Cats vs. Dogs Class dataset for multiple annotators¶
Imports¶
[1]:
from sklearn.preprocessing import LabelBinarizer
from sklearn.preprocessing import OneHotEncoder
from scipy.stats import mode
import numpy as np
def ook(t):
lb = LabelBinarizer()
y_ook = lb.fit_transform(t)
if len(np.unique(t))==2:
y_ook = np.concatenate((1-y_ook.astype(bool), y_ook), axis = 1)
return y_ook
[2]:
import tensorflow_datasets as tfds
import tensorflow as tf
import keras
from keras.models import Sequential,Model
from keras.layers import Dense,Conv2D,Flatten,MaxPooling2D,GlobalAveragePooling2D
from keras.utils.vis_utils import plot_model
from tensorflow.keras import regularizers
import numpy as np
import matplotlib.pyplot as plt
import scipy as sp
import cv2
import os
import time
import sys
[3]:
# from google.colab import drive
# drive.mount('/content/drive')
[4]:
# os.chdir('/content/drive/Shareddrives/Multiple Anotators/CrowdLayer/Notebooks')
# cwd = os.getcwd()
# sys.path.append("../Models")
# from Multiple_Annotators_C import MultipleAnnotators_Classification
#import sys
#sys.path.insert(1, '../input/multiple-annotators-c/')
#os.chdir('/Multiple Anotators-c/')
#cwd = os.getcwd()
#sys.path.append('/input/multiple-annotators-c')
#from Multiple_Annotators_C import MultipleAnnotators_Classification
# seed_value= 12321
# from numpy.random import seed
# seed(seed_value)
# tf.random.set_seed(seed_value)
Download and Prepare the Dataset¶
We will use the Cats vs Dogs dataset and we can load it via Tensorflow Datasets. The images are labeled 0 for cats and 1 for dogs.
Multiple annotators model¶
[5]:
validation_data = tf.data.experimental.load('/kaggle/input/catsvsdog-ma/cats_dogs_Te')
train_data_MA = tf.data.experimental.load('/kaggle/input/catsvsdog-ma/cats_dogs_MA_Tr_1')
2023-02-14 21:49:59.295511: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2023-02-14 21:49:59.462904: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2023-02-14 21:49:59.463771: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2023-02-14 21:49:59.464974: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations: AVX2 AVX512F FMA
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2023-02-14 21:49:59.470654: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2023-02-14 21:49:59.471346: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2023-02-14 21:49:59.472030: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2023-02-14 21:50:01.493675: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2023-02-14 21:50:01.494509: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2023-02-14 21:50:01.495200: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2023-02-14 21:50:01.495814: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 15401 MB memory: -> device: 0, name: Tesla P100-PCIE-16GB, pci bus id: 0000:00:04.0, compute capability: 6.0
[6]:
image_count = tf.data.experimental.cardinality(train_data_MA).numpy() # los datos de training son 18610 usar subconjunto de 5000
image_count
[6]:
18610
[7]:
image_count1 = tf.data.experimental.cardinality(validation_data).numpy() # los datos de training son 18610
image_count1
[7]:
4652
[8]:
#X_test = [validation_data[i][0] for i in range(image_count1)]
#Y_true_test = [validation_data[i][1] for i in range(image_count1)]
#Y_true_test = np.asarray([aux[1].numpy() for aux in validation_data])
#X_test = np.asarray([aux[0].numpy() for aux in validation_data])
[9]:
image_count
[9]:
18610
[10]:
val_size = int(image_count * 0.2)
train_ds_MA = train_data_MA.skip(val_size)
val_ds_MA = train_data_MA.take(val_size)
[11]:
batch_size = 100
train_batches_MA = train_ds_MA.shuffle(1024).batch(batch_size)
val_batches_MA = val_ds_MA.shuffle(1024).batch(batch_size)
test_batches_MA = validation_data.shuffle(1024).batch(batch_size)
[12]:
image_count = tf.data.experimental.cardinality(train_ds_MA).numpy() # los datos de training son 18610 usar subconjunto de 5000
image_count
[12]:
14888
[13]:
image_count_val = tf.data.experimental.cardinality(val_ds_MA).numpy() # los datos de training son 18610 usar subconjunto de 5000
image_count_val
[13]:
3722
[ ]:
[14]:
from sklearn.metrics import classification_report
i = 0
fig, ax = plt.subplots(1, 4)
for image, label, label2 in train_batches_MA.take(4):
# predictedLabel = int(predictions[i] >= 0.5)
# print(label2)
ax[i].axis('off')
# ax[i].set_title(classNames[label[i]])
ax[i].imshow(image[0])
i += 1
for j in range(label2.shape[1]):
print('annotator',j+1)
print(classification_report(label ,label2[:,j]))
plt.show()
2023-02-14 21:50:02.257325: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:185] None of the MLIR Optimization Passes are enabled (registered 2)
2023-02-14 21:50:17.128965: I tensorflow/core/kernels/data/shuffle_dataset_op.cc:175] Filling up shuffle buffer (this may take a while): 1 of 1024
2023-02-14 21:50:21.116166: I tensorflow/core/kernels/data/shuffle_dataset_op.cc:228] Shuffle buffer filled.
annotator 1
precision recall f1-score support
0 0.71 0.76 0.73 45
1 0.79 0.75 0.77 55
accuracy 0.75 100
macro avg 0.75 0.75 0.75 100
weighted avg 0.75 0.75 0.75 100
annotator 2
precision recall f1-score support
0 0.91 0.87 0.89 45
1 0.89 0.93 0.91 55
accuracy 0.90 100
macro avg 0.90 0.90 0.90 100
weighted avg 0.90 0.90 0.90 100
annotator 3
precision recall f1-score support
0 0.64 0.51 0.57 45
1 0.66 0.76 0.71 55
accuracy 0.65 100
macro avg 0.65 0.64 0.64 100
weighted avg 0.65 0.65 0.64 100
annotator 4
precision recall f1-score support
0 0.83 0.22 0.35 45
1 0.60 0.96 0.74 55
accuracy 0.63 100
macro avg 0.72 0.59 0.55 100
weighted avg 0.71 0.63 0.57 100
annotator 5
precision recall f1-score support
0 0.87 0.89 0.88 45
1 0.91 0.89 0.90 55
accuracy 0.89 100
macro avg 0.89 0.89 0.89 100
weighted avg 0.89 0.89 0.89 100
annotator 1
precision recall f1-score support
0 0.83 0.70 0.76 50
1 0.74 0.86 0.80 50
accuracy 0.78 100
macro avg 0.79 0.78 0.78 100
weighted avg 0.79 0.78 0.78 100
annotator 2
precision recall f1-score support
0 0.91 0.78 0.84 50
1 0.81 0.92 0.86 50
accuracy 0.85 100
macro avg 0.86 0.85 0.85 100
weighted avg 0.86 0.85 0.85 100
annotator 3
precision recall f1-score support
0 0.75 0.54 0.63 50
1 0.64 0.82 0.72 50
accuracy 0.68 100
macro avg 0.70 0.68 0.67 100
weighted avg 0.70 0.68 0.67 100
annotator 4
precision recall f1-score support
0 0.91 0.20 0.33 50
1 0.55 0.98 0.71 50
accuracy 0.59 100
macro avg 0.73 0.59 0.52 100
weighted avg 0.73 0.59 0.52 100
annotator 5
precision recall f1-score support
0 0.86 0.88 0.87 50
1 0.88 0.86 0.87 50
accuracy 0.87 100
macro avg 0.87 0.87 0.87 100
weighted avg 0.87 0.87 0.87 100
annotator 1
precision recall f1-score support
0 0.75 0.76 0.76 51
1 0.75 0.73 0.74 49
accuracy 0.75 100
macro avg 0.75 0.75 0.75 100
weighted avg 0.75 0.75 0.75 100
annotator 2
precision recall f1-score support
0 0.98 0.86 0.92 51
1 0.87 0.98 0.92 49
accuracy 0.92 100
macro avg 0.93 0.92 0.92 100
weighted avg 0.93 0.92 0.92 100
annotator 3
precision recall f1-score support
0 0.69 0.61 0.65 51
1 0.64 0.71 0.67 49
accuracy 0.66 100
macro avg 0.66 0.66 0.66 100
weighted avg 0.66 0.66 0.66 100
annotator 4
precision recall f1-score support
0 0.92 0.22 0.35 51
1 0.55 0.98 0.70 49
accuracy 0.59 100
macro avg 0.73 0.60 0.52 100
weighted avg 0.73 0.59 0.52 100
annotator 5
precision recall f1-score support
0 0.90 0.88 0.89 51
1 0.88 0.90 0.89 49
accuracy 0.89 100
macro avg 0.89 0.89 0.89 100
weighted avg 0.89 0.89 0.89 100
annotator 1
precision recall f1-score support
0 0.70 0.73 0.72 52
1 0.70 0.67 0.68 48
accuracy 0.70 100
macro avg 0.70 0.70 0.70 100
weighted avg 0.70 0.70 0.70 100
annotator 2
precision recall f1-score support
0 0.96 0.88 0.92 52
1 0.88 0.96 0.92 48
accuracy 0.92 100
macro avg 0.92 0.92 0.92 100
weighted avg 0.92 0.92 0.92 100
annotator 3
precision recall f1-score support
0 0.73 0.63 0.68 52
1 0.65 0.75 0.70 48
accuracy 0.69 100
macro avg 0.69 0.69 0.69 100
weighted avg 0.70 0.69 0.69 100
annotator 4
precision recall f1-score support
0 0.88 0.27 0.41 52
1 0.55 0.96 0.70 48
accuracy 0.60 100
macro avg 0.71 0.61 0.55 100
weighted avg 0.72 0.60 0.55 100
annotator 5
precision recall f1-score support
0 0.94 0.90 0.92 52
1 0.90 0.94 0.92 48
accuracy 0.92 100
macro avg 0.92 0.92 0.92 100
weighted avg 0.92 0.92 0.92 100
Build the classifier from multiple annotators¶
[15]:
import tensorflow_datasets as tfds
import tensorflow as tf
import time
from tensorflow.keras import regularizers
import keras
from keras.models import Sequential,Model
from keras.layers import Dense,Conv2D,Flatten,MaxPooling2D,GlobalAveragePooling2D
from keras.utils.vis_utils import plot_model
class MultipleAnnotators_Classification():
def __init__(self, output_dim, num_annotators, q= 0.0001):
self.K = output_dim
self.R = num_annotators
self.q = q
#self.callbacks #=callbacks
#self.l1_param=l1_param
#self.l2_param=l1_param
def CrowdLayer(self, input):
#x = keras.layers.Dense(self.R + self.K, , activation='tanh')(input)
output_cla = keras.layers.Dense(self.K, activation='softmax')(input)
output_ann = keras.layers.Dense(self.R, activation='sigmoid')(input)
output = keras.layers.Concatenate()([output_cla, output_ann])
return output
#RCDNN
# def loss(self):
# def custom_loss(y_true, y_pred):
# # print(y_true,y_pred)
# pred = y_pred[:, :self.K]
# pred = tf.clip_by_value(pred, clip_value_min=1e-9, clip_value_max=1-1e-9) #estabilidad numerica de la funcion de costo
# ann_ = y_pred[:, self.K:]
# Y_true = tf.one_hot(tf.cast(y_true, dtype=tf.int32), depth=self.K, axis=1)
# Y_hat = tf.repeat(tf.expand_dims(pred,-1), self.R, axis = -1)
# p_logreg = tf.math.reduce_prod(tf.math.pow(Y_hat, Y_true), axis=1)
# temp1 = ann_*tf.math.log(p_logreg)
# temp2 = (1 - ann_)*tf.math.log(1/self.K)*tf.reduce_sum(Y_true,axis=1)
# # temp2 = (tf.ones(tf.shape(ann_)) - ann_)*tf.math.log(1/K)
# # print(tf.reduce_mean(Y_true,axis=1).numpy())
# return -tf.math.reduce_sum((temp1 + temp2))
# return custom_loss
def loss(self):
def custom_loss(y_true, y_pred):
# print(y_true,y_pred)
# q = 0.1
pred = y_pred[:, :self.K]
pred = tf.clip_by_value(pred, clip_value_min=1e-9, clip_value_max=1)
ann_ = y_pred[:, self.K:]
# ann_ = tf.clip_by_value(ann_, clip_value_min=1e-9, clip_value_max=1-1e-9)
Y_true = tf.one_hot(tf.cast(y_true, dtype=tf.int32), depth=self.K, axis=1)
Y_hat = tf.repeat(tf.expand_dims(pred,-1), self.R, axis = -1)
p_gcce = Y_true*(1 - Y_hat**self.q)/self.q
temp1 = ann_*tf.math.reduce_sum(p_gcce, axis=1)
temp2 = (1 - ann_)*(1-(1/self.K)**self.q)/self.q*tf.reduce_sum(Y_true,axis=1)
return tf.math.reduce_sum((temp1 + temp2))
return custom_loss
@tf.function
def train_step(self, x, Y, y):
with tf.GradientTape() as tape:
logits = self.model(x, training=True)
loss_value = self.loss_fn(Y, logits)
grads = tape.gradient(loss_value, self.model.trainable_weights)
self.optimizer.apply_gradients(zip(grads, self.model.trainable_weights))
self.train_acc_metric.update_state(y, logits[:, :self.K])
return loss_value
@tf.function
def test_step(self, x, y):
val_logits = self.model(x, training=False)
self.val_acc_metric.update_state(y, val_logits[:,:self.K])
def fit(self, model, Data_tr, Data_Val, epochs):
self.model = model
#++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
# Instantiate an optimizer.
self.optimizer = tf.keras.optimizers.RMSprop(learning_rate=1e-4)
#self.optimizer = tf.keras.optimizers.Adam(learning_rate=1e-3)
#self.optimizer = tf.keras.optimizers.SGD(learning_rate=1e-4, clipnorm=1.0)
#++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
# Instantiate a loss function.
self.loss_fn = self.loss()
self.train_acc_metric = keras.metrics.SparseCategoricalAccuracy()
self.val_acc_metric = keras.metrics.SparseCategoricalAccuracy()
train_loss = np.zeros(epochs)
train_accur = np.zeros(epochs)
val_accur = np.zeros(epochs)
val_loss = np.zeros(epochs)
for epoch in range(epochs):
print("\nStart of epoch %d" % (epoch,))
start_time = time.time()
# Iterate over the batches of the dataset.
for step, (x_batch_train, y_batch_train, Y_batch_train) in enumerate(Data_tr):
# print(y_batch_train, Y_batch_train)
loss_value = self.train_step(x_batch_train, Y_batch_train, y_batch_train)
# Log every 200 batches.
if step % 10 == 0:
train_acc = self.train_acc_metric.result()
print(
"Training loss (for one batch) at step %d: %.4f, Accuracy: %.4f"
% (step, float(loss_value), float(train_acc))
)
# print("Seen so far: %d samples" % ((step + 1) * batch_size))
# Run a validation loop at the end of each epoch.
for x_batch_val, y_batch_val,Y_batch_val in Data_Val:
val_logits = model(x_batch_val, training=False)
val_loss_value = self.loss_fn(Y_batch_val, val_logits)
self.val_acc_metric.update_state(y_batch_val, val_logits[:,:self.K])
# np.round(np.mean([model(x_batch_val, training= True) for sample in range(100)]), 2)
# Display metrics at the end of each epoch.
train_acc = self.train_acc_metric.result()
val_acc = self.val_acc_metric.result()
print('---- Training ----')
print("Training loss: %.4f" % (float(loss_value),))
print("Training acc over epoch: %.4f" % (float(train_acc),))
# Reset training metrics at the end of each epoch
self.train_acc_metric.reset_states()
self.val_acc_metric.reset_states()
train_loss[epoch] = float(loss_value)
train_accur[epoch] = float(train_acc)
val_accur[epoch] = float(val_acc)
val_loss[epoch] = float(val_loss_value)
print('---- Validation ----')
print("Validation loss: %.4f" % (float(val_loss_value),))
print("Validation acc: %.4f" % (float(val_acc),))
print("Time taken: %.2fs" % (time.time() - start_time))
fig, (ax1, ax2) = plt.subplots(1, 2)
fig.suptitle('Loss and accuracy')
ax1.plot(range(1,epochs+1),train_loss)
ax1.plot(range(1,epochs+1), val_loss)
ax2.plot(range(1,epochs+1),train_accur)
ax2.plot(range(1,epochs+1),val_accur)
#plt.figure(figsize=(16,9))
ax1.set(xlabel= 'Epoch', ylabel="Loss")
ax2.set(xlabel= 'Epoch',ylabel="Accuracy")
ax1.legend(['Training_loss', 'Validation_loss'])
ax2.legend(['Training', 'Validation'])
ax1.grid()
ax2.grid()
plt.show()
return self.model
def eval_model(self, Data):
self.val_acc_metric = keras.metrics.SparseCategoricalAccuracy()
for x_batch_val, y_batch_val in Data:
self.test_step(x_batch_val, y_batch_val)
val_acc = self.val_acc_metric.result()
self.val_acc_metric.reset_states()
return val_acc
[16]:
def custom_loss(y_true, y_pred):
# print(y_true,y_pred)
K = 2 #len(np.unique(y_true))
R = 5
q = 0.1
pred = y_pred[:, K]
pred = tf.clip_by_value(pred, clip_value_min=1e-9, clip_value_max=1)
ann_ = y_pred[:, K:]
# ann_ = tf.clip_by_value(ann_, clip_value_min=1e-9, clip_value_max=1-1e-9)
Y_true = tf.one_hot(tf.cast(y_true, dtype=tf.int32), depth=K, axis=1)
Y_hat = tf.repeat(tf.expand_dims(pred,-1), R, axis = -1)
p_gcce = Y_true*(1 - Y_hat**q)/q
temp1 = ann_*tf.math.reduce_sum(p_gcce, axis=1)
temp2 = (1 - ann_)*(1-(1/K)**q)/q*tf.reduce_sum(Y_true,axis=1)
return tf.math.reduce_sum((temp1 + temp2))
[17]:
MA = MultipleAnnotators_Classification(2, 5, 0.001)
def create_model():
l1 = 1e-2
# Block 1
inputs = keras.layers.Input(shape=(150, 150, 3), name='entrada')
x = keras.layers.BatchNormalization()(inputs)
x = keras.layers.Conv2D(32, (3, 3), activation="relu" , name="block1_conv1")(x)
x = keras.layers.BatchNormalization()(x)
x = keras.layers.MaxPooling2D((2, 2), strides=(2, 2), name="block1_pool")(x)
# Block 2
x = keras.layers.BatchNormalization()(x)
x = keras.layers.Conv2D(32, (3, 3), activation="relu", name="block2_conv1")(x)
x = keras.layers.BatchNormalization()(x)
#x = keras.layers.Dropout(0.2)(x)
x = keras.layers.MaxPooling2D((2, 2), strides=(2, 2), name="block2_pool")(x)
# Block 3
x = keras.layers.BatchNormalization()(x)
x = keras.layers.Conv2D(64, (3, 3), activation="relu", name="block3_conv1" )(x)
x = keras.layers.BatchNormalization()(x)
# x = keras.layers.Dropout(0.2)(x)
x = keras.layers.MaxPooling2D((2, 2), strides=(2, 2), name="block3_pool")(x)
# Block 4
x = keras.layers.BatchNormalization()(x)
x = keras.layers.Conv2D(64, (3, 3), activation="relu", name="block4_conv1")(x)
x = keras.layers.BatchNormalization()(x)
x = keras.layers.MaxPooling2D((2, 2), strides=(2, 2), name="block4_pool")(x)
#x = keras.layers.Dropout(0.2)(x)
#x = keras.layers.GlobalAveragePooling2D()(x)
x = keras.layers.Flatten()(x)
#x = keras.layers.Dropout(0.5)(x)
x = keras.layers.BatchNormalization()(x)
x = keras.layers.Dense(128, activation='relu')(x)
x = keras.layers.BatchNormalization()(x)
x = keras.layers.Dropout(0.5)(x)
output = MA.CrowdLayer(x)
model = keras.Model(inputs=inputs,outputs=output)
return model
[18]:
from sklearn.metrics import classification_report, balanced_accuracy_score, roc_auc_score
from sklearn.metrics import normalized_mutual_info_score, mutual_info_score, adjusted_mutual_info_score
#model = create_model()
K=2
R=5
val_q = [0.0001] #0.2, 0.4, 0.6, 0.8]
NUM_RUNS = 10
N_EPOCHS =50
ACC = np.zeros(NUM_RUNS)
AUC = np.zeros(NUM_RUNS)
AUCSK = np.zeros(NUM_RUNS)
MI = np.zeros(NUM_RUNS)
NMI = np.zeros(NUM_RUNS)
AMI = np.zeros(NUM_RUNS)
BACC = np.zeros(NUM_RUNS)
BACC1 = []
MI1 = []
NMI1 =[]
AMI1 = []
AUCSK1 = []
val_acc = np.zeros(NUM_RUNS)
for i in range(NUM_RUNS):
MA = MultipleAnnotators_Classification(2, 5, val_q[0])
model = create_model()
model = MA.fit(model, train_batches_MA, val_batches_MA, N_EPOCHS)
#model = MA.fit(model, Data_train_MA, N_EPOCHS)
ACC[i] = MA.eval_model(test_batches_MA)
print("===== Q: %.4f" % (float(val_q[0]),))
print("Validation acc: %.4f" % (float(ACC[i]),))
#AUC =======================
val_AUC_metric = tf.keras.metrics.AUC( from_logits = True)
for x_batch_val, y_batch_val in test_batches_MA:
val_logits = model(x_batch_val.numpy(), training=False)
# tf.print(y_batch_val)
val_AUC_metric.update_state(y_batch_val, val_logits[:,:K].numpy().argmax(axis=1).astype('float')) #val_logits[:,Y.shape[1]:].argmax(axis=1).astype('float'))
AUCSK1.append(roc_auc_score(ook(y_batch_val.numpy().ravel()), val_logits[:,:K].numpy()))
BACC1.append(balanced_accuracy_score(y_batch_val.numpy().squeeze(), val_logits[:,:K].numpy().argmax(axis=1).squeeze(), adjusted=True))
MI1.append(mutual_info_score(y_batch_val.numpy().squeeze(), val_logits[:,:K].numpy().argmax(axis=1).squeeze()))
NMI1.append(normalized_mutual_info_score(y_batch_val.numpy().squeeze(), val_logits[:,:K].numpy().argmax(axis=1).squeeze()))
AMI1.append(normalized_mutual_info_score(y_batch_val.numpy().squeeze(), val_logits[:,:K].numpy().argmax(axis=1).squeeze()))
val_AUC = val_AUC_metric.result()
val_AUC_metric.reset_states()
val_AUC = val_AUC.numpy()
print("Validation AUC: %.4f" % (float(val_AUC),))
AUC[i] = val_AUC
#===================================================
# balanced. Accurcy
BACC[i] = np.array(BACC1).mean() # balanced_accuracy_score(Y_true_test.squeeze(), val_logits[:,:K].numpy().argmax(axis=1).squeeze(), adjusted=True)
print("Validation Balanced_ACC: %.4f" % (float(BACC[i])))
#MI
MI[i] = np.array(MI1).mean() #mutual_info_score(Y_true_test.squeeze(), val_logits[:,:K].numpy().argmax(axis=1).squeeze())
print("Validation MI: %.4f" % (float(MI[i]),))
NMI[i] = np.array(NMI1).mean() #normalized_mutual_info_score(Y_true_test.squeeze(), val_logits[:,:K].numpy().argmax(axis=1).squeeze())
print("Validation Normalized MI: %.4f" % (float(NMI[i]),))
AMI[i]= np.array(AMI1).mean() #adjusted_mutual_info_score(Y_true_test.squeeze(), val_logits[:,:K].numpy().argmax(axis=1).squeeze())
print("Validation Adjusted MI: %.4f" % (float(AMI[i]),))
AUCSK[i] = np.array(AUCSK1).mean()
print("Validation aUc_Sklearn: %.4f" % (float(AUCSK[i]),))
import pandas as pd
df = pd.DataFrame(ACC)
#df.to_csv('/content/CatDogs_MA_InceptionV3.csv',index=False) # save to notebook output
Start of epoch 0
2023-02-14 21:50:26.911913: I tensorflow/stream_executor/cuda/cuda_dnn.cc:369] Loaded cuDNN version 8005
Training loss (for one batch) at step 0: 527.5652, Accuracy: 0.4400
Training loss (for one batch) at step 10: 485.7293, Accuracy: 0.5664
Training loss (for one batch) at step 20: 470.1454, Accuracy: 0.5643
Training loss (for one batch) at step 30: 428.4604, Accuracy: 0.5665
Training loss (for one batch) at step 40: 460.2888, Accuracy: 0.5710
Training loss (for one batch) at step 50: 446.4178, Accuracy: 0.5696
Training loss (for one batch) at step 60: 445.7569, Accuracy: 0.5716
Training loss (for one batch) at step 70: 419.4080, Accuracy: 0.5735
Training loss (for one batch) at step 80: 450.2828, Accuracy: 0.5765
Training loss (for one batch) at step 90: 448.1821, Accuracy: 0.5760
Training loss (for one batch) at step 100: 469.8572, Accuracy: 0.5758
Training loss (for one batch) at step 110: 421.2848, Accuracy: 0.5786
Training loss (for one batch) at step 120: 415.3198, Accuracy: 0.5806
Training loss (for one batch) at step 130: 422.9765, Accuracy: 0.5824
Training loss (for one batch) at step 140: 421.0312, Accuracy: 0.5812
---- Training ----
Training loss: 350.7345
Training acc over epoch: 0.5824
---- Validation ----
Validation loss: 101.4941
Validation acc: 0.5134
Time taken: 69.16s
Start of epoch 1
Training loss (for one batch) at step 0: 414.7354, Accuracy: 0.6000
Training loss (for one batch) at step 10: 368.4087, Accuracy: 0.6127
Training loss (for one batch) at step 20: 399.2114, Accuracy: 0.6190
Training loss (for one batch) at step 30: 409.7503, Accuracy: 0.6155
Training loss (for one batch) at step 40: 374.2572, Accuracy: 0.6195
Training loss (for one batch) at step 50: 390.4643, Accuracy: 0.6225
Training loss (for one batch) at step 60: 383.4763, Accuracy: 0.6238
Training loss (for one batch) at step 70: 412.1743, Accuracy: 0.6234
Training loss (for one batch) at step 80: 414.9368, Accuracy: 0.6248
Training loss (for one batch) at step 90: 373.5606, Accuracy: 0.6266
Training loss (for one batch) at step 100: 397.1875, Accuracy: 0.6267
Training loss (for one batch) at step 110: 386.2385, Accuracy: 0.6250
Training loss (for one batch) at step 120: 381.1939, Accuracy: 0.6228
Training loss (for one batch) at step 130: 368.0622, Accuracy: 0.6244
Training loss (for one batch) at step 140: 382.5007, Accuracy: 0.6243
---- Training ----
Training loss: 352.9120
Training acc over epoch: 0.6250
---- Validation ----
Validation loss: 79.5263
Validation acc: 0.5306
Time taken: 10.53s
Start of epoch 2
Training loss (for one batch) at step 0: 375.4754, Accuracy: 0.6500
Training loss (for one batch) at step 10: 403.0332, Accuracy: 0.6327
Training loss (for one batch) at step 20: 378.3479, Accuracy: 0.6305
Training loss (for one batch) at step 30: 414.4794, Accuracy: 0.6365
Training loss (for one batch) at step 40: 398.7999, Accuracy: 0.6380
Training loss (for one batch) at step 50: 381.3060, Accuracy: 0.6376
Training loss (for one batch) at step 60: 384.7852, Accuracy: 0.6408
Training loss (for one batch) at step 70: 367.2447, Accuracy: 0.6406
Training loss (for one batch) at step 80: 396.3054, Accuracy: 0.6412
Training loss (for one batch) at step 90: 358.3841, Accuracy: 0.6390
Training loss (for one batch) at step 100: 363.4851, Accuracy: 0.6410
Training loss (for one batch) at step 110: 362.2993, Accuracy: 0.6422
Training loss (for one batch) at step 120: 349.9545, Accuracy: 0.6455
Training loss (for one batch) at step 130: 353.6843, Accuracy: 0.6447
Training loss (for one batch) at step 140: 369.4059, Accuracy: 0.6448
---- Training ----
Training loss: 306.4173
Training acc over epoch: 0.6452
---- Validation ----
Validation loss: 71.6043
Validation acc: 0.6905
Time taken: 10.33s
Start of epoch 3
Training loss (for one batch) at step 0: 361.0397, Accuracy: 0.7700
Training loss (for one batch) at step 10: 375.9297, Accuracy: 0.6682
Training loss (for one batch) at step 20: 348.9310, Accuracy: 0.6719
Training loss (for one batch) at step 30: 400.0352, Accuracy: 0.6655
Training loss (for one batch) at step 40: 342.6651, Accuracy: 0.6629
Training loss (for one batch) at step 50: 402.3806, Accuracy: 0.6661
Training loss (for one batch) at step 60: 357.3140, Accuracy: 0.6648
Training loss (for one batch) at step 70: 367.7959, Accuracy: 0.6641
Training loss (for one batch) at step 80: 379.3242, Accuracy: 0.6647
Training loss (for one batch) at step 90: 360.5052, Accuracy: 0.6644
Training loss (for one batch) at step 100: 380.4537, Accuracy: 0.6665
Training loss (for one batch) at step 110: 369.4744, Accuracy: 0.6677
Training loss (for one batch) at step 120: 338.3556, Accuracy: 0.6683
Training loss (for one batch) at step 130: 358.0639, Accuracy: 0.6685
Training loss (for one batch) at step 140: 386.4206, Accuracy: 0.6680
---- Training ----
Training loss: 318.3352
Training acc over epoch: 0.6669
---- Validation ----
Validation loss: 71.0933
Validation acc: 0.7106
Time taken: 10.45s
Start of epoch 4
Training loss (for one batch) at step 0: 365.2827, Accuracy: 0.6900
Training loss (for one batch) at step 10: 369.3956, Accuracy: 0.6782
Training loss (for one batch) at step 20: 334.1480, Accuracy: 0.6795
Training loss (for one batch) at step 30: 364.9940, Accuracy: 0.6755
Training loss (for one batch) at step 40: 377.7835, Accuracy: 0.6737
Training loss (for one batch) at step 50: 331.5173, Accuracy: 0.6776
Training loss (for one batch) at step 60: 338.0648, Accuracy: 0.6772
Training loss (for one batch) at step 70: 356.2871, Accuracy: 0.6776
Training loss (for one batch) at step 80: 346.5218, Accuracy: 0.6783
Training loss (for one batch) at step 90: 335.0776, Accuracy: 0.6758
Training loss (for one batch) at step 100: 326.9565, Accuracy: 0.6742
Training loss (for one batch) at step 110: 358.7886, Accuracy: 0.6729
Training loss (for one batch) at step 120: 334.4588, Accuracy: 0.6731
Training loss (for one batch) at step 130: 358.6262, Accuracy: 0.6732
Training loss (for one batch) at step 140: 329.9813, Accuracy: 0.6743
---- Training ----
Training loss: 288.3139
Training acc over epoch: 0.6753
---- Validation ----
Validation loss: 78.8003
Validation acc: 0.7034
Time taken: 9.91s
Start of epoch 5
Training loss (for one batch) at step 0: 327.2693, Accuracy: 0.6800
Training loss (for one batch) at step 10: 336.1891, Accuracy: 0.6936
Training loss (for one batch) at step 20: 342.4390, Accuracy: 0.6957
Training loss (for one batch) at step 30: 333.1637, Accuracy: 0.6868
Training loss (for one batch) at step 40: 323.9966, Accuracy: 0.6861
Training loss (for one batch) at step 50: 313.2051, Accuracy: 0.6867
Training loss (for one batch) at step 60: 347.4801, Accuracy: 0.6885
Training loss (for one batch) at step 70: 322.6429, Accuracy: 0.6923
Training loss (for one batch) at step 80: 350.3183, Accuracy: 0.6946
Training loss (for one batch) at step 90: 341.9393, Accuracy: 0.6938
Training loss (for one batch) at step 100: 328.0479, Accuracy: 0.6921
Training loss (for one batch) at step 110: 336.2398, Accuracy: 0.6923
Training loss (for one batch) at step 120: 344.7343, Accuracy: 0.6941
Training loss (for one batch) at step 130: 344.7416, Accuracy: 0.6939
Training loss (for one batch) at step 140: 337.3373, Accuracy: 0.6962
---- Training ----
Training loss: 295.1274
Training acc over epoch: 0.6953
---- Validation ----
Validation loss: 75.8963
Validation acc: 0.7007
Time taken: 9.73s
Start of epoch 6
Training loss (for one batch) at step 0: 328.4754, Accuracy: 0.7500
Training loss (for one batch) at step 10: 345.5945, Accuracy: 0.7036
Training loss (for one batch) at step 20: 336.1838, Accuracy: 0.6986
Training loss (for one batch) at step 30: 334.5241, Accuracy: 0.6977
Training loss (for one batch) at step 40: 343.0800, Accuracy: 0.7020
Training loss (for one batch) at step 50: 329.4874, Accuracy: 0.7051
Training loss (for one batch) at step 60: 319.4249, Accuracy: 0.7061
Training loss (for one batch) at step 70: 323.9439, Accuracy: 0.7068
Training loss (for one batch) at step 80: 328.1810, Accuracy: 0.7046
Training loss (for one batch) at step 90: 329.0745, Accuracy: 0.7065
Training loss (for one batch) at step 100: 336.4267, Accuracy: 0.7055
Training loss (for one batch) at step 110: 314.7868, Accuracy: 0.7051
Training loss (for one batch) at step 120: 332.8512, Accuracy: 0.7048
Training loss (for one batch) at step 130: 344.6004, Accuracy: 0.7060
Training loss (for one batch) at step 140: 324.1347, Accuracy: 0.7069
---- Training ----
Training loss: 295.2740
Training acc over epoch: 0.7078
---- Validation ----
Validation loss: 71.8922
Validation acc: 0.7383
Time taken: 9.66s
Start of epoch 7
Training loss (for one batch) at step 0: 321.7647, Accuracy: 0.6800
Training loss (for one batch) at step 10: 310.0202, Accuracy: 0.6936
Training loss (for one batch) at step 20: 309.3121, Accuracy: 0.7124
Training loss (for one batch) at step 30: 344.8871, Accuracy: 0.7052
Training loss (for one batch) at step 40: 317.7033, Accuracy: 0.7129
Training loss (for one batch) at step 50: 328.6810, Accuracy: 0.7169
Training loss (for one batch) at step 60: 314.9474, Accuracy: 0.7203
Training loss (for one batch) at step 70: 311.1190, Accuracy: 0.7235
Training loss (for one batch) at step 80: 336.7076, Accuracy: 0.7217
Training loss (for one batch) at step 90: 338.0130, Accuracy: 0.7199
Training loss (for one batch) at step 100: 335.6043, Accuracy: 0.7190
Training loss (for one batch) at step 110: 311.4397, Accuracy: 0.7214
Training loss (for one batch) at step 120: 327.8146, Accuracy: 0.7214
Training loss (for one batch) at step 130: 313.3764, Accuracy: 0.7210
Training loss (for one batch) at step 140: 324.4893, Accuracy: 0.7225
---- Training ----
Training loss: 295.1404
Training acc over epoch: 0.7215
---- Validation ----
Validation loss: 69.7238
Validation acc: 0.6964
Time taken: 9.50s
Start of epoch 8
Training loss (for one batch) at step 0: 313.9289, Accuracy: 0.7600
Training loss (for one batch) at step 10: 301.8256, Accuracy: 0.7227
Training loss (for one batch) at step 20: 302.5697, Accuracy: 0.7219
Training loss (for one batch) at step 30: 309.3948, Accuracy: 0.7239
Training loss (for one batch) at step 40: 351.4895, Accuracy: 0.7278
Training loss (for one batch) at step 50: 305.1110, Accuracy: 0.7296
Training loss (for one batch) at step 60: 315.3144, Accuracy: 0.7290
Training loss (for one batch) at step 70: 322.4496, Accuracy: 0.7287
Training loss (for one batch) at step 80: 307.2949, Accuracy: 0.7283
Training loss (for one batch) at step 90: 332.2927, Accuracy: 0.7246
Training loss (for one batch) at step 100: 317.0076, Accuracy: 0.7243
Training loss (for one batch) at step 110: 315.9072, Accuracy: 0.7261
Training loss (for one batch) at step 120: 304.7992, Accuracy: 0.7261
Training loss (for one batch) at step 130: 321.2657, Accuracy: 0.7256
Training loss (for one batch) at step 140: 311.0507, Accuracy: 0.7284
---- Training ----
Training loss: 289.8389
Training acc over epoch: 0.7280
---- Validation ----
Validation loss: 83.8228
Validation acc: 0.7227
Time taken: 9.77s
Start of epoch 9
Training loss (for one batch) at step 0: 297.6439, Accuracy: 0.7400
Training loss (for one batch) at step 10: 301.3385, Accuracy: 0.7464
Training loss (for one batch) at step 20: 335.2910, Accuracy: 0.7633
Training loss (for one batch) at step 30: 308.5423, Accuracy: 0.7587
Training loss (for one batch) at step 40: 330.3203, Accuracy: 0.7573
Training loss (for one batch) at step 50: 316.8153, Accuracy: 0.7557
Training loss (for one batch) at step 60: 318.7785, Accuracy: 0.7556
Training loss (for one batch) at step 70: 311.4369, Accuracy: 0.7548
Training loss (for one batch) at step 80: 314.5453, Accuracy: 0.7523
Training loss (for one batch) at step 90: 322.3553, Accuracy: 0.7531
Training loss (for one batch) at step 100: 290.3810, Accuracy: 0.7530
Training loss (for one batch) at step 110: 309.8307, Accuracy: 0.7516
Training loss (for one batch) at step 120: 298.1726, Accuracy: 0.7510
Training loss (for one batch) at step 130: 303.1388, Accuracy: 0.7511
Training loss (for one batch) at step 140: 312.4652, Accuracy: 0.7518
---- Training ----
Training loss: 298.6415
Training acc over epoch: 0.7521
---- Validation ----
Validation loss: 68.2228
Validation acc: 0.6773
Time taken: 9.58s
Start of epoch 10
Training loss (for one batch) at step 0: 303.8081, Accuracy: 0.7700
Training loss (for one batch) at step 10: 315.6491, Accuracy: 0.7582
Training loss (for one batch) at step 20: 299.7039, Accuracy: 0.7510
Training loss (for one batch) at step 30: 318.7493, Accuracy: 0.7523
Training loss (for one batch) at step 40: 296.9448, Accuracy: 0.7505
Training loss (for one batch) at step 50: 305.3838, Accuracy: 0.7506
Training loss (for one batch) at step 60: 317.8546, Accuracy: 0.7493
Training loss (for one batch) at step 70: 307.6823, Accuracy: 0.7541
Training loss (for one batch) at step 80: 313.1857, Accuracy: 0.7548
Training loss (for one batch) at step 90: 313.6532, Accuracy: 0.7521
Training loss (for one batch) at step 100: 322.8497, Accuracy: 0.7526
Training loss (for one batch) at step 110: 311.6450, Accuracy: 0.7502
Training loss (for one batch) at step 120: 318.1743, Accuracy: 0.7528
Training loss (for one batch) at step 130: 297.5695, Accuracy: 0.7536
Training loss (for one batch) at step 140: 312.5851, Accuracy: 0.7545
---- Training ----
Training loss: 270.7544
Training acc over epoch: 0.7548
---- Validation ----
Validation loss: 69.5605
Validation acc: 0.7184
Time taken: 9.60s
Start of epoch 11
Training loss (for one batch) at step 0: 295.1462, Accuracy: 0.8100
Training loss (for one batch) at step 10: 302.8084, Accuracy: 0.7591
Training loss (for one batch) at step 20: 303.0034, Accuracy: 0.7652
Training loss (for one batch) at step 30: 311.3145, Accuracy: 0.7545
Training loss (for one batch) at step 40: 286.4696, Accuracy: 0.7615
Training loss (for one batch) at step 50: 294.6967, Accuracy: 0.7659
Training loss (for one batch) at step 60: 275.4273, Accuracy: 0.7667
Training loss (for one batch) at step 70: 309.5190, Accuracy: 0.7666
Training loss (for one batch) at step 80: 301.8839, Accuracy: 0.7659
Training loss (for one batch) at step 90: 303.8592, Accuracy: 0.7624
Training loss (for one batch) at step 100: 305.8885, Accuracy: 0.7607
Training loss (for one batch) at step 110: 294.8497, Accuracy: 0.7617
Training loss (for one batch) at step 120: 300.2937, Accuracy: 0.7614
Training loss (for one batch) at step 130: 281.0685, Accuracy: 0.7625
Training loss (for one batch) at step 140: 296.7326, Accuracy: 0.7612
---- Training ----
Training loss: 266.6598
Training acc over epoch: 0.7616
---- Validation ----
Validation loss: 74.9823
Validation acc: 0.7375
Time taken: 9.69s
Start of epoch 12
Training loss (for one batch) at step 0: 296.6288, Accuracy: 0.7300
Training loss (for one batch) at step 10: 291.0009, Accuracy: 0.7855
Training loss (for one batch) at step 20: 288.2358, Accuracy: 0.7790
Training loss (for one batch) at step 30: 306.8449, Accuracy: 0.7710
Training loss (for one batch) at step 40: 273.7520, Accuracy: 0.7729
Training loss (for one batch) at step 50: 276.0591, Accuracy: 0.7747
Training loss (for one batch) at step 60: 296.8549, Accuracy: 0.7738
Training loss (for one batch) at step 70: 317.8672, Accuracy: 0.7724
Training loss (for one batch) at step 80: 298.5279, Accuracy: 0.7709
Training loss (for one batch) at step 90: 297.2271, Accuracy: 0.7697
Training loss (for one batch) at step 100: 304.9115, Accuracy: 0.7693
Training loss (for one batch) at step 110: 310.8136, Accuracy: 0.7685
Training loss (for one batch) at step 120: 298.3556, Accuracy: 0.7679
Training loss (for one batch) at step 130: 287.6582, Accuracy: 0.7679
Training loss (for one batch) at step 140: 296.9242, Accuracy: 0.7675
---- Training ----
Training loss: 271.1704
Training acc over epoch: 0.7677
---- Validation ----
Validation loss: 64.7361
Validation acc: 0.7270
Time taken: 9.56s
Start of epoch 13
Training loss (for one batch) at step 0: 287.1867, Accuracy: 0.7700
Training loss (for one batch) at step 10: 297.4686, Accuracy: 0.7927
Training loss (for one batch) at step 20: 281.0555, Accuracy: 0.7800
Training loss (for one batch) at step 30: 293.9285, Accuracy: 0.7826
Training loss (for one batch) at step 40: 283.0506, Accuracy: 0.7824
Training loss (for one batch) at step 50: 296.8868, Accuracy: 0.7839
Training loss (for one batch) at step 60: 287.4199, Accuracy: 0.7830
Training loss (for one batch) at step 70: 286.9984, Accuracy: 0.7837
Training loss (for one batch) at step 80: 296.7943, Accuracy: 0.7805
Training loss (for one batch) at step 90: 296.5591, Accuracy: 0.7799
Training loss (for one batch) at step 100: 283.2963, Accuracy: 0.7788
Training loss (for one batch) at step 110: 290.9077, Accuracy: 0.7788
Training loss (for one batch) at step 120: 291.3411, Accuracy: 0.7793
Training loss (for one batch) at step 130: 306.5003, Accuracy: 0.7773
Training loss (for one batch) at step 140: 291.8774, Accuracy: 0.7775
---- Training ----
Training loss: 258.3038
Training acc over epoch: 0.7783
---- Validation ----
Validation loss: 65.8365
Validation acc: 0.7187
Time taken: 9.49s
Start of epoch 14
Training loss (for one batch) at step 0: 283.7563, Accuracy: 0.7800
Training loss (for one batch) at step 10: 290.0991, Accuracy: 0.7809
Training loss (for one batch) at step 20: 278.3028, Accuracy: 0.7800
Training loss (for one batch) at step 30: 290.3202, Accuracy: 0.7887
Training loss (for one batch) at step 40: 276.7822, Accuracy: 0.7898
Training loss (for one batch) at step 50: 282.2297, Accuracy: 0.7927
Training loss (for one batch) at step 60: 290.2224, Accuracy: 0.7905
Training loss (for one batch) at step 70: 299.0808, Accuracy: 0.7899
Training loss (for one batch) at step 80: 293.3069, Accuracy: 0.7900
Training loss (for one batch) at step 90: 286.8884, Accuracy: 0.7878
Training loss (for one batch) at step 100: 297.0613, Accuracy: 0.7867
Training loss (for one batch) at step 110: 290.0004, Accuracy: 0.7879
Training loss (for one batch) at step 120: 289.4159, Accuracy: 0.7888
Training loss (for one batch) at step 130: 270.9205, Accuracy: 0.7908
Training loss (for one batch) at step 140: 289.3708, Accuracy: 0.7902
---- Training ----
Training loss: 249.7713
Training acc over epoch: 0.7900
---- Validation ----
Validation loss: 65.4737
Validation acc: 0.7257
Time taken: 9.51s
Start of epoch 15
Training loss (for one batch) at step 0: 281.0325, Accuracy: 0.8500
Training loss (for one batch) at step 10: 275.9476, Accuracy: 0.7973
Training loss (for one batch) at step 20: 289.4301, Accuracy: 0.7871
Training loss (for one batch) at step 30: 284.5569, Accuracy: 0.7884
Training loss (for one batch) at step 40: 284.0876, Accuracy: 0.7944
Training loss (for one batch) at step 50: 304.0462, Accuracy: 0.7988
Training loss (for one batch) at step 60: 277.5830, Accuracy: 0.7984
Training loss (for one batch) at step 70: 283.1639, Accuracy: 0.7987
Training loss (for one batch) at step 80: 273.1905, Accuracy: 0.7946
Training loss (for one batch) at step 90: 286.8107, Accuracy: 0.7933
Training loss (for one batch) at step 100: 280.2020, Accuracy: 0.7908
Training loss (for one batch) at step 110: 281.0310, Accuracy: 0.7903
Training loss (for one batch) at step 120: 297.7278, Accuracy: 0.7867
Training loss (for one batch) at step 130: 274.7339, Accuracy: 0.7871
Training loss (for one batch) at step 140: 260.2668, Accuracy: 0.7866
---- Training ----
Training loss: 265.2728
Training acc over epoch: 0.7869
---- Validation ----
Validation loss: 71.5178
Validation acc: 0.7192
Time taken: 9.63s
Start of epoch 16
Training loss (for one batch) at step 0: 274.4103, Accuracy: 0.7500
Training loss (for one batch) at step 10: 300.0254, Accuracy: 0.8018
Training loss (for one batch) at step 20: 276.3546, Accuracy: 0.7933
Training loss (for one batch) at step 30: 287.9843, Accuracy: 0.7955
Training loss (for one batch) at step 40: 285.2507, Accuracy: 0.7944
Training loss (for one batch) at step 50: 269.1497, Accuracy: 0.7955
Training loss (for one batch) at step 60: 270.6111, Accuracy: 0.7964
Training loss (for one batch) at step 70: 291.2901, Accuracy: 0.7969
Training loss (for one batch) at step 80: 294.9871, Accuracy: 0.7925
Training loss (for one batch) at step 90: 264.6625, Accuracy: 0.7922
Training loss (for one batch) at step 100: 298.0137, Accuracy: 0.7938
Training loss (for one batch) at step 110: 282.0747, Accuracy: 0.7940
Training loss (for one batch) at step 120: 287.8063, Accuracy: 0.7960
Training loss (for one batch) at step 130: 283.6217, Accuracy: 0.7955
Training loss (for one batch) at step 140: 289.8761, Accuracy: 0.7946
---- Training ----
Training loss: 256.7251
Training acc over epoch: 0.7947
---- Validation ----
Validation loss: 64.4297
Validation acc: 0.7018
Time taken: 9.80s
Start of epoch 17
Training loss (for one batch) at step 0: 292.0485, Accuracy: 0.7600
Training loss (for one batch) at step 10: 279.7513, Accuracy: 0.8082
Training loss (for one batch) at step 20: 275.5767, Accuracy: 0.8076
Training loss (for one batch) at step 30: 266.7609, Accuracy: 0.8035
Training loss (for one batch) at step 40: 273.1532, Accuracy: 0.8024
Training loss (for one batch) at step 50: 278.0567, Accuracy: 0.8063
Training loss (for one batch) at step 60: 276.4338, Accuracy: 0.8072
Training loss (for one batch) at step 70: 291.9246, Accuracy: 0.8056
Training loss (for one batch) at step 80: 297.4688, Accuracy: 0.8032
Training loss (for one batch) at step 90: 289.9101, Accuracy: 0.8013
Training loss (for one batch) at step 100: 279.5326, Accuracy: 0.8017
Training loss (for one batch) at step 110: 288.1198, Accuracy: 0.8021
Training loss (for one batch) at step 120: 274.8012, Accuracy: 0.8028
Training loss (for one batch) at step 130: 268.0208, Accuracy: 0.8035
Training loss (for one batch) at step 140: 273.4841, Accuracy: 0.8027
---- Training ----
Training loss: 250.7330
Training acc over epoch: 0.8023
---- Validation ----
Validation loss: 68.7349
Validation acc: 0.7370
Time taken: 9.47s
Start of epoch 18
Training loss (for one batch) at step 0: 280.5630, Accuracy: 0.8000
Training loss (for one batch) at step 10: 273.3553, Accuracy: 0.7873
Training loss (for one batch) at step 20: 278.0201, Accuracy: 0.8019
Training loss (for one batch) at step 30: 272.9178, Accuracy: 0.7997
Training loss (for one batch) at step 40: 254.3600, Accuracy: 0.8049
Training loss (for one batch) at step 50: 264.8229, Accuracy: 0.8092
Training loss (for one batch) at step 60: 291.8868, Accuracy: 0.8036
Training loss (for one batch) at step 70: 286.1880, Accuracy: 0.8035
Training loss (for one batch) at step 80: 286.6671, Accuracy: 0.8048
Training loss (for one batch) at step 90: 269.4072, Accuracy: 0.8032
Training loss (for one batch) at step 100: 269.3436, Accuracy: 0.8033
Training loss (for one batch) at step 110: 271.5280, Accuracy: 0.8031
Training loss (for one batch) at step 120: 265.4930, Accuracy: 0.8017
Training loss (for one batch) at step 130: 279.0914, Accuracy: 0.8018
Training loss (for one batch) at step 140: 269.4370, Accuracy: 0.8011
---- Training ----
Training loss: 247.9927
Training acc over epoch: 0.8012
---- Validation ----
Validation loss: 71.3467
Validation acc: 0.7284
Time taken: 9.66s
Start of epoch 19
Training loss (for one batch) at step 0: 267.3128, Accuracy: 0.8300
Training loss (for one batch) at step 10: 270.8720, Accuracy: 0.8245
Training loss (for one batch) at step 20: 273.5750, Accuracy: 0.8224
Training loss (for one batch) at step 30: 276.4769, Accuracy: 0.8171
Training loss (for one batch) at step 40: 295.6112, Accuracy: 0.8163
Training loss (for one batch) at step 50: 264.0731, Accuracy: 0.8165
Training loss (for one batch) at step 60: 276.1700, Accuracy: 0.8167
Training loss (for one batch) at step 70: 280.4786, Accuracy: 0.8166
Training loss (for one batch) at step 80: 284.1524, Accuracy: 0.8125
Training loss (for one batch) at step 90: 293.8797, Accuracy: 0.8105
Training loss (for one batch) at step 100: 257.7791, Accuracy: 0.8101
Training loss (for one batch) at step 110: 283.2241, Accuracy: 0.8095
Training loss (for one batch) at step 120: 273.7171, Accuracy: 0.8106
Training loss (for one batch) at step 130: 276.6753, Accuracy: 0.8087
Training loss (for one batch) at step 140: 269.7658, Accuracy: 0.8094
---- Training ----
Training loss: 248.0699
Training acc over epoch: 0.8096
---- Validation ----
Validation loss: 61.3913
Validation acc: 0.7289
Time taken: 9.60s
Start of epoch 20
Training loss (for one batch) at step 0: 267.8854, Accuracy: 0.8000
Training loss (for one batch) at step 10: 270.6834, Accuracy: 0.8218
Training loss (for one batch) at step 20: 276.9342, Accuracy: 0.8124
Training loss (for one batch) at step 30: 242.9749, Accuracy: 0.8158
Training loss (for one batch) at step 40: 284.6319, Accuracy: 0.8161
Training loss (for one batch) at step 50: 259.6818, Accuracy: 0.8210
Training loss (for one batch) at step 60: 270.6411, Accuracy: 0.8197
Training loss (for one batch) at step 70: 272.0410, Accuracy: 0.8186
Training loss (for one batch) at step 80: 267.1286, Accuracy: 0.8162
Training loss (for one batch) at step 90: 278.6115, Accuracy: 0.8149
Training loss (for one batch) at step 100: 256.2336, Accuracy: 0.8149
Training loss (for one batch) at step 110: 268.8988, Accuracy: 0.8148
Training loss (for one batch) at step 120: 258.3041, Accuracy: 0.8144
Training loss (for one batch) at step 130: 265.0509, Accuracy: 0.8134
Training loss (for one batch) at step 140: 275.9163, Accuracy: 0.8137
---- Training ----
Training loss: 232.4071
Training acc over epoch: 0.8132
---- Validation ----
Validation loss: 74.1707
Validation acc: 0.7133
Time taken: 9.69s
Start of epoch 21
Training loss (for one batch) at step 0: 276.4044, Accuracy: 0.7800
Training loss (for one batch) at step 10: 261.0562, Accuracy: 0.8236
Training loss (for one batch) at step 20: 263.4520, Accuracy: 0.8257
Training loss (for one batch) at step 30: 279.7070, Accuracy: 0.8216
Training loss (for one batch) at step 40: 256.7255, Accuracy: 0.8217
Training loss (for one batch) at step 50: 252.4158, Accuracy: 0.8241
Training loss (for one batch) at step 60: 270.0327, Accuracy: 0.8215
Training loss (for one batch) at step 70: 276.1600, Accuracy: 0.8238
Training loss (for one batch) at step 80: 269.4739, Accuracy: 0.8235
Training loss (for one batch) at step 90: 257.7021, Accuracy: 0.8190
Training loss (for one batch) at step 100: 255.9322, Accuracy: 0.8178
Training loss (for one batch) at step 110: 269.9975, Accuracy: 0.8181
Training loss (for one batch) at step 120: 278.6348, Accuracy: 0.8176
Training loss (for one batch) at step 130: 254.9585, Accuracy: 0.8167
Training loss (for one batch) at step 140: 266.3175, Accuracy: 0.8163
---- Training ----
Training loss: 250.0136
Training acc over epoch: 0.8165
---- Validation ----
Validation loss: 71.6924
Validation acc: 0.7391
Time taken: 9.62s
Start of epoch 22
Training loss (for one batch) at step 0: 248.2838, Accuracy: 0.9000
Training loss (for one batch) at step 10: 249.2325, Accuracy: 0.8264
Training loss (for one batch) at step 20: 281.3386, Accuracy: 0.8176
Training loss (for one batch) at step 30: 276.5588, Accuracy: 0.8190
Training loss (for one batch) at step 40: 262.4187, Accuracy: 0.8178
Training loss (for one batch) at step 50: 252.6470, Accuracy: 0.8214
Training loss (for one batch) at step 60: 264.4310, Accuracy: 0.8208
Training loss (for one batch) at step 70: 262.0877, Accuracy: 0.8245
Training loss (for one batch) at step 80: 285.3459, Accuracy: 0.8200
Training loss (for one batch) at step 90: 266.0043, Accuracy: 0.8188
Training loss (for one batch) at step 100: 251.2785, Accuracy: 0.8177
Training loss (for one batch) at step 110: 266.6462, Accuracy: 0.8184
Training loss (for one batch) at step 120: 253.9668, Accuracy: 0.8165
Training loss (for one batch) at step 130: 272.4949, Accuracy: 0.8166
Training loss (for one batch) at step 140: 258.9280, Accuracy: 0.8167
---- Training ----
Training loss: 220.2854
Training acc over epoch: 0.8166
---- Validation ----
Validation loss: 73.4674
Validation acc: 0.7308
Time taken: 9.73s
Start of epoch 23
Training loss (for one batch) at step 0: 256.3245, Accuracy: 0.9100
Training loss (for one batch) at step 10: 253.5680, Accuracy: 0.8564
Training loss (for one batch) at step 20: 252.2861, Accuracy: 0.8367
Training loss (for one batch) at step 30: 260.5576, Accuracy: 0.8300
Training loss (for one batch) at step 40: 256.0079, Accuracy: 0.8210
Training loss (for one batch) at step 50: 267.9693, Accuracy: 0.8257
Training loss (for one batch) at step 60: 250.6280, Accuracy: 0.8259
Training loss (for one batch) at step 70: 264.6138, Accuracy: 0.8270
Training loss (for one batch) at step 80: 274.8406, Accuracy: 0.8252
Training loss (for one batch) at step 90: 263.6708, Accuracy: 0.8240
Training loss (for one batch) at step 100: 248.9697, Accuracy: 0.8236
Training loss (for one batch) at step 110: 251.1908, Accuracy: 0.8236
Training loss (for one batch) at step 120: 259.8826, Accuracy: 0.8227
Training loss (for one batch) at step 130: 273.4259, Accuracy: 0.8221
Training loss (for one batch) at step 140: 284.1017, Accuracy: 0.8208
---- Training ----
Training loss: 215.5548
Training acc over epoch: 0.8221
---- Validation ----
Validation loss: 75.0286
Validation acc: 0.7362
Time taken: 9.51s
Start of epoch 24
Training loss (for one batch) at step 0: 240.5153, Accuracy: 0.8700
Training loss (for one batch) at step 10: 259.7513, Accuracy: 0.8127
Training loss (for one batch) at step 20: 276.9256, Accuracy: 0.8195
Training loss (for one batch) at step 30: 251.5533, Accuracy: 0.8184
Training loss (for one batch) at step 40: 249.6874, Accuracy: 0.8217
Training loss (for one batch) at step 50: 253.3733, Accuracy: 0.8259
Training loss (for one batch) at step 60: 255.6430, Accuracy: 0.8277
Training loss (for one batch) at step 70: 269.4500, Accuracy: 0.8283
Training loss (for one batch) at step 80: 249.9317, Accuracy: 0.8281
Training loss (for one batch) at step 90: 252.6238, Accuracy: 0.8259
Training loss (for one batch) at step 100: 255.3096, Accuracy: 0.8245
Training loss (for one batch) at step 110: 256.0580, Accuracy: 0.8244
Training loss (for one batch) at step 120: 253.1473, Accuracy: 0.8255
Training loss (for one batch) at step 130: 262.8895, Accuracy: 0.8235
Training loss (for one batch) at step 140: 256.2113, Accuracy: 0.8240
---- Training ----
Training loss: 244.2797
Training acc over epoch: 0.8240
---- Validation ----
Validation loss: 71.2011
Validation acc: 0.7254
Time taken: 10.35s
Start of epoch 25
Training loss (for one batch) at step 0: 250.4025, Accuracy: 0.8200
Training loss (for one batch) at step 10: 251.6705, Accuracy: 0.8336
Training loss (for one batch) at step 20: 255.6863, Accuracy: 0.8338
Training loss (for one batch) at step 30: 258.7211, Accuracy: 0.8239
Training loss (for one batch) at step 40: 251.2184, Accuracy: 0.8224
Training loss (for one batch) at step 50: 249.0692, Accuracy: 0.8282
Training loss (for one batch) at step 60: 245.7644, Accuracy: 0.8280
Training loss (for one batch) at step 70: 281.7910, Accuracy: 0.8263
Training loss (for one batch) at step 80: 248.8053, Accuracy: 0.8278
Training loss (for one batch) at step 90: 249.5143, Accuracy: 0.8241
Training loss (for one batch) at step 100: 266.8668, Accuracy: 0.8248
Training loss (for one batch) at step 110: 250.7866, Accuracy: 0.8259
Training loss (for one batch) at step 120: 246.7132, Accuracy: 0.8258
Training loss (for one batch) at step 130: 269.0662, Accuracy: 0.8257
Training loss (for one batch) at step 140: 267.2637, Accuracy: 0.8259
---- Training ----
Training loss: 227.9117
Training acc over epoch: 0.8254
---- Validation ----
Validation loss: 61.6761
Validation acc: 0.7294
Time taken: 9.58s
Start of epoch 26
Training loss (for one batch) at step 0: 251.1541, Accuracy: 0.7600
Training loss (for one batch) at step 10: 271.5143, Accuracy: 0.8464
Training loss (for one batch) at step 20: 255.7008, Accuracy: 0.8352
Training loss (for one batch) at step 30: 256.9074, Accuracy: 0.8268
Training loss (for one batch) at step 40: 232.6582, Accuracy: 0.8324
Training loss (for one batch) at step 50: 248.6599, Accuracy: 0.8367
Training loss (for one batch) at step 60: 240.2707, Accuracy: 0.8366
Training loss (for one batch) at step 70: 272.9339, Accuracy: 0.8368
Training loss (for one batch) at step 80: 237.2755, Accuracy: 0.8343
Training loss (for one batch) at step 90: 247.0536, Accuracy: 0.8332
Training loss (for one batch) at step 100: 264.4133, Accuracy: 0.8324
Training loss (for one batch) at step 110: 254.5302, Accuracy: 0.8329
Training loss (for one batch) at step 120: 241.8723, Accuracy: 0.8320
Training loss (for one batch) at step 130: 236.1336, Accuracy: 0.8322
Training loss (for one batch) at step 140: 248.4787, Accuracy: 0.8313
---- Training ----
Training loss: 221.9963
Training acc over epoch: 0.8311
---- Validation ----
Validation loss: 82.3368
Validation acc: 0.7243
Time taken: 9.47s
Start of epoch 27
Training loss (for one batch) at step 0: 246.6013, Accuracy: 0.8300
Training loss (for one batch) at step 10: 237.6495, Accuracy: 0.8173
Training loss (for one batch) at step 20: 252.1101, Accuracy: 0.8210
Training loss (for one batch) at step 30: 242.4967, Accuracy: 0.8216
Training loss (for one batch) at step 40: 218.8251, Accuracy: 0.8210
Training loss (for one batch) at step 50: 244.3240, Accuracy: 0.8318
Training loss (for one batch) at step 60: 234.6243, Accuracy: 0.8349
Training loss (for one batch) at step 70: 248.1142, Accuracy: 0.8324
Training loss (for one batch) at step 80: 266.1911, Accuracy: 0.8312
Training loss (for one batch) at step 90: 245.9972, Accuracy: 0.8318
Training loss (for one batch) at step 100: 239.7433, Accuracy: 0.8309
Training loss (for one batch) at step 110: 237.8589, Accuracy: 0.8320
Training loss (for one batch) at step 120: 237.3627, Accuracy: 0.8306
Training loss (for one batch) at step 130: 264.5100, Accuracy: 0.8304
Training loss (for one batch) at step 140: 238.8596, Accuracy: 0.8302
---- Training ----
Training loss: 213.5438
Training acc over epoch: 0.8297
---- Validation ----
Validation loss: 83.5850
Validation acc: 0.7391
Time taken: 9.52s
Start of epoch 28
Training loss (for one batch) at step 0: 240.8457, Accuracy: 0.8500
Training loss (for one batch) at step 10: 250.0671, Accuracy: 0.8309
Training loss (for one batch) at step 20: 259.5652, Accuracy: 0.8271
Training loss (for one batch) at step 30: 278.5872, Accuracy: 0.8294
Training loss (for one batch) at step 40: 265.2866, Accuracy: 0.8320
Training loss (for one batch) at step 50: 241.9479, Accuracy: 0.8320
Training loss (for one batch) at step 60: 249.6399, Accuracy: 0.8341
Training loss (for one batch) at step 70: 257.1518, Accuracy: 0.8337
Training loss (for one batch) at step 80: 254.5041, Accuracy: 0.8309
Training loss (for one batch) at step 90: 254.7278, Accuracy: 0.8301
Training loss (for one batch) at step 100: 242.3365, Accuracy: 0.8302
Training loss (for one batch) at step 110: 260.6128, Accuracy: 0.8307
Training loss (for one batch) at step 120: 236.3625, Accuracy: 0.8298
Training loss (for one batch) at step 130: 254.4282, Accuracy: 0.8304
Training loss (for one batch) at step 140: 264.2267, Accuracy: 0.8299
---- Training ----
Training loss: 208.5976
Training acc over epoch: 0.8297
---- Validation ----
Validation loss: 82.2375
Validation acc: 0.7265
Time taken: 9.68s
Start of epoch 29
Training loss (for one batch) at step 0: 234.9448, Accuracy: 0.9200
Training loss (for one batch) at step 10: 244.0874, Accuracy: 0.8418
Training loss (for one batch) at step 20: 257.8802, Accuracy: 0.8362
Training loss (for one batch) at step 30: 277.6575, Accuracy: 0.8355
Training loss (for one batch) at step 40: 233.8902, Accuracy: 0.8373
Training loss (for one batch) at step 50: 241.7343, Accuracy: 0.8388
Training loss (for one batch) at step 60: 239.6343, Accuracy: 0.8390
Training loss (for one batch) at step 70: 245.3801, Accuracy: 0.8407
Training loss (for one batch) at step 80: 261.7957, Accuracy: 0.8384
Training loss (for one batch) at step 90: 248.9822, Accuracy: 0.8375
Training loss (for one batch) at step 100: 236.4170, Accuracy: 0.8366
Training loss (for one batch) at step 110: 239.1794, Accuracy: 0.8375
Training loss (for one batch) at step 120: 234.6507, Accuracy: 0.8366
Training loss (for one batch) at step 130: 238.7095, Accuracy: 0.8351
Training loss (for one batch) at step 140: 225.8548, Accuracy: 0.8350
---- Training ----
Training loss: 218.6456
Training acc over epoch: 0.8355
---- Validation ----
Validation loss: 74.0508
Validation acc: 0.7300
Time taken: 9.57s
Start of epoch 30
Training loss (for one batch) at step 0: 253.9972, Accuracy: 0.7900
Training loss (for one batch) at step 10: 232.6485, Accuracy: 0.8282
Training loss (for one batch) at step 20: 242.5355, Accuracy: 0.8248
Training loss (for one batch) at step 30: 227.6811, Accuracy: 0.8303
Training loss (for one batch) at step 40: 214.0028, Accuracy: 0.8405
Training loss (for one batch) at step 50: 225.3032, Accuracy: 0.8412
Training loss (for one batch) at step 60: 232.4945, Accuracy: 0.8436
Training loss (for one batch) at step 70: 255.8896, Accuracy: 0.8401
Training loss (for one batch) at step 80: 256.5601, Accuracy: 0.8381
Training loss (for one batch) at step 90: 233.5290, Accuracy: 0.8380
Training loss (for one batch) at step 100: 245.9796, Accuracy: 0.8361
Training loss (for one batch) at step 110: 242.3031, Accuracy: 0.8377
Training loss (for one batch) at step 120: 239.8906, Accuracy: 0.8379
Training loss (for one batch) at step 130: 249.8137, Accuracy: 0.8368
Training loss (for one batch) at step 140: 234.6663, Accuracy: 0.8362
---- Training ----
Training loss: 215.4558
Training acc over epoch: 0.8360
---- Validation ----
Validation loss: 71.2096
Validation acc: 0.7176
Time taken: 9.59s
Start of epoch 31
Training loss (for one batch) at step 0: 244.8465, Accuracy: 0.8300
Training loss (for one batch) at step 10: 236.5500, Accuracy: 0.8627
Training loss (for one batch) at step 20: 247.4740, Accuracy: 0.8510
Training loss (for one batch) at step 30: 232.8262, Accuracy: 0.8458
Training loss (for one batch) at step 40: 222.6032, Accuracy: 0.8495
Training loss (for one batch) at step 50: 223.8422, Accuracy: 0.8512
Training loss (for one batch) at step 60: 228.9022, Accuracy: 0.8497
Training loss (for one batch) at step 70: 238.4008, Accuracy: 0.8501
Training loss (for one batch) at step 80: 238.5222, Accuracy: 0.8464
Training loss (for one batch) at step 90: 246.8759, Accuracy: 0.8435
Training loss (for one batch) at step 100: 225.1709, Accuracy: 0.8435
Training loss (for one batch) at step 110: 254.9440, Accuracy: 0.8447
Training loss (for one batch) at step 120: 223.5086, Accuracy: 0.8450
Training loss (for one batch) at step 130: 260.0201, Accuracy: 0.8444
Training loss (for one batch) at step 140: 226.6798, Accuracy: 0.8440
---- Training ----
Training loss: 196.1309
Training acc over epoch: 0.8438
---- Validation ----
Validation loss: 65.5344
Validation acc: 0.7303
Time taken: 9.76s
Start of epoch 32
Training loss (for one batch) at step 0: 245.6620, Accuracy: 0.8400
Training loss (for one batch) at step 10: 249.9153, Accuracy: 0.8227
Training loss (for one batch) at step 20: 239.4022, Accuracy: 0.8357
Training loss (for one batch) at step 30: 239.2310, Accuracy: 0.8361
Training loss (for one batch) at step 40: 232.2878, Accuracy: 0.8410
Training loss (for one batch) at step 50: 211.8647, Accuracy: 0.8445
Training loss (for one batch) at step 60: 235.3716, Accuracy: 0.8448
Training loss (for one batch) at step 70: 247.5717, Accuracy: 0.8454
Training loss (for one batch) at step 80: 236.4346, Accuracy: 0.8425
Training loss (for one batch) at step 90: 231.2220, Accuracy: 0.8398
Training loss (for one batch) at step 100: 229.5193, Accuracy: 0.8388
Training loss (for one batch) at step 110: 230.6603, Accuracy: 0.8398
Training loss (for one batch) at step 120: 247.0635, Accuracy: 0.8379
Training loss (for one batch) at step 130: 246.2391, Accuracy: 0.8386
Training loss (for one batch) at step 140: 219.3626, Accuracy: 0.8384
---- Training ----
Training loss: 226.0105
Training acc over epoch: 0.8399
---- Validation ----
Validation loss: 69.3452
Validation acc: 0.7329
Time taken: 11.97s
Start of epoch 33
Training loss (for one batch) at step 0: 244.7112, Accuracy: 0.7700
Training loss (for one batch) at step 10: 226.9756, Accuracy: 0.8664
Training loss (for one batch) at step 20: 231.2593, Accuracy: 0.8590
Training loss (for one batch) at step 30: 218.7529, Accuracy: 0.8500
Training loss (for one batch) at step 40: 231.0149, Accuracy: 0.8485
Training loss (for one batch) at step 50: 214.4648, Accuracy: 0.8486
Training loss (for one batch) at step 60: 229.7529, Accuracy: 0.8490
Training loss (for one batch) at step 70: 240.3711, Accuracy: 0.8486
Training loss (for one batch) at step 80: 241.3553, Accuracy: 0.8460
Training loss (for one batch) at step 90: 258.1209, Accuracy: 0.8456
Training loss (for one batch) at step 100: 236.4638, Accuracy: 0.8442
Training loss (for one batch) at step 110: 226.5936, Accuracy: 0.8457
Training loss (for one batch) at step 120: 239.6080, Accuracy: 0.8448
Training loss (for one batch) at step 130: 230.5999, Accuracy: 0.8455
Training loss (for one batch) at step 140: 229.2570, Accuracy: 0.8448
---- Training ----
Training loss: 207.7554
Training acc over epoch: 0.8440
---- Validation ----
Validation loss: 86.3289
Validation acc: 0.7289
Time taken: 9.61s
Start of epoch 34
Training loss (for one batch) at step 0: 229.9075, Accuracy: 0.8700
Training loss (for one batch) at step 10: 235.6371, Accuracy: 0.8691
Training loss (for one batch) at step 20: 221.1867, Accuracy: 0.8638
Training loss (for one batch) at step 30: 233.0702, Accuracy: 0.8597
Training loss (for one batch) at step 40: 223.8034, Accuracy: 0.8529
Training loss (for one batch) at step 50: 222.3977, Accuracy: 0.8539
Training loss (for one batch) at step 60: 228.8967, Accuracy: 0.8539
Training loss (for one batch) at step 70: 249.3415, Accuracy: 0.8534
Training loss (for one batch) at step 80: 225.5385, Accuracy: 0.8520
Training loss (for one batch) at step 90: 243.4591, Accuracy: 0.8482
Training loss (for one batch) at step 100: 243.5622, Accuracy: 0.8491
Training loss (for one batch) at step 110: 227.2881, Accuracy: 0.8493
Training loss (for one batch) at step 120: 267.4191, Accuracy: 0.8484
Training loss (for one batch) at step 130: 216.9229, Accuracy: 0.8473
Training loss (for one batch) at step 140: 225.6045, Accuracy: 0.8462
---- Training ----
Training loss: 204.3867
Training acc over epoch: 0.8459
---- Validation ----
Validation loss: 72.7294
Validation acc: 0.7268
Time taken: 9.77s
Start of epoch 35
Training loss (for one batch) at step 0: 208.4686, Accuracy: 0.8600
Training loss (for one batch) at step 10: 226.6248, Accuracy: 0.8509
Training loss (for one batch) at step 20: 229.0807, Accuracy: 0.8476
Training loss (for one batch) at step 30: 251.1629, Accuracy: 0.8471
Training loss (for one batch) at step 40: 218.7491, Accuracy: 0.8480
Training loss (for one batch) at step 50: 237.4094, Accuracy: 0.8504
Training loss (for one batch) at step 60: 220.0597, Accuracy: 0.8505
Training loss (for one batch) at step 70: 219.8678, Accuracy: 0.8524
Training loss (for one batch) at step 80: 225.7566, Accuracy: 0.8498
Training loss (for one batch) at step 90: 249.7518, Accuracy: 0.8487
Training loss (for one batch) at step 100: 236.8051, Accuracy: 0.8474
Training loss (for one batch) at step 110: 230.3651, Accuracy: 0.8469
Training loss (for one batch) at step 120: 251.5175, Accuracy: 0.8460
Training loss (for one batch) at step 130: 233.2234, Accuracy: 0.8469
Training loss (for one batch) at step 140: 217.2608, Accuracy: 0.8473
---- Training ----
Training loss: 193.6242
Training acc over epoch: 0.8471
---- Validation ----
Validation loss: 82.7324
Validation acc: 0.7407
Time taken: 9.62s
Start of epoch 36
Training loss (for one batch) at step 0: 236.0006, Accuracy: 0.7600
Training loss (for one batch) at step 10: 275.4685, Accuracy: 0.8336
Training loss (for one batch) at step 20: 224.0850, Accuracy: 0.8405
Training loss (for one batch) at step 30: 229.4396, Accuracy: 0.8484
Training loss (for one batch) at step 40: 214.2181, Accuracy: 0.8507
Training loss (for one batch) at step 50: 219.6051, Accuracy: 0.8543
Training loss (for one batch) at step 60: 210.5523, Accuracy: 0.8559
Training loss (for one batch) at step 70: 221.7872, Accuracy: 0.8545
Training loss (for one batch) at step 80: 252.8681, Accuracy: 0.8517
Training loss (for one batch) at step 90: 248.5285, Accuracy: 0.8485
Training loss (for one batch) at step 100: 220.0254, Accuracy: 0.8479
Training loss (for one batch) at step 110: 218.4499, Accuracy: 0.8478
Training loss (for one batch) at step 120: 236.8250, Accuracy: 0.8470
Training loss (for one batch) at step 130: 222.1794, Accuracy: 0.8462
Training loss (for one batch) at step 140: 225.6872, Accuracy: 0.8468
---- Training ----
Training loss: 213.6448
Training acc over epoch: 0.8459
---- Validation ----
Validation loss: 78.9477
Validation acc: 0.7233
Time taken: 9.55s
Start of epoch 37
Training loss (for one batch) at step 0: 249.1841, Accuracy: 0.8000
Training loss (for one batch) at step 10: 258.0826, Accuracy: 0.8455
Training loss (for one batch) at step 20: 233.1591, Accuracy: 0.8490
Training loss (for one batch) at step 30: 234.6354, Accuracy: 0.8506
Training loss (for one batch) at step 40: 216.6036, Accuracy: 0.8534
Training loss (for one batch) at step 50: 226.4022, Accuracy: 0.8569
Training loss (for one batch) at step 60: 227.7011, Accuracy: 0.8559
Training loss (for one batch) at step 70: 235.0992, Accuracy: 0.8558
Training loss (for one batch) at step 80: 239.4027, Accuracy: 0.8523
Training loss (for one batch) at step 90: 240.1903, Accuracy: 0.8533
Training loss (for one batch) at step 100: 224.4450, Accuracy: 0.8511
Training loss (for one batch) at step 110: 206.4951, Accuracy: 0.8518
Training loss (for one batch) at step 120: 238.6525, Accuracy: 0.8526
Training loss (for one batch) at step 130: 212.2846, Accuracy: 0.8520
Training loss (for one batch) at step 140: 224.5153, Accuracy: 0.8513
---- Training ----
Training loss: 195.0289
Training acc over epoch: 0.8501
---- Validation ----
Validation loss: 71.0064
Validation acc: 0.7268
Time taken: 9.74s
Start of epoch 38
Training loss (for one batch) at step 0: 217.5391, Accuracy: 0.8000
Training loss (for one batch) at step 10: 233.3684, Accuracy: 0.8418
Training loss (for one batch) at step 20: 221.3058, Accuracy: 0.8419
Training loss (for one batch) at step 30: 239.0179, Accuracy: 0.8371
Training loss (for one batch) at step 40: 217.2572, Accuracy: 0.8454
Training loss (for one batch) at step 50: 224.2989, Accuracy: 0.8502
Training loss (for one batch) at step 60: 222.0026, Accuracy: 0.8502
Training loss (for one batch) at step 70: 226.7914, Accuracy: 0.8520
Training loss (for one batch) at step 80: 225.1154, Accuracy: 0.8498
Training loss (for one batch) at step 90: 238.7160, Accuracy: 0.8502
Training loss (for one batch) at step 100: 223.8884, Accuracy: 0.8500
Training loss (for one batch) at step 110: 229.3535, Accuracy: 0.8499
Training loss (for one batch) at step 120: 239.7208, Accuracy: 0.8475
Training loss (for one batch) at step 130: 250.9081, Accuracy: 0.8473
Training loss (for one batch) at step 140: 213.8529, Accuracy: 0.8477
---- Training ----
Training loss: 198.9168
Training acc over epoch: 0.8476
---- Validation ----
Validation loss: 89.4679
Validation acc: 0.7292
Time taken: 9.56s
Start of epoch 39
Training loss (for one batch) at step 0: 239.5202, Accuracy: 0.8400
Training loss (for one batch) at step 10: 201.8464, Accuracy: 0.8527
Training loss (for one batch) at step 20: 215.8803, Accuracy: 0.8605
Training loss (for one batch) at step 30: 218.5889, Accuracy: 0.8600
Training loss (for one batch) at step 40: 224.1362, Accuracy: 0.8595
Training loss (for one batch) at step 50: 224.3408, Accuracy: 0.8584
Training loss (for one batch) at step 60: 207.1462, Accuracy: 0.8590
Training loss (for one batch) at step 70: 223.2149, Accuracy: 0.8562
Training loss (for one batch) at step 80: 210.5287, Accuracy: 0.8547
Training loss (for one batch) at step 90: 233.2205, Accuracy: 0.8538
Training loss (for one batch) at step 100: 215.8994, Accuracy: 0.8533
Training loss (for one batch) at step 110: 204.4661, Accuracy: 0.8545
Training loss (for one batch) at step 120: 240.0645, Accuracy: 0.8539
Training loss (for one batch) at step 130: 223.5328, Accuracy: 0.8534
Training loss (for one batch) at step 140: 214.7520, Accuracy: 0.8525
---- Training ----
Training loss: 192.4134
Training acc over epoch: 0.8522
---- Validation ----
Validation loss: 101.1248
Validation acc: 0.7464
Time taken: 9.55s
Start of epoch 40
Training loss (for one batch) at step 0: 225.3490, Accuracy: 0.8500
Training loss (for one batch) at step 10: 228.7551, Accuracy: 0.8673
Training loss (for one batch) at step 20: 215.2202, Accuracy: 0.8643
Training loss (for one batch) at step 30: 197.7697, Accuracy: 0.8565
Training loss (for one batch) at step 40: 216.7826, Accuracy: 0.8559
Training loss (for one batch) at step 50: 210.9516, Accuracy: 0.8563
Training loss (for one batch) at step 60: 185.4324, Accuracy: 0.8605
Training loss (for one batch) at step 70: 221.6328, Accuracy: 0.8597
Training loss (for one batch) at step 80: 237.9002, Accuracy: 0.8560
Training loss (for one batch) at step 90: 243.1780, Accuracy: 0.8542
Training loss (for one batch) at step 100: 220.1061, Accuracy: 0.8536
Training loss (for one batch) at step 110: 216.0082, Accuracy: 0.8541
Training loss (for one batch) at step 120: 216.7163, Accuracy: 0.8530
Training loss (for one batch) at step 130: 221.8288, Accuracy: 0.8538
Training loss (for one batch) at step 140: 243.2987, Accuracy: 0.8538
---- Training ----
Training loss: 187.5336
Training acc over epoch: 0.8530
---- Validation ----
Validation loss: 65.3169
Validation acc: 0.7399
Time taken: 12.26s
Start of epoch 41
Training loss (for one batch) at step 0: 223.1430, Accuracy: 0.8300
Training loss (for one batch) at step 10: 214.4745, Accuracy: 0.8609
Training loss (for one batch) at step 20: 224.4166, Accuracy: 0.8576
Training loss (for one batch) at step 30: 204.4060, Accuracy: 0.8558
Training loss (for one batch) at step 40: 229.6153, Accuracy: 0.8541
Training loss (for one batch) at step 50: 203.5903, Accuracy: 0.8608
Training loss (for one batch) at step 60: 200.2915, Accuracy: 0.8625
Training loss (for one batch) at step 70: 231.5151, Accuracy: 0.8601
Training loss (for one batch) at step 80: 213.8078, Accuracy: 0.8581
Training loss (for one batch) at step 90: 223.4111, Accuracy: 0.8568
Training loss (for one batch) at step 100: 249.9630, Accuracy: 0.8551
Training loss (for one batch) at step 110: 232.8295, Accuracy: 0.8560
Training loss (for one batch) at step 120: 204.2507, Accuracy: 0.8560
Training loss (for one batch) at step 130: 211.0639, Accuracy: 0.8555
Training loss (for one batch) at step 140: 205.2897, Accuracy: 0.8550
---- Training ----
Training loss: 205.8790
Training acc over epoch: 0.8546
---- Validation ----
Validation loss: 73.7562
Validation acc: 0.7340
Time taken: 11.61s
Start of epoch 42
Training loss (for one batch) at step 0: 210.7261, Accuracy: 0.8800
Training loss (for one batch) at step 10: 219.5418, Accuracy: 0.8618
Training loss (for one batch) at step 20: 223.3268, Accuracy: 0.8610
Training loss (for one batch) at step 30: 224.7014, Accuracy: 0.8555
Training loss (for one batch) at step 40: 207.7954, Accuracy: 0.8571
Training loss (for one batch) at step 50: 207.5413, Accuracy: 0.8590
Training loss (for one batch) at step 60: 212.0487, Accuracy: 0.8610
Training loss (for one batch) at step 70: 233.3741, Accuracy: 0.8596
Training loss (for one batch) at step 80: 220.9644, Accuracy: 0.8560
Training loss (for one batch) at step 90: 242.7449, Accuracy: 0.8536
Training loss (for one batch) at step 100: 204.8577, Accuracy: 0.8532
Training loss (for one batch) at step 110: 208.1953, Accuracy: 0.8541
Training loss (for one batch) at step 120: 236.1350, Accuracy: 0.8555
Training loss (for one batch) at step 130: 226.7918, Accuracy: 0.8550
Training loss (for one batch) at step 140: 190.4578, Accuracy: 0.8548
---- Training ----
Training loss: 212.8306
Training acc over epoch: 0.8559
---- Validation ----
Validation loss: 78.1958
Validation acc: 0.7380
Time taken: 12.03s
Start of epoch 43
Training loss (for one batch) at step 0: 211.8441, Accuracy: 0.8000
Training loss (for one batch) at step 10: 211.3328, Accuracy: 0.8682
Training loss (for one batch) at step 20: 216.5611, Accuracy: 0.8638
Training loss (for one batch) at step 30: 215.6792, Accuracy: 0.8619
Training loss (for one batch) at step 40: 214.7200, Accuracy: 0.8598
Training loss (for one batch) at step 50: 220.4537, Accuracy: 0.8616
Training loss (for one batch) at step 60: 217.9123, Accuracy: 0.8623
Training loss (for one batch) at step 70: 219.7946, Accuracy: 0.8631
Training loss (for one batch) at step 80: 205.6727, Accuracy: 0.8596
Training loss (for one batch) at step 90: 214.1844, Accuracy: 0.8595
Training loss (for one batch) at step 100: 240.4062, Accuracy: 0.8562
Training loss (for one batch) at step 110: 209.3367, Accuracy: 0.8571
Training loss (for one batch) at step 120: 217.5429, Accuracy: 0.8581
Training loss (for one batch) at step 130: 217.9002, Accuracy: 0.8579
Training loss (for one batch) at step 140: 182.7216, Accuracy: 0.8568
---- Training ----
Training loss: 191.6472
Training acc over epoch: 0.8566
---- Validation ----
Validation loss: 81.1752
Validation acc: 0.7294
Time taken: 11.67s
Start of epoch 44
Training loss (for one batch) at step 0: 213.2601, Accuracy: 0.7500
Training loss (for one batch) at step 10: 210.4605, Accuracy: 0.8609
Training loss (for one batch) at step 20: 219.0646, Accuracy: 0.8552
Training loss (for one batch) at step 30: 211.3838, Accuracy: 0.8539
Training loss (for one batch) at step 40: 210.8233, Accuracy: 0.8556
Training loss (for one batch) at step 50: 201.3317, Accuracy: 0.8618
Training loss (for one batch) at step 60: 219.7709, Accuracy: 0.8589
Training loss (for one batch) at step 70: 217.9709, Accuracy: 0.8586
Training loss (for one batch) at step 80: 218.9336, Accuracy: 0.8581
Training loss (for one batch) at step 90: 237.6157, Accuracy: 0.8555
Training loss (for one batch) at step 100: 195.1134, Accuracy: 0.8571
Training loss (for one batch) at step 110: 226.8462, Accuracy: 0.8574
Training loss (for one batch) at step 120: 203.3912, Accuracy: 0.8579
Training loss (for one batch) at step 130: 211.7243, Accuracy: 0.8567
Training loss (for one batch) at step 140: 199.8425, Accuracy: 0.8562
---- Training ----
Training loss: 187.9834
Training acc over epoch: 0.8557
---- Validation ----
Validation loss: 76.3640
Validation acc: 0.7351
Time taken: 12.68s
Start of epoch 45
Training loss (for one batch) at step 0: 222.3538, Accuracy: 0.8700
Training loss (for one batch) at step 10: 210.8752, Accuracy: 0.8573
Training loss (for one batch) at step 20: 199.0697, Accuracy: 0.8533
Training loss (for one batch) at step 30: 203.8229, Accuracy: 0.8606
Training loss (for one batch) at step 40: 189.1152, Accuracy: 0.8620
Training loss (for one batch) at step 50: 197.7253, Accuracy: 0.8625
Training loss (for one batch) at step 60: 211.7287, Accuracy: 0.8621
Training loss (for one batch) at step 70: 244.4792, Accuracy: 0.8586
Training loss (for one batch) at step 80: 223.9814, Accuracy: 0.8593
Training loss (for one batch) at step 90: 222.8166, Accuracy: 0.8582
Training loss (for one batch) at step 100: 202.9300, Accuracy: 0.8578
Training loss (for one batch) at step 110: 176.4690, Accuracy: 0.8581
Training loss (for one batch) at step 120: 204.1123, Accuracy: 0.8588
Training loss (for one batch) at step 130: 212.3854, Accuracy: 0.8586
Training loss (for one batch) at step 140: 220.5320, Accuracy: 0.8562
---- Training ----
Training loss: 194.5763
Training acc over epoch: 0.8561
---- Validation ----
Validation loss: 75.9550
Validation acc: 0.7157
Time taken: 9.59s
Start of epoch 46
Training loss (for one batch) at step 0: 230.0925, Accuracy: 0.8000
Training loss (for one batch) at step 10: 200.1562, Accuracy: 0.8473
Training loss (for one batch) at step 20: 223.3355, Accuracy: 0.8543
Training loss (for one batch) at step 30: 220.0371, Accuracy: 0.8519
Training loss (for one batch) at step 40: 194.4322, Accuracy: 0.8578
Training loss (for one batch) at step 50: 203.2601, Accuracy: 0.8600
Training loss (for one batch) at step 60: 201.8698, Accuracy: 0.8585
Training loss (for one batch) at step 70: 214.8760, Accuracy: 0.8576
Training loss (for one batch) at step 80: 220.6109, Accuracy: 0.8559
Training loss (for one batch) at step 90: 207.3334, Accuracy: 0.8562
Training loss (for one batch) at step 100: 190.1530, Accuracy: 0.8570
Training loss (for one batch) at step 110: 199.3675, Accuracy: 0.8579
Training loss (for one batch) at step 120: 221.2017, Accuracy: 0.8579
Training loss (for one batch) at step 130: 213.6779, Accuracy: 0.8588
Training loss (for one batch) at step 140: 217.9612, Accuracy: 0.8593
---- Training ----
Training loss: 176.6621
Training acc over epoch: 0.8587
---- Validation ----
Validation loss: 82.8633
Validation acc: 0.7356
Time taken: 9.56s
Start of epoch 47
Training loss (for one batch) at step 0: 213.9652, Accuracy: 0.7900
Training loss (for one batch) at step 10: 210.8337, Accuracy: 0.8582
Training loss (for one batch) at step 20: 227.5936, Accuracy: 0.8562
Training loss (for one batch) at step 30: 215.8772, Accuracy: 0.8561
Training loss (for one batch) at step 40: 205.1736, Accuracy: 0.8576
Training loss (for one batch) at step 50: 203.0273, Accuracy: 0.8625
Training loss (for one batch) at step 60: 211.5575, Accuracy: 0.8618
Training loss (for one batch) at step 70: 193.8993, Accuracy: 0.8617
Training loss (for one batch) at step 80: 197.0710, Accuracy: 0.8604
Training loss (for one batch) at step 90: 215.0128, Accuracy: 0.8592
Training loss (for one batch) at step 100: 213.6631, Accuracy: 0.8588
Training loss (for one batch) at step 110: 195.7118, Accuracy: 0.8601
Training loss (for one batch) at step 120: 203.1730, Accuracy: 0.8601
Training loss (for one batch) at step 130: 196.2987, Accuracy: 0.8598
Training loss (for one batch) at step 140: 194.9056, Accuracy: 0.8600
---- Training ----
Training loss: 181.0541
Training acc over epoch: 0.8599
---- Validation ----
Validation loss: 77.7909
Validation acc: 0.7491
Time taken: 9.53s
Start of epoch 48
Training loss (for one batch) at step 0: 197.4661, Accuracy: 0.8900
Training loss (for one batch) at step 10: 194.8199, Accuracy: 0.8709
Training loss (for one batch) at step 20: 219.4585, Accuracy: 0.8676
Training loss (for one batch) at step 30: 215.6848, Accuracy: 0.8616
Training loss (for one batch) at step 40: 222.6736, Accuracy: 0.8632
Training loss (for one batch) at step 50: 209.0684, Accuracy: 0.8645
Training loss (for one batch) at step 60: 192.6278, Accuracy: 0.8664
Training loss (for one batch) at step 70: 219.3259, Accuracy: 0.8611
Training loss (for one batch) at step 80: 226.5875, Accuracy: 0.8598
Training loss (for one batch) at step 90: 214.3805, Accuracy: 0.8589
Training loss (for one batch) at step 100: 207.9332, Accuracy: 0.8583
Training loss (for one batch) at step 110: 194.4805, Accuracy: 0.8584
Training loss (for one batch) at step 120: 207.2030, Accuracy: 0.8586
Training loss (for one batch) at step 130: 201.6017, Accuracy: 0.8595
Training loss (for one batch) at step 140: 209.0617, Accuracy: 0.8589
---- Training ----
Training loss: 181.4627
Training acc over epoch: 0.8591
---- Validation ----
Validation loss: 96.0079
Validation acc: 0.7303
Time taken: 9.56s
Start of epoch 49
Training loss (for one batch) at step 0: 204.8411, Accuracy: 0.8500
Training loss (for one batch) at step 10: 199.5325, Accuracy: 0.8709
Training loss (for one batch) at step 20: 204.9743, Accuracy: 0.8643
Training loss (for one batch) at step 30: 215.7510, Accuracy: 0.8648
Training loss (for one batch) at step 40: 185.2607, Accuracy: 0.8661
Training loss (for one batch) at step 50: 191.7678, Accuracy: 0.8694
Training loss (for one batch) at step 60: 191.7196, Accuracy: 0.8666
Training loss (for one batch) at step 70: 210.0619, Accuracy: 0.8627
Training loss (for one batch) at step 80: 198.7667, Accuracy: 0.8621
Training loss (for one batch) at step 90: 213.3179, Accuracy: 0.8600
Training loss (for one batch) at step 100: 203.1328, Accuracy: 0.8593
Training loss (for one batch) at step 110: 218.2767, Accuracy: 0.8590
Training loss (for one batch) at step 120: 214.2911, Accuracy: 0.8588
Training loss (for one batch) at step 130: 203.7892, Accuracy: 0.8591
Training loss (for one batch) at step 140: 208.8095, Accuracy: 0.8593
---- Training ----
Training loss: 182.7482
Training acc over epoch: 0.8588
---- Validation ----
Validation loss: 70.6561
Validation acc: 0.7364
Time taken: 9.60s
===== Q: 0.0001
Validation acc: 0.7496
Validation AUC: 0.7471
Validation Balanced_ACC: 0.4953
Validation MI: 0.1444
Validation Normalized MI: 0.2161
Validation Adjusted MI: 0.2161
Validation aUc_Sklearn: 0.8370
Start of epoch 0
Training loss (for one batch) at step 0: 532.5384, Accuracy: 0.4900
Training loss (for one batch) at step 10: 488.6142, Accuracy: 0.5136
Training loss (for one batch) at step 20: 499.7857, Accuracy: 0.5181
Training loss (for one batch) at step 30: 515.0516, Accuracy: 0.5252
Training loss (for one batch) at step 40: 469.2635, Accuracy: 0.5290
Training loss (for one batch) at step 50: 474.1211, Accuracy: 0.5355
Training loss (for one batch) at step 60: 507.0574, Accuracy: 0.5382
Training loss (for one batch) at step 70: 437.6133, Accuracy: 0.5414
Training loss (for one batch) at step 80: 412.1067, Accuracy: 0.5457
Training loss (for one batch) at step 90: 427.4181, Accuracy: 0.5479
Training loss (for one batch) at step 100: 413.1493, Accuracy: 0.5527
Training loss (for one batch) at step 110: 414.5948, Accuracy: 0.5549
Training loss (for one batch) at step 120: 467.1740, Accuracy: 0.5559
Training loss (for one batch) at step 130: 445.4490, Accuracy: 0.5578
Training loss (for one batch) at step 140: 465.0882, Accuracy: 0.5594
---- Training ----
Training loss: 381.7527
Training acc over epoch: 0.5582
---- Validation ----
Validation loss: 91.4688
Validation acc: 0.5134
Time taken: 12.93s
Start of epoch 1
Training loss (for one batch) at step 0: 423.3222, Accuracy: 0.5700
Training loss (for one batch) at step 10: 386.0050, Accuracy: 0.6082
Training loss (for one batch) at step 20: 406.3389, Accuracy: 0.6119
Training loss (for one batch) at step 30: 416.2047, Accuracy: 0.6013
Training loss (for one batch) at step 40: 389.9232, Accuracy: 0.5963
Training loss (for one batch) at step 50: 394.0270, Accuracy: 0.5912
Training loss (for one batch) at step 60: 439.0099, Accuracy: 0.5926
Training loss (for one batch) at step 70: 395.7939, Accuracy: 0.5918
Training loss (for one batch) at step 80: 395.9757, Accuracy: 0.5954
Training loss (for one batch) at step 90: 400.1547, Accuracy: 0.5965
Training loss (for one batch) at step 100: 395.6791, Accuracy: 0.5950
Training loss (for one batch) at step 110: 368.7736, Accuracy: 0.5976
Training loss (for one batch) at step 120: 407.2421, Accuracy: 0.5986
Training loss (for one batch) at step 130: 381.8151, Accuracy: 0.5998
Training loss (for one batch) at step 140: 386.4744, Accuracy: 0.6011
---- Training ----
Training loss: 342.1550
Training acc over epoch: 0.5997
---- Validation ----
Validation loss: 77.7988
Validation acc: 0.5742
Time taken: 9.74s
Start of epoch 2
Training loss (for one batch) at step 0: 395.3844, Accuracy: 0.6000
Training loss (for one batch) at step 10: 382.1066, Accuracy: 0.6309
Training loss (for one batch) at step 20: 350.0018, Accuracy: 0.6281
Training loss (for one batch) at step 30: 375.0786, Accuracy: 0.6194
Training loss (for one batch) at step 40: 386.4919, Accuracy: 0.6115
Training loss (for one batch) at step 50: 400.8813, Accuracy: 0.6086
Training loss (for one batch) at step 60: 363.8419, Accuracy: 0.6121
Training loss (for one batch) at step 70: 365.3380, Accuracy: 0.6114
Training loss (for one batch) at step 80: 373.2495, Accuracy: 0.6109
Training loss (for one batch) at step 90: 348.4840, Accuracy: 0.6107
Training loss (for one batch) at step 100: 378.3018, Accuracy: 0.6110
Training loss (for one batch) at step 110: 382.3780, Accuracy: 0.6105
Training loss (for one batch) at step 120: 389.9186, Accuracy: 0.6088
Training loss (for one batch) at step 130: 375.7372, Accuracy: 0.6089
Training loss (for one batch) at step 140: 376.8884, Accuracy: 0.6105
---- Training ----
Training loss: 315.8708
Training acc over epoch: 0.6106
---- Validation ----
Validation loss: 76.7956
Validation acc: 0.6421
Time taken: 10.87s
Start of epoch 3
Training loss (for one batch) at step 0: 386.4293, Accuracy: 0.5500
Training loss (for one batch) at step 10: 345.0205, Accuracy: 0.6109
Training loss (for one batch) at step 20: 357.3569, Accuracy: 0.6114
Training loss (for one batch) at step 30: 350.7261, Accuracy: 0.6255
Training loss (for one batch) at step 40: 339.4309, Accuracy: 0.6249
Training loss (for one batch) at step 50: 377.9559, Accuracy: 0.6263
Training loss (for one batch) at step 60: 375.5185, Accuracy: 0.6246
Training loss (for one batch) at step 70: 340.7684, Accuracy: 0.6272
Training loss (for one batch) at step 80: 383.8318, Accuracy: 0.6277
Training loss (for one batch) at step 90: 354.1663, Accuracy: 0.6278
Training loss (for one batch) at step 100: 347.8271, Accuracy: 0.6262
Training loss (for one batch) at step 110: 349.0966, Accuracy: 0.6265
Training loss (for one batch) at step 120: 352.9415, Accuracy: 0.6286
Training loss (for one batch) at step 130: 353.2911, Accuracy: 0.6271
Training loss (for one batch) at step 140: 348.8654, Accuracy: 0.6276
---- Training ----
Training loss: 313.5492
Training acc over epoch: 0.6276
---- Validation ----
Validation loss: 70.5024
Validation acc: 0.6574
Time taken: 10.99s
Start of epoch 4
Training loss (for one batch) at step 0: 352.2755, Accuracy: 0.5700
Training loss (for one batch) at step 10: 336.2549, Accuracy: 0.6355
Training loss (for one batch) at step 20: 352.6798, Accuracy: 0.6357
Training loss (for one batch) at step 30: 344.6459, Accuracy: 0.6371
Training loss (for one batch) at step 40: 333.4243, Accuracy: 0.6373
Training loss (for one batch) at step 50: 342.2039, Accuracy: 0.6394
Training loss (for one batch) at step 60: 337.2793, Accuracy: 0.6415
Training loss (for one batch) at step 70: 350.7221, Accuracy: 0.6401
Training loss (for one batch) at step 80: 348.6115, Accuracy: 0.6377
Training loss (for one batch) at step 90: 375.6487, Accuracy: 0.6385
Training loss (for one batch) at step 100: 330.1545, Accuracy: 0.6387
Training loss (for one batch) at step 110: 352.1969, Accuracy: 0.6391
Training loss (for one batch) at step 120: 344.3016, Accuracy: 0.6421
Training loss (for one batch) at step 130: 329.2963, Accuracy: 0.6422
Training loss (for one batch) at step 140: 322.1857, Accuracy: 0.6430
---- Training ----
Training loss: 293.7799
Training acc over epoch: 0.6422
---- Validation ----
Validation loss: 70.3065
Validation acc: 0.6795
Time taken: 10.71s
Start of epoch 5
Training loss (for one batch) at step 0: 338.7166, Accuracy: 0.6700
Training loss (for one batch) at step 10: 336.9220, Accuracy: 0.6700
Training loss (for one batch) at step 20: 324.5417, Accuracy: 0.6690
Training loss (for one batch) at step 30: 336.1305, Accuracy: 0.6594
Training loss (for one batch) at step 40: 330.5879, Accuracy: 0.6554
Training loss (for one batch) at step 50: 318.3790, Accuracy: 0.6606
Training loss (for one batch) at step 60: 343.9391, Accuracy: 0.6644
Training loss (for one batch) at step 70: 332.9477, Accuracy: 0.6652
Training loss (for one batch) at step 80: 322.4262, Accuracy: 0.6664
Training loss (for one batch) at step 90: 331.2088, Accuracy: 0.6673
Training loss (for one batch) at step 100: 337.5417, Accuracy: 0.6639
Training loss (for one batch) at step 110: 327.0276, Accuracy: 0.6647
Training loss (for one batch) at step 120: 330.8496, Accuracy: 0.6645
Training loss (for one batch) at step 130: 335.1422, Accuracy: 0.6638
Training loss (for one batch) at step 140: 329.7738, Accuracy: 0.6656
---- Training ----
Training loss: 297.7567
Training acc over epoch: 0.6656
---- Validation ----
Validation loss: 66.9397
Validation acc: 0.6838
Time taken: 10.80s
Start of epoch 6
Training loss (for one batch) at step 0: 317.3223, Accuracy: 0.7000
Training loss (for one batch) at step 10: 330.2586, Accuracy: 0.6636
Training loss (for one batch) at step 20: 333.4373, Accuracy: 0.6700
Training loss (for one batch) at step 30: 341.3629, Accuracy: 0.6668
Training loss (for one batch) at step 40: 318.1258, Accuracy: 0.6717
Training loss (for one batch) at step 50: 338.6136, Accuracy: 0.6761
Training loss (for one batch) at step 60: 327.3754, Accuracy: 0.6774
Training loss (for one batch) at step 70: 333.1983, Accuracy: 0.6780
Training loss (for one batch) at step 80: 310.3654, Accuracy: 0.6815
Training loss (for one batch) at step 90: 332.8737, Accuracy: 0.6798
Training loss (for one batch) at step 100: 308.6673, Accuracy: 0.6821
Training loss (for one batch) at step 110: 328.4141, Accuracy: 0.6826
Training loss (for one batch) at step 120: 328.3063, Accuracy: 0.6838
Training loss (for one batch) at step 130: 325.1132, Accuracy: 0.6850
Training loss (for one batch) at step 140: 321.8087, Accuracy: 0.6859
---- Training ----
Training loss: 278.9861
Training acc over epoch: 0.6857
---- Validation ----
Validation loss: 72.5834
Validation acc: 0.6797
Time taken: 10.98s
Start of epoch 7
Training loss (for one batch) at step 0: 324.7319, Accuracy: 0.6700
Training loss (for one batch) at step 10: 320.4777, Accuracy: 0.6873
Training loss (for one batch) at step 20: 317.8574, Accuracy: 0.6948
Training loss (for one batch) at step 30: 319.7660, Accuracy: 0.6913
Training loss (for one batch) at step 40: 319.2584, Accuracy: 0.6888
Training loss (for one batch) at step 50: 321.5404, Accuracy: 0.6914
Training loss (for one batch) at step 60: 316.7865, Accuracy: 0.6923
Training loss (for one batch) at step 70: 323.0349, Accuracy: 0.6982
Training loss (for one batch) at step 80: 328.0501, Accuracy: 0.6990
Training loss (for one batch) at step 90: 317.0677, Accuracy: 0.6973
Training loss (for one batch) at step 100: 316.0226, Accuracy: 0.6998
Training loss (for one batch) at step 110: 329.0569, Accuracy: 0.7011
Training loss (for one batch) at step 120: 328.1413, Accuracy: 0.7009
Training loss (for one batch) at step 130: 311.4375, Accuracy: 0.6993
Training loss (for one batch) at step 140: 315.8752, Accuracy: 0.7000
---- Training ----
Training loss: 278.0114
Training acc over epoch: 0.6987
---- Validation ----
Validation loss: 68.0495
Validation acc: 0.6929
Time taken: 11.93s
Start of epoch 8
Training loss (for one batch) at step 0: 303.3510, Accuracy: 0.7300
Training loss (for one batch) at step 10: 305.7903, Accuracy: 0.6945
Training loss (for one batch) at step 20: 315.8214, Accuracy: 0.7195
Training loss (for one batch) at step 30: 314.0969, Accuracy: 0.7255
Training loss (for one batch) at step 40: 303.2455, Accuracy: 0.7271
Training loss (for one batch) at step 50: 329.9886, Accuracy: 0.7243
Training loss (for one batch) at step 60: 303.2504, Accuracy: 0.7280
Training loss (for one batch) at step 70: 306.6991, Accuracy: 0.7280
Training loss (for one batch) at step 80: 309.1677, Accuracy: 0.7263
Training loss (for one batch) at step 90: 331.5750, Accuracy: 0.7245
Training loss (for one batch) at step 100: 329.2792, Accuracy: 0.7215
Training loss (for one batch) at step 110: 312.7532, Accuracy: 0.7228
Training loss (for one batch) at step 120: 315.2220, Accuracy: 0.7238
Training loss (for one batch) at step 130: 321.9098, Accuracy: 0.7237
Training loss (for one batch) at step 140: 308.0268, Accuracy: 0.7236
---- Training ----
Training loss: 291.8036
Training acc over epoch: 0.7225
---- Validation ----
Validation loss: 67.5904
Validation acc: 0.6943
Time taken: 11.33s
Start of epoch 9
Training loss (for one batch) at step 0: 317.4150, Accuracy: 0.6600
Training loss (for one batch) at step 10: 307.0728, Accuracy: 0.7236
Training loss (for one batch) at step 20: 292.4958, Accuracy: 0.7205
Training loss (for one batch) at step 30: 301.7903, Accuracy: 0.7174
Training loss (for one batch) at step 40: 298.4632, Accuracy: 0.7259
Training loss (for one batch) at step 50: 297.8342, Accuracy: 0.7276
Training loss (for one batch) at step 60: 293.4085, Accuracy: 0.7289
Training loss (for one batch) at step 70: 327.3109, Accuracy: 0.7301
Training loss (for one batch) at step 80: 326.4135, Accuracy: 0.7296
Training loss (for one batch) at step 90: 321.7564, Accuracy: 0.7297
Training loss (for one batch) at step 100: 301.1256, Accuracy: 0.7306
Training loss (for one batch) at step 110: 302.3408, Accuracy: 0.7302
Training loss (for one batch) at step 120: 309.2495, Accuracy: 0.7305
Training loss (for one batch) at step 130: 287.2078, Accuracy: 0.7285
Training loss (for one batch) at step 140: 298.7309, Accuracy: 0.7276
---- Training ----
Training loss: 286.6552
Training acc over epoch: 0.7276
---- Validation ----
Validation loss: 78.5559
Validation acc: 0.6937
Time taken: 11.27s
Start of epoch 10
Training loss (for one batch) at step 0: 303.0883, Accuracy: 0.8300
Training loss (for one batch) at step 10: 294.1454, Accuracy: 0.7436
Training loss (for one batch) at step 20: 312.7238, Accuracy: 0.7424
Training loss (for one batch) at step 30: 312.7216, Accuracy: 0.7355
Training loss (for one batch) at step 40: 313.3586, Accuracy: 0.7295
Training loss (for one batch) at step 50: 298.4822, Accuracy: 0.7382
Training loss (for one batch) at step 60: 302.1435, Accuracy: 0.7387
Training loss (for one batch) at step 70: 321.3716, Accuracy: 0.7379
Training loss (for one batch) at step 80: 317.8369, Accuracy: 0.7377
Training loss (for one batch) at step 90: 305.5776, Accuracy: 0.7359
Training loss (for one batch) at step 100: 307.7971, Accuracy: 0.7362
Training loss (for one batch) at step 110: 302.8350, Accuracy: 0.7361
Training loss (for one batch) at step 120: 291.5175, Accuracy: 0.7352
Training loss (for one batch) at step 130: 303.6716, Accuracy: 0.7350
Training loss (for one batch) at step 140: 295.2291, Accuracy: 0.7326
---- Training ----
Training loss: 272.1477
Training acc over epoch: 0.7341
---- Validation ----
Validation loss: 71.7400
Validation acc: 0.6886
Time taken: 13.93s
Start of epoch 11
Training loss (for one batch) at step 0: 305.0111, Accuracy: 0.6900
Training loss (for one batch) at step 10: 303.1106, Accuracy: 0.7518
Training loss (for one batch) at step 20: 303.5155, Accuracy: 0.7610
Training loss (for one batch) at step 30: 292.7544, Accuracy: 0.7600
Training loss (for one batch) at step 40: 299.2800, Accuracy: 0.7512
Training loss (for one batch) at step 50: 282.3928, Accuracy: 0.7527
Training loss (for one batch) at step 60: 296.7701, Accuracy: 0.7513
Training loss (for one batch) at step 70: 309.7339, Accuracy: 0.7506
Training loss (for one batch) at step 80: 305.1432, Accuracy: 0.7481
Training loss (for one batch) at step 90: 301.9145, Accuracy: 0.7491
Training loss (for one batch) at step 100: 309.7827, Accuracy: 0.7469
Training loss (for one batch) at step 110: 291.7575, Accuracy: 0.7473
Training loss (for one batch) at step 120: 302.0829, Accuracy: 0.7470
Training loss (for one batch) at step 130: 301.5794, Accuracy: 0.7460
Training loss (for one batch) at step 140: 313.4924, Accuracy: 0.7457
---- Training ----
Training loss: 264.4988
Training acc over epoch: 0.7454
---- Validation ----
Validation loss: 69.1021
Validation acc: 0.6913
Time taken: 12.10s
Start of epoch 12
Training loss (for one batch) at step 0: 289.6073, Accuracy: 0.7500
Training loss (for one batch) at step 10: 287.1545, Accuracy: 0.7545
Training loss (for one batch) at step 20: 302.9305, Accuracy: 0.7586
Training loss (for one batch) at step 30: 318.5050, Accuracy: 0.7513
Training loss (for one batch) at step 40: 293.1088, Accuracy: 0.7520
Training loss (for one batch) at step 50: 297.7626, Accuracy: 0.7553
Training loss (for one batch) at step 60: 308.1187, Accuracy: 0.7564
Training loss (for one batch) at step 70: 278.6101, Accuracy: 0.7561
Training loss (for one batch) at step 80: 309.9602, Accuracy: 0.7562
Training loss (for one batch) at step 90: 287.4146, Accuracy: 0.7566
Training loss (for one batch) at step 100: 297.7722, Accuracy: 0.7548
Training loss (for one batch) at step 110: 302.0871, Accuracy: 0.7551
Training loss (for one batch) at step 120: 294.8892, Accuracy: 0.7559
Training loss (for one batch) at step 130: 304.1132, Accuracy: 0.7539
Training loss (for one batch) at step 140: 301.6970, Accuracy: 0.7538
---- Training ----
Training loss: 262.4887
Training acc over epoch: 0.7540
---- Validation ----
Validation loss: 69.9440
Validation acc: 0.7074
Time taken: 11.85s
Start of epoch 13
Training loss (for one batch) at step 0: 293.5333, Accuracy: 0.7600
Training loss (for one batch) at step 10: 295.0630, Accuracy: 0.7745
Training loss (for one batch) at step 20: 287.0632, Accuracy: 0.7705
Training loss (for one batch) at step 30: 306.0687, Accuracy: 0.7684
Training loss (for one batch) at step 40: 285.9910, Accuracy: 0.7690
Training loss (for one batch) at step 50: 326.7936, Accuracy: 0.7720
Training loss (for one batch) at step 60: 289.9844, Accuracy: 0.7710
Training loss (for one batch) at step 70: 300.4421, Accuracy: 0.7721
Training loss (for one batch) at step 80: 302.0107, Accuracy: 0.7711
Training loss (for one batch) at step 90: 289.7980, Accuracy: 0.7699
Training loss (for one batch) at step 100: 300.9894, Accuracy: 0.7655
Training loss (for one batch) at step 110: 296.8793, Accuracy: 0.7659
Training loss (for one batch) at step 120: 274.0785, Accuracy: 0.7669
Training loss (for one batch) at step 130: 287.1812, Accuracy: 0.7663
Training loss (for one batch) at step 140: 317.5496, Accuracy: 0.7651
---- Training ----
Training loss: 268.2782
Training acc over epoch: 0.7662
---- Validation ----
Validation loss: 69.5460
Validation acc: 0.7166
Time taken: 18.23s
Start of epoch 14
Training loss (for one batch) at step 0: 277.2243, Accuracy: 0.7500
Training loss (for one batch) at step 10: 300.0375, Accuracy: 0.7691
Training loss (for one batch) at step 20: 295.4369, Accuracy: 0.7643
Training loss (for one batch) at step 30: 289.8055, Accuracy: 0.7626
Training loss (for one batch) at step 40: 293.4218, Accuracy: 0.7639
Training loss (for one batch) at step 50: 292.4017, Accuracy: 0.7673
Training loss (for one batch) at step 60: 297.9947, Accuracy: 0.7718
Training loss (for one batch) at step 70: 289.0736, Accuracy: 0.7697
Training loss (for one batch) at step 80: 305.4175, Accuracy: 0.7699
Training loss (for one batch) at step 90: 300.5384, Accuracy: 0.7674
Training loss (for one batch) at step 100: 287.5411, Accuracy: 0.7663
Training loss (for one batch) at step 110: 288.9365, Accuracy: 0.7672
Training loss (for one batch) at step 120: 279.3419, Accuracy: 0.7669
Training loss (for one batch) at step 130: 282.7158, Accuracy: 0.7664
Training loss (for one batch) at step 140: 280.2452, Accuracy: 0.7655
---- Training ----
Training loss: 253.5590
Training acc over epoch: 0.7662
---- Validation ----
Validation loss: 72.9656
Validation acc: 0.7042
Time taken: 10.51s
Start of epoch 15
Training loss (for one batch) at step 0: 300.2415, Accuracy: 0.7300
Training loss (for one batch) at step 10: 279.8725, Accuracy: 0.7827
Training loss (for one batch) at step 20: 286.4525, Accuracy: 0.7852
Training loss (for one batch) at step 30: 288.8819, Accuracy: 0.7890
Training loss (for one batch) at step 40: 292.2672, Accuracy: 0.7898
Training loss (for one batch) at step 50: 282.0624, Accuracy: 0.7894
Training loss (for one batch) at step 60: 284.5005, Accuracy: 0.7857
Training loss (for one batch) at step 70: 285.2954, Accuracy: 0.7849
Training loss (for one batch) at step 80: 290.9615, Accuracy: 0.7842
Training loss (for one batch) at step 90: 308.6555, Accuracy: 0.7807
Training loss (for one batch) at step 100: 290.6957, Accuracy: 0.7803
Training loss (for one batch) at step 110: 279.9492, Accuracy: 0.7785
Training loss (for one batch) at step 120: 280.5481, Accuracy: 0.7774
Training loss (for one batch) at step 130: 288.9973, Accuracy: 0.7763
Training loss (for one batch) at step 140: 279.2231, Accuracy: 0.7767
---- Training ----
Training loss: 254.4610
Training acc over epoch: 0.7765
---- Validation ----
Validation loss: 67.7294
Validation acc: 0.7211
Time taken: 11.29s
Start of epoch 16
Training loss (for one batch) at step 0: 285.6526, Accuracy: 0.7000
Training loss (for one batch) at step 10: 279.4961, Accuracy: 0.7891
Training loss (for one batch) at step 20: 293.5388, Accuracy: 0.7881
Training loss (for one batch) at step 30: 301.9880, Accuracy: 0.7855
Training loss (for one batch) at step 40: 268.9919, Accuracy: 0.7800
Training loss (for one batch) at step 50: 275.0904, Accuracy: 0.7825
Training loss (for one batch) at step 60: 280.5791, Accuracy: 0.7844
Training loss (for one batch) at step 70: 283.7652, Accuracy: 0.7825
Training loss (for one batch) at step 80: 267.1924, Accuracy: 0.7816
Training loss (for one batch) at step 90: 289.0630, Accuracy: 0.7798
Training loss (for one batch) at step 100: 291.7070, Accuracy: 0.7782
Training loss (for one batch) at step 110: 283.8939, Accuracy: 0.7793
Training loss (for one batch) at step 120: 275.0883, Accuracy: 0.7796
Training loss (for one batch) at step 130: 279.4808, Accuracy: 0.7808
Training loss (for one batch) at step 140: 291.6949, Accuracy: 0.7787
---- Training ----
Training loss: 248.3508
Training acc over epoch: 0.7789
---- Validation ----
Validation loss: 79.3065
Validation acc: 0.7308
Time taken: 9.67s
Start of epoch 17
Training loss (for one batch) at step 0: 278.9569, Accuracy: 0.7800
Training loss (for one batch) at step 10: 273.6313, Accuracy: 0.7927
Training loss (for one batch) at step 20: 281.0540, Accuracy: 0.7852
Training loss (for one batch) at step 30: 292.5652, Accuracy: 0.7890
Training loss (for one batch) at step 40: 272.3949, Accuracy: 0.7868
Training loss (for one batch) at step 50: 276.1077, Accuracy: 0.7914
Training loss (for one batch) at step 60: 266.1566, Accuracy: 0.7928
Training loss (for one batch) at step 70: 289.3810, Accuracy: 0.7918
Training loss (for one batch) at step 80: 283.2122, Accuracy: 0.7906
Training loss (for one batch) at step 90: 271.7021, Accuracy: 0.7904
Training loss (for one batch) at step 100: 271.2239, Accuracy: 0.7900
Training loss (for one batch) at step 110: 284.2550, Accuracy: 0.7905
Training loss (for one batch) at step 120: 309.3036, Accuracy: 0.7894
Training loss (for one batch) at step 130: 274.8307, Accuracy: 0.7884
Training loss (for one batch) at step 140: 279.1962, Accuracy: 0.7884
---- Training ----
Training loss: 243.1614
Training acc over epoch: 0.7880
---- Validation ----
Validation loss: 61.6628
Validation acc: 0.7219
Time taken: 9.88s
Start of epoch 18
Training loss (for one batch) at step 0: 276.8687, Accuracy: 0.7400
Training loss (for one batch) at step 10: 278.1779, Accuracy: 0.7755
Training loss (for one batch) at step 20: 281.2620, Accuracy: 0.7848
Training loss (for one batch) at step 30: 284.5782, Accuracy: 0.7852
Training loss (for one batch) at step 40: 284.5167, Accuracy: 0.7866
Training loss (for one batch) at step 50: 292.7481, Accuracy: 0.7933
Training loss (for one batch) at step 60: 271.7957, Accuracy: 0.7961
Training loss (for one batch) at step 70: 292.5124, Accuracy: 0.7961
Training loss (for one batch) at step 80: 262.5511, Accuracy: 0.7923
Training loss (for one batch) at step 90: 293.8504, Accuracy: 0.7916
Training loss (for one batch) at step 100: 285.4265, Accuracy: 0.7880
Training loss (for one batch) at step 110: 284.6936, Accuracy: 0.7893
Training loss (for one batch) at step 120: 281.6410, Accuracy: 0.7901
Training loss (for one batch) at step 130: 283.2319, Accuracy: 0.7902
Training loss (for one batch) at step 140: 271.5083, Accuracy: 0.7893
---- Training ----
Training loss: 261.6610
Training acc over epoch: 0.7887
---- Validation ----
Validation loss: 68.7282
Validation acc: 0.7423
Time taken: 9.61s
Start of epoch 19
Training loss (for one batch) at step 0: 274.9865, Accuracy: 0.7500
Training loss (for one batch) at step 10: 290.3083, Accuracy: 0.7982
Training loss (for one batch) at step 20: 258.6560, Accuracy: 0.8000
Training loss (for one batch) at step 30: 295.1824, Accuracy: 0.8035
Training loss (for one batch) at step 40: 276.7542, Accuracy: 0.7968
Training loss (for one batch) at step 50: 261.4685, Accuracy: 0.8031
Training loss (for one batch) at step 60: 252.8735, Accuracy: 0.8028
Training loss (for one batch) at step 70: 305.6752, Accuracy: 0.7999
Training loss (for one batch) at step 80: 265.6022, Accuracy: 0.7990
Training loss (for one batch) at step 90: 285.9725, Accuracy: 0.7982
Training loss (for one batch) at step 100: 268.9459, Accuracy: 0.7963
Training loss (for one batch) at step 110: 275.7827, Accuracy: 0.7968
Training loss (for one batch) at step 120: 277.7347, Accuracy: 0.7960
Training loss (for one batch) at step 130: 284.6166, Accuracy: 0.7936
Training loss (for one batch) at step 140: 272.6837, Accuracy: 0.7938
---- Training ----
Training loss: 236.6186
Training acc over epoch: 0.7933
---- Validation ----
Validation loss: 56.6453
Validation acc: 0.7354
Time taken: 9.66s
Start of epoch 20
Training loss (for one batch) at step 0: 274.3875, Accuracy: 0.8200
Training loss (for one batch) at step 10: 280.5367, Accuracy: 0.8082
Training loss (for one batch) at step 20: 280.1284, Accuracy: 0.8014
Training loss (for one batch) at step 30: 262.6679, Accuracy: 0.8003
Training loss (for one batch) at step 40: 271.7087, Accuracy: 0.8039
Training loss (for one batch) at step 50: 281.2905, Accuracy: 0.8047
Training loss (for one batch) at step 60: 267.8870, Accuracy: 0.8074
Training loss (for one batch) at step 70: 282.7853, Accuracy: 0.8076
Training loss (for one batch) at step 80: 264.1942, Accuracy: 0.8051
Training loss (for one batch) at step 90: 275.5016, Accuracy: 0.8059
Training loss (for one batch) at step 100: 261.6238, Accuracy: 0.8031
Training loss (for one batch) at step 110: 250.2139, Accuracy: 0.8032
Training loss (for one batch) at step 120: 280.6247, Accuracy: 0.8019
Training loss (for one batch) at step 130: 263.3033, Accuracy: 0.8000
Training loss (for one batch) at step 140: 263.2477, Accuracy: 0.8006
---- Training ----
Training loss: 243.5570
Training acc over epoch: 0.8004
---- Validation ----
Validation loss: 77.5258
Validation acc: 0.7195
Time taken: 9.62s
Start of epoch 21
Training loss (for one batch) at step 0: 271.3231, Accuracy: 0.7800
Training loss (for one batch) at step 10: 262.4483, Accuracy: 0.8191
Training loss (for one batch) at step 20: 280.7091, Accuracy: 0.8157
Training loss (for one batch) at step 30: 270.2301, Accuracy: 0.8103
Training loss (for one batch) at step 40: 260.5597, Accuracy: 0.8117
Training loss (for one batch) at step 50: 259.7688, Accuracy: 0.8125
Training loss (for one batch) at step 60: 246.7004, Accuracy: 0.8095
Training loss (for one batch) at step 70: 278.2313, Accuracy: 0.8106
Training loss (for one batch) at step 80: 261.3978, Accuracy: 0.8077
Training loss (for one batch) at step 90: 281.8924, Accuracy: 0.8056
Training loss (for one batch) at step 100: 252.1762, Accuracy: 0.8065
Training loss (for one batch) at step 110: 248.3022, Accuracy: 0.8064
Training loss (for one batch) at step 120: 274.3501, Accuracy: 0.8075
Training loss (for one batch) at step 130: 282.2742, Accuracy: 0.8058
Training loss (for one batch) at step 140: 265.7295, Accuracy: 0.8040
---- Training ----
Training loss: 237.2026
Training acc over epoch: 0.8034
---- Validation ----
Validation loss: 78.2420
Validation acc: 0.7166
Time taken: 9.78s
Start of epoch 22
Training loss (for one batch) at step 0: 283.3798, Accuracy: 0.7700
Training loss (for one batch) at step 10: 267.3272, Accuracy: 0.8155
Training loss (for one batch) at step 20: 276.6802, Accuracy: 0.8105
Training loss (for one batch) at step 30: 274.0831, Accuracy: 0.8094
Training loss (for one batch) at step 40: 267.1344, Accuracy: 0.8166
Training loss (for one batch) at step 50: 278.5636, Accuracy: 0.8149
Training loss (for one batch) at step 60: 247.9444, Accuracy: 0.8141
Training loss (for one batch) at step 70: 255.2581, Accuracy: 0.8135
Training loss (for one batch) at step 80: 281.2744, Accuracy: 0.8111
Training loss (for one batch) at step 90: 278.5781, Accuracy: 0.8107
Training loss (for one batch) at step 100: 254.9485, Accuracy: 0.8088
Training loss (for one batch) at step 110: 264.4947, Accuracy: 0.8089
Training loss (for one batch) at step 120: 256.2567, Accuracy: 0.8083
Training loss (for one batch) at step 130: 262.6553, Accuracy: 0.8061
Training loss (for one batch) at step 140: 281.6485, Accuracy: 0.8070
---- Training ----
Training loss: 243.0540
Training acc over epoch: 0.8062
---- Validation ----
Validation loss: 81.6703
Validation acc: 0.7300
Time taken: 39.42s
Start of epoch 23
Training loss (for one batch) at step 0: 260.8897, Accuracy: 0.8400
Training loss (for one batch) at step 10: 270.6202, Accuracy: 0.8118
Training loss (for one batch) at step 20: 265.0498, Accuracy: 0.8105
Training loss (for one batch) at step 30: 259.1041, Accuracy: 0.8119
Training loss (for one batch) at step 40: 248.1390, Accuracy: 0.8144
Training loss (for one batch) at step 50: 253.2325, Accuracy: 0.8139
Training loss (for one batch) at step 60: 269.5449, Accuracy: 0.8111
Training loss (for one batch) at step 70: 258.1758, Accuracy: 0.8113
Training loss (for one batch) at step 80: 259.9524, Accuracy: 0.8102
Training loss (for one batch) at step 90: 271.6006, Accuracy: 0.8070
Training loss (for one batch) at step 100: 275.3140, Accuracy: 0.8069
Training loss (for one batch) at step 110: 252.1321, Accuracy: 0.8061
Training loss (for one batch) at step 120: 255.7211, Accuracy: 0.8051
Training loss (for one batch) at step 130: 269.1982, Accuracy: 0.8034
Training loss (for one batch) at step 140: 251.7784, Accuracy: 0.8043
---- Training ----
Training loss: 231.2586
Training acc over epoch: 0.8043
---- Validation ----
Validation loss: 69.5229
Validation acc: 0.7311
Time taken: 36.57s
Start of epoch 24
Training loss (for one batch) at step 0: 255.5729, Accuracy: 0.8400
Training loss (for one batch) at step 10: 264.6110, Accuracy: 0.8109
Training loss (for one batch) at step 20: 237.6397, Accuracy: 0.8310
Training loss (for one batch) at step 30: 273.6924, Accuracy: 0.8129
Training loss (for one batch) at step 40: 264.2789, Accuracy: 0.8139
Training loss (for one batch) at step 50: 271.2490, Accuracy: 0.8165
Training loss (for one batch) at step 60: 255.2498, Accuracy: 0.8180
Training loss (for one batch) at step 70: 249.7104, Accuracy: 0.8161
Training loss (for one batch) at step 80: 251.5654, Accuracy: 0.8152
Training loss (for one batch) at step 90: 274.5956, Accuracy: 0.8115
Training loss (for one batch) at step 100: 279.0344, Accuracy: 0.8112
Training loss (for one batch) at step 110: 255.0255, Accuracy: 0.8108
Training loss (for one batch) at step 120: 259.4402, Accuracy: 0.8102
Training loss (for one batch) at step 130: 271.2220, Accuracy: 0.8092
Training loss (for one batch) at step 140: 293.1538, Accuracy: 0.8082
---- Training ----
Training loss: 229.6865
Training acc over epoch: 0.8082
---- Validation ----
Validation loss: 69.8550
Validation acc: 0.7311
Time taken: 9.68s
Start of epoch 25
Training loss (for one batch) at step 0: 257.6564, Accuracy: 0.8200
Training loss (for one batch) at step 10: 244.9152, Accuracy: 0.8227
Training loss (for one batch) at step 20: 260.8322, Accuracy: 0.8229
Training loss (for one batch) at step 30: 265.3689, Accuracy: 0.8187
Training loss (for one batch) at step 40: 248.7107, Accuracy: 0.8195
Training loss (for one batch) at step 50: 248.1336, Accuracy: 0.8255
Training loss (for one batch) at step 60: 273.3561, Accuracy: 0.8230
Training loss (for one batch) at step 70: 239.0714, Accuracy: 0.8217
Training loss (for one batch) at step 80: 267.3216, Accuracy: 0.8215
Training loss (for one batch) at step 90: 260.4554, Accuracy: 0.8204
Training loss (for one batch) at step 100: 252.7559, Accuracy: 0.8175
Training loss (for one batch) at step 110: 266.0077, Accuracy: 0.8159
Training loss (for one batch) at step 120: 259.8048, Accuracy: 0.8164
Training loss (for one batch) at step 130: 274.8131, Accuracy: 0.8163
Training loss (for one batch) at step 140: 238.6626, Accuracy: 0.8145
---- Training ----
Training loss: 239.0369
Training acc over epoch: 0.8136
---- Validation ----
Validation loss: 80.2194
Validation acc: 0.7351
Time taken: 9.56s
Start of epoch 26
Training loss (for one batch) at step 0: 258.6690, Accuracy: 0.7700
Training loss (for one batch) at step 10: 261.2102, Accuracy: 0.8227
Training loss (for one batch) at step 20: 255.9427, Accuracy: 0.8229
Training loss (for one batch) at step 30: 268.3126, Accuracy: 0.8223
Training loss (for one batch) at step 40: 245.6063, Accuracy: 0.8215
Training loss (for one batch) at step 50: 247.6566, Accuracy: 0.8231
Training loss (for one batch) at step 60: 264.9247, Accuracy: 0.8254
Training loss (for one batch) at step 70: 245.0810, Accuracy: 0.8255
Training loss (for one batch) at step 80: 265.7842, Accuracy: 0.8221
Training loss (for one batch) at step 90: 277.6035, Accuracy: 0.8211
Training loss (for one batch) at step 100: 265.4823, Accuracy: 0.8185
Training loss (for one batch) at step 110: 255.6524, Accuracy: 0.8169
Training loss (for one batch) at step 120: 258.8064, Accuracy: 0.8168
Training loss (for one batch) at step 130: 253.4174, Accuracy: 0.8182
Training loss (for one batch) at step 140: 256.7956, Accuracy: 0.8172
---- Training ----
Training loss: 232.6860
Training acc over epoch: 0.8169
---- Validation ----
Validation loss: 68.4831
Validation acc: 0.7305
Time taken: 9.66s
Start of epoch 27
Training loss (for one batch) at step 0: 256.4757, Accuracy: 0.7300
Training loss (for one batch) at step 10: 258.6226, Accuracy: 0.8209
Training loss (for one batch) at step 20: 264.3828, Accuracy: 0.8176
Training loss (for one batch) at step 30: 257.5995, Accuracy: 0.8223
Training loss (for one batch) at step 40: 245.5201, Accuracy: 0.8212
Training loss (for one batch) at step 50: 247.7158, Accuracy: 0.8212
Training loss (for one batch) at step 60: 271.1698, Accuracy: 0.8174
Training loss (for one batch) at step 70: 254.2565, Accuracy: 0.8182
Training loss (for one batch) at step 80: 247.1605, Accuracy: 0.8177
Training loss (for one batch) at step 90: 252.7296, Accuracy: 0.8165
Training loss (for one batch) at step 100: 253.8263, Accuracy: 0.8171
Training loss (for one batch) at step 110: 250.9061, Accuracy: 0.8164
Training loss (for one batch) at step 120: 244.7705, Accuracy: 0.8157
Training loss (for one batch) at step 130: 271.2404, Accuracy: 0.8155
Training loss (for one batch) at step 140: 248.6124, Accuracy: 0.8155
---- Training ----
Training loss: 221.4033
Training acc over epoch: 0.8148
---- Validation ----
Validation loss: 71.9015
Validation acc: 0.7364
Time taken: 9.63s
Start of epoch 28
Training loss (for one batch) at step 0: 235.5808, Accuracy: 0.7800
Training loss (for one batch) at step 10: 256.4344, Accuracy: 0.8173
Training loss (for one batch) at step 20: 260.0185, Accuracy: 0.8224
Training loss (for one batch) at step 30: 242.0847, Accuracy: 0.8261
Training loss (for one batch) at step 40: 264.1365, Accuracy: 0.8198
Training loss (for one batch) at step 50: 255.4674, Accuracy: 0.8214
Training loss (for one batch) at step 60: 236.4243, Accuracy: 0.8244
Training loss (for one batch) at step 70: 254.1051, Accuracy: 0.8228
Training loss (for one batch) at step 80: 253.2704, Accuracy: 0.8225
Training loss (for one batch) at step 90: 248.9781, Accuracy: 0.8224
Training loss (for one batch) at step 100: 242.2585, Accuracy: 0.8201
Training loss (for one batch) at step 110: 242.6348, Accuracy: 0.8198
Training loss (for one batch) at step 120: 254.0946, Accuracy: 0.8201
Training loss (for one batch) at step 130: 227.2512, Accuracy: 0.8185
Training loss (for one batch) at step 140: 272.3870, Accuracy: 0.8191
---- Training ----
Training loss: 231.8638
Training acc over epoch: 0.8178
---- Validation ----
Validation loss: 61.4181
Validation acc: 0.7163
Time taken: 9.59s
Start of epoch 29
Training loss (for one batch) at step 0: 250.4533, Accuracy: 0.8300
Training loss (for one batch) at step 10: 235.7308, Accuracy: 0.8336
Training loss (for one batch) at step 20: 262.1500, Accuracy: 0.8286
Training loss (for one batch) at step 30: 229.8177, Accuracy: 0.8248
Training loss (for one batch) at step 40: 234.1809, Accuracy: 0.8178
Training loss (for one batch) at step 50: 234.8945, Accuracy: 0.8239
Training loss (for one batch) at step 60: 245.6645, Accuracy: 0.8257
Training loss (for one batch) at step 70: 270.1336, Accuracy: 0.8234
Training loss (for one batch) at step 80: 254.9946, Accuracy: 0.8246
Training loss (for one batch) at step 90: 255.4315, Accuracy: 0.8225
Training loss (for one batch) at step 100: 258.2914, Accuracy: 0.8213
Training loss (for one batch) at step 110: 248.0752, Accuracy: 0.8202
Training loss (for one batch) at step 120: 251.2256, Accuracy: 0.8206
Training loss (for one batch) at step 130: 259.2333, Accuracy: 0.8209
Training loss (for one batch) at step 140: 236.4722, Accuracy: 0.8201
---- Training ----
Training loss: 218.4847
Training acc over epoch: 0.8207
---- Validation ----
Validation loss: 76.3847
Validation acc: 0.7235
Time taken: 9.55s
Start of epoch 30
Training loss (for one batch) at step 0: 265.6779, Accuracy: 0.8200
Training loss (for one batch) at step 10: 255.6735, Accuracy: 0.8227
Training loss (for one batch) at step 20: 257.1455, Accuracy: 0.8157
Training loss (for one batch) at step 30: 250.5950, Accuracy: 0.8161
Training loss (for one batch) at step 40: 236.6871, Accuracy: 0.8254
Training loss (for one batch) at step 50: 251.0590, Accuracy: 0.8280
Training loss (for one batch) at step 60: 223.4849, Accuracy: 0.8318
Training loss (for one batch) at step 70: 245.4058, Accuracy: 0.8313
Training loss (for one batch) at step 80: 247.3261, Accuracy: 0.8310
Training loss (for one batch) at step 90: 247.9943, Accuracy: 0.8309
Training loss (for one batch) at step 100: 244.3882, Accuracy: 0.8294
Training loss (for one batch) at step 110: 248.4085, Accuracy: 0.8303
Training loss (for one batch) at step 120: 255.9235, Accuracy: 0.8282
Training loss (for one batch) at step 130: 251.2894, Accuracy: 0.8259
Training loss (for one batch) at step 140: 249.1193, Accuracy: 0.8257
---- Training ----
Training loss: 215.8392
Training acc over epoch: 0.8264
---- Validation ----
Validation loss: 78.0000
Validation acc: 0.7208
Time taken: 9.63s
Start of epoch 31
Training loss (for one batch) at step 0: 235.1634, Accuracy: 0.8600
Training loss (for one batch) at step 10: 246.7077, Accuracy: 0.8464
Training loss (for one batch) at step 20: 246.4282, Accuracy: 0.8343
Training loss (for one batch) at step 30: 248.3089, Accuracy: 0.8306
Training loss (for one batch) at step 40: 242.9275, Accuracy: 0.8293
Training loss (for one batch) at step 50: 255.3733, Accuracy: 0.8322
Training loss (for one batch) at step 60: 229.8822, Accuracy: 0.8325
Training loss (for one batch) at step 70: 245.4441, Accuracy: 0.8334
Training loss (for one batch) at step 80: 244.0942, Accuracy: 0.8304
Training loss (for one batch) at step 90: 237.0383, Accuracy: 0.8289
Training loss (for one batch) at step 100: 264.4553, Accuracy: 0.8267
Training loss (for one batch) at step 110: 254.1648, Accuracy: 0.8262
Training loss (for one batch) at step 120: 253.3862, Accuracy: 0.8254
Training loss (for one batch) at step 130: 242.9842, Accuracy: 0.8259
Training loss (for one batch) at step 140: 257.7800, Accuracy: 0.8257
---- Training ----
Training loss: 225.2136
Training acc over epoch: 0.8252
---- Validation ----
Validation loss: 74.8686
Validation acc: 0.7303
Time taken: 9.53s
Start of epoch 32
Training loss (for one batch) at step 0: 234.7693, Accuracy: 0.8400
Training loss (for one batch) at step 10: 256.4996, Accuracy: 0.8273
Training loss (for one batch) at step 20: 239.4987, Accuracy: 0.8367
Training loss (for one batch) at step 30: 249.8239, Accuracy: 0.8365
Training loss (for one batch) at step 40: 246.1911, Accuracy: 0.8356
Training loss (for one batch) at step 50: 228.7412, Accuracy: 0.8408
Training loss (for one batch) at step 60: 235.0481, Accuracy: 0.8384
Training loss (for one batch) at step 70: 226.4197, Accuracy: 0.8365
Training loss (for one batch) at step 80: 251.5651, Accuracy: 0.8336
Training loss (for one batch) at step 90: 227.3966, Accuracy: 0.8340
Training loss (for one batch) at step 100: 252.8682, Accuracy: 0.8296
Training loss (for one batch) at step 110: 250.3952, Accuracy: 0.8289
Training loss (for one batch) at step 120: 247.3540, Accuracy: 0.8294
Training loss (for one batch) at step 130: 247.0959, Accuracy: 0.8299
Training loss (for one batch) at step 140: 230.8268, Accuracy: 0.8297
---- Training ----
Training loss: 198.4410
Training acc over epoch: 0.8286
---- Validation ----
Validation loss: 73.4166
Validation acc: 0.7227
Time taken: 9.53s
Start of epoch 33
Training loss (for one batch) at step 0: 238.2776, Accuracy: 0.8100
Training loss (for one batch) at step 10: 216.8593, Accuracy: 0.8436
Training loss (for one batch) at step 20: 247.9523, Accuracy: 0.8400
Training loss (for one batch) at step 30: 247.5690, Accuracy: 0.8284
Training loss (for one batch) at step 40: 235.1917, Accuracy: 0.8268
Training loss (for one batch) at step 50: 239.1824, Accuracy: 0.8318
Training loss (for one batch) at step 60: 246.0228, Accuracy: 0.8302
Training loss (for one batch) at step 70: 225.9839, Accuracy: 0.8304
Training loss (for one batch) at step 80: 243.1880, Accuracy: 0.8283
Training loss (for one batch) at step 90: 228.3529, Accuracy: 0.8277
Training loss (for one batch) at step 100: 221.7016, Accuracy: 0.8262
Training loss (for one batch) at step 110: 227.8886, Accuracy: 0.8268
Training loss (for one batch) at step 120: 246.4049, Accuracy: 0.8265
Training loss (for one batch) at step 130: 237.8077, Accuracy: 0.8263
Training loss (for one batch) at step 140: 229.7881, Accuracy: 0.8260
---- Training ----
Training loss: 213.0684
Training acc over epoch: 0.8260
---- Validation ----
Validation loss: 78.0020
Validation acc: 0.7251
Time taken: 17.69s
Start of epoch 34
Training loss (for one batch) at step 0: 239.0599, Accuracy: 0.8600
Training loss (for one batch) at step 10: 244.0765, Accuracy: 0.8382
Training loss (for one batch) at step 20: 255.7596, Accuracy: 0.8357
Training loss (for one batch) at step 30: 257.1249, Accuracy: 0.8306
Training loss (for one batch) at step 40: 222.9073, Accuracy: 0.8320
Training loss (for one batch) at step 50: 220.3813, Accuracy: 0.8345
Training loss (for one batch) at step 60: 238.6107, Accuracy: 0.8316
Training loss (for one batch) at step 70: 230.7518, Accuracy: 0.8293
Training loss (for one batch) at step 80: 238.3144, Accuracy: 0.8289
Training loss (for one batch) at step 90: 243.2593, Accuracy: 0.8286
Training loss (for one batch) at step 100: 231.2838, Accuracy: 0.8276
Training loss (for one batch) at step 110: 240.6799, Accuracy: 0.8295
Training loss (for one batch) at step 120: 216.7725, Accuracy: 0.8287
Training loss (for one batch) at step 130: 234.7756, Accuracy: 0.8297
Training loss (for one batch) at step 140: 248.8953, Accuracy: 0.8284
---- Training ----
Training loss: 224.9543
Training acc over epoch: 0.8287
---- Validation ----
Validation loss: 70.5153
Validation acc: 0.7300
Time taken: 13.57s
Start of epoch 35
Training loss (for one batch) at step 0: 232.9356, Accuracy: 0.8600
Training loss (for one batch) at step 10: 227.8999, Accuracy: 0.8327
Training loss (for one batch) at step 20: 246.5025, Accuracy: 0.8362
Training loss (for one batch) at step 30: 225.8062, Accuracy: 0.8384
Training loss (for one batch) at step 40: 235.5363, Accuracy: 0.8344
Training loss (for one batch) at step 50: 228.9272, Accuracy: 0.8376
Training loss (for one batch) at step 60: 205.0694, Accuracy: 0.8361
Training loss (for one batch) at step 70: 238.6744, Accuracy: 0.8369
Training loss (for one batch) at step 80: 219.4124, Accuracy: 0.8333
Training loss (for one batch) at step 90: 245.6010, Accuracy: 0.8345
Training loss (for one batch) at step 100: 235.6540, Accuracy: 0.8341
Training loss (for one batch) at step 110: 232.8627, Accuracy: 0.8337
Training loss (for one batch) at step 120: 255.7673, Accuracy: 0.8336
Training loss (for one batch) at step 130: 239.2982, Accuracy: 0.8336
Training loss (for one batch) at step 140: 247.3188, Accuracy: 0.8332
---- Training ----
Training loss: 211.4240
Training acc over epoch: 0.8340
---- Validation ----
Validation loss: 95.1816
Validation acc: 0.7203
Time taken: 9.63s
Start of epoch 36
Training loss (for one batch) at step 0: 237.7467, Accuracy: 0.8200
Training loss (for one batch) at step 10: 226.8647, Accuracy: 0.8564
Training loss (for one batch) at step 20: 229.1485, Accuracy: 0.8462
Training loss (for one batch) at step 30: 239.4305, Accuracy: 0.8294
Training loss (for one batch) at step 40: 232.0672, Accuracy: 0.8320
Training loss (for one batch) at step 50: 236.5449, Accuracy: 0.8349
Training loss (for one batch) at step 60: 225.5422, Accuracy: 0.8354
Training loss (for one batch) at step 70: 231.6777, Accuracy: 0.8356
Training loss (for one batch) at step 80: 241.3332, Accuracy: 0.8332
Training loss (for one batch) at step 90: 220.7050, Accuracy: 0.8307
Training loss (for one batch) at step 100: 251.4076, Accuracy: 0.8298
Training loss (for one batch) at step 110: 239.8605, Accuracy: 0.8275
Training loss (for one batch) at step 120: 242.3238, Accuracy: 0.8286
Training loss (for one batch) at step 130: 258.0800, Accuracy: 0.8297
Training loss (for one batch) at step 140: 242.2086, Accuracy: 0.8291
---- Training ----
Training loss: 190.8828
Training acc over epoch: 0.8290
---- Validation ----
Validation loss: 86.1767
Validation acc: 0.7106
Time taken: 9.59s
Start of epoch 37
Training loss (for one batch) at step 0: 230.5379, Accuracy: 0.8200
Training loss (for one batch) at step 10: 214.9417, Accuracy: 0.8391
Training loss (for one batch) at step 20: 225.0596, Accuracy: 0.8400
Training loss (for one batch) at step 30: 236.1614, Accuracy: 0.8397
Training loss (for one batch) at step 40: 215.0872, Accuracy: 0.8363
Training loss (for one batch) at step 50: 226.5049, Accuracy: 0.8371
Training loss (for one batch) at step 60: 230.5347, Accuracy: 0.8397
Training loss (for one batch) at step 70: 248.4264, Accuracy: 0.8369
Training loss (for one batch) at step 80: 240.4931, Accuracy: 0.8347
Training loss (for one batch) at step 90: 227.0157, Accuracy: 0.8333
Training loss (for one batch) at step 100: 236.1184, Accuracy: 0.8321
Training loss (for one batch) at step 110: 233.4433, Accuracy: 0.8311
Training loss (for one batch) at step 120: 234.4109, Accuracy: 0.8314
Training loss (for one batch) at step 130: 242.4969, Accuracy: 0.8298
Training loss (for one batch) at step 140: 219.1986, Accuracy: 0.8299
---- Training ----
Training loss: 208.8055
Training acc over epoch: 0.8298
---- Validation ----
Validation loss: 75.9376
Validation acc: 0.7039
Time taken: 9.55s
Start of epoch 38
Training loss (for one batch) at step 0: 230.3796, Accuracy: 0.8000
Training loss (for one batch) at step 10: 217.6215, Accuracy: 0.8309
Training loss (for one batch) at step 20: 243.9621, Accuracy: 0.8348
Training loss (for one batch) at step 30: 227.5560, Accuracy: 0.8316
Training loss (for one batch) at step 40: 228.8665, Accuracy: 0.8322
Training loss (for one batch) at step 50: 235.3100, Accuracy: 0.8335
Training loss (for one batch) at step 60: 213.8705, Accuracy: 0.8364
Training loss (for one batch) at step 70: 224.1072, Accuracy: 0.8354
Training loss (for one batch) at step 80: 222.1025, Accuracy: 0.8348
Training loss (for one batch) at step 90: 248.2388, Accuracy: 0.8333
Training loss (for one batch) at step 100: 221.8412, Accuracy: 0.8332
Training loss (for one batch) at step 110: 217.2371, Accuracy: 0.8345
Training loss (for one batch) at step 120: 243.3704, Accuracy: 0.8336
Training loss (for one batch) at step 130: 230.4152, Accuracy: 0.8338
Training loss (for one batch) at step 140: 255.5522, Accuracy: 0.8326
---- Training ----
Training loss: 198.6040
Training acc over epoch: 0.8327
---- Validation ----
Validation loss: 81.3925
Validation acc: 0.7120
Time taken: 9.55s
Start of epoch 39
Training loss (for one batch) at step 0: 242.8118, Accuracy: 0.8600
Training loss (for one batch) at step 10: 240.4739, Accuracy: 0.8555
Training loss (for one batch) at step 20: 217.6738, Accuracy: 0.8495
Training loss (for one batch) at step 30: 252.6936, Accuracy: 0.8397
Training loss (for one batch) at step 40: 233.0319, Accuracy: 0.8366
Training loss (for one batch) at step 50: 217.0856, Accuracy: 0.8378
Training loss (for one batch) at step 60: 215.1580, Accuracy: 0.8372
Training loss (for one batch) at step 70: 219.6463, Accuracy: 0.8377
Training loss (for one batch) at step 80: 238.3092, Accuracy: 0.8378
Training loss (for one batch) at step 90: 258.0686, Accuracy: 0.8348
Training loss (for one batch) at step 100: 222.4865, Accuracy: 0.8339
Training loss (for one batch) at step 110: 222.9928, Accuracy: 0.8351
Training loss (for one batch) at step 120: 228.0995, Accuracy: 0.8340
Training loss (for one batch) at step 130: 217.0885, Accuracy: 0.8344
Training loss (for one batch) at step 140: 228.3335, Accuracy: 0.8351
---- Training ----
Training loss: 221.7710
Training acc over epoch: 0.8351
---- Validation ----
Validation loss: 86.8606
Validation acc: 0.7071
Time taken: 9.70s
Start of epoch 40
Training loss (for one batch) at step 0: 224.1778, Accuracy: 0.9000
Training loss (for one batch) at step 10: 222.8591, Accuracy: 0.8455
Training loss (for one batch) at step 20: 200.8625, Accuracy: 0.8490
Training loss (for one batch) at step 30: 201.2006, Accuracy: 0.8435
Training loss (for one batch) at step 40: 226.8645, Accuracy: 0.8415
Training loss (for one batch) at step 50: 229.5835, Accuracy: 0.8465
Training loss (for one batch) at step 60: 220.5945, Accuracy: 0.8470
Training loss (for one batch) at step 70: 236.1765, Accuracy: 0.8463
Training loss (for one batch) at step 80: 212.4132, Accuracy: 0.8462
Training loss (for one batch) at step 90: 246.5308, Accuracy: 0.8416
Training loss (for one batch) at step 100: 243.9339, Accuracy: 0.8417
Training loss (for one batch) at step 110: 249.2395, Accuracy: 0.8404
Training loss (for one batch) at step 120: 236.9610, Accuracy: 0.8388
Training loss (for one batch) at step 130: 228.3718, Accuracy: 0.8390
Training loss (for one batch) at step 140: 221.1117, Accuracy: 0.8385
---- Training ----
Training loss: 194.9970
Training acc over epoch: 0.8381
---- Validation ----
Validation loss: 70.7449
Validation acc: 0.7112
Time taken: 9.62s
Start of epoch 41
Training loss (for one batch) at step 0: 229.2253, Accuracy: 0.8600
Training loss (for one batch) at step 10: 216.8373, Accuracy: 0.8545
Training loss (for one batch) at step 20: 222.6333, Accuracy: 0.8405
Training loss (for one batch) at step 30: 219.3059, Accuracy: 0.8390
Training loss (for one batch) at step 40: 227.7141, Accuracy: 0.8390
Training loss (for one batch) at step 50: 205.0518, Accuracy: 0.8433
Training loss (for one batch) at step 60: 217.1050, Accuracy: 0.8446
Training loss (for one batch) at step 70: 205.7720, Accuracy: 0.8441
Training loss (for one batch) at step 80: 229.0675, Accuracy: 0.8406
Training loss (for one batch) at step 90: 220.0661, Accuracy: 0.8399
Training loss (for one batch) at step 100: 225.0611, Accuracy: 0.8400
Training loss (for one batch) at step 110: 213.3906, Accuracy: 0.8397
Training loss (for one batch) at step 120: 234.8103, Accuracy: 0.8398
Training loss (for one batch) at step 130: 229.9227, Accuracy: 0.8388
Training loss (for one batch) at step 140: 224.1520, Accuracy: 0.8366
---- Training ----
Training loss: 192.9002
Training acc over epoch: 0.8374
---- Validation ----
Validation loss: 91.8686
Validation acc: 0.7114
Time taken: 9.62s
Start of epoch 42
Training loss (for one batch) at step 0: 228.6310, Accuracy: 0.8200
Training loss (for one batch) at step 10: 206.9173, Accuracy: 0.8455
Training loss (for one batch) at step 20: 204.8218, Accuracy: 0.8467
Training loss (for one batch) at step 30: 226.7624, Accuracy: 0.8400
Training loss (for one batch) at step 40: 227.9356, Accuracy: 0.8383
Training loss (for one batch) at step 50: 222.6228, Accuracy: 0.8406
Training loss (for one batch) at step 60: 222.6140, Accuracy: 0.8395
Training loss (for one batch) at step 70: 210.3168, Accuracy: 0.8394
Training loss (for one batch) at step 80: 206.2185, Accuracy: 0.8380
Training loss (for one batch) at step 90: 233.1849, Accuracy: 0.8368
Training loss (for one batch) at step 100: 206.1390, Accuracy: 0.8345
Training loss (for one batch) at step 110: 222.8156, Accuracy: 0.8355
Training loss (for one batch) at step 120: 236.2708, Accuracy: 0.8364
Training loss (for one batch) at step 130: 241.0279, Accuracy: 0.8365
Training loss (for one batch) at step 140: 225.0012, Accuracy: 0.8374
---- Training ----
Training loss: 208.9243
Training acc over epoch: 0.8376
---- Validation ----
Validation loss: 78.3577
Validation acc: 0.7286
Time taken: 9.80s
Start of epoch 43
Training loss (for one batch) at step 0: 207.9833, Accuracy: 0.8400
Training loss (for one batch) at step 10: 219.3080, Accuracy: 0.8509
Training loss (for one batch) at step 20: 220.0188, Accuracy: 0.8429
Training loss (for one batch) at step 30: 225.8286, Accuracy: 0.8445
Training loss (for one batch) at step 40: 210.2229, Accuracy: 0.8385
Training loss (for one batch) at step 50: 217.5680, Accuracy: 0.8431
Training loss (for one batch) at step 60: 242.1019, Accuracy: 0.8382
Training loss (for one batch) at step 70: 260.2966, Accuracy: 0.8394
Training loss (for one batch) at step 80: 215.0033, Accuracy: 0.8407
Training loss (for one batch) at step 90: 211.0357, Accuracy: 0.8411
Training loss (for one batch) at step 100: 224.6381, Accuracy: 0.8414
Training loss (for one batch) at step 110: 240.2129, Accuracy: 0.8421
Training loss (for one batch) at step 120: 217.2250, Accuracy: 0.8418
Training loss (for one batch) at step 130: 219.7348, Accuracy: 0.8416
Training loss (for one batch) at step 140: 205.0614, Accuracy: 0.8404
---- Training ----
Training loss: 211.0566
Training acc over epoch: 0.8406
---- Validation ----
Validation loss: 74.5098
Validation acc: 0.7284
Time taken: 9.69s
Start of epoch 44
Training loss (for one batch) at step 0: 217.2118, Accuracy: 0.8900
Training loss (for one batch) at step 10: 229.5181, Accuracy: 0.8509
Training loss (for one batch) at step 20: 228.7163, Accuracy: 0.8543
Training loss (for one batch) at step 30: 234.9878, Accuracy: 0.8487
Training loss (for one batch) at step 40: 228.6249, Accuracy: 0.8446
Training loss (for one batch) at step 50: 204.3716, Accuracy: 0.8459
Training loss (for one batch) at step 60: 213.8173, Accuracy: 0.8461
Training loss (for one batch) at step 70: 229.2013, Accuracy: 0.8466
Training loss (for one batch) at step 80: 227.1932, Accuracy: 0.8446
Training loss (for one batch) at step 90: 227.3181, Accuracy: 0.8429
Training loss (for one batch) at step 100: 210.6842, Accuracy: 0.8440
Training loss (for one batch) at step 110: 206.9723, Accuracy: 0.8428
Training loss (for one batch) at step 120: 234.9632, Accuracy: 0.8420
Training loss (for one batch) at step 130: 222.9170, Accuracy: 0.8425
Training loss (for one batch) at step 140: 230.7695, Accuracy: 0.8422
---- Training ----
Training loss: 174.9993
Training acc over epoch: 0.8424
---- Validation ----
Validation loss: 82.3881
Validation acc: 0.7251
Time taken: 9.58s
Start of epoch 45
Training loss (for one batch) at step 0: 216.1512, Accuracy: 0.8400
Training loss (for one batch) at step 10: 224.4157, Accuracy: 0.8364
Training loss (for one batch) at step 20: 206.0892, Accuracy: 0.8424
Training loss (for one batch) at step 30: 211.1161, Accuracy: 0.8439
Training loss (for one batch) at step 40: 236.1564, Accuracy: 0.8395
Training loss (for one batch) at step 50: 224.7358, Accuracy: 0.8457
Training loss (for one batch) at step 60: 196.2570, Accuracy: 0.8448
Training loss (for one batch) at step 70: 235.5951, Accuracy: 0.8427
Training loss (for one batch) at step 80: 227.4509, Accuracy: 0.8431
Training loss (for one batch) at step 90: 210.9881, Accuracy: 0.8430
Training loss (for one batch) at step 100: 217.2804, Accuracy: 0.8414
Training loss (for one batch) at step 110: 231.4447, Accuracy: 0.8423
Training loss (for one batch) at step 120: 212.3268, Accuracy: 0.8432
Training loss (for one batch) at step 130: 217.6592, Accuracy: 0.8429
Training loss (for one batch) at step 140: 221.5043, Accuracy: 0.8429
---- Training ----
Training loss: 205.6389
Training acc over epoch: 0.8418
---- Validation ----
Validation loss: 76.0679
Validation acc: 0.7144
Time taken: 10.45s
Start of epoch 46
Training loss (for one batch) at step 0: 222.8471, Accuracy: 0.8600
Training loss (for one batch) at step 10: 211.2410, Accuracy: 0.8573
Training loss (for one batch) at step 20: 223.2019, Accuracy: 0.8529
Training loss (for one batch) at step 30: 221.7294, Accuracy: 0.8513
Training loss (for one batch) at step 40: 193.5117, Accuracy: 0.8493
Training loss (for one batch) at step 50: 229.3386, Accuracy: 0.8476
Training loss (for one batch) at step 60: 205.3208, Accuracy: 0.8489
Training loss (for one batch) at step 70: 219.5587, Accuracy: 0.8462
Training loss (for one batch) at step 80: 197.7336, Accuracy: 0.8457
Training loss (for one batch) at step 90: 216.8507, Accuracy: 0.8444
Training loss (for one batch) at step 100: 208.5492, Accuracy: 0.8428
Training loss (for one batch) at step 110: 209.7371, Accuracy: 0.8416
Training loss (for one batch) at step 120: 204.9613, Accuracy: 0.8424
Training loss (for one batch) at step 130: 199.6359, Accuracy: 0.8431
Training loss (for one batch) at step 140: 229.7995, Accuracy: 0.8414
---- Training ----
Training loss: 183.8128
Training acc over epoch: 0.8424
---- Validation ----
Validation loss: 79.5987
Validation acc: 0.7265
Time taken: 9.59s
Start of epoch 47
Training loss (for one batch) at step 0: 201.0368, Accuracy: 0.8900
Training loss (for one batch) at step 10: 215.0016, Accuracy: 0.8527
Training loss (for one batch) at step 20: 198.9877, Accuracy: 0.8448
Training loss (for one batch) at step 30: 220.2661, Accuracy: 0.8448
Training loss (for one batch) at step 40: 219.1730, Accuracy: 0.8420
Training loss (for one batch) at step 50: 213.4567, Accuracy: 0.8459
Training loss (for one batch) at step 60: 216.2872, Accuracy: 0.8480
Training loss (for one batch) at step 70: 209.5771, Accuracy: 0.8468
Training loss (for one batch) at step 80: 241.7383, Accuracy: 0.8446
Training loss (for one batch) at step 90: 256.3990, Accuracy: 0.8441
Training loss (for one batch) at step 100: 220.1964, Accuracy: 0.8431
Training loss (for one batch) at step 110: 220.0380, Accuracy: 0.8428
Training loss (for one batch) at step 120: 207.1185, Accuracy: 0.8429
Training loss (for one batch) at step 130: 238.6928, Accuracy: 0.8425
Training loss (for one batch) at step 140: 203.9113, Accuracy: 0.8423
---- Training ----
Training loss: 197.6594
Training acc over epoch: 0.8431
---- Validation ----
Validation loss: 60.0447
Validation acc: 0.7117
Time taken: 9.60s
Start of epoch 48
Training loss (for one batch) at step 0: 243.9871, Accuracy: 0.7700
Training loss (for one batch) at step 10: 209.3907, Accuracy: 0.8564
Training loss (for one batch) at step 20: 227.6386, Accuracy: 0.8600
Training loss (for one batch) at step 30: 204.8152, Accuracy: 0.8571
Training loss (for one batch) at step 40: 226.9688, Accuracy: 0.8524
Training loss (for one batch) at step 50: 202.1578, Accuracy: 0.8559
Training loss (for one batch) at step 60: 193.7741, Accuracy: 0.8536
Training loss (for one batch) at step 70: 228.1769, Accuracy: 0.8523
Training loss (for one batch) at step 80: 225.6700, Accuracy: 0.8478
Training loss (for one batch) at step 90: 211.7131, Accuracy: 0.8477
Training loss (for one batch) at step 100: 205.3952, Accuracy: 0.8475
Training loss (for one batch) at step 110: 191.2701, Accuracy: 0.8467
Training loss (for one batch) at step 120: 213.7761, Accuracy: 0.8483
Training loss (for one batch) at step 130: 221.8164, Accuracy: 0.8482
Training loss (for one batch) at step 140: 199.3681, Accuracy: 0.8479
---- Training ----
Training loss: 199.3873
Training acc over epoch: 0.8477
---- Validation ----
Validation loss: 69.5440
Validation acc: 0.7208
Time taken: 9.59s
Start of epoch 49
Training loss (for one batch) at step 0: 212.3693, Accuracy: 0.8300
Training loss (for one batch) at step 10: 223.0607, Accuracy: 0.8482
Training loss (for one batch) at step 20: 205.1521, Accuracy: 0.8610
Training loss (for one batch) at step 30: 202.5189, Accuracy: 0.8587
Training loss (for one batch) at step 40: 189.2282, Accuracy: 0.8600
Training loss (for one batch) at step 50: 197.0823, Accuracy: 0.8602
Training loss (for one batch) at step 60: 217.6607, Accuracy: 0.8579
Training loss (for one batch) at step 70: 202.2034, Accuracy: 0.8558
Training loss (for one batch) at step 80: 210.1555, Accuracy: 0.8541
Training loss (for one batch) at step 90: 218.9495, Accuracy: 0.8519
Training loss (for one batch) at step 100: 201.6199, Accuracy: 0.8492
Training loss (for one batch) at step 110: 214.3871, Accuracy: 0.8504
Training loss (for one batch) at step 120: 217.6900, Accuracy: 0.8500
Training loss (for one batch) at step 130: 209.3839, Accuracy: 0.8489
Training loss (for one batch) at step 140: 220.0328, Accuracy: 0.8487
---- Training ----
Training loss: 185.0838
Training acc over epoch: 0.8489
---- Validation ----
Validation loss: 62.7616
Validation acc: 0.7149
Time taken: 9.65s
===== Q: 0.0001
Validation acc: 0.7418
Validation AUC: 0.7391
Validation Balanced_ACC: 0.4875
Validation MI: 0.1415
Validation Normalized MI: 0.2121
Validation Adjusted MI: 0.2121
Validation aUc_Sklearn: 0.8273
Start of epoch 0
Training loss (for one batch) at step 0: 568.8857, Accuracy: 0.4200
Training loss (for one batch) at step 10: 451.6623, Accuracy: 0.5155
Training loss (for one batch) at step 20: 446.0349, Accuracy: 0.5333
Training loss (for one batch) at step 30: 532.7630, Accuracy: 0.5335
Training loss (for one batch) at step 40: 497.5677, Accuracy: 0.5305
Training loss (for one batch) at step 50: 432.1443, Accuracy: 0.5367
Training loss (for one batch) at step 60: 450.0684, Accuracy: 0.5392
Training loss (for one batch) at step 70: 471.9392, Accuracy: 0.5425
Training loss (for one batch) at step 80: 431.8223, Accuracy: 0.5479
Training loss (for one batch) at step 90: 456.9954, Accuracy: 0.5488
Training loss (for one batch) at step 100: 417.6610, Accuracy: 0.5505
Training loss (for one batch) at step 110: 415.3194, Accuracy: 0.5516
Training loss (for one batch) at step 120: 414.7607, Accuracy: 0.5542
Training loss (for one batch) at step 130: 442.2986, Accuracy: 0.5556
Training loss (for one batch) at step 140: 416.3670, Accuracy: 0.5577
---- Training ----
Training loss: 377.5881
Training acc over epoch: 0.5596
---- Validation ----
Validation loss: 80.6350
Validation acc: 0.5134
Time taken: 13.23s
Start of epoch 1
Training loss (for one batch) at step 0: 440.1349, Accuracy: 0.5200
Training loss (for one batch) at step 10: 413.3390, Accuracy: 0.6218
Training loss (for one batch) at step 20: 413.7313, Accuracy: 0.6157
Training loss (for one batch) at step 30: 396.4484, Accuracy: 0.6116
Training loss (for one batch) at step 40: 395.4899, Accuracy: 0.6080
Training loss (for one batch) at step 50: 384.6578, Accuracy: 0.6059
Training loss (for one batch) at step 60: 397.2538, Accuracy: 0.6049
Training loss (for one batch) at step 70: 405.0442, Accuracy: 0.6051
Training loss (for one batch) at step 80: 402.7356, Accuracy: 0.6075
Training loss (for one batch) at step 90: 386.2458, Accuracy: 0.6080
Training loss (for one batch) at step 100: 399.7001, Accuracy: 0.6070
Training loss (for one batch) at step 110: 379.1230, Accuracy: 0.6080
Training loss (for one batch) at step 120: 411.8099, Accuracy: 0.6081
Training loss (for one batch) at step 130: 397.9330, Accuracy: 0.6091
Training loss (for one batch) at step 140: 388.2307, Accuracy: 0.6087
---- Training ----
Training loss: 331.2840
Training acc over epoch: 0.6107
---- Validation ----
Validation loss: 86.0560
Validation acc: 0.5261
Time taken: 9.50s
Start of epoch 2
Training loss (for one batch) at step 0: 387.5569, Accuracy: 0.6200
Training loss (for one batch) at step 10: 379.3525, Accuracy: 0.6264
Training loss (for one batch) at step 20: 352.2607, Accuracy: 0.6310
Training loss (for one batch) at step 30: 380.0068, Accuracy: 0.6294
Training loss (for one batch) at step 40: 388.8598, Accuracy: 0.6312
Training loss (for one batch) at step 50: 351.8120, Accuracy: 0.6339
Training loss (for one batch) at step 60: 368.2588, Accuracy: 0.6320
Training loss (for one batch) at step 70: 376.6803, Accuracy: 0.6358
Training loss (for one batch) at step 80: 387.9629, Accuracy: 0.6328
Training loss (for one batch) at step 90: 391.5165, Accuracy: 0.6300
Training loss (for one batch) at step 100: 380.8597, Accuracy: 0.6331
Training loss (for one batch) at step 110: 362.5974, Accuracy: 0.6328
Training loss (for one batch) at step 120: 371.5395, Accuracy: 0.6305
Training loss (for one batch) at step 130: 365.3699, Accuracy: 0.6326
Training loss (for one batch) at step 140: 378.5998, Accuracy: 0.6342
---- Training ----
Training loss: 340.2219
Training acc over epoch: 0.6334
---- Validation ----
Validation loss: 76.6787
Validation acc: 0.6784
Time taken: 9.64s
Start of epoch 3
Training loss (for one batch) at step 0: 362.3145, Accuracy: 0.6600
Training loss (for one batch) at step 10: 357.5943, Accuracy: 0.6455
Training loss (for one batch) at step 20: 370.5651, Accuracy: 0.6348
Training loss (for one batch) at step 30: 353.3396, Accuracy: 0.6413
Training loss (for one batch) at step 40: 360.1042, Accuracy: 0.6463
Training loss (for one batch) at step 50: 341.5472, Accuracy: 0.6453
Training loss (for one batch) at step 60: 361.8873, Accuracy: 0.6461
Training loss (for one batch) at step 70: 344.2211, Accuracy: 0.6454
Training loss (for one batch) at step 80: 338.4622, Accuracy: 0.6478
Training loss (for one batch) at step 90: 348.9875, Accuracy: 0.6481
Training loss (for one batch) at step 100: 345.3639, Accuracy: 0.6465
Training loss (for one batch) at step 110: 343.4567, Accuracy: 0.6483
Training loss (for one batch) at step 120: 383.7917, Accuracy: 0.6491
Training loss (for one batch) at step 130: 362.2031, Accuracy: 0.6511
Training loss (for one batch) at step 140: 358.5007, Accuracy: 0.6520
---- Training ----
Training loss: 315.4431
Training acc over epoch: 0.6524
---- Validation ----
Validation loss: 74.4960
Validation acc: 0.6803
Time taken: 9.69s
Start of epoch 4
Training loss (for one batch) at step 0: 353.1662, Accuracy: 0.7100
Training loss (for one batch) at step 10: 315.5179, Accuracy: 0.6564
Training loss (for one batch) at step 20: 356.0611, Accuracy: 0.6595
Training loss (for one batch) at step 30: 342.7633, Accuracy: 0.6587
Training loss (for one batch) at step 40: 342.4004, Accuracy: 0.6620
Training loss (for one batch) at step 50: 332.2070, Accuracy: 0.6659
Training loss (for one batch) at step 60: 346.2701, Accuracy: 0.6702
Training loss (for one batch) at step 70: 339.8632, Accuracy: 0.6689
Training loss (for one batch) at step 80: 332.8863, Accuracy: 0.6721
Training loss (for one batch) at step 90: 338.4653, Accuracy: 0.6703
Training loss (for one batch) at step 100: 336.6691, Accuracy: 0.6704
Training loss (for one batch) at step 110: 355.5804, Accuracy: 0.6685
Training loss (for one batch) at step 120: 353.3295, Accuracy: 0.6668
Training loss (for one batch) at step 130: 329.4775, Accuracy: 0.6656
Training loss (for one batch) at step 140: 321.3842, Accuracy: 0.6650
---- Training ----
Training loss: 288.3111
Training acc over epoch: 0.6652
---- Validation ----
Validation loss: 78.0440
Validation acc: 0.6910
Time taken: 9.60s
Start of epoch 5
Training loss (for one batch) at step 0: 345.0480, Accuracy: 0.6900
Training loss (for one batch) at step 10: 331.3069, Accuracy: 0.7091
Training loss (for one batch) at step 20: 350.7400, Accuracy: 0.6990
Training loss (for one batch) at step 30: 337.7690, Accuracy: 0.6816
Training loss (for one batch) at step 40: 350.2240, Accuracy: 0.6805
Training loss (for one batch) at step 50: 331.9173, Accuracy: 0.6806
Training loss (for one batch) at step 60: 337.8227, Accuracy: 0.6849
Training loss (for one batch) at step 70: 333.6231, Accuracy: 0.6873
Training loss (for one batch) at step 80: 331.2105, Accuracy: 0.6864
Training loss (for one batch) at step 90: 326.0552, Accuracy: 0.6866
Training loss (for one batch) at step 100: 316.4421, Accuracy: 0.6867
Training loss (for one batch) at step 110: 336.8534, Accuracy: 0.6844
Training loss (for one batch) at step 120: 337.2903, Accuracy: 0.6836
Training loss (for one batch) at step 130: 333.8102, Accuracy: 0.6831
Training loss (for one batch) at step 140: 330.0198, Accuracy: 0.6808
---- Training ----
Training loss: 275.4659
Training acc over epoch: 0.6805
---- Validation ----
Validation loss: 64.0486
Validation acc: 0.6929
Time taken: 9.58s
Start of epoch 6
Training loss (for one batch) at step 0: 341.7885, Accuracy: 0.6600
Training loss (for one batch) at step 10: 317.0500, Accuracy: 0.6791
Training loss (for one batch) at step 20: 315.4883, Accuracy: 0.6957
Training loss (for one batch) at step 30: 322.5573, Accuracy: 0.6916
Training loss (for one batch) at step 40: 323.7074, Accuracy: 0.6871
Training loss (for one batch) at step 50: 332.8784, Accuracy: 0.6890
Training loss (for one batch) at step 60: 320.9678, Accuracy: 0.6877
Training loss (for one batch) at step 70: 325.7137, Accuracy: 0.6889
Training loss (for one batch) at step 80: 328.9455, Accuracy: 0.6858
Training loss (for one batch) at step 90: 332.9308, Accuracy: 0.6853
Training loss (for one batch) at step 100: 324.8896, Accuracy: 0.6871
Training loss (for one batch) at step 110: 312.8874, Accuracy: 0.6868
Training loss (for one batch) at step 120: 320.6367, Accuracy: 0.6877
Training loss (for one batch) at step 130: 339.0819, Accuracy: 0.6868
Training loss (for one batch) at step 140: 338.1364, Accuracy: 0.6882
---- Training ----
Training loss: 267.4569
Training acc over epoch: 0.6898
---- Validation ----
Validation loss: 64.6937
Validation acc: 0.7023
Time taken: 10.42s
Start of epoch 7
Training loss (for one batch) at step 0: 321.4506, Accuracy: 0.7500
Training loss (for one batch) at step 10: 324.3830, Accuracy: 0.7082
Training loss (for one batch) at step 20: 335.8668, Accuracy: 0.7086
Training loss (for one batch) at step 30: 341.3915, Accuracy: 0.7039
Training loss (for one batch) at step 40: 321.2319, Accuracy: 0.7017
Training loss (for one batch) at step 50: 304.8556, Accuracy: 0.7024
Training loss (for one batch) at step 60: 309.0815, Accuracy: 0.7026
Training loss (for one batch) at step 70: 328.6027, Accuracy: 0.7006
Training loss (for one batch) at step 80: 317.6494, Accuracy: 0.7025
Training loss (for one batch) at step 90: 314.4439, Accuracy: 0.7033
Training loss (for one batch) at step 100: 308.5929, Accuracy: 0.7028
Training loss (for one batch) at step 110: 305.4953, Accuracy: 0.7014
Training loss (for one batch) at step 120: 321.3786, Accuracy: 0.7017
Training loss (for one batch) at step 130: 308.8441, Accuracy: 0.7008
Training loss (for one batch) at step 140: 320.2889, Accuracy: 0.7015
---- Training ----
Training loss: 285.2229
Training acc over epoch: 0.7012
---- Validation ----
Validation loss: 69.4343
Validation acc: 0.6991
Time taken: 9.59s
Start of epoch 8
Training loss (for one batch) at step 0: 332.4197, Accuracy: 0.7200
Training loss (for one batch) at step 10: 307.3240, Accuracy: 0.7055
Training loss (for one batch) at step 20: 316.6884, Accuracy: 0.7076
Training loss (for one batch) at step 30: 311.5858, Accuracy: 0.7058
Training loss (for one batch) at step 40: 308.5185, Accuracy: 0.7049
Training loss (for one batch) at step 50: 313.3641, Accuracy: 0.7118
Training loss (for one batch) at step 60: 299.7479, Accuracy: 0.7118
Training loss (for one batch) at step 70: 307.5673, Accuracy: 0.7123
Training loss (for one batch) at step 80: 317.8106, Accuracy: 0.7125
Training loss (for one batch) at step 90: 335.4947, Accuracy: 0.7134
Training loss (for one batch) at step 100: 328.5099, Accuracy: 0.7129
Training loss (for one batch) at step 110: 317.0024, Accuracy: 0.7134
Training loss (for one batch) at step 120: 305.5397, Accuracy: 0.7147
Training loss (for one batch) at step 130: 314.6164, Accuracy: 0.7142
Training loss (for one batch) at step 140: 338.7523, Accuracy: 0.7138
---- Training ----
Training loss: 282.7435
Training acc over epoch: 0.7130
---- Validation ----
Validation loss: 66.9310
Validation acc: 0.7136
Time taken: 9.74s
Start of epoch 9
Training loss (for one batch) at step 0: 318.3720, Accuracy: 0.7300
Training loss (for one batch) at step 10: 313.7696, Accuracy: 0.7500
Training loss (for one batch) at step 20: 309.9187, Accuracy: 0.7324
Training loss (for one batch) at step 30: 314.7567, Accuracy: 0.7261
Training loss (for one batch) at step 40: 297.2915, Accuracy: 0.7263
Training loss (for one batch) at step 50: 299.2422, Accuracy: 0.7267
Training loss (for one batch) at step 60: 310.5887, Accuracy: 0.7280
Training loss (for one batch) at step 70: 315.0131, Accuracy: 0.7294
Training loss (for one batch) at step 80: 315.9003, Accuracy: 0.7259
Training loss (for one batch) at step 90: 288.9050, Accuracy: 0.7282
Training loss (for one batch) at step 100: 308.9812, Accuracy: 0.7253
Training loss (for one batch) at step 110: 304.1180, Accuracy: 0.7233
Training loss (for one batch) at step 120: 327.4302, Accuracy: 0.7221
Training loss (for one batch) at step 130: 313.0915, Accuracy: 0.7228
Training loss (for one batch) at step 140: 300.7927, Accuracy: 0.7240
---- Training ----
Training loss: 260.9564
Training acc over epoch: 0.7253
---- Validation ----
Validation loss: 73.4276
Validation acc: 0.7200
Time taken: 10.35s
Start of epoch 10
Training loss (for one batch) at step 0: 312.5225, Accuracy: 0.7300
Training loss (for one batch) at step 10: 297.5049, Accuracy: 0.7373
Training loss (for one batch) at step 20: 298.7780, Accuracy: 0.7329
Training loss (for one batch) at step 30: 318.7906, Accuracy: 0.7310
Training loss (for one batch) at step 40: 304.3469, Accuracy: 0.7310
Training loss (for one batch) at step 50: 305.5131, Accuracy: 0.7335
Training loss (for one batch) at step 60: 301.3974, Accuracy: 0.7334
Training loss (for one batch) at step 70: 308.1199, Accuracy: 0.7365
Training loss (for one batch) at step 80: 306.3313, Accuracy: 0.7320
Training loss (for one batch) at step 90: 298.6764, Accuracy: 0.7318
Training loss (for one batch) at step 100: 321.1740, Accuracy: 0.7319
Training loss (for one batch) at step 110: 312.0657, Accuracy: 0.7334
Training loss (for one batch) at step 120: 334.5854, Accuracy: 0.7325
Training loss (for one batch) at step 130: 302.5723, Accuracy: 0.7323
Training loss (for one batch) at step 140: 298.4282, Accuracy: 0.7326
---- Training ----
Training loss: 266.0488
Training acc over epoch: 0.7334
---- Validation ----
Validation loss: 65.9251
Validation acc: 0.7141
Time taken: 9.73s
Start of epoch 11
Training loss (for one batch) at step 0: 293.2030, Accuracy: 0.8100
Training loss (for one batch) at step 10: 291.8892, Accuracy: 0.7573
Training loss (for one batch) at step 20: 306.3955, Accuracy: 0.7519
Training loss (for one batch) at step 30: 309.8345, Accuracy: 0.7535
Training loss (for one batch) at step 40: 301.2954, Accuracy: 0.7512
Training loss (for one batch) at step 50: 318.3407, Accuracy: 0.7490
Training loss (for one batch) at step 60: 304.4682, Accuracy: 0.7525
Training loss (for one batch) at step 70: 301.5425, Accuracy: 0.7523
Training loss (for one batch) at step 80: 304.9781, Accuracy: 0.7531
Training loss (for one batch) at step 90: 309.9811, Accuracy: 0.7529
Training loss (for one batch) at step 100: 296.2306, Accuracy: 0.7524
Training loss (for one batch) at step 110: 288.0007, Accuracy: 0.7525
Training loss (for one batch) at step 120: 287.5788, Accuracy: 0.7507
Training loss (for one batch) at step 130: 301.3459, Accuracy: 0.7500
Training loss (for one batch) at step 140: 308.7523, Accuracy: 0.7499
---- Training ----
Training loss: 253.2676
Training acc over epoch: 0.7505
---- Validation ----
Validation loss: 70.4911
Validation acc: 0.7214
Time taken: 9.68s
Start of epoch 12
Training loss (for one batch) at step 0: 285.7964, Accuracy: 0.7500
Training loss (for one batch) at step 10: 286.6335, Accuracy: 0.7718
Training loss (for one batch) at step 20: 317.1285, Accuracy: 0.7571
Training loss (for one batch) at step 30: 299.2650, Accuracy: 0.7574
Training loss (for one batch) at step 40: 294.5296, Accuracy: 0.7585
Training loss (for one batch) at step 50: 286.6991, Accuracy: 0.7620
Training loss (for one batch) at step 60: 293.9570, Accuracy: 0.7634
Training loss (for one batch) at step 70: 290.3295, Accuracy: 0.7624
Training loss (for one batch) at step 80: 291.7386, Accuracy: 0.7593
Training loss (for one batch) at step 90: 298.1132, Accuracy: 0.7599
Training loss (for one batch) at step 100: 298.0750, Accuracy: 0.7610
Training loss (for one batch) at step 110: 290.3265, Accuracy: 0.7594
Training loss (for one batch) at step 120: 285.7820, Accuracy: 0.7585
Training loss (for one batch) at step 130: 309.8615, Accuracy: 0.7586
Training loss (for one batch) at step 140: 287.1933, Accuracy: 0.7578
---- Training ----
Training loss: 253.7690
Training acc over epoch: 0.7577
---- Validation ----
Validation loss: 77.1732
Validation acc: 0.7141
Time taken: 9.64s
Start of epoch 13
Training loss (for one batch) at step 0: 305.8049, Accuracy: 0.7300
Training loss (for one batch) at step 10: 285.2593, Accuracy: 0.7745
Training loss (for one batch) at step 20: 283.4224, Accuracy: 0.7714
Training loss (for one batch) at step 30: 294.5608, Accuracy: 0.7658
Training loss (for one batch) at step 40: 271.0587, Accuracy: 0.7663
Training loss (for one batch) at step 50: 296.1379, Accuracy: 0.7665
Training loss (for one batch) at step 60: 289.7236, Accuracy: 0.7651
Training loss (for one batch) at step 70: 298.9829, Accuracy: 0.7665
Training loss (for one batch) at step 80: 289.8520, Accuracy: 0.7658
Training loss (for one batch) at step 90: 316.5719, Accuracy: 0.7646
Training loss (for one batch) at step 100: 305.2875, Accuracy: 0.7647
Training loss (for one batch) at step 110: 301.4951, Accuracy: 0.7646
Training loss (for one batch) at step 120: 288.8124, Accuracy: 0.7640
Training loss (for one batch) at step 130: 297.0461, Accuracy: 0.7627
Training loss (for one batch) at step 140: 290.1013, Accuracy: 0.7627
---- Training ----
Training loss: 255.0035
Training acc over epoch: 0.7625
---- Validation ----
Validation loss: 66.1202
Validation acc: 0.7198
Time taken: 9.96s
Start of epoch 14
Training loss (for one batch) at step 0: 288.1241, Accuracy: 0.7400
Training loss (for one batch) at step 10: 293.3087, Accuracy: 0.7800
Training loss (for one batch) at step 20: 278.7477, Accuracy: 0.7824
Training loss (for one batch) at step 30: 293.9792, Accuracy: 0.7797
Training loss (for one batch) at step 40: 299.6009, Accuracy: 0.7785
Training loss (for one batch) at step 50: 284.5036, Accuracy: 0.7827
Training loss (for one batch) at step 60: 280.4675, Accuracy: 0.7808
Training loss (for one batch) at step 70: 291.1841, Accuracy: 0.7821
Training loss (for one batch) at step 80: 295.5735, Accuracy: 0.7825
Training loss (for one batch) at step 90: 286.6116, Accuracy: 0.7788
Training loss (for one batch) at step 100: 284.8466, Accuracy: 0.7765
Training loss (for one batch) at step 110: 284.9034, Accuracy: 0.7776
Training loss (for one batch) at step 120: 282.2497, Accuracy: 0.7785
Training loss (for one batch) at step 130: 287.7419, Accuracy: 0.7768
Training loss (for one batch) at step 140: 292.4550, Accuracy: 0.7757
---- Training ----
Training loss: 262.5373
Training acc over epoch: 0.7753
---- Validation ----
Validation loss: 70.8101
Validation acc: 0.7332
Time taken: 9.68s
Start of epoch 15
Training loss (for one batch) at step 0: 271.4908, Accuracy: 0.8300
Training loss (for one batch) at step 10: 288.4997, Accuracy: 0.7591
Training loss (for one batch) at step 20: 289.0241, Accuracy: 0.7833
Training loss (for one batch) at step 30: 286.8523, Accuracy: 0.7800
Training loss (for one batch) at step 40: 291.0082, Accuracy: 0.7815
Training loss (for one batch) at step 50: 272.8422, Accuracy: 0.7816
Training loss (for one batch) at step 60: 300.6828, Accuracy: 0.7802
Training loss (for one batch) at step 70: 297.7192, Accuracy: 0.7813
Training loss (for one batch) at step 80: 286.3366, Accuracy: 0.7833
Training loss (for one batch) at step 90: 291.8858, Accuracy: 0.7813
Training loss (for one batch) at step 100: 288.6480, Accuracy: 0.7820
Training loss (for one batch) at step 110: 273.4122, Accuracy: 0.7828
Training loss (for one batch) at step 120: 286.9872, Accuracy: 0.7821
Training loss (for one batch) at step 130: 266.3191, Accuracy: 0.7811
Training loss (for one batch) at step 140: 280.7135, Accuracy: 0.7809
---- Training ----
Training loss: 256.4079
Training acc over epoch: 0.7817
---- Validation ----
Validation loss: 73.2241
Validation acc: 0.7343
Time taken: 9.66s
Start of epoch 16
Training loss (for one batch) at step 0: 273.8684, Accuracy: 0.8400
Training loss (for one batch) at step 10: 275.1499, Accuracy: 0.7918
Training loss (for one batch) at step 20: 280.5081, Accuracy: 0.7957
Training loss (for one batch) at step 30: 294.6706, Accuracy: 0.7829
Training loss (for one batch) at step 40: 275.4331, Accuracy: 0.7832
Training loss (for one batch) at step 50: 293.3192, Accuracy: 0.7778
Training loss (for one batch) at step 60: 276.2329, Accuracy: 0.7789
Training loss (for one batch) at step 70: 291.7392, Accuracy: 0.7806
Training loss (for one batch) at step 80: 277.3940, Accuracy: 0.7804
Training loss (for one batch) at step 90: 291.4399, Accuracy: 0.7791
Training loss (for one batch) at step 100: 283.0084, Accuracy: 0.7789
Training loss (for one batch) at step 110: 282.6981, Accuracy: 0.7803
Training loss (for one batch) at step 120: 278.4494, Accuracy: 0.7810
Training loss (for one batch) at step 130: 272.0303, Accuracy: 0.7811
Training loss (for one batch) at step 140: 280.7204, Accuracy: 0.7805
---- Training ----
Training loss: 262.6554
Training acc over epoch: 0.7800
---- Validation ----
Validation loss: 67.2954
Validation acc: 0.7198
Time taken: 9.80s
Start of epoch 17
Training loss (for one batch) at step 0: 276.3493, Accuracy: 0.7900
Training loss (for one batch) at step 10: 273.3190, Accuracy: 0.7991
Training loss (for one batch) at step 20: 291.9644, Accuracy: 0.8029
Training loss (for one batch) at step 30: 284.1984, Accuracy: 0.7939
Training loss (for one batch) at step 40: 279.7314, Accuracy: 0.7898
Training loss (for one batch) at step 50: 276.8441, Accuracy: 0.7947
Training loss (for one batch) at step 60: 299.7368, Accuracy: 0.7925
Training loss (for one batch) at step 70: 287.6313, Accuracy: 0.7930
Training loss (for one batch) at step 80: 281.3807, Accuracy: 0.7911
Training loss (for one batch) at step 90: 276.4063, Accuracy: 0.7910
Training loss (for one batch) at step 100: 296.4628, Accuracy: 0.7919
Training loss (for one batch) at step 110: 289.6125, Accuracy: 0.7932
Training loss (for one batch) at step 120: 282.2740, Accuracy: 0.7934
Training loss (for one batch) at step 130: 283.9520, Accuracy: 0.7944
Training loss (for one batch) at step 140: 272.2465, Accuracy: 0.7933
---- Training ----
Training loss: 250.5517
Training acc over epoch: 0.7921
---- Validation ----
Validation loss: 80.6092
Validation acc: 0.7281
Time taken: 9.75s
Start of epoch 18
Training loss (for one batch) at step 0: 299.2094, Accuracy: 0.7100
Training loss (for one batch) at step 10: 284.2640, Accuracy: 0.7882
Training loss (for one batch) at step 20: 291.3737, Accuracy: 0.7910
Training loss (for one batch) at step 30: 265.8117, Accuracy: 0.7897
Training loss (for one batch) at step 40: 269.2454, Accuracy: 0.7966
Training loss (for one batch) at step 50: 265.5245, Accuracy: 0.7955
Training loss (for one batch) at step 60: 284.3410, Accuracy: 0.7975
Training loss (for one batch) at step 70: 288.6520, Accuracy: 0.7965
Training loss (for one batch) at step 80: 294.8955, Accuracy: 0.7946
Training loss (for one batch) at step 90: 285.7061, Accuracy: 0.7960
Training loss (for one batch) at step 100: 283.0970, Accuracy: 0.7951
Training loss (for one batch) at step 110: 270.5354, Accuracy: 0.7965
Training loss (for one batch) at step 120: 282.6263, Accuracy: 0.7956
Training loss (for one batch) at step 130: 286.4620, Accuracy: 0.7956
Training loss (for one batch) at step 140: 287.9868, Accuracy: 0.7955
---- Training ----
Training loss: 246.0517
Training acc over epoch: 0.7948
---- Validation ----
Validation loss: 66.9030
Validation acc: 0.7257
Time taken: 9.73s
Start of epoch 19
Training loss (for one batch) at step 0: 260.3656, Accuracy: 0.8400
Training loss (for one batch) at step 10: 266.1412, Accuracy: 0.8200
Training loss (for one batch) at step 20: 261.8364, Accuracy: 0.8005
Training loss (for one batch) at step 30: 284.2681, Accuracy: 0.7871
Training loss (for one batch) at step 40: 261.4339, Accuracy: 0.7937
Training loss (for one batch) at step 50: 280.3197, Accuracy: 0.7975
Training loss (for one batch) at step 60: 275.0101, Accuracy: 0.8000
Training loss (for one batch) at step 70: 273.3321, Accuracy: 0.7976
Training loss (for one batch) at step 80: 267.8221, Accuracy: 0.7973
Training loss (for one batch) at step 90: 272.1797, Accuracy: 0.7971
Training loss (for one batch) at step 100: 260.0889, Accuracy: 0.7985
Training loss (for one batch) at step 110: 256.8557, Accuracy: 0.7984
Training loss (for one batch) at step 120: 275.8348, Accuracy: 0.7988
Training loss (for one batch) at step 130: 273.0999, Accuracy: 0.7976
Training loss (for one batch) at step 140: 284.1077, Accuracy: 0.7962
---- Training ----
Training loss: 239.8792
Training acc over epoch: 0.7965
---- Validation ----
Validation loss: 59.4943
Validation acc: 0.7260
Time taken: 9.81s
Start of epoch 20
Training loss (for one batch) at step 0: 267.5208, Accuracy: 0.8400
Training loss (for one batch) at step 10: 296.7277, Accuracy: 0.8145
Training loss (for one batch) at step 20: 281.9499, Accuracy: 0.8062
Training loss (for one batch) at step 30: 280.5839, Accuracy: 0.8045
Training loss (for one batch) at step 40: 268.7167, Accuracy: 0.8022
Training loss (for one batch) at step 50: 251.4016, Accuracy: 0.8039
Training loss (for one batch) at step 60: 244.0307, Accuracy: 0.8066
Training loss (for one batch) at step 70: 262.3195, Accuracy: 0.8025
Training loss (for one batch) at step 80: 270.9123, Accuracy: 0.8005
Training loss (for one batch) at step 90: 259.1586, Accuracy: 0.7993
Training loss (for one batch) at step 100: 267.4345, Accuracy: 0.7975
Training loss (for one batch) at step 110: 257.5613, Accuracy: 0.7976
Training loss (for one batch) at step 120: 265.1866, Accuracy: 0.7979
Training loss (for one batch) at step 130: 261.2609, Accuracy: 0.7981
Training loss (for one batch) at step 140: 284.3525, Accuracy: 0.7986
---- Training ----
Training loss: 244.1778
Training acc over epoch: 0.7988
---- Validation ----
Validation loss: 73.2954
Validation acc: 0.7238
Time taken: 9.72s
Start of epoch 21
Training loss (for one batch) at step 0: 265.3951, Accuracy: 0.7400
Training loss (for one batch) at step 10: 280.5075, Accuracy: 0.8009
Training loss (for one batch) at step 20: 278.6241, Accuracy: 0.8000
Training loss (for one batch) at step 30: 251.6960, Accuracy: 0.8029
Training loss (for one batch) at step 40: 264.1543, Accuracy: 0.8007
Training loss (for one batch) at step 50: 255.2797, Accuracy: 0.8065
Training loss (for one batch) at step 60: 263.0266, Accuracy: 0.8052
Training loss (for one batch) at step 70: 296.2678, Accuracy: 0.8038
Training loss (for one batch) at step 80: 274.7296, Accuracy: 0.8042
Training loss (for one batch) at step 90: 294.4360, Accuracy: 0.8027
Training loss (for one batch) at step 100: 251.4995, Accuracy: 0.8038
Training loss (for one batch) at step 110: 266.3806, Accuracy: 0.8019
Training loss (for one batch) at step 120: 282.3898, Accuracy: 0.8014
Training loss (for one batch) at step 130: 258.4463, Accuracy: 0.8008
Training loss (for one batch) at step 140: 275.0091, Accuracy: 0.7994
---- Training ----
Training loss: 232.7555
Training acc over epoch: 0.7995
---- Validation ----
Validation loss: 70.3553
Validation acc: 0.7241
Time taken: 9.59s
Start of epoch 22
Training loss (for one batch) at step 0: 256.6317, Accuracy: 0.8700
Training loss (for one batch) at step 10: 267.0557, Accuracy: 0.8173
Training loss (for one batch) at step 20: 247.0319, Accuracy: 0.8219
Training loss (for one batch) at step 30: 250.3326, Accuracy: 0.8090
Training loss (for one batch) at step 40: 256.8799, Accuracy: 0.8093
Training loss (for one batch) at step 50: 265.6338, Accuracy: 0.8118
Training loss (for one batch) at step 60: 266.9166, Accuracy: 0.8110
Training loss (for one batch) at step 70: 285.6589, Accuracy: 0.8075
Training loss (for one batch) at step 80: 264.8353, Accuracy: 0.8043
Training loss (for one batch) at step 90: 276.4013, Accuracy: 0.8040
Training loss (for one batch) at step 100: 272.3308, Accuracy: 0.8050
Training loss (for one batch) at step 110: 279.8630, Accuracy: 0.8043
Training loss (for one batch) at step 120: 254.5966, Accuracy: 0.8061
Training loss (for one batch) at step 130: 258.0412, Accuracy: 0.8053
Training loss (for one batch) at step 140: 256.1932, Accuracy: 0.8045
---- Training ----
Training loss: 222.2882
Training acc over epoch: 0.8039
---- Validation ----
Validation loss: 73.0445
Validation acc: 0.7222
Time taken: 9.56s
Start of epoch 23
Training loss (for one batch) at step 0: 234.7218, Accuracy: 0.9000
Training loss (for one batch) at step 10: 264.3424, Accuracy: 0.8236
Training loss (for one batch) at step 20: 277.0273, Accuracy: 0.8329
Training loss (for one batch) at step 30: 249.5613, Accuracy: 0.8281
Training loss (for one batch) at step 40: 254.1422, Accuracy: 0.8212
Training loss (for one batch) at step 50: 251.4501, Accuracy: 0.8237
Training loss (for one batch) at step 60: 260.8612, Accuracy: 0.8231
Training loss (for one batch) at step 70: 242.9721, Accuracy: 0.8187
Training loss (for one batch) at step 80: 302.3999, Accuracy: 0.8157
Training loss (for one batch) at step 90: 254.4896, Accuracy: 0.8135
Training loss (for one batch) at step 100: 249.0810, Accuracy: 0.8157
Training loss (for one batch) at step 110: 248.4544, Accuracy: 0.8156
Training loss (for one batch) at step 120: 267.6320, Accuracy: 0.8136
Training loss (for one batch) at step 130: 253.3341, Accuracy: 0.8123
Training loss (for one batch) at step 140: 268.4758, Accuracy: 0.8113
---- Training ----
Training loss: 231.1224
Training acc over epoch: 0.8118
---- Validation ----
Validation loss: 67.9434
Validation acc: 0.7335
Time taken: 9.64s
Start of epoch 24
Training loss (for one batch) at step 0: 239.0714, Accuracy: 0.8400
Training loss (for one batch) at step 10: 255.4821, Accuracy: 0.8155
Training loss (for one batch) at step 20: 276.9843, Accuracy: 0.8167
Training loss (for one batch) at step 30: 270.6553, Accuracy: 0.8100
Training loss (for one batch) at step 40: 254.6047, Accuracy: 0.8083
Training loss (for one batch) at step 50: 261.0518, Accuracy: 0.8098
Training loss (for one batch) at step 60: 263.5450, Accuracy: 0.8146
Training loss (for one batch) at step 70: 279.2819, Accuracy: 0.8125
Training loss (for one batch) at step 80: 254.0916, Accuracy: 0.8114
Training loss (for one batch) at step 90: 246.0985, Accuracy: 0.8120
Training loss (for one batch) at step 100: 247.5107, Accuracy: 0.8119
Training loss (for one batch) at step 110: 265.1039, Accuracy: 0.8101
Training loss (for one batch) at step 120: 259.2791, Accuracy: 0.8105
Training loss (for one batch) at step 130: 244.1434, Accuracy: 0.8106
Training loss (for one batch) at step 140: 264.1385, Accuracy: 0.8097
---- Training ----
Training loss: 236.7735
Training acc over epoch: 0.8092
---- Validation ----
Validation loss: 60.7450
Validation acc: 0.7356
Time taken: 9.57s
Start of epoch 25
Training loss (for one batch) at step 0: 266.1420, Accuracy: 0.7800
Training loss (for one batch) at step 10: 242.3362, Accuracy: 0.8155
Training loss (for one batch) at step 20: 256.8047, Accuracy: 0.8152
Training loss (for one batch) at step 30: 250.4251, Accuracy: 0.8161
Training loss (for one batch) at step 40: 258.0813, Accuracy: 0.8195
Training loss (for one batch) at step 50: 244.1172, Accuracy: 0.8218
Training loss (for one batch) at step 60: 238.4470, Accuracy: 0.8198
Training loss (for one batch) at step 70: 278.9713, Accuracy: 0.8182
Training loss (for one batch) at step 80: 268.7217, Accuracy: 0.8169
Training loss (for one batch) at step 90: 265.1924, Accuracy: 0.8142
Training loss (for one batch) at step 100: 259.3540, Accuracy: 0.8119
Training loss (for one batch) at step 110: 289.2863, Accuracy: 0.8123
Training loss (for one batch) at step 120: 270.8283, Accuracy: 0.8118
Training loss (for one batch) at step 130: 280.7487, Accuracy: 0.8115
Training loss (for one batch) at step 140: 280.1719, Accuracy: 0.8107
---- Training ----
Training loss: 230.5517
Training acc over epoch: 0.8108
---- Validation ----
Validation loss: 67.4935
Validation acc: 0.7206
Time taken: 9.68s
Start of epoch 26
Training loss (for one batch) at step 0: 240.5555, Accuracy: 0.8500
Training loss (for one batch) at step 10: 235.8230, Accuracy: 0.8164
Training loss (for one batch) at step 20: 250.6454, Accuracy: 0.8162
Training loss (for one batch) at step 30: 258.3047, Accuracy: 0.8174
Training loss (for one batch) at step 40: 245.9718, Accuracy: 0.8183
Training loss (for one batch) at step 50: 242.2543, Accuracy: 0.8190
Training loss (for one batch) at step 60: 248.4176, Accuracy: 0.8198
Training loss (for one batch) at step 70: 258.6502, Accuracy: 0.8207
Training loss (for one batch) at step 80: 257.2397, Accuracy: 0.8189
Training loss (for one batch) at step 90: 259.1409, Accuracy: 0.8169
Training loss (for one batch) at step 100: 265.5749, Accuracy: 0.8163
Training loss (for one batch) at step 110: 249.0969, Accuracy: 0.8178
Training loss (for one batch) at step 120: 257.0253, Accuracy: 0.8156
Training loss (for one batch) at step 130: 260.8539, Accuracy: 0.8166
Training loss (for one batch) at step 140: 254.0574, Accuracy: 0.8154
---- Training ----
Training loss: 213.4419
Training acc over epoch: 0.8149
---- Validation ----
Validation loss: 76.8305
Validation acc: 0.7230
Time taken: 9.74s
Start of epoch 27
Training loss (for one batch) at step 0: 240.1873, Accuracy: 0.8800
Training loss (for one batch) at step 10: 259.5272, Accuracy: 0.8391
Training loss (for one batch) at step 20: 229.3839, Accuracy: 0.8443
Training loss (for one batch) at step 30: 256.2858, Accuracy: 0.8342
Training loss (for one batch) at step 40: 231.8938, Accuracy: 0.8332
Training loss (for one batch) at step 50: 236.6128, Accuracy: 0.8296
Training loss (for one batch) at step 60: 236.4161, Accuracy: 0.8261
Training loss (for one batch) at step 70: 265.6651, Accuracy: 0.8246
Training loss (for one batch) at step 80: 234.3083, Accuracy: 0.8217
Training loss (for one batch) at step 90: 276.8144, Accuracy: 0.8209
Training loss (for one batch) at step 100: 242.2392, Accuracy: 0.8210
Training loss (for one batch) at step 110: 249.0187, Accuracy: 0.8234
Training loss (for one batch) at step 120: 268.6803, Accuracy: 0.8229
Training loss (for one batch) at step 130: 259.5720, Accuracy: 0.8215
Training loss (for one batch) at step 140: 265.2687, Accuracy: 0.8201
---- Training ----
Training loss: 237.1714
Training acc over epoch: 0.8178
---- Validation ----
Validation loss: 70.0988
Validation acc: 0.7265
Time taken: 9.49s
Start of epoch 28
Training loss (for one batch) at step 0: 234.2666, Accuracy: 0.8700
Training loss (for one batch) at step 10: 246.9106, Accuracy: 0.8382
Training loss (for one batch) at step 20: 256.2159, Accuracy: 0.8343
Training loss (for one batch) at step 30: 242.4002, Accuracy: 0.8371
Training loss (for one batch) at step 40: 230.1869, Accuracy: 0.8344
Training loss (for one batch) at step 50: 246.8077, Accuracy: 0.8327
Training loss (for one batch) at step 60: 248.5719, Accuracy: 0.8333
Training loss (for one batch) at step 70: 256.4771, Accuracy: 0.8296
Training loss (for one batch) at step 80: 257.8198, Accuracy: 0.8264
Training loss (for one batch) at step 90: 236.5115, Accuracy: 0.8247
Training loss (for one batch) at step 100: 238.5842, Accuracy: 0.8226
Training loss (for one batch) at step 110: 237.7143, Accuracy: 0.8222
Training loss (for one batch) at step 120: 239.3602, Accuracy: 0.8239
Training loss (for one batch) at step 130: 253.2077, Accuracy: 0.8227
Training loss (for one batch) at step 140: 262.8368, Accuracy: 0.8205
---- Training ----
Training loss: 217.1011
Training acc over epoch: 0.8203
---- Validation ----
Validation loss: 83.4478
Validation acc: 0.7308
Time taken: 11.21s
Start of epoch 29
Training loss (for one batch) at step 0: 244.9474, Accuracy: 0.8200
Training loss (for one batch) at step 10: 245.4117, Accuracy: 0.8227
Training loss (for one batch) at step 20: 232.6878, Accuracy: 0.8276
Training loss (for one batch) at step 30: 245.6552, Accuracy: 0.8306
Training loss (for one batch) at step 40: 251.4379, Accuracy: 0.8276
Training loss (for one batch) at step 50: 254.8624, Accuracy: 0.8298
Training loss (for one batch) at step 60: 230.3929, Accuracy: 0.8282
Training loss (for one batch) at step 70: 230.9916, Accuracy: 0.8294
Training loss (for one batch) at step 80: 261.0179, Accuracy: 0.8267
Training loss (for one batch) at step 90: 244.9283, Accuracy: 0.8243
Training loss (for one batch) at step 100: 235.9061, Accuracy: 0.8240
Training loss (for one batch) at step 110: 242.1841, Accuracy: 0.8246
Training loss (for one batch) at step 120: 253.8230, Accuracy: 0.8244
Training loss (for one batch) at step 130: 230.8425, Accuracy: 0.8231
Training loss (for one batch) at step 140: 255.6741, Accuracy: 0.8223
---- Training ----
Training loss: 222.7404
Training acc over epoch: 0.8229
---- Validation ----
Validation loss: 77.6406
Validation acc: 0.7160
Time taken: 40.90s
Start of epoch 30
Training loss (for one batch) at step 0: 235.3695, Accuracy: 0.8000
Training loss (for one batch) at step 10: 246.9484, Accuracy: 0.8164
Training loss (for one batch) at step 20: 234.0510, Accuracy: 0.8229
Training loss (for one batch) at step 30: 271.7093, Accuracy: 0.8274
Training loss (for one batch) at step 40: 224.9515, Accuracy: 0.8305
Training loss (for one batch) at step 50: 237.2552, Accuracy: 0.8339
Training loss (for one batch) at step 60: 253.7531, Accuracy: 0.8310
Training loss (for one batch) at step 70: 257.6694, Accuracy: 0.8301
Training loss (for one batch) at step 80: 225.9790, Accuracy: 0.8284
Training loss (for one batch) at step 90: 221.7338, Accuracy: 0.8292
Training loss (for one batch) at step 100: 236.6535, Accuracy: 0.8281
Training loss (for one batch) at step 110: 237.2383, Accuracy: 0.8274
Training loss (for one batch) at step 120: 247.4409, Accuracy: 0.8266
Training loss (for one batch) at step 130: 239.9785, Accuracy: 0.8250
Training loss (for one batch) at step 140: 248.9273, Accuracy: 0.8243
---- Training ----
Training loss: 220.5516
Training acc over epoch: 0.8246
---- Validation ----
Validation loss: 76.0133
Validation acc: 0.7063
Time taken: 44.87s
Start of epoch 31
Training loss (for one batch) at step 0: 247.7379, Accuracy: 0.7900
Training loss (for one batch) at step 10: 245.2246, Accuracy: 0.8073
Training loss (for one batch) at step 20: 229.4580, Accuracy: 0.8243
Training loss (for one batch) at step 30: 234.1063, Accuracy: 0.8200
Training loss (for one batch) at step 40: 252.9295, Accuracy: 0.8168
Training loss (for one batch) at step 50: 252.5554, Accuracy: 0.8212
Training loss (for one batch) at step 60: 236.8541, Accuracy: 0.8233
Training loss (for one batch) at step 70: 230.5878, Accuracy: 0.8228
Training loss (for one batch) at step 80: 248.3631, Accuracy: 0.8227
Training loss (for one batch) at step 90: 224.5014, Accuracy: 0.8221
Training loss (for one batch) at step 100: 236.2500, Accuracy: 0.8216
Training loss (for one batch) at step 110: 234.8505, Accuracy: 0.8244
Training loss (for one batch) at step 120: 224.0596, Accuracy: 0.8244
Training loss (for one batch) at step 130: 257.0427, Accuracy: 0.8234
Training loss (for one batch) at step 140: 252.2448, Accuracy: 0.8226
---- Training ----
Training loss: 208.8132
Training acc over epoch: 0.8226
---- Validation ----
Validation loss: 68.3137
Validation acc: 0.7332
Time taken: 13.16s
Start of epoch 32
Training loss (for one batch) at step 0: 235.9049, Accuracy: 0.8500
Training loss (for one batch) at step 10: 250.2157, Accuracy: 0.8373
Training loss (for one batch) at step 20: 234.9003, Accuracy: 0.8252
Training loss (for one batch) at step 30: 224.0032, Accuracy: 0.8219
Training loss (for one batch) at step 40: 261.6929, Accuracy: 0.8293
Training loss (for one batch) at step 50: 249.5603, Accuracy: 0.8312
Training loss (for one batch) at step 60: 243.6052, Accuracy: 0.8279
Training loss (for one batch) at step 70: 252.3250, Accuracy: 0.8262
Training loss (for one batch) at step 80: 223.3544, Accuracy: 0.8233
Training loss (for one batch) at step 90: 236.4557, Accuracy: 0.8213
Training loss (for one batch) at step 100: 245.2966, Accuracy: 0.8216
Training loss (for one batch) at step 110: 267.2341, Accuracy: 0.8210
Training loss (for one batch) at step 120: 256.1924, Accuracy: 0.8210
Training loss (for one batch) at step 130: 243.9004, Accuracy: 0.8190
Training loss (for one batch) at step 140: 226.0442, Accuracy: 0.8207
---- Training ----
Training loss: 213.9327
Training acc over epoch: 0.8202
---- Validation ----
Validation loss: 56.3077
Validation acc: 0.7106
Time taken: 10.18s
Start of epoch 33
Training loss (for one batch) at step 0: 219.3780, Accuracy: 0.8800
Training loss (for one batch) at step 10: 239.4357, Accuracy: 0.8264
Training loss (for one batch) at step 20: 244.6727, Accuracy: 0.8343
Training loss (for one batch) at step 30: 234.3810, Accuracy: 0.8303
Training loss (for one batch) at step 40: 223.2774, Accuracy: 0.8356
Training loss (for one batch) at step 50: 246.5952, Accuracy: 0.8388
Training loss (for one batch) at step 60: 235.0320, Accuracy: 0.8352
Training loss (for one batch) at step 70: 261.8803, Accuracy: 0.8341
Training loss (for one batch) at step 80: 225.6315, Accuracy: 0.8342
Training loss (for one batch) at step 90: 240.7434, Accuracy: 0.8322
Training loss (for one batch) at step 100: 258.1422, Accuracy: 0.8308
Training loss (for one batch) at step 110: 221.1968, Accuracy: 0.8310
Training loss (for one batch) at step 120: 237.9480, Accuracy: 0.8293
Training loss (for one batch) at step 130: 255.7223, Accuracy: 0.8293
Training loss (for one batch) at step 140: 234.2554, Accuracy: 0.8274
---- Training ----
Training loss: 208.3088
Training acc over epoch: 0.8283
---- Validation ----
Validation loss: 82.6219
Validation acc: 0.7139
Time taken: 9.55s
Start of epoch 34
Training loss (for one batch) at step 0: 244.2167, Accuracy: 0.8200
Training loss (for one batch) at step 10: 227.8242, Accuracy: 0.8445
Training loss (for one batch) at step 20: 221.4904, Accuracy: 0.8400
Training loss (for one batch) at step 30: 233.6985, Accuracy: 0.8326
Training loss (for one batch) at step 40: 237.6233, Accuracy: 0.8329
Training loss (for one batch) at step 50: 213.4700, Accuracy: 0.8337
Training loss (for one batch) at step 60: 246.1398, Accuracy: 0.8354
Training loss (for one batch) at step 70: 224.5761, Accuracy: 0.8346
Training loss (for one batch) at step 80: 257.4429, Accuracy: 0.8343
Training loss (for one batch) at step 90: 232.5710, Accuracy: 0.8301
Training loss (for one batch) at step 100: 233.8082, Accuracy: 0.8293
Training loss (for one batch) at step 110: 256.8495, Accuracy: 0.8284
Training loss (for one batch) at step 120: 258.6192, Accuracy: 0.8279
Training loss (for one batch) at step 130: 241.3607, Accuracy: 0.8292
Training loss (for one batch) at step 140: 233.5746, Accuracy: 0.8289
---- Training ----
Training loss: 231.1290
Training acc over epoch: 0.8288
---- Validation ----
Validation loss: 81.9866
Validation acc: 0.7093
Time taken: 9.54s
Start of epoch 35
Training loss (for one batch) at step 0: 230.6310, Accuracy: 0.7900
Training loss (for one batch) at step 10: 242.6400, Accuracy: 0.8273
Training loss (for one batch) at step 20: 233.3724, Accuracy: 0.8281
Training loss (for one batch) at step 30: 249.6623, Accuracy: 0.8290
Training loss (for one batch) at step 40: 235.9851, Accuracy: 0.8317
Training loss (for one batch) at step 50: 232.9670, Accuracy: 0.8335
Training loss (for one batch) at step 60: 227.9044, Accuracy: 0.8333
Training loss (for one batch) at step 70: 225.8050, Accuracy: 0.8303
Training loss (for one batch) at step 80: 233.7613, Accuracy: 0.8278
Training loss (for one batch) at step 90: 241.3584, Accuracy: 0.8275
Training loss (for one batch) at step 100: 260.7972, Accuracy: 0.8266
Training loss (for one batch) at step 110: 229.2756, Accuracy: 0.8283
Training loss (for one batch) at step 120: 235.9443, Accuracy: 0.8298
Training loss (for one batch) at step 130: 208.9955, Accuracy: 0.8306
Training loss (for one batch) at step 140: 219.9080, Accuracy: 0.8296
---- Training ----
Training loss: 215.8701
Training acc over epoch: 0.8293
---- Validation ----
Validation loss: 71.5641
Validation acc: 0.7176
Time taken: 9.69s
Start of epoch 36
Training loss (for one batch) at step 0: 218.7443, Accuracy: 0.9100
Training loss (for one batch) at step 10: 239.7801, Accuracy: 0.8409
Training loss (for one batch) at step 20: 229.6436, Accuracy: 0.8357
Training loss (for one batch) at step 30: 226.6543, Accuracy: 0.8390
Training loss (for one batch) at step 40: 214.9633, Accuracy: 0.8410
Training loss (for one batch) at step 50: 219.3358, Accuracy: 0.8424
Training loss (for one batch) at step 60: 216.2000, Accuracy: 0.8443
Training loss (for one batch) at step 70: 230.2177, Accuracy: 0.8385
Training loss (for one batch) at step 80: 248.7017, Accuracy: 0.8357
Training loss (for one batch) at step 90: 248.3117, Accuracy: 0.8335
Training loss (for one batch) at step 100: 217.7673, Accuracy: 0.8335
Training loss (for one batch) at step 110: 228.0830, Accuracy: 0.8339
Training loss (for one batch) at step 120: 243.9439, Accuracy: 0.8330
Training loss (for one batch) at step 130: 226.9312, Accuracy: 0.8324
Training loss (for one batch) at step 140: 253.1754, Accuracy: 0.8318
---- Training ----
Training loss: 227.4523
Training acc over epoch: 0.8316
---- Validation ----
Validation loss: 78.3847
Validation acc: 0.7225
Time taken: 9.59s
Start of epoch 37
Training loss (for one batch) at step 0: 220.8493, Accuracy: 0.8700
Training loss (for one batch) at step 10: 223.6789, Accuracy: 0.8318
Training loss (for one batch) at step 20: 226.2724, Accuracy: 0.8424
Training loss (for one batch) at step 30: 226.7975, Accuracy: 0.8355
Training loss (for one batch) at step 40: 233.7830, Accuracy: 0.8337
Training loss (for one batch) at step 50: 232.5348, Accuracy: 0.8369
Training loss (for one batch) at step 60: 230.6130, Accuracy: 0.8369
Training loss (for one batch) at step 70: 237.2115, Accuracy: 0.8346
Training loss (for one batch) at step 80: 228.8734, Accuracy: 0.8327
Training loss (for one batch) at step 90: 241.5975, Accuracy: 0.8310
Training loss (for one batch) at step 100: 218.0959, Accuracy: 0.8313
Training loss (for one batch) at step 110: 219.0130, Accuracy: 0.8331
Training loss (for one batch) at step 120: 238.1155, Accuracy: 0.8318
Training loss (for one batch) at step 130: 226.5946, Accuracy: 0.8315
Training loss (for one batch) at step 140: 258.4516, Accuracy: 0.8300
---- Training ----
Training loss: 202.0920
Training acc over epoch: 0.8288
---- Validation ----
Validation loss: 79.1211
Validation acc: 0.7249
Time taken: 9.63s
Start of epoch 38
Training loss (for one batch) at step 0: 232.9817, Accuracy: 0.8500
Training loss (for one batch) at step 10: 253.2494, Accuracy: 0.8291
Training loss (for one batch) at step 20: 228.4005, Accuracy: 0.8390
Training loss (for one batch) at step 30: 229.8097, Accuracy: 0.8352
Training loss (for one batch) at step 40: 208.9173, Accuracy: 0.8351
Training loss (for one batch) at step 50: 222.1250, Accuracy: 0.8359
Training loss (for one batch) at step 60: 238.3908, Accuracy: 0.8348
Training loss (for one batch) at step 70: 211.8581, Accuracy: 0.8356
Training loss (for one batch) at step 80: 245.3698, Accuracy: 0.8325
Training loss (for one batch) at step 90: 245.6047, Accuracy: 0.8320
Training loss (for one batch) at step 100: 232.1185, Accuracy: 0.8304
Training loss (for one batch) at step 110: 232.0072, Accuracy: 0.8296
Training loss (for one batch) at step 120: 225.2266, Accuracy: 0.8310
Training loss (for one batch) at step 130: 242.8485, Accuracy: 0.8300
Training loss (for one batch) at step 140: 238.0581, Accuracy: 0.8294
---- Training ----
Training loss: 218.3212
Training acc over epoch: 0.8288
---- Validation ----
Validation loss: 92.4215
Validation acc: 0.7117
Time taken: 9.83s
Start of epoch 39
Training loss (for one batch) at step 0: 219.8919, Accuracy: 0.8500
Training loss (for one batch) at step 10: 228.6551, Accuracy: 0.8364
Training loss (for one batch) at step 20: 217.4753, Accuracy: 0.8386
Training loss (for one batch) at step 30: 241.8304, Accuracy: 0.8316
Training loss (for one batch) at step 40: 223.0713, Accuracy: 0.8346
Training loss (for one batch) at step 50: 234.5320, Accuracy: 0.8375
Training loss (for one batch) at step 60: 214.8170, Accuracy: 0.8398
Training loss (for one batch) at step 70: 222.2683, Accuracy: 0.8387
Training loss (for one batch) at step 80: 216.3513, Accuracy: 0.8380
Training loss (for one batch) at step 90: 230.6865, Accuracy: 0.8342
Training loss (for one batch) at step 100: 227.7464, Accuracy: 0.8307
Training loss (for one batch) at step 110: 223.0265, Accuracy: 0.8311
Training loss (for one batch) at step 120: 226.1981, Accuracy: 0.8317
Training loss (for one batch) at step 130: 224.6382, Accuracy: 0.8315
Training loss (for one batch) at step 140: 220.7974, Accuracy: 0.8314
---- Training ----
Training loss: 211.2923
Training acc over epoch: 0.8308
---- Validation ----
Validation loss: 83.3380
Validation acc: 0.7241
Time taken: 9.77s
Start of epoch 40
Training loss (for one batch) at step 0: 218.7774, Accuracy: 0.8400
Training loss (for one batch) at step 10: 223.0770, Accuracy: 0.8391
Training loss (for one batch) at step 20: 214.9090, Accuracy: 0.8414
Training loss (for one batch) at step 30: 255.1538, Accuracy: 0.8342
Training loss (for one batch) at step 40: 225.4198, Accuracy: 0.8334
Training loss (for one batch) at step 50: 232.3734, Accuracy: 0.8371
Training loss (for one batch) at step 60: 210.4738, Accuracy: 0.8367
Training loss (for one batch) at step 70: 216.6053, Accuracy: 0.8356
Training loss (for one batch) at step 80: 236.8077, Accuracy: 0.8341
Training loss (for one batch) at step 90: 259.1871, Accuracy: 0.8325
Training loss (for one batch) at step 100: 215.5978, Accuracy: 0.8325
Training loss (for one batch) at step 110: 221.0990, Accuracy: 0.8328
Training loss (for one batch) at step 120: 223.6682, Accuracy: 0.8331
Training loss (for one batch) at step 130: 251.3186, Accuracy: 0.8320
Training loss (for one batch) at step 140: 216.3574, Accuracy: 0.8304
---- Training ----
Training loss: 193.0591
Training acc over epoch: 0.8307
---- Validation ----
Validation loss: 77.2192
Validation acc: 0.7206
Time taken: 9.57s
Start of epoch 41
Training loss (for one batch) at step 0: 225.8048, Accuracy: 0.8400
Training loss (for one batch) at step 10: 210.4793, Accuracy: 0.8364
Training loss (for one batch) at step 20: 217.9125, Accuracy: 0.8395
Training loss (for one batch) at step 30: 216.2503, Accuracy: 0.8384
Training loss (for one batch) at step 40: 224.9591, Accuracy: 0.8405
Training loss (for one batch) at step 50: 236.3305, Accuracy: 0.8410
Training loss (for one batch) at step 60: 233.0443, Accuracy: 0.8423
Training loss (for one batch) at step 70: 226.3054, Accuracy: 0.8401
Training loss (for one batch) at step 80: 200.0298, Accuracy: 0.8399
Training loss (for one batch) at step 90: 228.8600, Accuracy: 0.8375
Training loss (for one batch) at step 100: 236.6974, Accuracy: 0.8377
Training loss (for one batch) at step 110: 219.9674, Accuracy: 0.8376
Training loss (for one batch) at step 120: 233.4905, Accuracy: 0.8376
Training loss (for one batch) at step 130: 229.5224, Accuracy: 0.8365
Training loss (for one batch) at step 140: 231.0572, Accuracy: 0.8358
---- Training ----
Training loss: 231.0610
Training acc over epoch: 0.8358
---- Validation ----
Validation loss: 76.8167
Validation acc: 0.7050
Time taken: 9.81s
Start of epoch 42
Training loss (for one batch) at step 0: 222.6756, Accuracy: 0.8300
Training loss (for one batch) at step 10: 206.8035, Accuracy: 0.8473
Training loss (for one batch) at step 20: 217.5736, Accuracy: 0.8433
Training loss (for one batch) at step 30: 242.1945, Accuracy: 0.8377
Training loss (for one batch) at step 40: 213.5858, Accuracy: 0.8390
Training loss (for one batch) at step 50: 201.6639, Accuracy: 0.8416
Training loss (for one batch) at step 60: 215.8832, Accuracy: 0.8433
Training loss (for one batch) at step 70: 221.5072, Accuracy: 0.8427
Training loss (for one batch) at step 80: 221.5431, Accuracy: 0.8391
Training loss (for one batch) at step 90: 226.3882, Accuracy: 0.8374
Training loss (for one batch) at step 100: 234.4612, Accuracy: 0.8371
Training loss (for one batch) at step 110: 205.6039, Accuracy: 0.8383
Training loss (for one batch) at step 120: 213.1352, Accuracy: 0.8355
Training loss (for one batch) at step 130: 221.9426, Accuracy: 0.8345
Training loss (for one batch) at step 140: 237.4054, Accuracy: 0.8335
---- Training ----
Training loss: 200.0415
Training acc over epoch: 0.8332
---- Validation ----
Validation loss: 73.9523
Validation acc: 0.7168
Time taken: 9.60s
Start of epoch 43
Training loss (for one batch) at step 0: 227.2273, Accuracy: 0.8300
Training loss (for one batch) at step 10: 199.8592, Accuracy: 0.8264
Training loss (for one batch) at step 20: 238.0352, Accuracy: 0.8267
Training loss (for one batch) at step 30: 208.7123, Accuracy: 0.8306
Training loss (for one batch) at step 40: 222.4830, Accuracy: 0.8376
Training loss (for one batch) at step 50: 229.0582, Accuracy: 0.8367
Training loss (for one batch) at step 60: 206.6049, Accuracy: 0.8374
Training loss (for one batch) at step 70: 213.3353, Accuracy: 0.8339
Training loss (for one batch) at step 80: 242.9052, Accuracy: 0.8332
Training loss (for one batch) at step 90: 199.1495, Accuracy: 0.8349
Training loss (for one batch) at step 100: 226.2308, Accuracy: 0.8326
Training loss (for one batch) at step 110: 230.3159, Accuracy: 0.8343
Training loss (for one batch) at step 120: 234.3087, Accuracy: 0.8354
Training loss (for one batch) at step 130: 238.2438, Accuracy: 0.8337
Training loss (for one batch) at step 140: 218.3726, Accuracy: 0.8332
---- Training ----
Training loss: 206.7480
Training acc over epoch: 0.8325
---- Validation ----
Validation loss: 82.4739
Validation acc: 0.7123
Time taken: 9.64s
Start of epoch 44
Training loss (for one batch) at step 0: 227.2998, Accuracy: 0.8100
Training loss (for one batch) at step 10: 207.2437, Accuracy: 0.8427
Training loss (for one batch) at step 20: 234.3904, Accuracy: 0.8438
Training loss (for one batch) at step 30: 208.8576, Accuracy: 0.8429
Training loss (for one batch) at step 40: 214.4279, Accuracy: 0.8429
Training loss (for one batch) at step 50: 217.6417, Accuracy: 0.8429
Training loss (for one batch) at step 60: 209.7145, Accuracy: 0.8443
Training loss (for one batch) at step 70: 216.6323, Accuracy: 0.8428
Training loss (for one batch) at step 80: 214.6248, Accuracy: 0.8430
Training loss (for one batch) at step 90: 237.5531, Accuracy: 0.8435
Training loss (for one batch) at step 100: 208.2892, Accuracy: 0.8410
Training loss (for one batch) at step 110: 212.3934, Accuracy: 0.8405
Training loss (for one batch) at step 120: 213.7562, Accuracy: 0.8399
Training loss (for one batch) at step 130: 231.2911, Accuracy: 0.8402
Training loss (for one batch) at step 140: 213.3674, Accuracy: 0.8387
---- Training ----
Training loss: 204.2462
Training acc over epoch: 0.8383
---- Validation ----
Validation loss: 83.9572
Validation acc: 0.7179
Time taken: 9.62s
Start of epoch 45
Training loss (for one batch) at step 0: 215.8687, Accuracy: 0.8300
Training loss (for one batch) at step 10: 206.1679, Accuracy: 0.8545
Training loss (for one batch) at step 20: 223.4741, Accuracy: 0.8410
Training loss (for one batch) at step 30: 218.1333, Accuracy: 0.8361
Training loss (for one batch) at step 40: 218.2756, Accuracy: 0.8446
Training loss (for one batch) at step 50: 207.5849, Accuracy: 0.8459
Training loss (for one batch) at step 60: 215.5799, Accuracy: 0.8479
Training loss (for one batch) at step 70: 216.3109, Accuracy: 0.8448
Training loss (for one batch) at step 80: 201.5624, Accuracy: 0.8422
Training loss (for one batch) at step 90: 223.1222, Accuracy: 0.8412
Training loss (for one batch) at step 100: 216.0837, Accuracy: 0.8407
Training loss (for one batch) at step 110: 217.2916, Accuracy: 0.8406
Training loss (for one batch) at step 120: 215.3176, Accuracy: 0.8403
Training loss (for one batch) at step 130: 206.8334, Accuracy: 0.8396
Training loss (for one batch) at step 140: 222.4428, Accuracy: 0.8387
---- Training ----
Training loss: 180.2562
Training acc over epoch: 0.8386
---- Validation ----
Validation loss: 80.9475
Validation acc: 0.7123
Time taken: 9.66s
Start of epoch 46
Training loss (for one batch) at step 0: 209.5163, Accuracy: 0.9000
Training loss (for one batch) at step 10: 208.1714, Accuracy: 0.8564
Training loss (for one batch) at step 20: 211.0677, Accuracy: 0.8529
Training loss (for one batch) at step 30: 224.6927, Accuracy: 0.8471
Training loss (for one batch) at step 40: 220.4998, Accuracy: 0.8490
Training loss (for one batch) at step 50: 204.5905, Accuracy: 0.8494
Training loss (for one batch) at step 60: 229.6257, Accuracy: 0.8467
Training loss (for one batch) at step 70: 221.1817, Accuracy: 0.8468
Training loss (for one batch) at step 80: 234.8339, Accuracy: 0.8437
Training loss (for one batch) at step 90: 244.1282, Accuracy: 0.8420
Training loss (for one batch) at step 100: 205.1720, Accuracy: 0.8429
Training loss (for one batch) at step 110: 194.1258, Accuracy: 0.8410
Training loss (for one batch) at step 120: 226.3641, Accuracy: 0.8400
Training loss (for one batch) at step 130: 236.3472, Accuracy: 0.8390
Training loss (for one batch) at step 140: 229.2845, Accuracy: 0.8402
---- Training ----
Training loss: 199.9718
Training acc over epoch: 0.8400
---- Validation ----
Validation loss: 97.3775
Validation acc: 0.7063
Time taken: 9.62s
Start of epoch 47
Training loss (for one batch) at step 0: 200.9306, Accuracy: 0.9100
Training loss (for one batch) at step 10: 226.3208, Accuracy: 0.8455
Training loss (for one batch) at step 20: 228.0918, Accuracy: 0.8357
Training loss (for one batch) at step 30: 214.4343, Accuracy: 0.8381
Training loss (for one batch) at step 40: 204.1142, Accuracy: 0.8390
Training loss (for one batch) at step 50: 216.8195, Accuracy: 0.8384
Training loss (for one batch) at step 60: 188.5444, Accuracy: 0.8390
Training loss (for one batch) at step 70: 199.5902, Accuracy: 0.8377
Training loss (for one batch) at step 80: 229.2637, Accuracy: 0.8385
Training loss (for one batch) at step 90: 206.7613, Accuracy: 0.8402
Training loss (for one batch) at step 100: 205.0843, Accuracy: 0.8405
Training loss (for one batch) at step 110: 205.5422, Accuracy: 0.8388
Training loss (for one batch) at step 120: 228.2941, Accuracy: 0.8383
Training loss (for one batch) at step 130: 240.8025, Accuracy: 0.8378
Training loss (for one batch) at step 140: 213.6078, Accuracy: 0.8387
---- Training ----
Training loss: 177.6363
Training acc over epoch: 0.8383
---- Validation ----
Validation loss: 89.8141
Validation acc: 0.7058
Time taken: 11.62s
Start of epoch 48
Training loss (for one batch) at step 0: 222.6773, Accuracy: 0.8500
Training loss (for one batch) at step 10: 227.4706, Accuracy: 0.8391
Training loss (for one batch) at step 20: 210.6111, Accuracy: 0.8467
Training loss (for one batch) at step 30: 195.9910, Accuracy: 0.8497
Training loss (for one batch) at step 40: 219.4763, Accuracy: 0.8495
Training loss (for one batch) at step 50: 201.2770, Accuracy: 0.8512
Training loss (for one batch) at step 60: 205.5363, Accuracy: 0.8497
Training loss (for one batch) at step 70: 205.8807, Accuracy: 0.8504
Training loss (for one batch) at step 80: 212.2473, Accuracy: 0.8459
Training loss (for one batch) at step 90: 223.6670, Accuracy: 0.8441
Training loss (for one batch) at step 100: 200.8794, Accuracy: 0.8433
Training loss (for one batch) at step 110: 220.6843, Accuracy: 0.8423
Training loss (for one batch) at step 120: 199.8413, Accuracy: 0.8430
Training loss (for one batch) at step 130: 206.0350, Accuracy: 0.8423
Training loss (for one batch) at step 140: 219.5604, Accuracy: 0.8409
---- Training ----
Training loss: 192.0632
Training acc over epoch: 0.8411
---- Validation ----
Validation loss: 72.6851
Validation acc: 0.7063
Time taken: 45.04s
Start of epoch 49
Training loss (for one batch) at step 0: 214.0029, Accuracy: 0.8900
Training loss (for one batch) at step 10: 201.8076, Accuracy: 0.8573
Training loss (for one batch) at step 20: 207.5621, Accuracy: 0.8490
Training loss (for one batch) at step 30: 200.5233, Accuracy: 0.8474
Training loss (for one batch) at step 40: 208.3105, Accuracy: 0.8420
Training loss (for one batch) at step 50: 196.7835, Accuracy: 0.8422
Training loss (for one batch) at step 60: 207.2898, Accuracy: 0.8434
Training loss (for one batch) at step 70: 240.8251, Accuracy: 0.8432
Training loss (for one batch) at step 80: 199.7949, Accuracy: 0.8423
Training loss (for one batch) at step 90: 232.1943, Accuracy: 0.8415
Training loss (for one batch) at step 100: 205.3019, Accuracy: 0.8401
Training loss (for one batch) at step 110: 224.3130, Accuracy: 0.8401
Training loss (for one batch) at step 120: 206.5830, Accuracy: 0.8412
Training loss (for one batch) at step 130: 207.0338, Accuracy: 0.8393
Training loss (for one batch) at step 140: 209.5956, Accuracy: 0.8385
---- Training ----
Training loss: 191.1318
Training acc over epoch: 0.8389
---- Validation ----
Validation loss: 75.7542
Validation acc: 0.7214
Time taken: 45.40s
===== Q: 0.0001
Validation acc: 0.7416
Validation AUC: 0.7391
Validation Balanced_ACC: 0.4842
Validation MI: 0.1399
Validation Normalized MI: 0.2092
Validation Adjusted MI: 0.2092
Validation aUc_Sklearn: 0.8249
Start of epoch 0
Training loss (for one batch) at step 0: 462.4945, Accuracy: 0.5500
Training loss (for one batch) at step 10: 481.6254, Accuracy: 0.5355
Training loss (for one batch) at step 20: 436.0072, Accuracy: 0.5414
Training loss (for one batch) at step 30: 472.8438, Accuracy: 0.5368
Training loss (for one batch) at step 40: 433.4752, Accuracy: 0.5434
Training loss (for one batch) at step 50: 411.9770, Accuracy: 0.5512
Training loss (for one batch) at step 60: 441.7998, Accuracy: 0.5564
Training loss (for one batch) at step 70: 438.5501, Accuracy: 0.5573
Training loss (for one batch) at step 80: 422.8438, Accuracy: 0.5531
Training loss (for one batch) at step 90: 447.9939, Accuracy: 0.5555
Training loss (for one batch) at step 100: 461.6626, Accuracy: 0.5605
Training loss (for one batch) at step 110: 440.8398, Accuracy: 0.5608
Training loss (for one batch) at step 120: 406.2740, Accuracy: 0.5661
Training loss (for one batch) at step 130: 404.3879, Accuracy: 0.5695
Training loss (for one batch) at step 140: 414.4918, Accuracy: 0.5716
---- Training ----
Training loss: 364.7257
Training acc over epoch: 0.5723
---- Validation ----
Validation loss: 115.5089
Validation acc: 0.5134
Time taken: 81.38s
Start of epoch 1
Training loss (for one batch) at step 0: 405.4017, Accuracy: 0.5500
Training loss (for one batch) at step 10: 415.0469, Accuracy: 0.5955
Training loss (for one batch) at step 20: 403.0074, Accuracy: 0.5976
Training loss (for one batch) at step 30: 406.2607, Accuracy: 0.6074
Training loss (for one batch) at step 40: 394.2004, Accuracy: 0.6098
Training loss (for one batch) at step 50: 394.2165, Accuracy: 0.6104
Training loss (for one batch) at step 60: 359.2870, Accuracy: 0.6092
Training loss (for one batch) at step 70: 401.5722, Accuracy: 0.6101
Training loss (for one batch) at step 80: 410.9602, Accuracy: 0.6121
Training loss (for one batch) at step 90: 371.4424, Accuracy: 0.6113
Training loss (for one batch) at step 100: 392.1620, Accuracy: 0.6106
Training loss (for one batch) at step 110: 405.6764, Accuracy: 0.6141
Training loss (for one batch) at step 120: 409.7331, Accuracy: 0.6160
Training loss (for one batch) at step 130: 391.8154, Accuracy: 0.6163
Training loss (for one batch) at step 140: 354.0526, Accuracy: 0.6171
---- Training ----
Training loss: 326.1150
Training acc over epoch: 0.6184
---- Validation ----
Validation loss: 99.2873
Validation acc: 0.5156
Time taken: 42.21s
Start of epoch 2
Training loss (for one batch) at step 0: 351.1504, Accuracy: 0.7200
Training loss (for one batch) at step 10: 359.7747, Accuracy: 0.6618
Training loss (for one batch) at step 20: 372.9870, Accuracy: 0.6452
Training loss (for one batch) at step 30: 382.2603, Accuracy: 0.6445
Training loss (for one batch) at step 40: 370.2717, Accuracy: 0.6393
Training loss (for one batch) at step 50: 408.9975, Accuracy: 0.6373
Training loss (for one batch) at step 60: 364.3759, Accuracy: 0.6387
Training loss (for one batch) at step 70: 343.9471, Accuracy: 0.6358
Training loss (for one batch) at step 80: 346.7945, Accuracy: 0.6333
Training loss (for one batch) at step 90: 377.0862, Accuracy: 0.6366
Training loss (for one batch) at step 100: 379.5439, Accuracy: 0.6368
Training loss (for one batch) at step 110: 333.3613, Accuracy: 0.6363
Training loss (for one batch) at step 120: 351.4202, Accuracy: 0.6379
Training loss (for one batch) at step 130: 372.4748, Accuracy: 0.6392
Training loss (for one batch) at step 140: 363.9186, Accuracy: 0.6392
---- Training ----
Training loss: 306.4356
Training acc over epoch: 0.6404
---- Validation ----
Validation loss: 86.4792
Validation acc: 0.6478
Time taken: 75.23s
Start of epoch 3
Training loss (for one batch) at step 0: 316.3073, Accuracy: 0.7200
Training loss (for one batch) at step 10: 329.0652, Accuracy: 0.6682
Training loss (for one batch) at step 20: 397.9362, Accuracy: 0.6771
Training loss (for one batch) at step 30: 354.6608, Accuracy: 0.6639
Training loss (for one batch) at step 40: 368.3846, Accuracy: 0.6583
Training loss (for one batch) at step 50: 329.7556, Accuracy: 0.6643
Training loss (for one batch) at step 60: 363.1759, Accuracy: 0.6659
Training loss (for one batch) at step 70: 348.4765, Accuracy: 0.6635
Training loss (for one batch) at step 80: 367.6892, Accuracy: 0.6605
Training loss (for one batch) at step 90: 369.4305, Accuracy: 0.6607
Training loss (for one batch) at step 100: 333.4050, Accuracy: 0.6645
Training loss (for one batch) at step 110: 342.3410, Accuracy: 0.6672
Training loss (for one batch) at step 120: 354.5013, Accuracy: 0.6679
Training loss (for one batch) at step 130: 344.5506, Accuracy: 0.6689
Training loss (for one batch) at step 140: 339.1610, Accuracy: 0.6691
---- Training ----
Training loss: 349.2352
Training acc over epoch: 0.6687
---- Validation ----
Validation loss: 68.8476
Validation acc: 0.7225
Time taken: 41.31s
Start of epoch 4
Training loss (for one batch) at step 0: 340.0846, Accuracy: 0.6300
Training loss (for one batch) at step 10: 350.6988, Accuracy: 0.6555
Training loss (for one batch) at step 20: 325.1117, Accuracy: 0.6657
Training loss (for one batch) at step 30: 341.2325, Accuracy: 0.6684
Training loss (for one batch) at step 40: 327.5476, Accuracy: 0.6751
Training loss (for one batch) at step 50: 328.1347, Accuracy: 0.6800
Training loss (for one batch) at step 60: 375.2064, Accuracy: 0.6816
Training loss (for one batch) at step 70: 367.6034, Accuracy: 0.6811
Training loss (for one batch) at step 80: 330.0310, Accuracy: 0.6848
Training loss (for one batch) at step 90: 333.6727, Accuracy: 0.6863
Training loss (for one batch) at step 100: 329.6299, Accuracy: 0.6853
Training loss (for one batch) at step 110: 334.2189, Accuracy: 0.6860
Training loss (for one batch) at step 120: 345.4995, Accuracy: 0.6861
Training loss (for one batch) at step 130: 364.2303, Accuracy: 0.6851
Training loss (for one batch) at step 140: 347.7419, Accuracy: 0.6855
---- Training ----
Training loss: 290.5633
Training acc over epoch: 0.6855
---- Validation ----
Validation loss: 76.0565
Validation acc: 0.7123
Time taken: 72.74s
Start of epoch 5
Training loss (for one batch) at step 0: 334.4956, Accuracy: 0.6800
Training loss (for one batch) at step 10: 330.0557, Accuracy: 0.7155
Training loss (for one batch) at step 20: 330.7626, Accuracy: 0.6957
Training loss (for one batch) at step 30: 342.0861, Accuracy: 0.7006
Training loss (for one batch) at step 40: 318.6769, Accuracy: 0.6978
Training loss (for one batch) at step 50: 339.4202, Accuracy: 0.6961
Training loss (for one batch) at step 60: 317.2092, Accuracy: 0.6997
Training loss (for one batch) at step 70: 340.0663, Accuracy: 0.7014
Training loss (for one batch) at step 80: 323.6572, Accuracy: 0.6990
Training loss (for one batch) at step 90: 336.8378, Accuracy: 0.7002
Training loss (for one batch) at step 100: 317.6731, Accuracy: 0.6995
Training loss (for one batch) at step 110: 319.2567, Accuracy: 0.6999
Training loss (for one batch) at step 120: 343.1408, Accuracy: 0.7014
Training loss (for one batch) at step 130: 299.9915, Accuracy: 0.7024
Training loss (for one batch) at step 140: 328.7273, Accuracy: 0.7038
---- Training ----
Training loss: 271.5647
Training acc over epoch: 0.7044
---- Validation ----
Validation loss: 66.6843
Validation acc: 0.7098
Time taken: 46.24s
Start of epoch 6
Training loss (for one batch) at step 0: 320.1594, Accuracy: 0.7300
Training loss (for one batch) at step 10: 330.8873, Accuracy: 0.7145
Training loss (for one batch) at step 20: 315.7589, Accuracy: 0.7033
Training loss (for one batch) at step 30: 335.0496, Accuracy: 0.7181
Training loss (for one batch) at step 40: 317.1343, Accuracy: 0.7134
Training loss (for one batch) at step 50: 322.8420, Accuracy: 0.7124
Training loss (for one batch) at step 60: 309.3087, Accuracy: 0.7116
Training loss (for one batch) at step 70: 339.3080, Accuracy: 0.7093
Training loss (for one batch) at step 80: 332.5640, Accuracy: 0.7114
Training loss (for one batch) at step 90: 333.9123, Accuracy: 0.7121
Training loss (for one batch) at step 100: 312.0687, Accuracy: 0.7135
Training loss (for one batch) at step 110: 326.7243, Accuracy: 0.7135
Training loss (for one batch) at step 120: 324.5206, Accuracy: 0.7136
Training loss (for one batch) at step 130: 299.7973, Accuracy: 0.7136
Training loss (for one batch) at step 140: 340.1161, Accuracy: 0.7155
---- Training ----
Training loss: 314.8104
Training acc over epoch: 0.7153
---- Validation ----
Validation loss: 68.3427
Validation acc: 0.7174
Time taken: 73.61s
Start of epoch 7
Training loss (for one batch) at step 0: 314.0048, Accuracy: 0.7400
Training loss (for one batch) at step 10: 306.2929, Accuracy: 0.7255
Training loss (for one batch) at step 20: 318.2923, Accuracy: 0.7129
Training loss (for one batch) at step 30: 330.7050, Accuracy: 0.7184
Training loss (for one batch) at step 40: 301.1106, Accuracy: 0.7285
Training loss (for one batch) at step 50: 328.3098, Accuracy: 0.7280
Training loss (for one batch) at step 60: 304.2132, Accuracy: 0.7311
Training loss (for one batch) at step 70: 333.9699, Accuracy: 0.7299
Training loss (for one batch) at step 80: 316.6306, Accuracy: 0.7286
Training loss (for one batch) at step 90: 310.9717, Accuracy: 0.7276
Training loss (for one batch) at step 100: 339.8084, Accuracy: 0.7277
Training loss (for one batch) at step 110: 302.9724, Accuracy: 0.7289
Training loss (for one batch) at step 120: 327.3542, Accuracy: 0.7297
Training loss (for one batch) at step 130: 315.7186, Accuracy: 0.7301
Training loss (for one batch) at step 140: 315.9429, Accuracy: 0.7287
---- Training ----
Training loss: 294.3959
Training acc over epoch: 0.7274
---- Validation ----
Validation loss: 72.1802
Validation acc: 0.7434
Time taken: 47.77s
Start of epoch 8
Training loss (for one batch) at step 0: 310.8565, Accuracy: 0.7300
Training loss (for one batch) at step 10: 293.7169, Accuracy: 0.7445
Training loss (for one batch) at step 20: 304.9658, Accuracy: 0.7348
Training loss (for one batch) at step 30: 310.4268, Accuracy: 0.7381
Training loss (for one batch) at step 40: 304.4953, Accuracy: 0.7429
Training loss (for one batch) at step 50: 309.0963, Accuracy: 0.7418
Training loss (for one batch) at step 60: 330.1755, Accuracy: 0.7451
Training loss (for one batch) at step 70: 307.6349, Accuracy: 0.7444
Training loss (for one batch) at step 80: 299.4986, Accuracy: 0.7435
Training loss (for one batch) at step 90: 307.6444, Accuracy: 0.7422
Training loss (for one batch) at step 100: 297.8293, Accuracy: 0.7414
Training loss (for one batch) at step 110: 294.5128, Accuracy: 0.7423
Training loss (for one batch) at step 120: 312.2430, Accuracy: 0.7438
Training loss (for one batch) at step 130: 304.7820, Accuracy: 0.7437
Training loss (for one batch) at step 140: 334.6964, Accuracy: 0.7438
---- Training ----
Training loss: 264.5877
Training acc over epoch: 0.7433
---- Validation ----
Validation loss: 61.4669
Validation acc: 0.7211
Time taken: 76.05s
Start of epoch 9
Training loss (for one batch) at step 0: 287.3747, Accuracy: 0.8200
Training loss (for one batch) at step 10: 298.2817, Accuracy: 0.7545
Training loss (for one batch) at step 20: 311.1218, Accuracy: 0.7495
Training loss (for one batch) at step 30: 317.3705, Accuracy: 0.7506
Training loss (for one batch) at step 40: 314.1895, Accuracy: 0.7490
Training loss (for one batch) at step 50: 298.9588, Accuracy: 0.7539
Training loss (for one batch) at step 60: 312.2599, Accuracy: 0.7510
Training loss (for one batch) at step 70: 302.9776, Accuracy: 0.7494
Training loss (for one batch) at step 80: 324.8993, Accuracy: 0.7510
Training loss (for one batch) at step 90: 306.8815, Accuracy: 0.7519
Training loss (for one batch) at step 100: 325.1688, Accuracy: 0.7510
Training loss (for one batch) at step 110: 303.2360, Accuracy: 0.7520
Training loss (for one batch) at step 120: 299.2087, Accuracy: 0.7532
Training loss (for one batch) at step 130: 301.7130, Accuracy: 0.7541
Training loss (for one batch) at step 140: 312.9973, Accuracy: 0.7540
---- Training ----
Training loss: 279.1200
Training acc over epoch: 0.7531
---- Validation ----
Validation loss: 68.8446
Validation acc: 0.7281
Time taken: 50.59s
Start of epoch 10
Training loss (for one batch) at step 0: 286.3283, Accuracy: 0.7500
Training loss (for one batch) at step 10: 299.8705, Accuracy: 0.7564
Training loss (for one batch) at step 20: 313.8468, Accuracy: 0.7624
Training loss (for one batch) at step 30: 284.4884, Accuracy: 0.7665
Training loss (for one batch) at step 40: 296.6488, Accuracy: 0.7683
Training loss (for one batch) at step 50: 303.2616, Accuracy: 0.7676
Training loss (for one batch) at step 60: 316.0962, Accuracy: 0.7692
Training loss (for one batch) at step 70: 305.5449, Accuracy: 0.7672
Training loss (for one batch) at step 80: 302.6543, Accuracy: 0.7649
Training loss (for one batch) at step 90: 308.8016, Accuracy: 0.7613
Training loss (for one batch) at step 100: 283.7222, Accuracy: 0.7621
Training loss (for one batch) at step 110: 300.6004, Accuracy: 0.7631
Training loss (for one batch) at step 120: 290.9043, Accuracy: 0.7628
Training loss (for one batch) at step 130: 283.5622, Accuracy: 0.7631
Training loss (for one batch) at step 140: 297.4100, Accuracy: 0.7632
---- Training ----
Training loss: 265.7136
Training acc over epoch: 0.7618
---- Validation ----
Validation loss: 74.1557
Validation acc: 0.7423
Time taken: 75.16s
Start of epoch 11
Training loss (for one batch) at step 0: 294.4209, Accuracy: 0.7900
Training loss (for one batch) at step 10: 289.8373, Accuracy: 0.7936
Training loss (for one batch) at step 20: 311.1196, Accuracy: 0.7767
Training loss (for one batch) at step 30: 328.1051, Accuracy: 0.7774
Training loss (for one batch) at step 40: 297.8694, Accuracy: 0.7749
Training loss (for one batch) at step 50: 278.7606, Accuracy: 0.7747
Training loss (for one batch) at step 60: 288.7660, Accuracy: 0.7777
Training loss (for one batch) at step 70: 297.8023, Accuracy: 0.7756
Training loss (for one batch) at step 80: 310.7255, Accuracy: 0.7741
Training loss (for one batch) at step 90: 312.9312, Accuracy: 0.7745
Training loss (for one batch) at step 100: 296.7748, Accuracy: 0.7733
Training loss (for one batch) at step 110: 299.2525, Accuracy: 0.7727
Training loss (for one batch) at step 120: 279.5258, Accuracy: 0.7736
Training loss (for one batch) at step 130: 287.1721, Accuracy: 0.7735
Training loss (for one batch) at step 140: 290.9178, Accuracy: 0.7726
---- Training ----
Training loss: 263.4150
Training acc over epoch: 0.7727
---- Validation ----
Validation loss: 73.2550
Validation acc: 0.7391
Time taken: 48.23s
Start of epoch 12
Training loss (for one batch) at step 0: 303.9907, Accuracy: 0.8100
Training loss (for one batch) at step 10: 293.8959, Accuracy: 0.8155
Training loss (for one batch) at step 20: 295.5364, Accuracy: 0.7952
Training loss (for one batch) at step 30: 292.0740, Accuracy: 0.7861
Training loss (for one batch) at step 40: 305.4835, Accuracy: 0.7810
Training loss (for one batch) at step 50: 282.4120, Accuracy: 0.7861
Training loss (for one batch) at step 60: 287.4642, Accuracy: 0.7893
Training loss (for one batch) at step 70: 288.0137, Accuracy: 0.7873
Training loss (for one batch) at step 80: 290.3631, Accuracy: 0.7832
Training loss (for one batch) at step 90: 310.0232, Accuracy: 0.7822
Training loss (for one batch) at step 100: 285.4058, Accuracy: 0.7813
Training loss (for one batch) at step 110: 295.1399, Accuracy: 0.7795
Training loss (for one batch) at step 120: 297.2738, Accuracy: 0.7804
Training loss (for one batch) at step 130: 302.7381, Accuracy: 0.7793
Training loss (for one batch) at step 140: 293.5962, Accuracy: 0.7794
---- Training ----
Training loss: 265.7573
Training acc over epoch: 0.7800
---- Validation ----
Validation loss: 72.0216
Validation acc: 0.7466
Time taken: 72.96s
Start of epoch 13
Training loss (for one batch) at step 0: 284.3481, Accuracy: 0.8500
Training loss (for one batch) at step 10: 288.7666, Accuracy: 0.7891
Training loss (for one batch) at step 20: 291.1651, Accuracy: 0.7871
Training loss (for one batch) at step 30: 277.1967, Accuracy: 0.7942
Training loss (for one batch) at step 40: 283.4806, Accuracy: 0.7912
Training loss (for one batch) at step 50: 298.0894, Accuracy: 0.7914
Training loss (for one batch) at step 60: 296.5533, Accuracy: 0.7936
Training loss (for one batch) at step 70: 275.7156, Accuracy: 0.7941
Training loss (for one batch) at step 80: 292.8945, Accuracy: 0.7894
Training loss (for one batch) at step 90: 284.6855, Accuracy: 0.7880
Training loss (for one batch) at step 100: 302.8197, Accuracy: 0.7846
Training loss (for one batch) at step 110: 293.3873, Accuracy: 0.7854
Training loss (for one batch) at step 120: 297.0097, Accuracy: 0.7845
Training loss (for one batch) at step 130: 294.6768, Accuracy: 0.7844
Training loss (for one batch) at step 140: 278.7377, Accuracy: 0.7830
---- Training ----
Training loss: 261.7491
Training acc over epoch: 0.7841
---- Validation ----
Validation loss: 68.2820
Validation acc: 0.7474
Time taken: 48.72s
Start of epoch 14
Training loss (for one batch) at step 0: 292.1087, Accuracy: 0.8000
Training loss (for one batch) at step 10: 304.4556, Accuracy: 0.7773
Training loss (for one batch) at step 20: 274.2166, Accuracy: 0.7771
Training loss (for one batch) at step 30: 283.0068, Accuracy: 0.7871
Training loss (for one batch) at step 40: 279.3158, Accuracy: 0.7883
Training loss (for one batch) at step 50: 280.7079, Accuracy: 0.7904
Training loss (for one batch) at step 60: 274.6526, Accuracy: 0.7923
Training loss (for one batch) at step 70: 297.0395, Accuracy: 0.7923
Training loss (for one batch) at step 80: 281.8358, Accuracy: 0.7870
Training loss (for one batch) at step 90: 292.2492, Accuracy: 0.7860
Training loss (for one batch) at step 100: 285.1976, Accuracy: 0.7859
Training loss (for one batch) at step 110: 287.4148, Accuracy: 0.7865
Training loss (for one batch) at step 120: 298.8933, Accuracy: 0.7858
Training loss (for one batch) at step 130: 278.1132, Accuracy: 0.7860
Training loss (for one batch) at step 140: 286.4241, Accuracy: 0.7856
---- Training ----
Training loss: 247.5695
Training acc over epoch: 0.7857
---- Validation ----
Validation loss: 63.6154
Validation acc: 0.7407
Time taken: 71.20s
Start of epoch 15
Training loss (for one batch) at step 0: 284.2998, Accuracy: 0.8000
Training loss (for one batch) at step 10: 261.7900, Accuracy: 0.7927
Training loss (for one batch) at step 20: 280.9203, Accuracy: 0.8043
Training loss (for one batch) at step 30: 273.1909, Accuracy: 0.8023
Training loss (for one batch) at step 40: 293.9917, Accuracy: 0.8005
Training loss (for one batch) at step 50: 277.6827, Accuracy: 0.8020
Training loss (for one batch) at step 60: 278.8696, Accuracy: 0.8003
Training loss (for one batch) at step 70: 269.2966, Accuracy: 0.7982
Training loss (for one batch) at step 80: 265.9207, Accuracy: 0.7998
Training loss (for one batch) at step 90: 286.8654, Accuracy: 0.7974
Training loss (for one batch) at step 100: 266.5216, Accuracy: 0.7948
Training loss (for one batch) at step 110: 268.6319, Accuracy: 0.7941
Training loss (for one batch) at step 120: 264.7807, Accuracy: 0.7942
Training loss (for one batch) at step 130: 261.4902, Accuracy: 0.7940
Training loss (for one batch) at step 140: 279.0629, Accuracy: 0.7915
---- Training ----
Training loss: 246.4611
Training acc over epoch: 0.7927
---- Validation ----
Validation loss: 59.2873
Validation acc: 0.7512
Time taken: 47.58s
Start of epoch 16
Training loss (for one batch) at step 0: 270.9504, Accuracy: 0.8800
Training loss (for one batch) at step 10: 278.6192, Accuracy: 0.8245
Training loss (for one batch) at step 20: 260.3029, Accuracy: 0.8095
Training loss (for one batch) at step 30: 264.5818, Accuracy: 0.8016
Training loss (for one batch) at step 40: 269.8613, Accuracy: 0.8061
Training loss (for one batch) at step 50: 293.1936, Accuracy: 0.8076
Training loss (for one batch) at step 60: 286.4384, Accuracy: 0.8084
Training loss (for one batch) at step 70: 256.3606, Accuracy: 0.8082
Training loss (for one batch) at step 80: 305.2526, Accuracy: 0.8033
Training loss (for one batch) at step 90: 290.1290, Accuracy: 0.8031
Training loss (for one batch) at step 100: 283.7592, Accuracy: 0.8021
Training loss (for one batch) at step 110: 274.7346, Accuracy: 0.8016
Training loss (for one batch) at step 120: 269.0453, Accuracy: 0.8016
Training loss (for one batch) at step 130: 283.1921, Accuracy: 0.8020
Training loss (for one batch) at step 140: 278.8521, Accuracy: 0.8013
---- Training ----
Training loss: 255.6921
Training acc over epoch: 0.8019
---- Validation ----
Validation loss: 66.1974
Validation acc: 0.7423
Time taken: 70.29s
Start of epoch 17
Training loss (for one batch) at step 0: 269.1211, Accuracy: 0.8700
Training loss (for one batch) at step 10: 273.1809, Accuracy: 0.8145
Training loss (for one batch) at step 20: 276.4555, Accuracy: 0.8071
Training loss (for one batch) at step 30: 279.8569, Accuracy: 0.8068
Training loss (for one batch) at step 40: 263.2839, Accuracy: 0.8080
Training loss (for one batch) at step 50: 263.5492, Accuracy: 0.8033
Training loss (for one batch) at step 60: 277.9813, Accuracy: 0.8067
Training loss (for one batch) at step 70: 272.4186, Accuracy: 0.8072
Training loss (for one batch) at step 80: 287.8450, Accuracy: 0.8079
Training loss (for one batch) at step 90: 277.7690, Accuracy: 0.8067
Training loss (for one batch) at step 100: 275.8005, Accuracy: 0.8050
Training loss (for one batch) at step 110: 265.0179, Accuracy: 0.8051
Training loss (for one batch) at step 120: 279.6172, Accuracy: 0.8073
Training loss (for one batch) at step 130: 263.2620, Accuracy: 0.8050
Training loss (for one batch) at step 140: 284.9210, Accuracy: 0.8040
---- Training ----
Training loss: 217.6964
Training acc over epoch: 0.8046
---- Validation ----
Validation loss: 83.0035
Validation acc: 0.7606
Time taken: 48.26s
Start of epoch 18
Training loss (for one batch) at step 0: 278.6023, Accuracy: 0.8000
Training loss (for one batch) at step 10: 258.4124, Accuracy: 0.8082
Training loss (for one batch) at step 20: 280.3420, Accuracy: 0.8195
Training loss (for one batch) at step 30: 259.1935, Accuracy: 0.8126
Training loss (for one batch) at step 40: 270.9853, Accuracy: 0.8115
Training loss (for one batch) at step 50: 261.8784, Accuracy: 0.8118
Training loss (for one batch) at step 60: 278.9285, Accuracy: 0.8167
Training loss (for one batch) at step 70: 284.1260, Accuracy: 0.8169
Training loss (for one batch) at step 80: 273.2464, Accuracy: 0.8119
Training loss (for one batch) at step 90: 275.1462, Accuracy: 0.8122
Training loss (for one batch) at step 100: 304.0707, Accuracy: 0.8111
Training loss (for one batch) at step 110: 274.3969, Accuracy: 0.8093
Training loss (for one batch) at step 120: 257.3253, Accuracy: 0.8097
Training loss (for one batch) at step 130: 257.0511, Accuracy: 0.8102
Training loss (for one batch) at step 140: 290.6767, Accuracy: 0.8091
---- Training ----
Training loss: 231.9300
Training acc over epoch: 0.8091
---- Validation ----
Validation loss: 65.0506
Validation acc: 0.7246
Time taken: 69.21s
Start of epoch 19
Training loss (for one batch) at step 0: 274.7234, Accuracy: 0.7700
Training loss (for one batch) at step 10: 272.6221, Accuracy: 0.8109
Training loss (for one batch) at step 20: 258.1388, Accuracy: 0.8014
Training loss (for one batch) at step 30: 264.5694, Accuracy: 0.8048
Training loss (for one batch) at step 40: 256.8817, Accuracy: 0.8090
Training loss (for one batch) at step 50: 283.4989, Accuracy: 0.8145
Training loss (for one batch) at step 60: 248.0944, Accuracy: 0.8149
Training loss (for one batch) at step 70: 292.2782, Accuracy: 0.8121
Training loss (for one batch) at step 80: 260.2818, Accuracy: 0.8100
Training loss (for one batch) at step 90: 276.4194, Accuracy: 0.8101
Training loss (for one batch) at step 100: 264.2451, Accuracy: 0.8096
Training loss (for one batch) at step 110: 259.6966, Accuracy: 0.8110
Training loss (for one batch) at step 120: 261.0137, Accuracy: 0.8119
Training loss (for one batch) at step 130: 271.9887, Accuracy: 0.8094
Training loss (for one batch) at step 140: 262.2770, Accuracy: 0.8097
---- Training ----
Training loss: 243.1116
Training acc over epoch: 0.8100
---- Validation ----
Validation loss: 72.0539
Validation acc: 0.7499
Time taken: 49.18s
Start of epoch 20
Training loss (for one batch) at step 0: 276.8091, Accuracy: 0.7900
Training loss (for one batch) at step 10: 233.3608, Accuracy: 0.8364
Training loss (for one batch) at step 20: 275.2937, Accuracy: 0.8200
Training loss (for one batch) at step 30: 275.8474, Accuracy: 0.8210
Training loss (for one batch) at step 40: 262.0216, Accuracy: 0.8246
Training loss (for one batch) at step 50: 249.8707, Accuracy: 0.8239
Training loss (for one batch) at step 60: 251.1090, Accuracy: 0.8216
Training loss (for one batch) at step 70: 272.6890, Accuracy: 0.8193
Training loss (for one batch) at step 80: 258.3679, Accuracy: 0.8151
Training loss (for one batch) at step 90: 272.8467, Accuracy: 0.8131
Training loss (for one batch) at step 100: 266.4882, Accuracy: 0.8115
Training loss (for one batch) at step 110: 255.7374, Accuracy: 0.8134
Training loss (for one batch) at step 120: 259.4840, Accuracy: 0.8129
Training loss (for one batch) at step 130: 249.7050, Accuracy: 0.8135
Training loss (for one batch) at step 140: 253.0890, Accuracy: 0.8135
---- Training ----
Training loss: 247.9153
Training acc over epoch: 0.8131
---- Validation ----
Validation loss: 80.1312
Validation acc: 0.7469
Time taken: 68.72s
Start of epoch 21
Training loss (for one batch) at step 0: 282.2611, Accuracy: 0.7800
Training loss (for one batch) at step 10: 273.5271, Accuracy: 0.8155
Training loss (for one batch) at step 20: 260.9685, Accuracy: 0.8133
Training loss (for one batch) at step 30: 255.9858, Accuracy: 0.8165
Training loss (for one batch) at step 40: 263.7656, Accuracy: 0.8205
Training loss (for one batch) at step 50: 248.1970, Accuracy: 0.8247
Training loss (for one batch) at step 60: 258.5424, Accuracy: 0.8259
Training loss (for one batch) at step 70: 258.5097, Accuracy: 0.8224
Training loss (for one batch) at step 80: 259.1600, Accuracy: 0.8193
Training loss (for one batch) at step 90: 273.1676, Accuracy: 0.8189
Training loss (for one batch) at step 100: 270.5992, Accuracy: 0.8173
Training loss (for one batch) at step 110: 286.5652, Accuracy: 0.8179
Training loss (for one batch) at step 120: 266.8365, Accuracy: 0.8177
Training loss (for one batch) at step 130: 267.1877, Accuracy: 0.8167
Training loss (for one batch) at step 140: 264.7686, Accuracy: 0.8170
---- Training ----
Training loss: 228.2524
Training acc over epoch: 0.8162
---- Validation ----
Validation loss: 86.3974
Validation acc: 0.7362
Time taken: 47.59s
Start of epoch 22
Training loss (for one batch) at step 0: 266.6479, Accuracy: 0.7900
Training loss (for one batch) at step 10: 251.6947, Accuracy: 0.8236
Training loss (for one batch) at step 20: 262.5891, Accuracy: 0.8267
Training loss (for one batch) at step 30: 266.7438, Accuracy: 0.8265
Training loss (for one batch) at step 40: 241.9616, Accuracy: 0.8298
Training loss (for one batch) at step 50: 245.8165, Accuracy: 0.8314
Training loss (for one batch) at step 60: 254.3049, Accuracy: 0.8295
Training loss (for one batch) at step 70: 265.8818, Accuracy: 0.8300
Training loss (for one batch) at step 80: 264.3921, Accuracy: 0.8286
Training loss (for one batch) at step 90: 270.7385, Accuracy: 0.8267
Training loss (for one batch) at step 100: 274.8617, Accuracy: 0.8261
Training loss (for one batch) at step 110: 263.4152, Accuracy: 0.8275
Training loss (for one batch) at step 120: 254.7430, Accuracy: 0.8255
Training loss (for one batch) at step 130: 256.1833, Accuracy: 0.8240
Training loss (for one batch) at step 140: 270.9845, Accuracy: 0.8235
---- Training ----
Training loss: 226.4933
Training acc over epoch: 0.8234
---- Validation ----
Validation loss: 74.0025
Validation acc: 0.7313
Time taken: 65.90s
Start of epoch 23
Training loss (for one batch) at step 0: 256.8888, Accuracy: 0.8300
Training loss (for one batch) at step 10: 232.4763, Accuracy: 0.8400
Training loss (for one batch) at step 20: 256.7532, Accuracy: 0.8286
Training loss (for one batch) at step 30: 247.1436, Accuracy: 0.8281
Training loss (for one batch) at step 40: 254.0002, Accuracy: 0.8271
Training loss (for one batch) at step 50: 268.9605, Accuracy: 0.8263
Training loss (for one batch) at step 60: 240.5142, Accuracy: 0.8280
Training loss (for one batch) at step 70: 247.0794, Accuracy: 0.8287
Training loss (for one batch) at step 80: 243.1646, Accuracy: 0.8262
Training loss (for one batch) at step 90: 269.1459, Accuracy: 0.8243
Training loss (for one batch) at step 100: 264.4463, Accuracy: 0.8238
Training loss (for one batch) at step 110: 267.0716, Accuracy: 0.8242
Training loss (for one batch) at step 120: 258.2694, Accuracy: 0.8253
Training loss (for one batch) at step 130: 261.3351, Accuracy: 0.8240
Training loss (for one batch) at step 140: 242.9196, Accuracy: 0.8241
---- Training ----
Training loss: 233.8724
Training acc over epoch: 0.8242
---- Validation ----
Validation loss: 65.2387
Validation acc: 0.7294
Time taken: 48.65s
Start of epoch 24
Training loss (for one batch) at step 0: 258.0302, Accuracy: 0.8000
Training loss (for one batch) at step 10: 244.7492, Accuracy: 0.8355
Training loss (for one batch) at step 20: 244.7027, Accuracy: 0.8290
Training loss (for one batch) at step 30: 243.3652, Accuracy: 0.8287
Training loss (for one batch) at step 40: 237.2431, Accuracy: 0.8320
Training loss (for one batch) at step 50: 250.9232, Accuracy: 0.8325
Training loss (for one batch) at step 60: 273.0603, Accuracy: 0.8305
Training loss (for one batch) at step 70: 259.3407, Accuracy: 0.8287
Training loss (for one batch) at step 80: 235.0057, Accuracy: 0.8294
Training loss (for one batch) at step 90: 283.1172, Accuracy: 0.8280
Training loss (for one batch) at step 100: 234.1767, Accuracy: 0.8285
Training loss (for one batch) at step 110: 256.9485, Accuracy: 0.8284
Training loss (for one batch) at step 120: 249.4432, Accuracy: 0.8283
Training loss (for one batch) at step 130: 252.4680, Accuracy: 0.8279
Training loss (for one batch) at step 140: 260.3544, Accuracy: 0.8276
---- Training ----
Training loss: 237.2687
Training acc over epoch: 0.8270
---- Validation ----
Validation loss: 55.4981
Validation acc: 0.7397
Time taken: 66.48s
Start of epoch 25
Training loss (for one batch) at step 0: 250.1915, Accuracy: 0.8000
Training loss (for one batch) at step 10: 252.2198, Accuracy: 0.8109
Training loss (for one batch) at step 20: 259.6033, Accuracy: 0.8252
Training loss (for one batch) at step 30: 264.4632, Accuracy: 0.8300
Training loss (for one batch) at step 40: 247.0715, Accuracy: 0.8280
Training loss (for one batch) at step 50: 238.4732, Accuracy: 0.8271
Training loss (for one batch) at step 60: 253.9532, Accuracy: 0.8236
Training loss (for one batch) at step 70: 238.6592, Accuracy: 0.8261
Training loss (for one batch) at step 80: 248.9244, Accuracy: 0.8256
Training loss (for one batch) at step 90: 275.2949, Accuracy: 0.8238
Training loss (for one batch) at step 100: 231.5262, Accuracy: 0.8225
Training loss (for one batch) at step 110: 250.6348, Accuracy: 0.8236
Training loss (for one batch) at step 120: 230.9099, Accuracy: 0.8237
Training loss (for one batch) at step 130: 234.4983, Accuracy: 0.8240
Training loss (for one batch) at step 140: 248.7874, Accuracy: 0.8228
---- Training ----
Training loss: 224.8908
Training acc over epoch: 0.8221
---- Validation ----
Validation loss: 64.3145
Validation acc: 0.7308
Time taken: 48.33s
Start of epoch 26
Training loss (for one batch) at step 0: 250.0240, Accuracy: 0.7800
Training loss (for one batch) at step 10: 237.6883, Accuracy: 0.8336
Training loss (for one batch) at step 20: 253.4129, Accuracy: 0.8367
Training loss (for one batch) at step 30: 229.0220, Accuracy: 0.8345
Training loss (for one batch) at step 40: 238.1877, Accuracy: 0.8371
Training loss (for one batch) at step 50: 248.6768, Accuracy: 0.8369
Training loss (for one batch) at step 60: 240.6119, Accuracy: 0.8387
Training loss (for one batch) at step 70: 250.4222, Accuracy: 0.8368
Training loss (for one batch) at step 80: 265.2087, Accuracy: 0.8330
Training loss (for one batch) at step 90: 244.8881, Accuracy: 0.8321
Training loss (for one batch) at step 100: 253.6809, Accuracy: 0.8331
Training loss (for one batch) at step 110: 246.7009, Accuracy: 0.8336
Training loss (for one batch) at step 120: 253.5293, Accuracy: 0.8338
Training loss (for one batch) at step 130: 255.3558, Accuracy: 0.8334
Training loss (for one batch) at step 140: 253.3228, Accuracy: 0.8334
---- Training ----
Training loss: 219.5867
Training acc over epoch: 0.8338
---- Validation ----
Validation loss: 70.0449
Validation acc: 0.7340
Time taken: 65.02s
Start of epoch 27
Training loss (for one batch) at step 0: 235.8341, Accuracy: 0.8300
Training loss (for one batch) at step 10: 244.8620, Accuracy: 0.8236
Training loss (for one batch) at step 20: 264.7985, Accuracy: 0.8319
Training loss (for one batch) at step 30: 233.3059, Accuracy: 0.8313
Training loss (for one batch) at step 40: 234.9283, Accuracy: 0.8295
Training loss (for one batch) at step 50: 222.8051, Accuracy: 0.8353
Training loss (for one batch) at step 60: 239.9344, Accuracy: 0.8343
Training loss (for one batch) at step 70: 258.9313, Accuracy: 0.8351
Training loss (for one batch) at step 80: 244.2281, Accuracy: 0.8338
Training loss (for one batch) at step 90: 250.7219, Accuracy: 0.8342
Training loss (for one batch) at step 100: 226.4685, Accuracy: 0.8349
Training loss (for one batch) at step 110: 244.0710, Accuracy: 0.8350
Training loss (for one batch) at step 120: 236.4039, Accuracy: 0.8360
Training loss (for one batch) at step 130: 256.5294, Accuracy: 0.8342
Training loss (for one batch) at step 140: 269.9176, Accuracy: 0.8336
---- Training ----
Training loss: 216.4017
Training acc over epoch: 0.8338
---- Validation ----
Validation loss: 70.8455
Validation acc: 0.7437
Time taken: 48.04s
Start of epoch 28
Training loss (for one batch) at step 0: 242.4442, Accuracy: 0.8300
Training loss (for one batch) at step 10: 253.9982, Accuracy: 0.8336
Training loss (for one batch) at step 20: 244.0049, Accuracy: 0.8381
Training loss (for one batch) at step 30: 244.6115, Accuracy: 0.8368
Training loss (for one batch) at step 40: 241.3944, Accuracy: 0.8398
Training loss (for one batch) at step 50: 234.6852, Accuracy: 0.8369
Training loss (for one batch) at step 60: 251.5306, Accuracy: 0.8370
Training loss (for one batch) at step 70: 238.6640, Accuracy: 0.8369
Training loss (for one batch) at step 80: 256.7857, Accuracy: 0.8352
Training loss (for one batch) at step 90: 249.2139, Accuracy: 0.8346
Training loss (for one batch) at step 100: 234.3451, Accuracy: 0.8312
Training loss (for one batch) at step 110: 228.6245, Accuracy: 0.8308
Training loss (for one batch) at step 120: 247.9938, Accuracy: 0.8297
Training loss (for one batch) at step 130: 244.8251, Accuracy: 0.8298
Training loss (for one batch) at step 140: 263.2701, Accuracy: 0.8289
---- Training ----
Training loss: 222.3003
Training acc over epoch: 0.8294
---- Validation ----
Validation loss: 70.5740
Validation acc: 0.7423
Time taken: 64.31s
Start of epoch 29
Training loss (for one batch) at step 0: 218.1247, Accuracy: 0.8600
Training loss (for one batch) at step 10: 252.7379, Accuracy: 0.8445
Training loss (for one batch) at step 20: 237.2572, Accuracy: 0.8433
Training loss (for one batch) at step 30: 238.6425, Accuracy: 0.8406
Training loss (for one batch) at step 40: 229.8781, Accuracy: 0.8405
Training loss (for one batch) at step 50: 238.0301, Accuracy: 0.8398
Training loss (for one batch) at step 60: 253.3215, Accuracy: 0.8385
Training loss (for one batch) at step 70: 251.5755, Accuracy: 0.8373
Training loss (for one batch) at step 80: 238.9077, Accuracy: 0.8362
Training loss (for one batch) at step 90: 241.8138, Accuracy: 0.8355
Training loss (for one batch) at step 100: 244.8103, Accuracy: 0.8352
Training loss (for one batch) at step 110: 232.0469, Accuracy: 0.8363
Training loss (for one batch) at step 120: 245.1599, Accuracy: 0.8356
Training loss (for one batch) at step 130: 232.6996, Accuracy: 0.8362
Training loss (for one batch) at step 140: 255.0106, Accuracy: 0.8335
---- Training ----
Training loss: 217.7722
Training acc over epoch: 0.8334
---- Validation ----
Validation loss: 57.9025
Validation acc: 0.7319
Time taken: 48.64s
Start of epoch 30
Training loss (for one batch) at step 0: 235.5676, Accuracy: 0.8500
Training loss (for one batch) at step 10: 227.9011, Accuracy: 0.8382
Training loss (for one batch) at step 20: 240.9100, Accuracy: 0.8424
Training loss (for one batch) at step 30: 230.4607, Accuracy: 0.8368
Training loss (for one batch) at step 40: 213.9492, Accuracy: 0.8395
Training loss (for one batch) at step 50: 223.1538, Accuracy: 0.8427
Training loss (for one batch) at step 60: 240.4632, Accuracy: 0.8448
Training loss (for one batch) at step 70: 247.7459, Accuracy: 0.8415
Training loss (for one batch) at step 80: 239.9848, Accuracy: 0.8396
Training loss (for one batch) at step 90: 237.7835, Accuracy: 0.8376
Training loss (for one batch) at step 100: 239.4716, Accuracy: 0.8370
Training loss (for one batch) at step 110: 229.3162, Accuracy: 0.8372
Training loss (for one batch) at step 120: 232.6068, Accuracy: 0.8359
Training loss (for one batch) at step 130: 239.0383, Accuracy: 0.8355
Training loss (for one batch) at step 140: 245.0959, Accuracy: 0.8335
---- Training ----
Training loss: 210.9928
Training acc over epoch: 0.8336
---- Validation ----
Validation loss: 80.2742
Validation acc: 0.7440
Time taken: 62.50s
Start of epoch 31
Training loss (for one batch) at step 0: 235.1917, Accuracy: 0.8600
Training loss (for one batch) at step 10: 235.4564, Accuracy: 0.8527
Training loss (for one batch) at step 20: 235.1737, Accuracy: 0.8386
Training loss (for one batch) at step 30: 230.6480, Accuracy: 0.8365
Training loss (for one batch) at step 40: 224.7778, Accuracy: 0.8376
Training loss (for one batch) at step 50: 211.9872, Accuracy: 0.8392
Training loss (for one batch) at step 60: 244.7687, Accuracy: 0.8377
Training loss (for one batch) at step 70: 242.8194, Accuracy: 0.8370
Training loss (for one batch) at step 80: 247.0181, Accuracy: 0.8364
Training loss (for one batch) at step 90: 234.0558, Accuracy: 0.8370
Training loss (for one batch) at step 100: 237.1889, Accuracy: 0.8363
Training loss (for one batch) at step 110: 245.5160, Accuracy: 0.8359
Training loss (for one batch) at step 120: 235.1613, Accuracy: 0.8340
Training loss (for one batch) at step 130: 236.0915, Accuracy: 0.8345
Training loss (for one batch) at step 140: 273.2734, Accuracy: 0.8349
---- Training ----
Training loss: 198.8916
Training acc over epoch: 0.8339
---- Validation ----
Validation loss: 69.5140
Validation acc: 0.7311
Time taken: 50.07s
Start of epoch 32
Training loss (for one batch) at step 0: 222.5908, Accuracy: 0.9000
Training loss (for one batch) at step 10: 221.8011, Accuracy: 0.8336
Training loss (for one batch) at step 20: 234.1125, Accuracy: 0.8395
Training loss (for one batch) at step 30: 220.9308, Accuracy: 0.8413
Training loss (for one batch) at step 40: 243.8415, Accuracy: 0.8402
Training loss (for one batch) at step 50: 236.5141, Accuracy: 0.8439
Training loss (for one batch) at step 60: 244.0346, Accuracy: 0.8444
Training loss (for one batch) at step 70: 241.0282, Accuracy: 0.8445
Training loss (for one batch) at step 80: 248.0688, Accuracy: 0.8441
Training loss (for one batch) at step 90: 251.8934, Accuracy: 0.8430
Training loss (for one batch) at step 100: 247.4602, Accuracy: 0.8426
Training loss (for one batch) at step 110: 225.8649, Accuracy: 0.8444
Training loss (for one batch) at step 120: 225.9892, Accuracy: 0.8440
Training loss (for one batch) at step 130: 255.0406, Accuracy: 0.8425
Training loss (for one batch) at step 140: 229.9544, Accuracy: 0.8427
---- Training ----
Training loss: 215.9789
Training acc over epoch: 0.8416
---- Validation ----
Validation loss: 80.6376
Validation acc: 0.7225
Time taken: 62.26s
Start of epoch 33
Training loss (for one batch) at step 0: 234.1671, Accuracy: 0.8300
Training loss (for one batch) at step 10: 226.8352, Accuracy: 0.8364
Training loss (for one batch) at step 20: 228.6739, Accuracy: 0.8390
Training loss (for one batch) at step 30: 234.8060, Accuracy: 0.8387
Training loss (for one batch) at step 40: 228.8894, Accuracy: 0.8439
Training loss (for one batch) at step 50: 214.6906, Accuracy: 0.8439
Training loss (for one batch) at step 60: 218.4252, Accuracy: 0.8457
Training loss (for one batch) at step 70: 232.5582, Accuracy: 0.8434
Training loss (for one batch) at step 80: 227.5370, Accuracy: 0.8425
Training loss (for one batch) at step 90: 256.1789, Accuracy: 0.8404
Training loss (for one batch) at step 100: 230.7294, Accuracy: 0.8390
Training loss (for one batch) at step 110: 207.5177, Accuracy: 0.8392
Training loss (for one batch) at step 120: 227.6081, Accuracy: 0.8392
Training loss (for one batch) at step 130: 237.9324, Accuracy: 0.8392
Training loss (for one batch) at step 140: 238.7729, Accuracy: 0.8398
---- Training ----
Training loss: 210.8249
Training acc over epoch: 0.8390
---- Validation ----
Validation loss: 68.1152
Validation acc: 0.7219
Time taken: 50.16s
Start of epoch 34
Training loss (for one batch) at step 0: 248.0710, Accuracy: 0.8400
Training loss (for one batch) at step 10: 228.9749, Accuracy: 0.8391
Training loss (for one batch) at step 20: 234.5841, Accuracy: 0.8533
Training loss (for one batch) at step 30: 220.7189, Accuracy: 0.8468
Training loss (for one batch) at step 40: 214.9964, Accuracy: 0.8434
Training loss (for one batch) at step 50: 232.7071, Accuracy: 0.8443
Training loss (for one batch) at step 60: 243.5460, Accuracy: 0.8426
Training loss (for one batch) at step 70: 225.7935, Accuracy: 0.8438
Training loss (for one batch) at step 80: 239.7450, Accuracy: 0.8438
Training loss (for one batch) at step 90: 235.9796, Accuracy: 0.8430
Training loss (for one batch) at step 100: 234.4512, Accuracy: 0.8425
Training loss (for one batch) at step 110: 240.6454, Accuracy: 0.8421
Training loss (for one batch) at step 120: 231.1975, Accuracy: 0.8403
Training loss (for one batch) at step 130: 230.1047, Accuracy: 0.8402
Training loss (for one batch) at step 140: 236.8565, Accuracy: 0.8397
---- Training ----
Training loss: 207.1141
Training acc over epoch: 0.8394
---- Validation ----
Validation loss: 71.6887
Validation acc: 0.7356
Time taken: 61.75s
Start of epoch 35
Training loss (for one batch) at step 0: 238.8915, Accuracy: 0.8100
Training loss (for one batch) at step 10: 221.9431, Accuracy: 0.8318
Training loss (for one batch) at step 20: 238.1560, Accuracy: 0.8452
Training loss (for one batch) at step 30: 230.5274, Accuracy: 0.8500
Training loss (for one batch) at step 40: 215.5167, Accuracy: 0.8478
Training loss (for one batch) at step 50: 202.4013, Accuracy: 0.8488
Training loss (for one batch) at step 60: 252.6227, Accuracy: 0.8479
Training loss (for one batch) at step 70: 221.8537, Accuracy: 0.8496
Training loss (for one batch) at step 80: 211.0092, Accuracy: 0.8465
Training loss (for one batch) at step 90: 213.9940, Accuracy: 0.8451
Training loss (for one batch) at step 100: 231.1083, Accuracy: 0.8426
Training loss (for one batch) at step 110: 207.6015, Accuracy: 0.8431
Training loss (for one batch) at step 120: 214.2682, Accuracy: 0.8440
Training loss (for one batch) at step 130: 240.0047, Accuracy: 0.8420
Training loss (for one batch) at step 140: 207.8425, Accuracy: 0.8423
---- Training ----
Training loss: 217.1189
Training acc over epoch: 0.8419
---- Validation ----
Validation loss: 76.8453
Validation acc: 0.7397
Time taken: 49.20s
Start of epoch 36
Training loss (for one batch) at step 0: 229.4817, Accuracy: 0.8400
Training loss (for one batch) at step 10: 225.8322, Accuracy: 0.8464
Training loss (for one batch) at step 20: 193.1220, Accuracy: 0.8471
Training loss (for one batch) at step 30: 229.3266, Accuracy: 0.8448
Training loss (for one batch) at step 40: 229.7551, Accuracy: 0.8493
Training loss (for one batch) at step 50: 211.7243, Accuracy: 0.8504
Training loss (for one batch) at step 60: 215.5340, Accuracy: 0.8479
Training loss (for one batch) at step 70: 230.1568, Accuracy: 0.8476
Training loss (for one batch) at step 80: 231.3604, Accuracy: 0.8456
Training loss (for one batch) at step 90: 228.7775, Accuracy: 0.8449
Training loss (for one batch) at step 100: 203.7249, Accuracy: 0.8431
Training loss (for one batch) at step 110: 247.6174, Accuracy: 0.8425
Training loss (for one batch) at step 120: 209.9360, Accuracy: 0.8428
Training loss (for one batch) at step 130: 235.9964, Accuracy: 0.8410
Training loss (for one batch) at step 140: 202.7510, Accuracy: 0.8409
---- Training ----
Training loss: 193.7022
Training acc over epoch: 0.8411
---- Validation ----
Validation loss: 65.9168
Validation acc: 0.7329
Time taken: 60.29s
Start of epoch 37
Training loss (for one batch) at step 0: 228.4591, Accuracy: 0.8700
Training loss (for one batch) at step 10: 216.3184, Accuracy: 0.8473
Training loss (for one batch) at step 20: 244.3288, Accuracy: 0.8481
Training loss (for one batch) at step 30: 226.4332, Accuracy: 0.8484
Training loss (for one batch) at step 40: 230.0277, Accuracy: 0.8529
Training loss (for one batch) at step 50: 216.0572, Accuracy: 0.8563
Training loss (for one batch) at step 60: 211.5734, Accuracy: 0.8570
Training loss (for one batch) at step 70: 227.8326, Accuracy: 0.8551
Training loss (for one batch) at step 80: 235.5448, Accuracy: 0.8535
Training loss (for one batch) at step 90: 208.9021, Accuracy: 0.8516
Training loss (for one batch) at step 100: 224.6111, Accuracy: 0.8485
Training loss (for one batch) at step 110: 234.5329, Accuracy: 0.8482
Training loss (for one batch) at step 120: 242.4199, Accuracy: 0.8472
Training loss (for one batch) at step 130: 238.3328, Accuracy: 0.8475
Training loss (for one batch) at step 140: 208.4721, Accuracy: 0.8475
---- Training ----
Training loss: 209.9401
Training acc over epoch: 0.8467
---- Validation ----
Validation loss: 59.2811
Validation acc: 0.7389
Time taken: 50.07s
Start of epoch 38
Training loss (for one batch) at step 0: 200.2514, Accuracy: 0.8600
Training loss (for one batch) at step 10: 218.9404, Accuracy: 0.8500
Training loss (for one batch) at step 20: 205.0095, Accuracy: 0.8486
Training loss (for one batch) at step 30: 213.7009, Accuracy: 0.8448
Training loss (for one batch) at step 40: 210.0462, Accuracy: 0.8451
Training loss (for one batch) at step 50: 212.5746, Accuracy: 0.8476
Training loss (for one batch) at step 60: 202.3568, Accuracy: 0.8472
Training loss (for one batch) at step 70: 234.1981, Accuracy: 0.8462
Training loss (for one batch) at step 80: 224.5644, Accuracy: 0.8463
Training loss (for one batch) at step 90: 209.1035, Accuracy: 0.8441
Training loss (for one batch) at step 100: 230.6502, Accuracy: 0.8421
Training loss (for one batch) at step 110: 227.3220, Accuracy: 0.8431
Training loss (for one batch) at step 120: 251.9786, Accuracy: 0.8437
Training loss (for one batch) at step 130: 236.6940, Accuracy: 0.8432
Training loss (for one batch) at step 140: 233.2983, Accuracy: 0.8440
---- Training ----
Training loss: 201.5356
Training acc over epoch: 0.8439
---- Validation ----
Validation loss: 88.3494
Validation acc: 0.7273
Time taken: 60.99s
Start of epoch 39
Training loss (for one batch) at step 0: 225.2239, Accuracy: 0.8400
Training loss (for one batch) at step 10: 206.5844, Accuracy: 0.8482
Training loss (for one batch) at step 20: 264.9331, Accuracy: 0.8448
Training loss (for one batch) at step 30: 216.8722, Accuracy: 0.8468
Training loss (for one batch) at step 40: 226.6909, Accuracy: 0.8432
Training loss (for one batch) at step 50: 206.7373, Accuracy: 0.8459
Training loss (for one batch) at step 60: 203.7588, Accuracy: 0.8485
Training loss (for one batch) at step 70: 229.0410, Accuracy: 0.8485
Training loss (for one batch) at step 80: 218.1293, Accuracy: 0.8465
Training loss (for one batch) at step 90: 256.6113, Accuracy: 0.8465
Training loss (for one batch) at step 100: 219.9882, Accuracy: 0.8473
Training loss (for one batch) at step 110: 211.8772, Accuracy: 0.8481
Training loss (for one batch) at step 120: 220.2987, Accuracy: 0.8474
Training loss (for one batch) at step 130: 247.4305, Accuracy: 0.8464
Training loss (for one batch) at step 140: 221.8912, Accuracy: 0.8468
---- Training ----
Training loss: 207.9988
Training acc over epoch: 0.8456
---- Validation ----
Validation loss: 83.2030
Validation acc: 0.7233
Time taken: 52.66s
Start of epoch 40
Training loss (for one batch) at step 0: 250.2291, Accuracy: 0.7900
Training loss (for one batch) at step 10: 197.3427, Accuracy: 0.8236
Training loss (for one batch) at step 20: 210.8969, Accuracy: 0.8305
Training loss (for one batch) at step 30: 233.7159, Accuracy: 0.8329
Training loss (for one batch) at step 40: 194.5068, Accuracy: 0.8383
Training loss (for one batch) at step 50: 208.8749, Accuracy: 0.8422
Training loss (for one batch) at step 60: 221.0687, Accuracy: 0.8428
Training loss (for one batch) at step 70: 215.2544, Accuracy: 0.8445
Training loss (for one batch) at step 80: 220.7688, Accuracy: 0.8442
Training loss (for one batch) at step 90: 235.6701, Accuracy: 0.8425
Training loss (for one batch) at step 100: 213.8069, Accuracy: 0.8420
Training loss (for one batch) at step 110: 210.0951, Accuracy: 0.8427
Training loss (for one batch) at step 120: 236.1404, Accuracy: 0.8439
Training loss (for one batch) at step 130: 217.4057, Accuracy: 0.8448
Training loss (for one batch) at step 140: 235.6296, Accuracy: 0.8440
---- Training ----
Training loss: 183.8158
Training acc over epoch: 0.8449
---- Validation ----
Validation loss: 67.1604
Validation acc: 0.7319
Time taken: 62.92s
Start of epoch 41
Training loss (for one batch) at step 0: 201.6018, Accuracy: 0.9100
Training loss (for one batch) at step 10: 218.7362, Accuracy: 0.8564
Training loss (for one batch) at step 20: 226.8827, Accuracy: 0.8505
Training loss (for one batch) at step 30: 199.4951, Accuracy: 0.8435
Training loss (for one batch) at step 40: 203.8652, Accuracy: 0.8454
Training loss (for one batch) at step 50: 215.1268, Accuracy: 0.8441
Training loss (for one batch) at step 60: 207.8360, Accuracy: 0.8470
Training loss (for one batch) at step 70: 223.5363, Accuracy: 0.8492
Training loss (for one batch) at step 80: 214.4162, Accuracy: 0.8483
Training loss (for one batch) at step 90: 244.9549, Accuracy: 0.8468
Training loss (for one batch) at step 100: 207.7923, Accuracy: 0.8460
Training loss (for one batch) at step 110: 209.3508, Accuracy: 0.8486
Training loss (for one batch) at step 120: 237.1217, Accuracy: 0.8474
Training loss (for one batch) at step 130: 223.8634, Accuracy: 0.8477
Training loss (for one batch) at step 140: 210.6277, Accuracy: 0.8474
---- Training ----
Training loss: 216.9761
Training acc over epoch: 0.8476
---- Validation ----
Validation loss: 85.7685
Validation acc: 0.7265
Time taken: 53.59s
Start of epoch 42
Training loss (for one batch) at step 0: 213.3884, Accuracy: 0.8300
Training loss (for one batch) at step 10: 209.0462, Accuracy: 0.8482
Training loss (for one batch) at step 20: 217.0547, Accuracy: 0.8476
Training loss (for one batch) at step 30: 214.4684, Accuracy: 0.8410
Training loss (for one batch) at step 40: 196.3888, Accuracy: 0.8415
Training loss (for one batch) at step 50: 215.5532, Accuracy: 0.8461
Training loss (for one batch) at step 60: 236.9741, Accuracy: 0.8475
Training loss (for one batch) at step 70: 199.7617, Accuracy: 0.8489
Training loss (for one batch) at step 80: 210.1101, Accuracy: 0.8480
Training loss (for one batch) at step 90: 210.8024, Accuracy: 0.8496
Training loss (for one batch) at step 100: 227.2229, Accuracy: 0.8472
Training loss (for one batch) at step 110: 207.6299, Accuracy: 0.8473
Training loss (for one batch) at step 120: 226.2038, Accuracy: 0.8469
Training loss (for one batch) at step 130: 219.2592, Accuracy: 0.8476
Training loss (for one batch) at step 140: 216.2121, Accuracy: 0.8470
---- Training ----
Training loss: 190.0255
Training acc over epoch: 0.8468
---- Validation ----
Validation loss: 69.8368
Validation acc: 0.7249
Time taken: 61.01s
Start of epoch 43
Training loss (for one batch) at step 0: 217.6565, Accuracy: 0.8100
Training loss (for one batch) at step 10: 216.9246, Accuracy: 0.8409
Training loss (for one batch) at step 20: 212.3205, Accuracy: 0.8443
Training loss (for one batch) at step 30: 200.5739, Accuracy: 0.8516
Training loss (for one batch) at step 40: 216.7517, Accuracy: 0.8507
Training loss (for one batch) at step 50: 222.8231, Accuracy: 0.8553
Training loss (for one batch) at step 60: 211.8153, Accuracy: 0.8544
Training loss (for one batch) at step 70: 223.2672, Accuracy: 0.8515
Training loss (for one batch) at step 80: 198.6165, Accuracy: 0.8489
Training loss (for one batch) at step 90: 223.4437, Accuracy: 0.8477
Training loss (for one batch) at step 100: 208.6168, Accuracy: 0.8466
Training loss (for one batch) at step 110: 223.4621, Accuracy: 0.8481
Training loss (for one batch) at step 120: 210.4071, Accuracy: 0.8472
Training loss (for one batch) at step 130: 216.0071, Accuracy: 0.8488
Training loss (for one batch) at step 140: 210.7066, Accuracy: 0.8469
---- Training ----
Training loss: 175.4065
Training acc over epoch: 0.8467
---- Validation ----
Validation loss: 106.1096
Validation acc: 0.7305
Time taken: 53.30s
Start of epoch 44
Training loss (for one batch) at step 0: 225.2161, Accuracy: 0.8000
Training loss (for one batch) at step 10: 233.6169, Accuracy: 0.8591
Training loss (for one batch) at step 20: 215.2918, Accuracy: 0.8619
Training loss (for one batch) at step 30: 216.5158, Accuracy: 0.8519
Training loss (for one batch) at step 40: 213.1790, Accuracy: 0.8549
Training loss (for one batch) at step 50: 207.2064, Accuracy: 0.8539
Training loss (for one batch) at step 60: 226.5098, Accuracy: 0.8521
Training loss (for one batch) at step 70: 231.9284, Accuracy: 0.8514
Training loss (for one batch) at step 80: 206.0839, Accuracy: 0.8517
Training loss (for one batch) at step 90: 205.9093, Accuracy: 0.8505
Training loss (for one batch) at step 100: 196.6962, Accuracy: 0.8496
Training loss (for one batch) at step 110: 216.0640, Accuracy: 0.8486
Training loss (for one batch) at step 120: 206.7236, Accuracy: 0.8489
Training loss (for one batch) at step 130: 204.4394, Accuracy: 0.8488
Training loss (for one batch) at step 140: 208.1075, Accuracy: 0.8490
---- Training ----
Training loss: 177.7242
Training acc over epoch: 0.8483
---- Validation ----
Validation loss: 73.8137
Validation acc: 0.7270
Time taken: 61.61s
Start of epoch 45
Training loss (for one batch) at step 0: 192.2081, Accuracy: 0.9200
Training loss (for one batch) at step 10: 207.4653, Accuracy: 0.8464
Training loss (for one batch) at step 20: 196.1846, Accuracy: 0.8510
Training loss (for one batch) at step 30: 182.6079, Accuracy: 0.8474
Training loss (for one batch) at step 40: 203.0189, Accuracy: 0.8454
Training loss (for one batch) at step 50: 193.7542, Accuracy: 0.8492
Training loss (for one batch) at step 60: 222.8609, Accuracy: 0.8479
Training loss (for one batch) at step 70: 213.1559, Accuracy: 0.8485
Training loss (for one batch) at step 80: 195.5713, Accuracy: 0.8479
Training loss (for one batch) at step 90: 199.6198, Accuracy: 0.8473
Training loss (for one batch) at step 100: 217.9117, Accuracy: 0.8456
Training loss (for one batch) at step 110: 205.0130, Accuracy: 0.8467
Training loss (for one batch) at step 120: 200.6017, Accuracy: 0.8451
Training loss (for one batch) at step 130: 202.5632, Accuracy: 0.8444
Training loss (for one batch) at step 140: 197.5107, Accuracy: 0.8432
---- Training ----
Training loss: 164.1999
Training acc over epoch: 0.8435
---- Validation ----
Validation loss: 68.9356
Validation acc: 0.7268
Time taken: 54.68s
Start of epoch 46
Training loss (for one batch) at step 0: 197.0468, Accuracy: 0.8300
Training loss (for one batch) at step 10: 213.0845, Accuracy: 0.8591
Training loss (for one batch) at step 20: 203.0128, Accuracy: 0.8586
Training loss (for one batch) at step 30: 226.7425, Accuracy: 0.8561
Training loss (for one batch) at step 40: 182.7890, Accuracy: 0.8571
Training loss (for one batch) at step 50: 211.7321, Accuracy: 0.8559
Training loss (for one batch) at step 60: 200.3920, Accuracy: 0.8577
Training loss (for one batch) at step 70: 202.4686, Accuracy: 0.8561
Training loss (for one batch) at step 80: 243.6484, Accuracy: 0.8540
Training loss (for one batch) at step 90: 204.9140, Accuracy: 0.8511
Training loss (for one batch) at step 100: 204.2000, Accuracy: 0.8500
Training loss (for one batch) at step 110: 228.0543, Accuracy: 0.8499
Training loss (for one batch) at step 120: 206.3541, Accuracy: 0.8498
Training loss (for one batch) at step 130: 215.6771, Accuracy: 0.8501
Training loss (for one batch) at step 140: 200.8600, Accuracy: 0.8494
---- Training ----
Training loss: 180.5217
Training acc over epoch: 0.8492
---- Validation ----
Validation loss: 76.8593
Validation acc: 0.7289
Time taken: 60.85s
Start of epoch 47
Training loss (for one batch) at step 0: 205.2282, Accuracy: 0.8800
Training loss (for one batch) at step 10: 225.3042, Accuracy: 0.8436
Training loss (for one batch) at step 20: 194.5104, Accuracy: 0.8476
Training loss (for one batch) at step 30: 207.6927, Accuracy: 0.8484
Training loss (for one batch) at step 40: 195.5227, Accuracy: 0.8478
Training loss (for one batch) at step 50: 205.1373, Accuracy: 0.8510
Training loss (for one batch) at step 60: 182.7279, Accuracy: 0.8539
Training loss (for one batch) at step 70: 210.3254, Accuracy: 0.8510
Training loss (for one batch) at step 80: 199.3923, Accuracy: 0.8510
Training loss (for one batch) at step 90: 192.8136, Accuracy: 0.8529
Training loss (for one batch) at step 100: 193.6104, Accuracy: 0.8520
Training loss (for one batch) at step 110: 215.2464, Accuracy: 0.8512
Training loss (for one batch) at step 120: 204.0931, Accuracy: 0.8521
Training loss (for one batch) at step 130: 198.4651, Accuracy: 0.8525
Training loss (for one batch) at step 140: 195.2698, Accuracy: 0.8513
---- Training ----
Training loss: 201.9470
Training acc over epoch: 0.8510
---- Validation ----
Validation loss: 94.6841
Validation acc: 0.7265
Time taken: 54.15s
Start of epoch 48
Training loss (for one batch) at step 0: 210.6264, Accuracy: 0.8900
Training loss (for one batch) at step 10: 203.2037, Accuracy: 0.8527
Training loss (for one batch) at step 20: 200.8780, Accuracy: 0.8576
Training loss (for one batch) at step 30: 219.5609, Accuracy: 0.8506
Training loss (for one batch) at step 40: 196.9256, Accuracy: 0.8507
Training loss (for one batch) at step 50: 201.9439, Accuracy: 0.8510
Training loss (for one batch) at step 60: 216.0611, Accuracy: 0.8492
Training loss (for one batch) at step 70: 214.3053, Accuracy: 0.8499
Training loss (for one batch) at step 80: 186.2735, Accuracy: 0.8488
Training loss (for one batch) at step 90: 259.5701, Accuracy: 0.8467
Training loss (for one batch) at step 100: 190.9406, Accuracy: 0.8457
Training loss (for one batch) at step 110: 191.0650, Accuracy: 0.8461
Training loss (for one batch) at step 120: 235.3501, Accuracy: 0.8470
Training loss (for one batch) at step 130: 202.3817, Accuracy: 0.8480
Training loss (for one batch) at step 140: 235.8335, Accuracy: 0.8479
---- Training ----
Training loss: 164.7656
Training acc over epoch: 0.8482
---- Validation ----
Validation loss: 87.9655
Validation acc: 0.7311
Time taken: 58.62s
Start of epoch 49
Training loss (for one batch) at step 0: 204.5627, Accuracy: 0.7800
Training loss (for one batch) at step 10: 221.2933, Accuracy: 0.8436
Training loss (for one batch) at step 20: 188.6167, Accuracy: 0.8571
Training loss (for one batch) at step 30: 200.0119, Accuracy: 0.8519
Training loss (for one batch) at step 40: 191.6080, Accuracy: 0.8507
Training loss (for one batch) at step 50: 193.1429, Accuracy: 0.8533
Training loss (for one batch) at step 60: 215.9621, Accuracy: 0.8551
Training loss (for one batch) at step 70: 210.9236, Accuracy: 0.8548
Training loss (for one batch) at step 80: 216.9274, Accuracy: 0.8519
Training loss (for one batch) at step 90: 226.3201, Accuracy: 0.8513
Training loss (for one batch) at step 100: 202.2559, Accuracy: 0.8508
Training loss (for one batch) at step 110: 207.3746, Accuracy: 0.8507
Training loss (for one batch) at step 120: 194.7182, Accuracy: 0.8505
Training loss (for one batch) at step 130: 200.0363, Accuracy: 0.8513
Training loss (for one batch) at step 140: 201.0230, Accuracy: 0.8497
---- Training ----
Training loss: 189.4398
Training acc over epoch: 0.8508
---- Validation ----
Validation loss: 65.8887
Validation acc: 0.7311
Time taken: 57.65s
===== Q: 0.0001
Validation acc: 0.7425
Validation AUC: 0.7401
Validation Balanced_ACC: 0.4830
Validation MI: 0.1387
Validation Normalized MI: 0.2073
Validation Adjusted MI: 0.2073
Validation aUc_Sklearn: 0.8270
Start of epoch 0
2023-02-14 23:10:29.300861: I tensorflow/core/kernels/data/shuffle_dataset_op.cc:175] Filling up shuffle buffer (this may take a while): 1 of 1024
2023-02-14 23:10:32.539571: I tensorflow/core/kernels/data/shuffle_dataset_op.cc:228] Shuffle buffer filled.
Training loss (for one batch) at step 0: 436.6236, Accuracy: 0.5900
Training loss (for one batch) at step 10: 435.5614, Accuracy: 0.5491
Training loss (for one batch) at step 20: 410.9731, Accuracy: 0.5462
Training loss (for one batch) at step 30: 471.3681, Accuracy: 0.5542
Training loss (for one batch) at step 40: 433.4162, Accuracy: 0.5566
Training loss (for one batch) at step 50: 428.1716, Accuracy: 0.5569
Training loss (for one batch) at step 60: 433.5032, Accuracy: 0.5631
Training loss (for one batch) at step 70: 429.7946, Accuracy: 0.5599
Training loss (for one batch) at step 80: 430.2086, Accuracy: 0.5647
Training loss (for one batch) at step 90: 456.7353, Accuracy: 0.5653
Training loss (for one batch) at step 100: 419.2536, Accuracy: 0.5691
Training loss (for one batch) at step 110: 395.0050, Accuracy: 0.5692
Training loss (for one batch) at step 120: 414.5247, Accuracy: 0.5704
Training loss (for one batch) at step 130: 382.7275, Accuracy: 0.5730
Training loss (for one batch) at step 140: 388.0818, Accuracy: 0.5755
---- Training ----
Training loss: 394.8313
Training acc over epoch: 0.5761
---- Validation ----
Validation loss: 87.6995
Validation acc: 0.5134
Time taken: 68.22s
Start of epoch 1
Training loss (for one batch) at step 0: 396.7151, Accuracy: 0.5700
Training loss (for one batch) at step 10: 374.0270, Accuracy: 0.6045
Training loss (for one batch) at step 20: 392.8275, Accuracy: 0.6124
Training loss (for one batch) at step 30: 391.8384, Accuracy: 0.6077
Training loss (for one batch) at step 40: 415.2708, Accuracy: 0.6054
Training loss (for one batch) at step 50: 399.8800, Accuracy: 0.6075
Training loss (for one batch) at step 60: 363.1733, Accuracy: 0.6087
Training loss (for one batch) at step 70: 391.1283, Accuracy: 0.6106
Training loss (for one batch) at step 80: 401.3207, Accuracy: 0.6131
Training loss (for one batch) at step 90: 393.3836, Accuracy: 0.6168
Training loss (for one batch) at step 100: 388.7330, Accuracy: 0.6165
Training loss (for one batch) at step 110: 378.7025, Accuracy: 0.6186
Training loss (for one batch) at step 120: 394.8909, Accuracy: 0.6190
Training loss (for one batch) at step 130: 392.1635, Accuracy: 0.6191
Training loss (for one batch) at step 140: 382.8720, Accuracy: 0.6195
---- Training ----
Training loss: 355.5984
Training acc over epoch: 0.6204
---- Validation ----
Validation loss: 73.7167
Validation acc: 0.5228
Time taken: 55.17s
Start of epoch 2
Training loss (for one batch) at step 0: 373.6795, Accuracy: 0.6200
Training loss (for one batch) at step 10: 349.8850, Accuracy: 0.6491
Training loss (for one batch) at step 20: 401.0105, Accuracy: 0.6429
Training loss (for one batch) at step 30: 401.1235, Accuracy: 0.6432
Training loss (for one batch) at step 40: 349.6929, Accuracy: 0.6488
Training loss (for one batch) at step 50: 365.1029, Accuracy: 0.6455
Training loss (for one batch) at step 60: 361.0728, Accuracy: 0.6472
Training loss (for one batch) at step 70: 377.0344, Accuracy: 0.6513
Training loss (for one batch) at step 80: 362.3386, Accuracy: 0.6511
Training loss (for one batch) at step 90: 372.0692, Accuracy: 0.6488
Training loss (for one batch) at step 100: 357.8613, Accuracy: 0.6489
Training loss (for one batch) at step 110: 364.9118, Accuracy: 0.6495
Training loss (for one batch) at step 120: 392.6961, Accuracy: 0.6491
Training loss (for one batch) at step 130: 370.6182, Accuracy: 0.6492
Training loss (for one batch) at step 140: 366.3666, Accuracy: 0.6502
---- Training ----
Training loss: 333.8393
Training acc over epoch: 0.6494
---- Validation ----
Validation loss: 76.3876
Validation acc: 0.6531
Time taken: 69.05s
Start of epoch 3
Training loss (for one batch) at step 0: 336.2198, Accuracy: 0.7600
Training loss (for one batch) at step 10: 365.8463, Accuracy: 0.6836
Training loss (for one batch) at step 20: 341.6383, Accuracy: 0.6771
Training loss (for one batch) at step 30: 350.0477, Accuracy: 0.6616
Training loss (for one batch) at step 40: 380.0253, Accuracy: 0.6600
Training loss (for one batch) at step 50: 359.6744, Accuracy: 0.6627
Training loss (for one batch) at step 60: 342.7433, Accuracy: 0.6656
Training loss (for one batch) at step 70: 334.3157, Accuracy: 0.6672
Training loss (for one batch) at step 80: 344.7623, Accuracy: 0.6712
Training loss (for one batch) at step 90: 360.6187, Accuracy: 0.6725
Training loss (for one batch) at step 100: 347.4454, Accuracy: 0.6698
Training loss (for one batch) at step 110: 358.5741, Accuracy: 0.6716
Training loss (for one batch) at step 120: 357.0551, Accuracy: 0.6699
Training loss (for one batch) at step 130: 350.0647, Accuracy: 0.6708
Training loss (for one batch) at step 140: 338.0910, Accuracy: 0.6701
---- Training ----
Training loss: 308.0896
Training acc over epoch: 0.6698
---- Validation ----
Validation loss: 69.7564
Validation acc: 0.7120
Time taken: 46.18s
Start of epoch 4
Training loss (for one batch) at step 0: 328.9666, Accuracy: 0.7100
Training loss (for one batch) at step 10: 339.1441, Accuracy: 0.6745
Training loss (for one batch) at step 20: 337.8909, Accuracy: 0.6714
Training loss (for one batch) at step 30: 352.2290, Accuracy: 0.6774
Training loss (for one batch) at step 40: 364.9357, Accuracy: 0.6863
Training loss (for one batch) at step 50: 339.1195, Accuracy: 0.6908
Training loss (for one batch) at step 60: 340.5474, Accuracy: 0.6887
Training loss (for one batch) at step 70: 347.6205, Accuracy: 0.6859
Training loss (for one batch) at step 80: 354.3122, Accuracy: 0.6844
Training loss (for one batch) at step 90: 345.6548, Accuracy: 0.6857
Training loss (for one batch) at step 100: 349.2248, Accuracy: 0.6868
Training loss (for one batch) at step 110: 332.4956, Accuracy: 0.6885
Training loss (for one batch) at step 120: 339.8990, Accuracy: 0.6877
Training loss (for one batch) at step 130: 334.7568, Accuracy: 0.6888
Training loss (for one batch) at step 140: 312.8282, Accuracy: 0.6893
---- Training ----
Training loss: 308.9268
Training acc over epoch: 0.6896
---- Validation ----
Validation loss: 65.5789
Validation acc: 0.7243
Time taken: 69.48s
Start of epoch 5
Training loss (for one batch) at step 0: 351.3552, Accuracy: 0.7400
Training loss (for one batch) at step 10: 332.9081, Accuracy: 0.7027
Training loss (for one batch) at step 20: 335.3109, Accuracy: 0.6990
Training loss (for one batch) at step 30: 334.2821, Accuracy: 0.7052
Training loss (for one batch) at step 40: 316.3886, Accuracy: 0.7066
Training loss (for one batch) at step 50: 349.9482, Accuracy: 0.7043
Training loss (for one batch) at step 60: 348.0242, Accuracy: 0.6980
Training loss (for one batch) at step 70: 337.8733, Accuracy: 0.6987
Training loss (for one batch) at step 80: 347.6678, Accuracy: 0.7000
Training loss (for one batch) at step 90: 338.2899, Accuracy: 0.6981
Training loss (for one batch) at step 100: 352.0109, Accuracy: 0.6972
Training loss (for one batch) at step 110: 336.4189, Accuracy: 0.6977
Training loss (for one batch) at step 120: 329.9515, Accuracy: 0.7005
Training loss (for one batch) at step 130: 310.9571, Accuracy: 0.7009
Training loss (for one batch) at step 140: 320.9157, Accuracy: 0.7012
---- Training ----
Training loss: 293.2957
Training acc over epoch: 0.7012
---- Validation ----
Validation loss: 73.1645
Validation acc: 0.7297
Time taken: 55.36s
Start of epoch 6
Training loss (for one batch) at step 0: 315.5915, Accuracy: 0.7400
Training loss (for one batch) at step 10: 350.5636, Accuracy: 0.7182
Training loss (for one batch) at step 20: 317.6315, Accuracy: 0.7171
Training loss (for one batch) at step 30: 319.8180, Accuracy: 0.7194
Training loss (for one batch) at step 40: 350.3785, Accuracy: 0.7173
Training loss (for one batch) at step 50: 318.8316, Accuracy: 0.7182
Training loss (for one batch) at step 60: 331.2031, Accuracy: 0.7144
Training loss (for one batch) at step 70: 311.3257, Accuracy: 0.7158
Training loss (for one batch) at step 80: 350.2624, Accuracy: 0.7147
Training loss (for one batch) at step 90: 308.1411, Accuracy: 0.7133
Training loss (for one batch) at step 100: 338.3254, Accuracy: 0.7120
Training loss (for one batch) at step 110: 315.9450, Accuracy: 0.7123
Training loss (for one batch) at step 120: 332.0000, Accuracy: 0.7112
Training loss (for one batch) at step 130: 302.6914, Accuracy: 0.7145
Training loss (for one batch) at step 140: 302.8059, Accuracy: 0.7140
---- Training ----
Training loss: 296.1591
Training acc over epoch: 0.7139
---- Validation ----
Validation loss: 70.5558
Validation acc: 0.7343
Time taken: 67.16s
Start of epoch 7
Training loss (for one batch) at step 0: 303.1763, Accuracy: 0.8000
Training loss (for one batch) at step 10: 339.5162, Accuracy: 0.7300
Training loss (for one batch) at step 20: 306.1322, Accuracy: 0.7329
Training loss (for one batch) at step 30: 325.7250, Accuracy: 0.7274
Training loss (for one batch) at step 40: 319.2345, Accuracy: 0.7361
Training loss (for one batch) at step 50: 322.6181, Accuracy: 0.7335
Training loss (for one batch) at step 60: 324.2099, Accuracy: 0.7346
Training loss (for one batch) at step 70: 300.6813, Accuracy: 0.7330
Training loss (for one batch) at step 80: 307.7044, Accuracy: 0.7340
Training loss (for one batch) at step 90: 312.5745, Accuracy: 0.7344
Training loss (for one batch) at step 100: 321.3698, Accuracy: 0.7321
Training loss (for one batch) at step 110: 299.6636, Accuracy: 0.7342
Training loss (for one batch) at step 120: 309.2267, Accuracy: 0.7342
Training loss (for one batch) at step 130: 298.1317, Accuracy: 0.7344
Training loss (for one batch) at step 140: 312.5305, Accuracy: 0.7351
---- Training ----
Training loss: 281.9247
Training acc over epoch: 0.7354
---- Validation ----
Validation loss: 73.4071
Validation acc: 0.7198
Time taken: 49.16s
Start of epoch 8
Training loss (for one batch) at step 0: 332.0253, Accuracy: 0.7500
Training loss (for one batch) at step 10: 312.9711, Accuracy: 0.7564
Training loss (for one batch) at step 20: 306.9419, Accuracy: 0.7533
Training loss (for one batch) at step 30: 308.1533, Accuracy: 0.7484
Training loss (for one batch) at step 40: 292.7788, Accuracy: 0.7495
Training loss (for one batch) at step 50: 301.1482, Accuracy: 0.7531
Training loss (for one batch) at step 60: 334.4452, Accuracy: 0.7503
Training loss (for one batch) at step 70: 297.8073, Accuracy: 0.7493
Training loss (for one batch) at step 80: 335.2913, Accuracy: 0.7490
Training loss (for one batch) at step 90: 295.9787, Accuracy: 0.7468
Training loss (for one batch) at step 100: 311.6848, Accuracy: 0.7457
Training loss (for one batch) at step 110: 308.6985, Accuracy: 0.7447
Training loss (for one batch) at step 120: 331.5086, Accuracy: 0.7436
Training loss (for one batch) at step 130: 306.9093, Accuracy: 0.7431
Training loss (for one batch) at step 140: 307.9696, Accuracy: 0.7425
---- Training ----
Training loss: 267.6394
Training acc over epoch: 0.7438
---- Validation ----
Validation loss: 69.5071
Validation acc: 0.7397
Time taken: 67.83s
Start of epoch 9
Training loss (for one batch) at step 0: 307.4138, Accuracy: 0.7200
Training loss (for one batch) at step 10: 305.3198, Accuracy: 0.7618
Training loss (for one batch) at step 20: 324.4145, Accuracy: 0.7610
Training loss (for one batch) at step 30: 290.7440, Accuracy: 0.7671
Training loss (for one batch) at step 40: 315.3165, Accuracy: 0.7676
Training loss (for one batch) at step 50: 308.7544, Accuracy: 0.7714
Training loss (for one batch) at step 60: 307.5600, Accuracy: 0.7680
Training loss (for one batch) at step 70: 312.9109, Accuracy: 0.7680
Training loss (for one batch) at step 80: 304.0013, Accuracy: 0.7643
Training loss (for one batch) at step 90: 307.1109, Accuracy: 0.7627
Training loss (for one batch) at step 100: 300.6502, Accuracy: 0.7610
Training loss (for one batch) at step 110: 301.0058, Accuracy: 0.7615
Training loss (for one batch) at step 120: 323.5319, Accuracy: 0.7617
Training loss (for one batch) at step 130: 303.5409, Accuracy: 0.7615
Training loss (for one batch) at step 140: 313.2774, Accuracy: 0.7598
---- Training ----
Training loss: 270.2176
Training acc over epoch: 0.7595
---- Validation ----
Validation loss: 64.7523
Validation acc: 0.7394
Time taken: 53.70s
Start of epoch 10
Training loss (for one batch) at step 0: 293.1106, Accuracy: 0.8100
Training loss (for one batch) at step 10: 292.2770, Accuracy: 0.7818
Training loss (for one batch) at step 20: 332.5087, Accuracy: 0.7790
Training loss (for one batch) at step 30: 306.5567, Accuracy: 0.7755
Training loss (for one batch) at step 40: 305.6478, Accuracy: 0.7766
Training loss (for one batch) at step 50: 295.3223, Accuracy: 0.7798
Training loss (for one batch) at step 60: 301.2914, Accuracy: 0.7779
Training loss (for one batch) at step 70: 313.5097, Accuracy: 0.7759
Training loss (for one batch) at step 80: 295.1678, Accuracy: 0.7769
Training loss (for one batch) at step 90: 309.8512, Accuracy: 0.7756
Training loss (for one batch) at step 100: 313.2135, Accuracy: 0.7732
Training loss (for one batch) at step 110: 299.1552, Accuracy: 0.7750
Training loss (for one batch) at step 120: 297.9593, Accuracy: 0.7746
Training loss (for one batch) at step 130: 306.1283, Accuracy: 0.7730
Training loss (for one batch) at step 140: 291.9704, Accuracy: 0.7730
---- Training ----
Training loss: 264.5132
Training acc over epoch: 0.7720
---- Validation ----
Validation loss: 72.0313
Validation acc: 0.7579
Time taken: 62.59s
Start of epoch 11
Training loss (for one batch) at step 0: 295.4385, Accuracy: 0.7800
Training loss (for one batch) at step 10: 301.2710, Accuracy: 0.7873
Training loss (for one batch) at step 20: 289.2612, Accuracy: 0.7852
Training loss (for one batch) at step 30: 305.8139, Accuracy: 0.7826
Training loss (for one batch) at step 40: 304.6992, Accuracy: 0.7788
Training loss (for one batch) at step 50: 297.2608, Accuracy: 0.7780
Training loss (for one batch) at step 60: 316.0440, Accuracy: 0.7836
Training loss (for one batch) at step 70: 306.9107, Accuracy: 0.7818
Training loss (for one batch) at step 80: 308.2156, Accuracy: 0.7816
Training loss (for one batch) at step 90: 290.6125, Accuracy: 0.7785
Training loss (for one batch) at step 100: 298.0476, Accuracy: 0.7773
Training loss (for one batch) at step 110: 292.1115, Accuracy: 0.7800
Training loss (for one batch) at step 120: 290.1648, Accuracy: 0.7778
Training loss (for one batch) at step 130: 302.0177, Accuracy: 0.7757
Training loss (for one batch) at step 140: 304.9943, Accuracy: 0.7749
---- Training ----
Training loss: 247.7040
Training acc over epoch: 0.7753
---- Validation ----
Validation loss: 71.1890
Validation acc: 0.7654
Time taken: 58.10s
Start of epoch 12
Training loss (for one batch) at step 0: 296.2700, Accuracy: 0.7800
Training loss (for one batch) at step 10: 284.1924, Accuracy: 0.7982
Training loss (for one batch) at step 20: 301.7405, Accuracy: 0.7852
Training loss (for one batch) at step 30: 305.6154, Accuracy: 0.7865
Training loss (for one batch) at step 40: 305.4304, Accuracy: 0.7876
Training loss (for one batch) at step 50: 306.1944, Accuracy: 0.7890
Training loss (for one batch) at step 60: 285.4800, Accuracy: 0.7836
Training loss (for one batch) at step 70: 280.3773, Accuracy: 0.7848
Training loss (for one batch) at step 80: 304.5646, Accuracy: 0.7833
Training loss (for one batch) at step 90: 291.0331, Accuracy: 0.7832
Training loss (for one batch) at step 100: 315.4756, Accuracy: 0.7825
Training loss (for one batch) at step 110: 315.1168, Accuracy: 0.7829
Training loss (for one batch) at step 120: 312.0225, Accuracy: 0.7809
Training loss (for one batch) at step 130: 279.3751, Accuracy: 0.7797
Training loss (for one batch) at step 140: 295.1747, Accuracy: 0.7804
---- Training ----
Training loss: 260.6504
Training acc over epoch: 0.7798
---- Validation ----
Validation loss: 62.6049
Validation acc: 0.7512
Time taken: 60.11s
Start of epoch 13
Training loss (for one batch) at step 0: 292.3410, Accuracy: 0.8400
Training loss (for one batch) at step 10: 280.6440, Accuracy: 0.7991
Training loss (for one batch) at step 20: 285.8524, Accuracy: 0.7919
Training loss (for one batch) at step 30: 302.8146, Accuracy: 0.7935
Training loss (for one batch) at step 40: 278.5385, Accuracy: 0.7941
Training loss (for one batch) at step 50: 286.1875, Accuracy: 0.7976
Training loss (for one batch) at step 60: 281.8354, Accuracy: 0.7989
Training loss (for one batch) at step 70: 293.0889, Accuracy: 0.7970
Training loss (for one batch) at step 80: 294.8672, Accuracy: 0.7933
Training loss (for one batch) at step 90: 299.2309, Accuracy: 0.7921
Training loss (for one batch) at step 100: 269.1779, Accuracy: 0.7926
Training loss (for one batch) at step 110: 306.9193, Accuracy: 0.7920
Training loss (for one batch) at step 120: 302.1734, Accuracy: 0.7921
Training loss (for one batch) at step 130: 314.5122, Accuracy: 0.7918
Training loss (for one batch) at step 140: 300.9662, Accuracy: 0.7909
---- Training ----
Training loss: 265.0974
Training acc over epoch: 0.7906
---- Validation ----
Validation loss: 65.3173
Validation acc: 0.7477
Time taken: 63.96s
Start of epoch 14
Training loss (for one batch) at step 0: 297.4746, Accuracy: 0.7700
Training loss (for one batch) at step 10: 274.2362, Accuracy: 0.7827
Training loss (for one batch) at step 20: 278.2871, Accuracy: 0.7914
Training loss (for one batch) at step 30: 287.4460, Accuracy: 0.7900
Training loss (for one batch) at step 40: 276.1657, Accuracy: 0.7934
Training loss (for one batch) at step 50: 280.8741, Accuracy: 0.7996
Training loss (for one batch) at step 60: 280.3805, Accuracy: 0.8020
Training loss (for one batch) at step 70: 299.2661, Accuracy: 0.7993
Training loss (for one batch) at step 80: 284.0093, Accuracy: 0.7988
Training loss (for one batch) at step 90: 307.7014, Accuracy: 0.7982
Training loss (for one batch) at step 100: 293.5166, Accuracy: 0.7986
Training loss (for one batch) at step 110: 285.0305, Accuracy: 0.8003
Training loss (for one batch) at step 120: 290.3971, Accuracy: 0.8013
Training loss (for one batch) at step 130: 298.2986, Accuracy: 0.8021
Training loss (for one batch) at step 140: 290.5678, Accuracy: 0.8010
---- Training ----
Training loss: 266.7722
Training acc over epoch: 0.8010
---- Validation ----
Validation loss: 78.6809
Validation acc: 0.7238
Time taken: 58.54s
Start of epoch 15
Training loss (for one batch) at step 0: 289.3987, Accuracy: 0.8200
Training loss (for one batch) at step 10: 263.7898, Accuracy: 0.8045
Training loss (for one batch) at step 20: 264.7597, Accuracy: 0.7981
Training loss (for one batch) at step 30: 285.3677, Accuracy: 0.7913
Training loss (for one batch) at step 40: 289.4695, Accuracy: 0.7934
Training loss (for one batch) at step 50: 275.6686, Accuracy: 0.8000
Training loss (for one batch) at step 60: 265.1435, Accuracy: 0.8020
Training loss (for one batch) at step 70: 299.9144, Accuracy: 0.8027
Training loss (for one batch) at step 80: 270.1911, Accuracy: 0.8019
Training loss (for one batch) at step 90: 306.9959, Accuracy: 0.7984
Training loss (for one batch) at step 100: 282.6654, Accuracy: 0.7971
Training loss (for one batch) at step 110: 270.1904, Accuracy: 0.7992
Training loss (for one batch) at step 120: 290.6110, Accuracy: 0.7990
Training loss (for one batch) at step 130: 280.2569, Accuracy: 0.7988
Training loss (for one batch) at step 140: 276.3566, Accuracy: 0.7978
---- Training ----
Training loss: 246.3596
Training acc over epoch: 0.7988
---- Validation ----
Validation loss: 69.0349
Validation acc: 0.7324
Time taken: 68.72s
Start of epoch 16
Training loss (for one batch) at step 0: 282.3566, Accuracy: 0.8000
Training loss (for one batch) at step 10: 283.1825, Accuracy: 0.8173
Training loss (for one batch) at step 20: 301.1286, Accuracy: 0.8148
Training loss (for one batch) at step 30: 306.3073, Accuracy: 0.8119
Training loss (for one batch) at step 40: 290.4330, Accuracy: 0.8095
Training loss (for one batch) at step 50: 268.0714, Accuracy: 0.8112
Training loss (for one batch) at step 60: 268.8825, Accuracy: 0.8100
Training loss (for one batch) at step 70: 284.6416, Accuracy: 0.8090
Training loss (for one batch) at step 80: 295.9355, Accuracy: 0.8093
Training loss (for one batch) at step 90: 291.3719, Accuracy: 0.8085
Training loss (for one batch) at step 100: 277.4905, Accuracy: 0.8086
Training loss (for one batch) at step 110: 282.2858, Accuracy: 0.8076
Training loss (for one batch) at step 120: 284.9858, Accuracy: 0.8076
Training loss (for one batch) at step 130: