train_acc: 0.5245 | val_loss: 1.4318 | val_acc: 0.5112 | test_acc: 0.5059 | Time: 2.5070 s
>>> Epoch [ 638/1000]
train_loss: 1.4024 | train_acc: 0.5245 | val_loss: 1.4318 | val_acc: 0.5112 | test_acc: 0.5059 | Time: 2.5172 s
>>> Epoch [ 639/1000]
train_loss: 1.4023 | train_acc: 0.5246 | val_loss: 1.4317 | val_acc: 0.5112 | test_acc: 0.5060 | Time: 2.5465 s
>>> Epoch [ 640/1000]
train_loss: 1.4022 | train_acc: 0.5246 | val_loss: 1.4317 | val_acc: 0.5112 | test_acc: 0.5060 | Time: 2.1917 s
>>> Epoch [ 641/1000]
train_loss: 1.4021 | train_acc: 0.5246 | val_loss: 1.4316 | val_acc: 0.5114 | test_acc: 0.5060 | Time: 2.4200 s
>>> Epoch [ 642/1000]
train_loss: 1.4020 | train_acc: 0.5246 | val_loss: 1.4315 | val_acc: 0.5114 | test_acc: 0.5060 | Time: 2.6322 s
>>> Epoch [ 643/1000]
train_loss: 1.4019 | train_acc: 0.5246 | val_loss: 1.4315 | val_acc: 0.5114 | test_acc: 0.5060 | Time: 2.5997 s
>>> Epoch [ 644/1000]
train_loss: 1.4018 | train_acc: 0.5247 | val_loss: 1.4314 | val_acc: 0.5114 | test_acc: 0.5061 | Time: 2.6109 s
>>> Epoch [ 645/1000]
train_loss: 1.4017 | train_acc: 0.5247 | val_loss: 1.4314 | val_acc: 0.5112 | test_acc: 0.5061 | Time: 2.2952 s
>>> Epoch [ 646/1000]
train_loss: 1.4016 | train_acc: 0.5247 | val_loss: 1.4313 | val_acc: 0.5112 | test_acc: 0.5062 | Time: 2.6711 s
>>> Epoch [ 647/1000]
train_loss: 1.4016 | train_acc: 0.5247 | val_loss: 1.4313 | val_acc: 0.5110 | test_acc: 0.5062 | Time: 2.2971 s
>>> Epoch [ 648/1000]
train_loss: 1.4015 | train_acc: 0.5246 | val_loss: 1.4312 | val_acc: 0.5110 | test_acc: 0.5063 | Time: 2.2083 s
>>> Epoch [ 649/1000]
train_loss: 1.4014 | train_acc: 0.5247 | val_loss: 1.4311 | val_acc: 0.5112 | test_acc: 0.5062 | Time: 2.4922 s
>>> Epoch [ 650/1000]
train_loss: 1.4013 | train_acc: 0.5247 | val_loss: 1.4311 | val_acc: 0.5112 | test_acc: 0.5061 | Time: 2.2587 s
>>> Epoch [ 651/1000]
train_loss: 1.4012 | train_acc: 0.5247 | val_loss: 1.4310 | val_acc: 0.5112 | test_acc: 0.5062 | Time: 2.3123 s
>>> Epoch [ 652/1000]
train_loss: 1.4011 | train_acc: 0.5247 | val_loss: 1.4310 | val_acc: 0.5112 | test_acc: 0.5062 | Time: 2.6198 s
>>> Epoch [ 653/1000]
train_loss: 1.4010 | train_acc: 0.5248 | val_loss: 1.4309 | val_acc: 0.5112 | test_acc: 0.5063 | Time: 2.6159 s
>>> Epoch [ 654/1000]
train_loss: 1.4009 | train_acc: 0.5248 | val_loss: 1.4309 | val_acc: 0.5112 | test_acc: 0.5064 | Time: 2.5510 s
>>> Epoch [ 655/1000]
train_loss: 1.4009 | train_acc: 0.5248 | val_loss: 1.4308 | val_acc: 0.5112 | test_acc: 0.5064 | Time: 2.5634 s
>>> Epoch [ 656/1000]
train_loss: 1.4008 | train_acc: 0.5248 | val_loss: 1.4308 | val_acc: 0.5112 | test_acc: 0.5065 | Time: 2.1222 s
>>> Epoch [ 657/1000]
train_loss: 1.4007 | train_acc: 0.5248 | val_loss: 1.4307 | val_acc: 0.5112 | test_acc: 0.5065 | Time: 2.4833 s
>>> Epoch [ 658/1000]
train_loss: 1.4006 | train_acc: 0.5248 | val_loss: 1.4307 | val_acc: 0.5112 | test_acc: 0.5064 | Time: 2.5473 s
>>> Epoch [ 659/1000]
train_loss: 1.4005 | train_acc: 0.5248 | val_loss: 1.4306 | val_acc: 0.5112 | test_acc: 0.5064 | Time: 2.3729 s
>>> Epoch [ 660/1000]
train_loss: 1.4004 | train_acc: 0.5249 | val_loss: 1.4305 | val_acc: 0.5112 | test_acc: 0.5065 | Time: 2.4296 s
>>> Epoch [ 661/1000]
train_loss: 1.4003 | train_acc: 0.5249 | val_loss: 1.4305 | val_acc: 0.5112 | test_acc: 0.5066 | Time: 2.6101 s
>>> Epoch [ 662/1000]
train_loss: 1.4002 | train_acc: 0.5249 | val_loss: 1.4304 | val_acc: 0.5112 | test_acc: 0.5066 | Time: 2.5430 s
>>> Epoch [ 663/1000]
train_loss: 1.4002 | train_acc: 0.5249 | val_loss: 1.4304 | val_acc: 0.5114 | test_acc: 0.5069 | Time: 2.5649 s
>>> Epoch [ 664/1000]
train_loss: 1.4001 | train_acc: 0.5249 | val_loss: 1.4303 | val_acc: 0.5114 | test_acc: 0.5069 | Time: 2.7339 s
>>> Epoch [ 665/1000]
train_loss: 1.4000 | train_acc: 0.5250 | val_loss: 1.4303 | val_acc: 0.5112 | test_acc: 0.5070 | Time: 2.3229 s
>>> Epoch [ 666/1000]
train_loss: 1.3999 | train_acc: 0.5250 | val_loss: 1.4302 | val_acc: 0.5112 | test_acc: 0.5071 | Time: 2.4977 s
>>> Epoch [ 667/1000]
train_loss: 1.3998 | train_acc: 0.5250 | val_loss: 1.4302 | val_acc: 0.5112 | test_acc: 0.5071 | Time: 2.2709 s
>>> Epoch [ 668/1000]
train_loss: 1.3997 | train_acc: 0.5250 | val_loss: 1.4301 | val_acc: 0.5112 | test_acc: 0.5071 | Time: 2.3154 s
>>> Epoch [ 669/1000]
train_loss: 1.3996 | train_acc: 0.5251 | val_loss: 1.4301 | val_acc: 0.5112 | test_acc: 0.5071 | Time: 2.2986 s
>>> Epoch [ 670/1000]
train_loss: 1.3996 | train_acc: 0.5252 | val_loss: 1.4300 | val_acc: 0.5110 | test_acc: 0.5071 | Time: 2.5538 s
>>> Epoch [ 671/1000]
train_loss: 1.3995 | train_acc: 0.5253 | val_loss: 1.4300 | val_acc: 0.5110 | test_acc: 0.5072 | Time: 2.6699 s
>>> Epoch [ 672/1000]
train_loss: 1.3994 | train_acc: 0.5253 | val_loss: 1.4299 | val_acc: 0.5114 | test_acc: 0.5072 | Time: 2.6010 s
>>> Epoch [ 673/1000]
train_loss: 1.3993 | train_acc: 0.5254 | val_loss: 1.4298 | val_acc: 0.5116 | test_acc: 0.5072 | Time: 2.5206 s
>>> Epoch [ 674/1000]
train_loss: 1.3992 | train_acc: 0.5253 | val_loss: 1.4298 | val_acc: 0.5114 | test_acc: 0.5075 | Time: 2.6932 s
>>> Epoch [ 675/1000]
train_loss: 1.3991 | train_acc: 0.5254 | val_loss: 1.4297 | val_acc: 0.5114 | test_acc: 0.5075 | Time: 2.4777 s
>>> Epoch [ 676/1000]
train_loss: 1.3991 | train_acc: 0.5253 | val_loss: 1.4297 | val_acc: 0.5116 | test_acc: 0.5074 | Time: 2.7529 s
>>> Epoch [ 677/1000]
train_loss: 1.3990 | train_acc: 0.5254 | val_loss: 1.4296 | val_acc: 0.5116 | test_acc: 0.5074 | Time: 2.7102 s
>>> Epoch [ 678/1000]
train_loss: 1.3989 | train_acc: 0.5254 | val_loss: 1.4296 | val_acc: 0.5116 | test_acc: 0.5075 | Time: 2.6427 s
>>> Epoch [ 679/1000]
train_loss: 1.3988 | train_acc: 0.5254 | val_loss: 1.4295 | val_acc: 0.5116 | test_acc: 0.5076 | Time: 2.2773 s
>>> Epoch [ 680/1000]
train_loss: 1.3987 | train_acc: 0.5255 | val_loss: 1.4295 | val_acc: 0.5118 | test_acc: 0.5076 | Time: 2.3118 s
>>> Epoch [ 681/1000]
train_loss: 1.3986 | train_acc: 0.5255 | val_loss: 1.4294 | val_acc: 0.5118 | test_acc: 0.5076 | Time: 2.4259 s
>>> Epoch [ 682/1000]
train_loss: 1.3986 | train_acc: 0.5255 | val_loss: 1.4294 | val_acc: 0.5118 | test_acc: 0.5076 | Time: 2.4732 s
>>> Epoch [ 683/1000]
train_loss: 1.3985 | train_acc: 0.5255 | val_loss: 1.4293 | val_acc: 0.5118 | test_acc: 0.5076 | Time: 2.4352 s
>>> Epoch [ 684/1000]
train_loss: 1.3984 | train_acc: 0.5256 | val_loss: 1.4293 | val_acc: 0.5118 | test_acc: 0.5077 | Time: 2.4974 s
>>> Epoch [ 685/1000]
train_loss: 1.3983 | train_acc: 0.5256 | val_loss: 1.4292 | val_acc: 0.5120 | test_acc: 0.5076 | Time: 2.7193 s
>>> Epoch [ 686/1000]
train_loss: 1.3982 | train_acc: 0.5256 | val_loss: 1.4292 | val_acc: 0.5120 | test_acc: 0.5075 | Time: 2.5667 s
>>> Epoch [ 687/1000]
train_loss: 1.3981 | train_acc: 0.5257 | val_loss: 1.4291 | val_acc: 0.5118 | test_acc: 0.5075 | Time: 2.4385 s
>>> Epoch [ 688/1000]
train_loss: 1.3981 | train_acc: 0.5257 | val_loss: 1.4291 | val_acc: 0.5118 | test_acc: 0.5075 | Time: 2.3199 s
>>> Epoch [ 689/1000]
train_loss: 1.3980 | train_acc: 0.5257 | val_loss: 1.4290 | val_acc: 0.5116 | test_acc: 0.5074 | Time: 2.2352 s
>>> Epoch [ 690/1000]
train_loss: 1.3979 | train_acc: 0.5258 | val_loss: 1.4290 | val_acc: 0.5116 | test_acc: 0.5073 | Time: 2.4361 s
>>> Epoch [ 691/1000]
train_loss: 1.3978 | train_acc: 0.5258 | val_loss: 1.4289 | val_acc: 0.5114 | test_acc: 0.5073 | Time: 2.3542 s
>>> Epoch [ 692/1000]
train_loss: 1.3977 | train_acc: 0.5258 | val_loss: 1.4289 | val_acc: 0.5114 | test_acc: 0.5074 | Time: 2.3838 s
>>> Epoch [ 693/1000]
train_loss: 1.3976 | train_acc: 0.5259 | val_loss: 1.4288 | val_acc: 0.5112 | test_acc: 0.5074 | Time: 2.6274 s
>>> Epoch [ 694/1000]
train_loss: 1.3976 | train_acc: 0.5259 | val_loss: 1.4288 | val_acc: 0.5114 | test_acc: 0.5073 | Time: 2.8260 s
>>> Epoch [ 695/1000]
train_loss: 1.3975 | train_acc: 0.5258 | val_loss: 1.4287 | val_acc: 0.5116 | test_acc: 0.5074 | Time: 2.5566 s
>>> Epoch [ 696/1000]
train_loss: 1.3974 | train_acc: 0.5258 | val_loss: 1.4287 | val_acc: 0.5116 | test_acc: 0.5073 | Time: 2.6325 s
>>> Epoch [ 697/1000]
train_loss: 1.3973 | train_acc: 0.5258 | val_loss: 1.4286 | val_acc: 0.5116 | test_acc: 0.5074 | Time: 2.5964 s
>>> Epoch [ 698/1000]
train_loss: 1.3972 | train_acc: 0.5259 | val_loss: 1.4286 | val_acc: 0.5116 | test_acc: 0.5074 | Time: 2.7208 s
>>> Epoch [ 699/1000]
train_loss: 1.3971 | train_acc: 0.5259 | val_loss: 1.4285 | val_acc: 0.5118 | test_acc: 0.5074 | Time: 2.2744 s
>>> Epoch [ 700/1000]
train_loss: 1.3971 | train_acc: 0.5259 | val_loss: 1.4285 | val_acc: 0.5118 | test_acc: 0.5072 | Time: 2.4218 s
>>> Epoch [ 701/1000]
train_loss: 1.3970 | train_acc: 0.5259 | val_loss: 1.4284 | val_acc: 0.5120 | test_acc: 0.5072 | Time: 2.4052 s
>>> Epoch [ 702/1000]
train_loss: 1.3969 | train_acc: 0.5260 | val_loss: 1.4284 | val_acc: 0.5120 | test_acc: 0.5072 | Time: 2.4151 s
>>> Epoch [ 703/1000]
train_loss: 1.3968 | train_acc: 0.5260 | val_loss: 1.4283 | val_acc: 0.5120 | test_acc: 0.5072 | Time: 2.4457 s
>>> Epoch [ 704/1000]
train_loss: 1.3967 | train_acc: 0.5261 | val_loss: 1.4283 | val_acc: 0.5120 | test_acc: 0.5071 | Time: 2.3799 s
>>> Epoch [ 705/1000]
train_loss: 1.3967 | train_acc: 0.5262 | val_loss: 1.4282 | val_acc: 0.5120 | test_acc: 0.5071 | Time: 2.4702 s
>>> Epoch [ 706/1000]
train_loss: 1.3966 | train_acc: 0.5262 | val_loss: 1.4282 | val_acc: 0.5122 | test_acc: 0.5071 | Time: 2.6376 s
>>> Epoch [ 707/1000]
train_loss: 1.3965 | train_acc: 0.5263 | val_loss: 1.4281 | val_acc: 0.5122 | test_acc: 0.5072 | Time: 2.7242 s
>>> Epoch [ 708/1000]
train_loss: 1.3964 | train_acc: 0.5262 | val_loss: 1.4281 | val_acc: 0.5122 | test_acc: 0.5072 | Time: 2.6998 s
>>> Epoch [ 709/1000]
train_loss: 1.3963 | train_acc: 0.5263 | val_loss: 1.4280 | val_acc: 0.5122 | test_acc: 0.5072 | Time: 2.7023 s
>>> Epoch [ 710/1000]
train_loss: 1.3963 | train_acc: 0.5263 | val_loss: 1.4280 | val_acc: 0.5122 | test_acc: 0.5072 | Time: 2.5783 s
>>> Epoch [ 711/1000]
train_loss: 1.3962 | train_acc: 0.5264 | val_loss: 1.4279 | val_acc: 0.5124 | test_acc: 0.5072 | Time: 2.6327 s
>>> Epoch [ 712/1000]
train_loss: 1.3961 | train_acc: 0.5264 | val_loss: 1.4279 | val_acc: 0.5124 | test_acc: 0.5072 | Time: 2.6836 s
>>> Epoch [ 713/1000]
train_loss: 1.3960 | train_acc: 0.5265 | val_loss: 1.4278 | val_acc: 0.5124 | test_acc: 0.5072 | Time: 2.7467 s
>>> Epoch [ 714/1000]
train_loss: 1.3959 | train_acc: 0.5265 | val_loss: 1.4278 | val_acc: 0.5124 | test_acc: 0.5073 | Time: 2.7785 s
>>> Epoch [ 715/1000]
train_loss: 1.3959 | train_acc: 0.5264 | val_loss: 1.4277 | val_acc: 0.5124 | test_acc: 0.5073 | Time: 2.5651 s
>>> Epoch [ 716/1000]
train_loss: 1.3958 | train_acc: 0.5264 | val_loss: 1.4277 | val_acc: 0.5126 | test_acc: 0.5073 | Time: 2.6489 s
>>> Epoch [ 717/1000]
train_loss: 1.3957 | train_acc: 0.5264 | val_loss: 1.4276 | val_acc: 0.5128 | test_acc: 0.5075 | Time: 2.5907 s
>>> Epoch [ 718/1000]
train_loss: 1.3956 | train_acc: 0.5264 | val_loss: 1.4276 | val_acc: 0.5128 | test_acc: 0.5075 | Time: 2.8546 s
>>> Epoch [ 719/1000]
train_loss: 1.3955 | train_acc: 0.5264 | val_loss: 1.4275 | val_acc: 0.5130 | test_acc: 0.5076 | Time: 2.6765 s
>>> Epoch [ 720/1000]
train_loss: 1.3955 | train_acc: 0.5264 | val_loss: 1.4275 | val_acc: 0.5132 | test_acc: 0.5076 | Time: 2.4922 s
>>> Epoch [ 721/1000]
train_loss: 1.3954 | train_acc: 0.5264 | val_loss: 1.4274 | val_acc: 0.5132 | test_acc: 0.5075 | Time: 2.6372 s
>>> Epoch [ 722/1000]
train_loss: 1.3953 | train_acc: 0.5264 | val_loss: 1.4274 | val_acc: 0.5134 | test_acc: 0.5076 | Time: 2.6191 s
>>> Epoch [ 723/1000]
train_loss: 1.3952 | train_acc: 0.5265 | val_loss: 1.4273 | val_acc: 0.5136 | test_acc: 0.5076 | Time: 2.4978 s
>>> Epoch [ 724/1000]
train_loss: 1.3951 | train_acc: 0.5266 | val_loss: 1.4273 | val_acc: 0.5136 | test_acc: 0.5076 | Time: 2.3405 s
>>> Epoch [ 725/1000]
train_loss: 1.3951 | train_acc: 0.5266 | val_loss: 1.4272 | val_acc: 0.5136 | test_acc: 0.5076 | Time: 2.6832 s
>>> Epoch [ 726/1000]
train_loss: 1.3950 | train_acc: 0.5266 | val_loss: 1.4272 | val_acc: 0.5138 | test_acc: 0.5077 | Time: 2.4620 s
>>> Epoch [ 727/1000]
train_loss: 1.3949 | train_acc: 0.5266 | val_loss: 1.4271 | val_acc: 0.5140 | test_acc: 0.5078 | Time: 2.5293 s
>>> Epoch [ 728/1000]
train_loss: 1.3948 | train_acc: 0.5266 | val_loss: 1.4271 | val_acc: 0.5142 | test_acc: 0.5078 | Time: 2.5208 s
>>> Epoch [ 729/1000]
train_loss: 1.3948 | train_acc: 0.5265 | val_loss: 1.4270 | val_acc: 0.5144 | test_acc: 0.5079 | Time: 2.6161 s
>>> Epoch [ 730/1000]
train_loss: 1.3947 | train_acc: 0.5266 | val_loss: 1.4270 | val_acc: 0.5144 | test_acc: 0.5080 | Time: 2.4275 s
>>> Epoch [ 731/1000]
train_loss: 1.3946 | train_acc: 0.5266 | val_loss: 1.4269 | val_acc: 0.5144 | test_acc: 0.5081 | Time: 2.8258 s
>>> Epoch [ 732/1000]
train_loss: 1.3945 | train_acc: 0.5265 | val_loss: 1.4269 | val_acc: 0.5144 | test_acc: 0.5081 | Time: 2.9498 s
>>> Epoch [ 733/1000]
train_loss: 1.3944 | train_acc: 0.5266 | val_loss: 1.4268 | val_acc: 0.5144 | test_acc: 0.5081 | Time: 2.8196 s
>>> Epoch [ 734/1000]
train_loss: 1.3944 | train_acc: 0.5267 | val_loss: 1.4268 | val_acc: 0.5144 | test_acc: 0.5083 | Time: 2.7711 s
>>> Epoch [ 735/1000]
train_loss: 1.3943 | train_acc: 0.5266 | val_loss: 1.4268 | val_acc: 0.5142 | test_acc: 0.5083 | Time: 2.8288 s
>>> Epoch [ 736/1000]
train_loss: 1.3942 | train_acc: 0.5266 | val_loss: 1.4267 | val_acc: 0.5144 | test_acc: 0.5084 | Time: 2.6906 s
>>> Epoch [ 737/1000]
train_loss: 1.3941 | train_acc: 0.5265 | val_loss: 1.4267 | val_acc: 0.5142 | test_acc: 0.5084 | Time: 2.9329 s
>>> Epoch [ 738/1000]
train_loss: 1.3941 | train_acc: 0.5266 | val_loss: 1.4266 | val_acc: 0.5142 | test_acc: 0.5084 | Time: 2.7265 s
>>> Epoch [ 739/1000]
train_loss: 1.3940 | train_acc: 0.5267 | val_loss: 1.4266 | val_acc: 0.5142 | test_acc: 0.5085 | Time: 2.5761 s
>>> Epoch [ 740/1000]
train_loss: 1.3939 | train_acc: 0.5267 | val_loss: 1.4265 | val_acc: 0.5142 | test_acc: 0.5085 | Time: 2.6334 s
>>> Epoch [ 741/1000]
train_loss: 1.3938 | train_acc: 0.5268 | val_loss: 1.4265 | val_acc: 0.5142 | test_acc: 0.5085 | Time: 2.8396 s
>>> Epoch [ 742/1000]
train_loss: 1.3938 | train_acc: 0.5268 | val_loss: 1.4264 | val_acc: 0.5142 | test_acc: 0.5085 | Time: 2.7003 s
>>> Epoch [ 743/1000]
train_loss: 1.3937 | train_acc: 0.5268 | val_loss: 1.4264 | val_acc: 0.5142 | test_acc: 0.5084 | Time: 2.8621 s
>>> Epoch [ 744/1000]
train_loss: 1.3936 | train_acc: 0.5268 | val_loss: 1.4263 | val_acc: 0.5142 | test_acc: 0.5084 | Time: 2.6849 s
>>> Epoch [ 745/1000]
train_loss: 1.3935 | train_acc: 0.5269 | val_loss: 1.4263 | val_acc: 0.5142 | test_acc: 0.5083 | Time: 2.6732 s
>>> Epoch [ 746/1000]
train_loss: 1.3934 | train_acc: 0.5269 | val_loss: 1.4262 | val_acc: 0.5144 | test_acc: 0.5083 | Time: 2.7495 s
>>> Epoch [ 747/1000]
train_loss: 1.3934 | train_acc: 0.5269 | val_loss: 1.4262 | val_acc: 0.5144 | test_acc: 0.5084 | Time: 2.6106 s
>>> Epoch [ 748/1000]
train_loss: 1.3933 | train_acc: 0.5269 | val_loss: 1.4261 | val_acc: 0.5144 | test_acc: 0.5083 | Time: 2.7903 s
>>> Epoch [ 749/1000]
train_loss: 1.3932 | train_acc: 0.5269 | val_loss: 1.4261 | val_acc: 0.5146 | test_acc: 0.5083 | Time: 2.7545 s
>>> Epoch [ 750/1000]
train_loss: 1.3931 | train_acc: 0.5269 | val_loss: 1.4261 | val_acc: 0.5146 | test_acc: 0.5084 | Time: 2.6036 s
>>> Epoch [ 751/1000]
train_loss: 1.3931 | train_acc: 0.5270 | val_loss: 1.4260 | val_acc: 0.5146 | test_acc: 0.5085 | Time: 2.6638 s
>>> Epoch [ 752/1000]
train_loss: 1.3930 | train_acc: 0.5270 | val_loss: 1.4260 | val_acc: 0.5146 | test_acc: 0.5085 | Time: 2.9192 s
>>> Epoch [ 753/1000]
train_loss: 1.3929 | train_acc: 0.5271 | val_loss: 1.4259 | val_acc: 0.5148 | test_acc: 0.5085 | Time: 2.7386 s
>>> Epoch [ 754/1000]
train_loss: 1.3928 | train_acc: 0.5271 | val_loss: 1.4259 | val_acc: 0.5148 | test_acc: 0.5085 | Time: 2.7780 s
>>> Epoch [ 755/1000]
train_loss: 1.3928 | train_acc: 0.5271 | val_loss: 1.4258 | val_acc: 0.5148 | test_acc: 0.5083 | Time: 2.9508 s
>>> Epoch [ 756/1000]
train_loss: 1.3927 | train_acc: 0.5272 | val_loss: 1.4258 | val_acc: 0.5148 | test_acc: 0.5083 | Time: 2.9250 s
>>> Epoch [ 757/1000]
train_loss: 1.3926 | train_acc: 0.5272 | val_loss: 1.4257 | val_acc: 0.5148 | test_acc: 0.5083 | Time: 2.6949 s
>>> Epoch [ 758/1000]
train_loss: 1.3925 | train_acc: 0.5273 | val_loss: 1.4257 | val_acc: 0.5148 | test_acc: 0.5083 | Time: 2.9116 s
>>> Epoch [ 759/1000]
train_loss: 1.3925 | train_acc: 0.5273 | val_loss: 1.4256 | val_acc: 0.5148 | test_acc: 0.5082 | Time: 2.8253 s
>>> Epoch [ 760/1000]
train_loss: 1.3924 | train_acc: 0.5274 | val_loss: 1.4256 | val_acc: 0.5148 | test_acc: 0.5082 | Time: 2.8045 s
>>> Epoch [ 761/1000]
train_loss: 1.3923 | train_acc: 0.5274 | val_loss: 1.4256 | val_acc: 0.5148 | test_acc: 0.5082 | Time: 2.7230 s
>>> Epoch [ 762/1000]
train_loss: 1.3922 | train_acc: 0.5274 | val_loss: 1.4255 | val_acc: 0.5148 | test_acc: 0.5082 | Time: 2.7303 s
>>> Epoch [ 763/1000]
train_loss: 1.3922 | train_acc: 0.5275 | val_loss: 1.4255 | val_acc: 0.5148 | test_acc: 0.5082 | Time: 2.7852 s
>>> Epoch [ 764/1000]
train_loss: 1.3921 | train_acc: 0.5275 | val_loss: 1.4254 | val_acc: 0.5148 | test_acc: 0.5082 | Time: 2.8626 s
>>> Epoch [ 765/1000]
train_loss: 1.3920 | train_acc: 0.5275 | val_loss: 1.4254 | val_acc: 0.5148 | test_acc: 0.5082 | Time: 2.7940 s
>>> Epoch [ 766/1000]
train_loss: 1.3919 | train_acc: 0.5275 | val_loss: 1.4253 | val_acc: 0.5146 | test_acc: 0.5082 | Time: 2.7936 s
>>> Epoch [ 767/1000]
train_loss: 1.3919 | train_acc: 0.5276 | val_loss: 1.4253 | val_acc: 0.5146 | test_acc: 0.5083 | Time: 2.7404 s
>>> Epoch [ 768/1000]
train_loss: 1.3918 | train_acc: 0.5276 | val_loss: 1.4252 | val_acc: 0.5148 | test_acc: 0.5083 | Time: 2.6965 s
>>> Epoch [ 769/1000]
train_loss: 1.3917 | train_acc: 0.5275 | val_loss: 1.4252 | val_acc: 0.5148 | test_acc: 0.5083 | Time: 2.6309 s
>>> Epoch [ 770/1000]
train_loss: 1.3916 | train_acc: 0.5276 | val_loss: 1.4252 | val_acc: 0.5146 | test_acc: 0.5082 | Time: 2.6540 s
>>> Epoch [ 771/1000]
train_loss: 1.3916 | train_acc: 0.5277 | val_loss: 1.4251 | val_acc: 0.5146 | test_acc: 0.5083 | Time: 2.8277 s
>>> Epoch [ 772/1000]
train_loss: 1.3915 | train_acc: 0.5278 | val_loss: 1.4251 | val_acc: 0.5146 | test_acc: 0.5084 | Time: 2.7695 s
>>> Epoch [ 773/1000]
train_loss: 1.3914 | train_acc: 0.5278 | val_loss: 1.4250 | val_acc: 0.5144 | test_acc: 0.5085 | Time: 2.7054 s
>>> Epoch [ 774/1000]
train_loss: 1.3914 | train_acc: 0.5279 | val_loss: 1.4250 | val_acc: 0.5146 | test_acc: 0.5084 | Time: 2.8579 s
>>> Epoch [ 775/1000]
train_loss: 1.3913 | train_acc: 0.5279 | val_loss: 1.4249 | val_acc: 0.5146 | test_acc: 0.5084 | Time: 2.8307 s
>>> Epoch [ 776/1000]
train_loss: 1.3912 | train_acc: 0.5280 | val_loss: 1.4249 | val_acc: 0.5146 | test_acc: 0.5084 | Time: 2.6375 s
>>> Epoch [ 777/1000]
train_loss: 1.3911 | train_acc: 0.5280 | val_loss: 1.4248 | val_acc: 0.5146 | test_acc: 0.5084 | Time: 2.6783 s
>>> Epoch [ 778/1000]
train_loss: 1.3911 | train_acc: 0.5280 | val_loss: 1.4248 | val_acc: 0.5148 | test_acc: 0.5084 | Time: 2.7380 s
>>> Epoch [ 779/1000]
train_loss: 1.3910 | train_acc: 0.5281 | val_loss: 1.4248 | val_acc: 0.5146 | test_acc: 0.5084 | Time: 2.7877 s
>>> Epoch [ 780/1000]
train_loss: 1.3909 | train_acc: 0.5281 | val_loss: 1.4247 | val_acc: 0.5148 | test_acc: 0.5083 | Time: 2.8233 s
>>> Epoch [ 781/1000]
train_loss: 1.3908 | train_acc: 0.5282 | val_loss: 1.4247 | val_acc: 0.5148 | test_acc: 0.5083 | Time: 2.7288 s
>>> Epoch [ 782/1000]
train_loss: 1.3908 | train_acc: 0.5283 | val_loss: 1.4246 | val_acc: 0.5148 | test_acc: 0.5083 | Time: 2.6222 s
>>> Epoch [ 783/1000]
train_loss: 1.3907 | train_acc: 0.5284 | val_loss: 1.4246 | val_acc: 0.5148 | test_acc: 0.5083 | Time: 2.7331 s
>>> Epoch [ 784/1000]
train_loss: 1.3906 | train_acc: 0.5284 | val_loss: 1.4245 | val_acc: 0.5150 | test_acc: 0.5083 | Time: 2.6877 s
>>> Epoch [ 785/1000]
train_loss: 1.3906 | train_acc: 0.5285 | val_loss: 1.4245 | val_acc: 0.5148 | test_acc: 0.5083 | Time: 2.6791 s
>>> Epoch [ 786/1000]
train_loss: 1.3905 | train_acc: 0.5285 | val_loss: 1.4245 | val_acc: 0.5150 | test_acc: 0.5084 | Time: 2.6294 s
>>> Epoch [ 787/1000]
train_loss: 1.3904 | train_acc: 0.5286 | val_loss: 1.4244 | val_acc: 0.5150 | test_acc: 0.5084 | Time: 2.7849 s
>>> Epoch [ 788/1000]
train_loss: 1.3903 | train_acc: 0.5286 | val_loss: 1.4244 | val_acc: 0.5150 | test_acc: 0.5086 | Time: 2.6915 s
>>> Epoch [ 789/1000]
train_loss: 1.3903 | train_acc: 0.5286 | val_loss: 1.4243 | val_acc: 0.5154 | test_acc: 0.5086 | Time: 2.7698 s
>>> Epoch [ 790/1000]
train_loss: 1.3902 | train_acc: 0.5286 | val_loss: 1.4243 | val_acc: 0.5154 | test_acc: 0.5087 | Time: 2.5927 s
>>> Epoch [ 791/1000]
train_loss: 1.3901 | train_acc: 0.5287 | val_loss: 1.4242 | val_acc: 0.5154 | test_acc: 0.5087 | Time: 2.7010 s
>>> Epoch [ 792/1000]
train_loss: 1.3901 | train_acc: 0.5287 | val_loss: 1.4242 | val_acc: 0.5154 | test_acc: 0.5087 | Time: 2.7643 s
>>> Epoch [ 793/1000]
train_loss: 1.3900 | train_acc: 0.5287 | val_loss: 1.4242 | val_acc: 0.5154 | test_acc: 0.5088 | Time: 2.7868 s
>>> Epoch [ 794/1000]
train_loss: 1.3899 | train_acc: 0.5288 | val_loss: 1.4241 | val_acc: 0.5156 | test_acc: 0.5088 | Time: 2.4301 s
>>> Epoch [ 795/1000]
train_loss: 1.3898 | train_acc: 0.5288 | val_loss: 1.4241 | val_acc: 0.5158 | test_acc: 0.5087 | Time: 2.4601 s
>>> Epoch [ 796/1000]
train_loss: 1.3898 | train_acc: 0.5288 | val_loss: 1.4240 | val_acc: 0.5160 | test_acc: 0.5087 | Time: 2.4727 s
>>> Epoch [ 797/1000]
train_loss: 1.3897 | train_acc: 0.5288 | val_loss: 1.4240 | val_acc: 0.5160 | test_acc: 0.5088 | Time: 2.6354 s
>>> Epoch [ 798/1000]
train_loss: 1.3896 | train_acc: 0.5288 | val_loss: 1.4239 | val_acc: 0.5160 | test_acc: 0.5089 | Time: 2.5021 s
>>> Epoch [ 799/1000]
train_loss: 1.3896 | train_acc: 0.5287 | val_loss: 1.4239 | val_acc: 0.5160 | test_acc: 0.5090 | Time: 2.4929 s
>>> Epoch [ 800/1000]
train_loss: 1.3895 | train_acc: 0.5287 | val_loss: 1.4239 | val_acc: 0.5158 | test_acc: 0.5092 | Time: 2.5737 s
>>> Epoch [ 801/1000]
train_loss: 1.3894 | train_acc: 0.5288 | val_loss: 1.4238 | val_acc: 0.5158 | test_acc: 0.5092 | Time: 2.6297 s
>>> Epoch [ 802/1000]
train_loss: 1.3893 | train_acc: 0.5288 | val_loss: 1.4238 | val_acc: 0.5158 | test_acc: 0.5093 | Time: 2.5989 s
>>> Epoch [ 803/1000]
train_loss: 1.3893 | train_acc: 0.5288 | val_loss: 1.4237 | val_acc: 0.5160 | test_acc: 0.5095 | Time: 2.7176 s
>>> Epoch [ 804/1000]
train_loss: 1.3892 | train_acc: 0.5288 | val_loss: 1.4237 | val_acc: 0.5160 | test_acc: 0.5094 | Time: 2.8044 s
>>> Epoch [ 805/1000]
train_loss: 1.3891 | train_acc: 0.5289 | val_loss: 1.4237 | val_acc: 0.5160 | test_acc: 0.5096 | Time: 2.6260 s
>>> Epoch [ 806/1000]
train_loss: 1.3891 | train_acc: 0.5289 | val_loss: 1.4236 | val_acc: 0.5160 | test_acc: 0.5097 | Time: 2.8200 s
>>> Epoch [ 807/1000]
train_loss: 1.3890 | train_acc: 0.5289 | val_loss: 1.4236 | val_acc: 0.5160 | test_acc: 0.5099 | Time: 2.7850 s
>>> Epoch [ 808/1000]
train_loss: 1.3889 | train_acc: 0.5289 | val_loss: 1.4235 | val_acc: 0.5160 | test_acc: 0.5098 | Time: 2.7487 s
>>> Epoch [ 809/1000]
train_loss: 1.3888 | train_acc: 0.5288 | val_loss: 1.4235 | val_acc: 0.5160 | test_acc: 0.5100 | Time: 2.4090 s
>>> Epoch [ 810/1000]
train_loss: 1.3888 | train_acc: 0.5288 | val_loss: 1.4234 | val_acc: 0.5162 | test_acc: 0.5100 | Time: 2.7345 s
>>> Epoch [ 811/1000]
train_loss: 1.3887 | train_acc: 0.5289 | val_loss: 1.4234 | val_acc: 0.5162 | test_acc: 0.5100 | Time: 2.7264 s
>>> Epoch [ 812/1000]
train_loss: 1.3886 | train_acc: 0.5290 | val_loss: 1.4234 | val_acc: 0.5162 | test_acc: 0.5100 | Time: 2.7395 s
>>> Epoch [ 813/1000]
train_loss: 1.3886 | train_acc: 0.5290 | val_loss: 1.4233 | val_acc: 0.5162 | test_acc: 0.5100 | Time: 3.0580 s
>>> Epoch [ 814/1000]
train_loss: 1.3885 | train_acc: 0.5290 | val_loss: 1.4233 | val_acc: 0.5162 | test_acc: 0.5100 | Time: 2.9297 s
>>> Epoch [ 815/1000]
train_loss: 1.3884 | train_acc: 0.5290 | val_loss: 1.4232 | val_acc: 0.5162 | test_acc: 0.5100 | Time: 2.8134 s
>>> Epoch [ 816/1000]
train_loss: 1.3884 | train_acc: 0.5290 | val_loss: 1.4232 | val_acc: 0.5160 | test_acc: 0.5100 | Time: 2.9254 s
>>> Epoch [ 817/1000]
train_loss: 1.3883 | train_acc: 0.5290 | val_loss: 1.4232 | val_acc: 0.5158 | test_acc: 0.5099 | Time: 2.8712 s
>>> Epoch [ 818/1000]
train_loss: 1.3882 | train_acc: 0.5291 | val_loss: 1.4231 | val_acc: 0.5158 | test_acc: 0.5099 | Time: 2.8186 s
>>> Epoch [ 819/1000]
train_loss: 1.3881 | train_acc: 0.5291 | val_loss: 1.4231 | val_acc: 0.5160 | test_acc: 0.5099 | Time: 2.7678 s
>>> Epoch [ 820/1000]
train_loss: 1.3881 | train_acc: 0.5291 | val_loss: 1.4230 | val_acc: 0.5160 | test_acc: 0.5100 | Time: 2.9160 s
>>> Epoch [ 821/1000]
train_loss: 1.3880 | train_acc: 0.5290 | val_loss: 1.4230 | val_acc: 0.5160 | test_acc: 0.5100 | Time: 2.7148 s
>>> Epoch [ 822/1000]
train_loss: 1.3879 | train_acc: 0.5291 | val_loss: 1.4230 | val_acc: 0.5160 | test_acc: 0.5100 | Time: 2.6620 s
>>> Epoch [ 823/1000]
train_loss: 1.3879 | train_acc: 0.5291 | val_loss: 1.4229 | val_acc: 0.5160 | test_acc: 0.5101 | Time: 2.8819 s
>>> Epoch [ 824/1000]
train_loss: 1.3878 | train_acc: 0.5291 | val_loss: 1.4229 | val_acc: 0.5160 | test_acc: 0.5101 | Time: 2.8792 s
>>> Epoch [ 825/1000]
train_loss: 1.3877 | train_acc: 0.5292 | val_loss: 1.4228 | val_acc: 0.5160 | test_acc: 0.5100 | Time: 2.6159 s
>>> Epoch [ 826/1000]
train_loss: 1.3877 | train_acc: 0.5292 | val_loss: 1.4228 | val_acc: 0.5160 | test_acc: 0.5099 | Time: 2.6084 s
>>> Epoch [ 827/1000]
train_loss: 1.3876 | train_acc: 0.5291 | val_loss: 1.4228 | val_acc: 0.5162 | test_acc: 0.5099 | Time: 2.6861 s
>>> Epoch [ 828/1000]
train_loss: 1.3875 | train_acc: 0.5292 | val_loss: 1.4227 | val_acc: 0.5162 | test_acc: 0.5100 | Time: 2.6678 s
>>> Epoch [ 829/1000]
train_loss: 1.3875 | train_acc: 0.5293 | val_loss: 1.4227 | val_acc: 0.5162 | test_acc: 0.5102 | Time: 2.6722 s
>>> Epoch [ 830/1000]
train_loss: 1.3874 | train_acc: 0.5293 | val_loss: 1.4226 | val_acc: 0.5162 | test_acc: 0.5102 | Time: 2.7033 s
>>> Epoch [ 831/1000]
train_loss: 1.3873 | train_acc: 0.5293 | val_loss: 1.4226 | val_acc: 0.5162 | test_acc: 0.5103 | Time: 2.5873 s
>>> Epoch [ 832/1000]
train_loss: 1.3873 | train_acc: 0.5292 | val_loss: 1.4226 | val_acc: 0.5162 | test_acc: 0.5104 | Time: 2.5539 s
>>> Epoch [ 833/1000]
train_loss: 1.3872 | train_acc: 0.5293 | val_loss: 1.4225 | val_acc: 0.5162 | test_acc: 0.5103 | Time: 2.7090 s
>>> Epoch [ 834/1000]
train_loss: 1.3871 | train_acc: 0.5293 | val_loss: 1.4225 | val_acc: 0.5160 | test_acc: 0.5104 | Time: 2.6319 s
>>> Epoch [ 835/1000]
train_loss: 1.3871 | train_acc: 0.5293 | val_loss: 1.4224 | val_acc: 0.5160 | test_acc: 0.5104 | Time: 2.5415 s
>>> Epoch [ 836/1000]
train_loss: 1.3870 | train_acc: 0.5294 | val_loss: 1.4224 | val_acc: 0.5160 | test_acc: 0.5104 | Time: 2.6999 s
>>> Epoch [ 837/1000]
train_loss: 1.3869 | train_acc: 0.5294 | val_loss: 1.4224 | val_acc: 0.5160 | test_acc: 0.5103 | Time: 2.6253 s
>>> Epoch [ 838/1000]
train_loss: 1.3868 | train_acc: 0.5294 | val_loss: 1.4223 | val_acc: 0.5158 | test_acc: 0.5102 | Time: 2.7886 s
>>> Epoch [ 839/1000]
train_loss: 1.3868 | train_acc: 0.5294 | val_loss: 1.4223 | val_acc: 0.5156 | test_acc: 0.5102 | Time: 2.7121 s
>>> Epoch [ 840/1000]
train_loss: 1.3867 | train_acc: 0.5294 | val_loss: 1.4222 | val_acc: 0.5156 | test_acc: 0.5102 | Time: 2.9626 s
>>> Epoch [ 841/1000]
train_loss: 1.3866 | train_acc: 0.5294 | val_loss: 1.4222 | val_acc: 0.5154 | test_acc: 0.5102 | Time: 2.7632 s
>>> Epoch [ 842/1000]
train_loss: 1.3866 | train_acc: 0.5294 | val_loss: 1.4222 | val_acc: 0.5156 | test_acc: 0.5103 | Time: 3.0745 s
>>> Epoch [ 843/1000]
train_loss: 1.3865 | train_acc: 0.5294 | val_loss: 1.4221 | val_acc: 0.5156 | test_acc: 0.5103 | Time: 2.9616 s
>>> Epoch [ 844/1000]
train_loss: 1.3864 | train_acc: 0.5294 | val_loss: 1.4221 | val_acc: 0.5156 | test_acc: 0.5103 | Time: 2.8519 s
>>> Epoch [ 845/1000]
train_loss: 1.3864 | train_acc: 0.5294 | val_loss: 1.4220 | val_acc: 0.5156 | test_acc: 0.5102 | Time: 2.5895 s
>>> Epoch [ 846/1000]
train_loss: 1.3863 | train_acc: 0.5294 | val_loss: 1.4220 | val_acc: 0.5160 | test_acc: 0.5101 | Time: 2.6842 s
>>> Epoch [ 847/1000]
train_loss: 1.3862 | train_acc: 0.5294 | val_loss: 1.4220 | val_acc: 0.5160 | test_acc: 0.5101 | Time: 2.8211 s
>>> Epoch [ 848/1000]
train_loss: 1.3862 | train_acc: 0.5294 | val_loss: 1.4219 | val_acc: 0.5158 | test_acc: 0.5101 | Time: 2.8434 s
>>> Epoch [ 849/1000]
train_loss: 1.3861 | train_acc: 0.5294 | val_loss: 1.4219 | val_acc: 0.5158 | test_acc: 0.5101 | Time: 2.7387 s
>>> Epoch [ 850/1000]
train_loss: 1.3860 | train_acc: 0.5295 | val_loss: 1.4218 | val_acc: 0.5158 | test_acc: 0.5105 | Time: 2.9300 s
>>> Epoch [ 851/1000]
train_loss: 1.3860 | train_acc: 0.5295 | val_loss: 1.4218 | val_acc: 0.5158 | test_acc: 0.5106 | Time: 3.0177 s
>>> Epoch [ 852/1000]
train_loss: 1.3859 | train_acc: 0.5295 | val_loss: 1.4218 | val_acc: 0.5158 | test_acc: 0.5106 | Time: 2.7939 s
>>> Epoch [ 853/1000]
train_loss: 1.3858 | train_acc: 0.5295 | val_loss: 1.4217 | val_acc: 0.5158 | test_acc: 0.5107 | Time: 2.7852 s
>>> Epoch [ 854/1000]
train_loss: 1.3858 | train_acc: 0.5295 | val_loss: 1.4217 | val_acc: 0.5162 | test_acc: 0.5107 | Time: 2.8614 s
>>> Epoch [ 855/1000]
train_loss: 1.3857 | train_acc: 0.5296 | val_loss: 1.4217 | val_acc: 0.5160 | test_acc: 0.5107 | Time: 2.8440 s
>>> Epoch [ 856/1000]
train_loss: 1.3856 | train_acc: 0.5296 | val_loss: 1.4216 | val_acc: 0.5162 | test_acc: 0.5107 | Time: 2.7934 s
>>> Epoch [ 857/1000]
train_loss: 1.3856 | train_acc: 0.5296 | val_loss: 1.4216 | val_acc: 0.5162 | test_acc: 0.5107 | Time: 2.6049 s
>>> Epoch [ 858/1000]
train_loss: 1.3855 | train_acc: 0.5296 | val_loss: 1.4215 | val_acc: 0.5162 | test_acc: 0.5107 | Time: 2.7939 s
>>> Epoch [ 859/1000]
train_loss: 1.3854 | train_acc: 0.5297 | val_loss: 1.4215 | val_acc: 0.5160 | test_acc: 0.5107 | Time: 2.6688 s
>>> Epoch [ 860/1000]
train_loss: 1.3854 | train_acc: 0.5297 | val_loss: 1.4215 | val_acc: 0.5158 | test_acc: 0.5108 | Time: 2.7206 s
>>> Epoch [ 861/1000]
train_loss: 1.3853 | train_acc: 0.5297 | val_loss: 1.4214 | val_acc: 0.5158 | test_acc: 0.5108 | Time: 2.6713 s
>>> Epoch [ 862/1000]
train_loss: 1.3853 | train_acc: 0.5297 | val_loss: 1.4214 | val_acc: 0.5160 | test_acc: 0.5108 | Time: 2.7248 s
>>> Epoch [ 863/1000]
train_loss: 1.3852 | train_acc: 0.5297 | val_loss: 1.4213 | val_acc: 0.5160 | test_acc: 0.5109 | Time: 2.5575 s
>>> Epoch [ 864/1000]
train_loss: 1.3851 | train_acc: 0.5298 | val_loss: 1.4213 | val_acc: 0.5160 | test_acc: 0.5110 | Time: 2.9065 s
>>> Epoch [ 865/1000]
train_loss: 1.3851 | train_acc: 0.5297 | val_loss: 1.4213 | val_acc: 0.5162 | test_acc: 0.5110 | Time: 2.6927 s
>>> Epoch [ 866/1000]
train_loss: 1.3850 | train_acc: 0.5297 | val_loss: 1.4212 | val_acc: 0.5162 | test_acc: 0.5111 | Time: 2.6054 s
>>> Epoch [ 867/1000]
train_loss: 1.3849 | train_acc: 0.5297 | val_loss: 1.4212 | val_acc: 0.5162 | test_acc: 0.5111 | Time: 3.0785 s
>>> Epoch [ 868/1000]
train_loss: 1.3849 | train_acc: 0.5297 | val_loss: 1.4212 | val_acc: 0.5160 | test_acc: 0.5110 | Time: 2.5272 s
>>> Epoch [ 869/1000]
train_loss: 1.3848 | train_acc: 0.5297 | val_loss: 1.4211 | val_acc: 0.5160 | test_acc: 0.5109 | Time: 2.9264 s
>>> Epoch [ 870/1000]
train_loss: 1.3847 | train_acc: 0.5297 | val_loss: 1.4211 | val_acc: 0.5162 | test_acc: 0.5110 | Time: 2.8906 s
>>> Epoch [ 871/1000]
train_loss: 1.3847 | train_acc: 0.5297 | val_loss: 1.4210 | val_acc: 0.5162 | test_acc: 0.5109 | Time: 3.1445 s
>>> Epoch [ 872/1000]
train_loss: 1.3846 | train_acc: 0.5298 | val_loss: 1.4210 | val_acc: 0.5166 | test_acc: 0.5110 | Time: 2.9127 s
>>> Epoch [ 873/1000]
train_loss: 1.3845 | train_acc: 0.5298 | val_loss: 1.4210 | val_acc: 0.5164 | test_acc: 0.5111 | Time: 2.7230 s
>>> Epoch [ 874/1000]
train_loss: 1.3845 | train_acc: 0.5298 | val_loss: 1.4209 | val_acc: 0.5164 | test_acc: 0.5111 | Time: 2.7017 s
>>> Epoch [ 875/1000]
train_loss: 1.3844 | train_acc: 0.5299 | val_loss: 1.4209 | val_acc: 0.5164 | test_acc: 0.5111 | Time: 2.7022 s
>>> Epoch [ 876/1000]
train_loss: 1.3843 | train_acc: 0.5298 | val_loss: 1.4209 | val_acc: 0.5164 | test_acc: 0.5113 | Time: 2.7611 s
>>> Epoch [ 877/1000]
train_loss: 1.3843 | train_acc: 0.5298 | val_loss: 1.4208 | val_acc: 0.5164 | test_acc: 0.5113 | Time: 3.1586 s
>>> Epoch [ 878/1000]
train_loss: 1.3842 | train_acc: 0.5298 | val_loss: 1.4208 | val_acc: 0.5164 | test_acc: 0.5112 | Time: 3.0463 s
>>> Epoch [ 879/1000]
train_loss: 1.3841 | train_acc: 0.5298 | val_loss: 1.4207 | val_acc: 0.5164 | test_acc: 0.5112 | Time: 3.0941 s
>>> Epoch [ 880/1000]
train_loss: 1.3841 | train_acc: 0.5298 | val_loss: 1.4207 | val_acc: 0.5164 | test_acc: 0.5112 | Time: 3.0276 s
>>> Epoch [ 881/1000]
train_loss: 1.3840 | train_acc: 0.5298 | val_loss: 1.4207 | val_acc: 0.5162 | test_acc: 0.5114 | Time: 2.7102 s
>>> Epoch [ 882/1000]
train_loss: 1.3840 | train_acc: 0.5299 | val_loss: 1.4206 | val_acc: 0.5164 | test_acc: 0.5114 | Time: 3.0647 s
>>> Epoch [ 883/1000]
train_loss: 1.3839 | train_acc: 0.5299 | val_loss: 1.4206 | val_acc: 0.5164 | test_acc: 0.5114 | Time: 3.3317 s
>>> Epoch [ 884/1000]
train_loss: 1.3838 | train_acc: 0.5299 | val_loss: 1.4206 | val_acc: 0.5164 | test_acc: 0.5114 | Time: 2.8529 s
>>> Epoch [ 885/1000]
train_loss: 1.3838 | train_acc: 0.5299 | val_loss: 1.4205 | val_acc: 0.5164 | test_acc: 0.5114 | Time: 2.5889 s
>>> Epoch [ 886/1000]
train_loss: 1.3837 | train_acc: 0.5299 | val_loss: 1.4205 | val_acc: 0.5164 | test_acc: 0.5112 | Time: 2.6624 s
>>> Epoch [ 887/1000]
train_loss: 1.3836 | train_acc: 0.5299 | val_loss: 1.4205 | val_acc: 0.5164 | test_acc: 0.5112 | Time: 2.5605 s
>>> Epoch [ 888/1000]
train_loss: 1.3836 | train_acc: 0.5299 | val_loss: 1.4204 | val_acc: 0.5166 | test_acc: 0.5112 | Time: 2.7601 s
>>> Epoch [ 889/1000]
train_loss: 1.3835 | train_acc: 0.5299 | val_loss: 1.4204 | val_acc: 0.5166 | test_acc: 0.5112 | Time: 2.5630 s
>>> Epoch [ 890/1000]
train_loss: 1.3834 | train_acc: 0.5298 | val_loss: 1.4203 | val_acc: 0.5164 | test_acc: 0.5112 | Time: 2.6665 s
>>> Epoch [ 891/1000]
train_loss: 1.3834 | train_acc: 0.5299 | val_loss: 1.4203 | val_acc: 0.5162 | test_acc: 0.5112 | Time: 3.1401 s
>>> Epoch [ 892/1000]
train_loss: 1.3833 | train_acc: 0.5299 | val_loss: 1.4203 | val_acc: 0.5164 | test_acc: 0.5112 | Time: 2.9923 s
>>> Epoch [ 893/1000]
train_loss: 1.3833 | train_acc: 0.5298 | val_loss: 1.4202 | val_acc: 0.5164 | test_acc: 0.5113 | Time: 3.0657 s
>>> Epoch [ 894/1000]
train_loss: 1.3832 | train_acc: 0.5298 | val_loss: 1.4202 | val_acc: 0.5164 | test_acc: 0.5113 | Time: 2.7625 s
>>> Epoch [ 895/1000]
train_loss: 1.3831 | train_acc: 0.5298 | val_loss: 1.4202 | val_acc: 0.5164 | test_acc: 0.5113 | Time: 3.0853 s
>>> Epoch [ 896/1000]
train_loss: 1.3831 | train_acc: 0.5299 | val_loss: 1.4201 | val_acc: 0.5164 | test_acc: 0.5113 | Time: 3.0973 s
>>> Epoch [ 897/1000]
train_loss: 1.3830 | train_acc: 0.5298 | val_loss: 1.4201 | val_acc: 0.5164 | test_acc: 0.5112 | Time: 2.9162 s
>>> Epoch [ 898/1000]
train_loss: 1.3829 | train_acc: 0.5298 | val_loss: 1.4201 | val_acc: 0.5164 | test_acc: 0.5112 | Time: 2.9089 s
>>> Epoch [ 899/1000]
train_loss: 1.3829 | train_acc: 0.5298 | val_loss: 1.4200 | val_acc: 0.5166 | test_acc: 0.5113 | Time: 2.9594 s
>>> Epoch [ 900/1000]
train_loss: 1.3828 | train_acc: 0.5299 | val_loss: 1.4200 | val_acc: 0.5166 | test_acc: 0.5113 | Time: 2.7037 s
>>> Epoch [ 901/1000]
train_loss: 1.3827 | train_acc: 0.5299 | val_loss: 1.4199 | val_acc: 0.5168 | test_acc: 0.5113 | Time: 2.7487 s
>>> Epoch [ 902/1000]
train_loss: 1.3827 | train_acc: 0.5299 | val_loss: 1.4199 | val_acc: 0.5168 | test_acc: 0.5114 | Time: 2.9660 s
>>> Epoch [ 903/1000]
train_loss: 1.3826 | train_acc: 0.5299 | val_loss: 1.4199 | val_acc: 0.5168 | test_acc: 0.5115 | Time: 2.9100 s
>>> Epoch [ 904/1000]
train_loss: 1.3826 | train_acc: 0.5300 | val_loss: 1.4198 | val_acc: 0.5168 | test_acc: 0.5115 | Time: 2.9006 s
>>> Epoch [ 905/1000]
train_loss: 1.3825 | train_acc: 0.5300 | val_loss: 1.4198 | val_acc: 0.5170 | test_acc: 0.5115 | Time: 3.0481 s
>>> Epoch [ 906/1000]
train_loss: 1.3824 | train_acc: 0.5300 | val_loss: 1.4198 | val_acc: 0.5172 | test_acc: 0.5116 | Time: 3.0015 s
>>> Epoch [ 907/1000]
train_loss: 1.3824 | train_acc: 0.5300 | val_loss: 1.4197 | val_acc: 0.5172 | test_acc: 0.5116 | Time: 2.8209 s
>>> Epoch [ 908/1000]
train_loss: 1.3823 | train_acc: 0.5301 | val_loss: 1.4197 | val_acc: 0.5172 | test_acc: 0.5114 | Time: 2.7361 s
>>> Epoch [ 909/1000]
train_loss: 1.3822 | train_acc: 0.5301 | val_loss: 1.4197 | val_acc: 0.5172 | test_acc: 0.5114 | Time: 2.5528 s
>>> Epoch [ 910/1000]
train_loss: 1.3822 | train_acc: 0.5301 | val_loss: 1.4196 | val_acc: 0.5172 | test_acc: 0.5114 | Time: 2.7505 s
>>> Epoch [ 911/1000]
train_loss: 1.3821 | train_acc: 0.5302 | val_loss: 1.4196 | val_acc: 0.5174 | test_acc: 0.5114 | Time: 2.7806 s
>>> Epoch [ 912/1000]
train_loss: 1.3821 | train_acc: 0.5302 | val_loss: 1.4196 | val_acc: 0.5172 | test_acc: 0.5113 | Time: 3.0912 s
>>> Epoch [ 913/1000]
train_loss: 1.3820 | train_acc: 0.5302 | val_loss: 1.4195 | val_acc: 0.5172 | test_acc: 0.5113 | Time: 3.0349 s
>>> Epoch [ 914/1000]
train_loss: 1.3819 | train_acc: 0.5302 | val_loss: 1.4195 | val_acc: 0.5170 | test_acc: 0.5112 | Time: 2.8306 s
>>> Epoch [ 915/1000]
train_loss: 1.3819 | train_acc: 0.5302 | val_loss: 1.4195 | val_acc: 0.5170 | test_acc: 0.5112 | Time: 2.8185 s
>>> Epoch [ 916/1000]
train_loss: 1.3818 | train_acc: 0.5302 | val_loss: 1.4194 | val_acc: 0.5170 | test_acc: 0.5111 | Time: 2.8115 s
>>> Epoch [ 917/1000]
train_loss: 1.3818 | train_acc: 0.5302 | val_loss: 1.4194 | val_acc: 0.5170 | test_acc: 0.5111 | Time: 2.9335 s
>>> Epoch [ 918/1000]
train_loss: 1.3817 | train_acc: 0.5302 | val_loss: 1.4193 | val_acc: 0.5170 | test_acc: 0.5111 | Time: 2.9805 s
>>> Epoch [ 919/1000]
train_loss: 1.3816 | train_acc: 0.5301 | val_loss: 1.4193 | val_acc: 0.5170 | test_acc: 0.5111 | Time: 3.0642 s
>>> Epoch [ 920/1000]
train_loss: 1.3816 | train_acc: 0.5301 | val_loss: 1.4193 | val_acc: 0.5170 | test_acc: 0.5112 | Time: 2.9298 s
>>> Epoch [ 921/1000]
train_loss: 1.3815 | train_acc: 0.5302 | val_loss: 1.4192 | val_acc: 0.5170 | test_acc: 0.5111 | Time: 2.7154 s
>>> Epoch [ 922/1000]
train_loss: 1.3814 | train_acc: 0.5302 | val_loss: 1.4192 | val_acc: 0.5168 | test_acc: 0.5112 | Time: 2.9157 s
>>> Epoch [ 923/1000]
train_loss: 1.3814 | train_acc: 0.5302 | val_loss: 1.4192 | val_acc: 0.5168 | test_acc: 0.5112 | Time: 3.0489 s
>>> Epoch [ 924/1000]
train_loss: 1.3813 | train_acc: 0.5302 | val_loss: 1.4191 | val_acc: 0.5168 | test_acc: 0.5112 | Time: 3.2695 s
>>> Epoch [ 925/1000]
train_loss: 1.3813 | train_acc: 0.5303 | val_loss: 1.4191 | val_acc: 0.5168 | test_acc: 0.5112 | Time: 2.9982 s
>>> Epoch [ 926/1000]
train_loss: 1.3812 | train_acc: 0.5303 | val_loss: 1.4191 | val_acc: 0.5168 | test_acc: 0.5111 | Time: 2.6932 s
>>> Epoch [ 927/1000]
train_loss: 1.3811 | train_acc: 0.5303 | val_loss: 1.4190 | val_acc: 0.5168 | test_acc: 0.5113 | Time: 2.7564 s
>>> Epoch [ 928/1000]
train_loss: 1.3811 | train_acc: 0.5303 | val_loss: 1.4190 | val_acc: 0.5168 | test_acc: 0.5114 | Time: 2.7471 s
>>> Epoch [ 929/1000]
train_loss: 1.3810 | train_acc: 0.5304 | val_loss: 1.4190 | val_acc: 0.5170 | test_acc: 0.5115 | Time: 2.6558 s
>>> Epoch [ 930/1000]
train_loss: 1.3810 | train_acc: 0.5304 | val_loss: 1.4189 | val_acc: 0.5170 | test_acc: 0.5115 | Time: 2.6472 s
>>> Epoch [ 931/1000]
train_loss: 1.3809 | train_acc: 0.5305 | val_loss: 1.4189 | val_acc: 0.5172 | test_acc: 0.5115 | Time: 2.7806 s
>>> Epoch [ 932/1000]
train_loss: 1.3808 | train_acc: 0.5305 | val_loss: 1.4189 | val_acc: 0.5172 | test_acc: 0.5114 | Time: 3.1080 s
>>> Epoch [ 933/1000]
train_loss: 1.3808 | train_acc: 0.5306 | val_loss: 1.4188 | val_acc: 0.5172 | test_acc: 0.5114 | Time: 3.0336 s
>>> Epoch [ 934/1000]
train_loss: 1.3807 | train_acc: 0.5307 | val_loss: 1.4188 | val_acc: 0.5172 | test_acc: 0.5114 | Time: 3.0634 s
>>> Epoch [ 935/1000]
train_loss: 1.3807 | train_acc: 0.5308 | val_loss: 1.4188 | val_acc: 0.5172 | test_acc: 0.5114 | Time: 2.6965 s
>>> Epoch [ 936/1000]
train_loss: 1.3806 | train_acc: 0.5307 | val_loss: 1.4187 | val_acc: 0.5172 | test_acc: 0.5114 | Time: 2.7844 s
>>> Epoch [ 937/1000]
train_loss: 1.3805 | train_acc: 0.5308 | val_loss: 1.4187 | val_acc: 0.5172 | test_acc: 0.5115 | Time: 2.7782 s
>>> Epoch [ 938/1000]
train_loss: 1.3805 | train_acc: 0.5308 | val_loss: 1.4187 | val_acc: 0.5172 | test_acc: 0.5116 | Time: 2.9871 s
>>> Epoch [ 939/1000]
train_loss: 1.3804 | train_acc: 0.5308 | val_loss: 1.4186 | val_acc: 0.5174 | test_acc: 0.5116 | Time: 2.8751 s
>>> Epoch [ 940/1000]
train_loss: 1.3804 | train_acc: 0.5308 | val_loss: 1.4186 | val_acc: 0.5174 | test_acc: 0.5116 | Time: 3.2225 s
>>> Epoch [ 941/1000]
train_loss: 1.3803 | train_acc: 0.5308 | val_loss: 1.4186 | val_acc: 0.5174 | test_acc: 0.5116 | Time: 2.8714 s
>>> Epoch [ 942/1000]
train_loss: 1.3802 | train_acc: 0.5309 | val_loss: 1.4185 | val_acc: 0.5174 | test_acc: 0.5117 | Time: 2.7487 s
>>> Epoch [ 943/1000]
train_loss: 1.3802 | train_acc: 0.5309 | val_loss: 1.4185 | val_acc: 0.5176 | test_acc: 0.5117 | Time: 3.0214 s
>>> Epoch [ 944/1000]
train_loss: 1.3801 | train_acc: 0.5309 | val_loss: 1.4185 | val_acc: 0.5174 | test_acc: 0.5116 | Time: 2.9991 s
>>> Epoch [ 945/1000]
train_loss: 1.3801 | train_acc: 0.5309 | val_loss: 1.4184 | val_acc: 0.5178 | test_acc: 0.5114 | Time: 3.0479 s
>>> Epoch [ 946/1000]
train_loss: 1.3800 | train_acc: 0.5309 | val_loss: 1.4184 | val_acc: 0.5174 | test_acc: 0.5114 | Time: 3.0763 s
>>> Epoch [ 947/1000]
train_loss: 1.3799 | train_acc: 0.5309 | val_loss: 1.4184 | val_acc: 0.5174 | test_acc: 0.5114 | Time: 2.8379 s
>>> Epoch [ 948/1000]
train_loss: 1.3799 | train_acc: 0.5309 | val_loss: 1.4183 | val_acc: 0.5174 | test_acc: 0.5114 | Time: 2.7086 s
>>> Epoch [ 949/1000]
train_loss: 1.3798 | train_acc: 0.5309 | val_loss: 1.4183 | val_acc: 0.5174 | test_acc: 0.5115 | Time: 2.8775 s
>>> Epoch [ 950/1000]
train_loss: 1.3798 | train_acc: 0.5309 | val_loss: 1.4183 | val_acc: 0.5174 | test_acc: 0.5117 | Time: 3.2422 s
>>> Epoch [ 951/1000]
train_loss: 1.3797 | train_acc: 0.5309 | val_loss: 1.4182 | val_acc: 0.5174 | test_acc: 0.5117 | Time: 3.1708 s
>>> Epoch [ 952/1000]
train_loss: 1.3796 | train_acc: 0.5309 | val_loss: 1.4182 | val_acc: 0.5174 | test_acc: 0.5116 | Time: 2.8849 s
>>> Epoch [ 953/1000]
train_loss: 1.3796 | train_acc: 0.5310 | val_loss: 1.4182 | val_acc: 0.5174 | test_acc: 0.5118 | Time: 3.1474 s
>>> Epoch [ 954/1000]
train_loss: 1.3795 | train_acc: 0.5310 | val_loss: 1.4181 | val_acc: 0.5174 | test_acc: 0.5118 | Time: 2.9445 s
>>> Epoch [ 955/1000]
train_loss: 1.3795 | train_acc: 0.5310 | val_loss: 1.4181 | val_acc: 0.5174 | test_acc: 0.5118 | Time: 3.0302 s
>>> Epoch [ 956/1000]
train_loss: 1.3794 | train_acc: 0.5310 | val_loss: 1.4181 | val_acc: 0.5174 | test_acc: 0.5119 | Time: 3.1848 s
>>> Epoch [ 957/1000]
train_loss: 1.3793 | train_acc: 0.5310 | val_loss: 1.4180 | val_acc: 0.5174 | test_acc: 0.5119 | Time: 3.0249 s
>>> Epoch [ 958/1000]
train_loss: 1.3793 | train_acc: 0.5310 | val_loss: 1.4180 | val_acc: 0.5174 | test_acc: 0.5120 | Time: 3.0234 s
>>> Epoch [ 959/1000]
train_loss: 1.3792 | train_acc: 0.5310 | val_loss: 1.4180 | val_acc: 0.5174 | test_acc: 0.5120 | Time: 3.0496 s
>>> Epoch [ 960/1000]
train_loss: 1.3792 | train_acc: 0.5311 | val_loss: 1.4179 | val_acc: 0.5174 | test_acc: 0.5121 | Time: 3.1072 s
>>> Epoch [ 961/1000]
train_loss: 1.3791 | train_acc: 0.5311 | val_loss: 1.4179 | val_acc: 0.5172 | test_acc: 0.5121 | Time: 3.0331 s
>>> Epoch [ 962/1000]
train_loss: 1.3790 | train_acc: 0.5311 | val_loss: 1.4179 | val_acc: 0.5172 | test_acc: 0.5122 | Time: 2.8247 s
>>> Epoch [ 963/1000]
train_loss: 1.3790 | train_acc: 0.5312 | val_loss: 1.4178 | val_acc: 0.5172 | test_acc: 0.5121 | Time: 2.7955 s
>>> Epoch [ 964/1000]
train_loss: 1.3789 | train_acc: 0.5312 | val_loss: 1.4178 | val_acc: 0.5172 | test_acc: 0.5121 | Time: 2.8195 s
>>> Epoch [ 965/1000]
train_loss: 1.3789 | train_acc: 0.5312 | val_loss: 1.4178 | val_acc: 0.5172 | test_acc: 0.5121 | Time: 2.6079 s
>>> Epoch [ 966/1000]
train_loss: 1.3788 | train_acc: 0.5312 | val_loss: 1.4177 | val_acc: 0.5172 | test_acc: 0.5122 | Time: 2.8692 s
>>> Epoch [ 967/1000]
train_loss: 1.3788 | train_acc: 0.5312 | val_loss: 1.4177 | val_acc: 0.5172 | test_acc: 0.5122 | Time: 3.0964 s
>>> Epoch [ 968/1000]
train_loss: 1.3787 | train_acc: 0.5312 | val_loss: 1.4177 | val_acc: 0.5174 | test_acc: 0.5121 | Time: 2.9394 s
>>> Epoch [ 969/1000]
train_loss: 1.3786 | train_acc: 0.5313 | val_loss: 1.4176 | val_acc: 0.5174 | test_acc: 0.5121 | Time: 3.1907 s
>>> Epoch [ 970/1000]
train_loss: 1.3786 | train_acc: 0.5314 | val_loss: 1.4176 | val_acc: 0.5174 | test_acc: 0.5121 | Time: 2.8092 s
>>> Epoch [ 971/1000]
train_loss: 1.3785 | train_acc: 0.5314 | val_loss: 1.4176 | val_acc: 0.5174 | test_acc: 0.5123 | Time: 2.7609 s
>>> Epoch [ 972/1000]
train_loss: 1.3785 | train_acc: 0.5313 | val_loss: 1.4175 | val_acc: 0.5176 | test_acc: 0.5123 | Time: 2.8558 s
>>> Epoch [ 973/1000]
train_loss: 1.3784 | train_acc: 0.5314 | val_loss: 1.4175 | val_acc: 0.5176 | test_acc: 0.5123 | Time: 2.9303 s
>>> Epoch [ 974/1000]
train_loss: 1.3783 | train_acc: 0.5314 | val_loss: 1.4175 | val_acc: 0.5178 | test_acc: 0.5123 | Time: 3.3065 s
>>> Epoch [ 975/1000]
train_loss: 1.3783 | train_acc: 0.5315 | val_loss: 1.4174 | val_acc: 0.5178 | test_acc: 0.5123 | Time: 3.1295 s
>>> Epoch [ 976/1000]
train_loss: 1.3782 | train_acc: 0.5315 | val_loss: 1.4174 | val_acc: 0.5178 | test_acc: 0.5123 | Time: 2.9724 s
>>> Epoch [ 977/1000]
train_loss: 1.3782 | train_acc: 0.5316 | val_loss: 1.4174 | val_acc: 0.5178 | test_acc: 0.5123 | Time: 2.9853 s
>>> Epoch [ 978/1000]
train_loss: 1.3781 | train_acc: 0.5316 | val_loss: 1.4173 | val_acc: 0.5178 | test_acc: 0.5122 | Time: 2.9436 s
>>> Epoch [ 979/1000]
train_loss: 1.3781 | train_acc: 0.5316 | val_loss: 1.4173 | val_acc: 0.5178 | test_acc: 0.5122 | Time: 3.0172 s
>>> Epoch [ 980/1000]
train_loss: 1.3780 | train_acc: 0.5316 | val_loss: 1.4173 | val_acc: 0.5178 | test_acc: 0.5122 | Time: 2.9130 s
>>> Epoch [ 981/1000]
train_loss: 1.3779 | train_acc: 0.5316 | val_loss: 1.4172 | val_acc: 0.5180 | test_acc: 0.5122 | Time: 2.7543 s
>>> Epoch [ 982/1000]
train_loss: 1.3779 | train_acc: 0.5317 | val_loss: 1.4172 | val_acc: 0.5178 | test_acc: 0.5122 | Time: 2.9984 s
>>> Epoch [ 983/1000]
train_loss: 1.3778 | train_acc: 0.5316 | val_loss: 1.4172 | val_acc: 0.5178 | test_acc: 0.5121 | Time: 3.0055 s
>>> Epoch [ 984/1000]
train_loss: 1.3778 | train_acc: 0.5317 | val_loss: 1.4172 | val_acc: 0.5178 | test_acc: 0.5122 | Time: 3.2516 s
>>> Epoch [ 985/1000]
train_loss: 1.3777 | train_acc: 0.5317 | val_loss: 1.4171 | val_acc: 0.5178 | test_acc: 0.5122 | Time: 3.1754 s
>>> Epoch [ 986/1000]
train_loss: 1.3777 | train_acc: 0.5318 | val_loss: 1.4171 | val_acc: 0.5178 | test_acc: 0.5121 | Time: 2.9085 s
>>> Epoch [ 987/1000]
train_loss: 1.3776 | train_acc: 0.5318 | val_loss: 1.4171 | val_acc: 0.5178 | test_acc: 0.5121 | Time: 2.9172 s
>>> Epoch [ 988/1000]
train_loss: 1.3775 | train_acc: 0.5318 | val_loss: 1.4170 | val_acc: 0.5178 | test_acc: 0.5121 | Time: 3.1061 s
>>> Epoch [ 989/1000]
train_loss: 1.3775 | train_acc: 0.5318 | val_loss: 1.4170 | val_acc: 0.5178 | test_acc: 0.5121 | Time: 3.0654 s
>>> Epoch [ 990/1000]
train_loss: 1.3774 | train_acc: 0.5318 | val_loss: 1.4170 | val_acc: 0.5178 | test_acc: 0.5121 | Time: 3.0486 s
>>> Epoch [ 991/1000]
train_loss: 1.3774 | train_acc: 0.5318 | val_loss: 1.4169 | val_acc: 0.5178 | test_acc: 0.5122 | Time: 2.8268 s
>>> Epoch [ 992/1000]
train_loss: 1.3773 | train_acc: 0.5318 | val_loss: 1.4169 | val_acc: 0.5176 | test_acc: 0.5123 | Time: 2.7783 s
>>> Epoch [ 993/1000]
train_loss: 1.3773 | train_acc: 0.5318 | val_loss: 1.4169 | val_acc: 0.5176 | test_acc: 0.5123 | Time: 2.9798 s
>>> Epoch [ 994/1000]
train_loss: 1.3772 | train_acc: 0.5319 | val_loss: 1.4168 | val_acc: 0.5176 | test_acc: 0.5123 | Time: 2.8486 s
>>> Epoch [ 995/1000]
train_loss: 1.3771 | train_acc: 0.5319 | val_loss: 1.4168 | val_acc: 0.5176 | test_acc: 0.5124 | Time: 2.9455 s
>>> Epoch [ 996/1000]
train_loss: 1.3771 | train_acc: 0.5319 | val_loss: 1.4168 | val_acc: 0.5176 | test_acc: 0.5124 | Time: 3.2298 s
>>> Epoch [ 997/1000]
train_loss: 1.3770 | train_acc: 0.5320 | val_loss: 1.4167 | val_acc: 0.5174 | test_acc: 0.5124 | Time: 2.9922 s
>>> Epoch [ 998/1000]
train_loss: 1.3770 | train_acc: 0.5321 | val_loss: 1.4167 | val_acc: 0.5174 | test_acc: 0.5124 | Time: 3.0042 s
>>> Epoch [ 999/1000]
train_loss: 1.3769 | train_acc: 0.5321 | val_loss: 1.4167 | val_acc: 0.5174 | test_acc: 0.5125 | Time: 2.8979 s
In [21]:
prediction_numbers = tf_lc.predict(test_hog)
prediction_classes = []
num_test_images = cifar10.test.data.shape[0]
for i in range(num_test_images):
prediction_classes.append(cifar10.classes[int(prediction_numbers[i])])
In [22]:
cifar10.plot_images(cifar10.test.images[:50], cifar10.test.class_names[:50], cls_pred=prediction_classes[:50],
nrows=5, ncols=10, fig_size=(20,50), fontsize=35, convert=False)
Out[22]:
True
In [23]:
test_accuracy = tf_lc.print_accuracy(test_hog, cifar10.test.one_hot_labels, cifar10.test.class_labels)
print('Accuracy of the linear classifier on test dataset: %.4f' %test_accuracy)
Accuracy of the linear classifier on test dataset: 0.5125
In [24]:
tf_lc.print_classification_results(test_hog, cifar10.test.one_hot_labels, cifar10.test.class_labels,
test_class_names=cifar10.classes, normalize=True)
Confusion matrix, without normalization
[[568 42 84 12 32 12 38 22 153 37]
[ 41 634 14 11 16 8 48 18 85 125]
[ 95 35 338 64 82 131 125 53 50 27]
[ 47 46 70 239 102 180 144 75 35 62]
[ 47 34 51 69 393 61 142 102 54 47]
[ 14 23 84 134 71 416 103 106 23 26]
[ 15 32 41 33 52 50 713 30 19 15]
[ 32 23 53 39 100 106 44 540 18 45]
[115 104 30 8 15 5 24 19 601 79]
[ 29 79 13 20 22 22 19 35 78 683]]
Normalized confusion matrix
[[ 0.568 0.042 0.084 0.012 0.032 0.012 0.038 0.022 0.153 0.037]
[ 0.041 0.634 0.014 0.011 0.016 0.008 0.048 0.018 0.085 0.125]
[ 0.095 0.035 0.338 0.064 0.082 0.131 0.125 0.053 0.05 0.027]
[ 0.047 0.046 0.07 0.239 0.102 0.18 0.144 0.075 0.035 0.062]
[ 0.047 0.034 0.051 0.069 0.393 0.061 0.142 0.102 0.054 0.047]
[ 0.014 0.023 0.084 0.134 0.071 0.416 0.103 0.106 0.023 0.026]
[ 0.015 0.032 0.041 0.033 0.052 0.05 0.713 0.03 0.019 0.015]
[ 0.032 0.023 0.053 0.039 0.1 0.106 0.044 0.54 0.018 0.045]
[ 0.115 0.104 0.03 0.008 0.015 0.005 0.024 0.019 0.601 0.079]
[ 0.029 0.079 0.013 0.02 0.022 0.022 0.019 0.035 0.078 0.683]]
Detailed classification report
precision recall f1-score support
airplane 0.57 0.57 0.57 1000
automobile 0.60 0.63 0.62 1000
bird 0.43 0.34 0.38 1000
cat 0.38 0.24 0.29 1000
deer 0.44 0.39 0.42 1000
dog 0.42 0.42 0.42 1000
frog 0.51 0.71 0.59 1000
horse 0.54 0.54 0.54 1000
ship 0.54 0.60 0.57 1000
truck 0.60 0.68 0.64 1000
avg / total 0.50 0.51 0.50 10000
In [25]:
tf_lc.close()
In [26]:
def output_HTML(read_file, output_file):
from nbconvert import HTMLExporter
import codecs
import nbformat
exporter = HTMLExporter()
output_notebook = nbformat.read(read_file, as_version=4)
print()
output, resources = exporter.from_notebook_node(output_notebook)
codecs.open(output_file, 'w', encoding='utf-8').write(output)
In [27]:
%%javascript
var notebook = IPython.notebook
notebook.save_notebook()
In [30]:
%%javascript
var kernel = IPython.notebook.kernel;
var thename = window.document.getElementById("notebook_name").innerHTML;
var command = "theNotebook = " + "'"+thename+"'";
kernel.execute(command);
In [31]:
current_file = './' + theNotebook + '.ipynb'
output_file = log_dir + str(file_no).zfill(2) + '_exp_no_' + str(exp_no).zfill(3) + '_' + theNotebook + '.html'
print('Current file: ' + str(current_file))
print('Output file: ' + str(output_file))
file_utils.mkdir_p(log_dir)
output_HTML(current_file, output_file)
Current file: ./10_Tensorflow_Linear_Classifier_HOG_features_CIFAR_10_Website.ipynb
Output file: ../logs/cifar10/10_tf_linear_hog/exp_no_105/10_exp_no_105_10_Tensorflow_Linear_Classifier_HOG_features_CIFAR_10_Website.html
In [ ]:
Content source: sonapraneeth-a/object-classification
Similar notebooks: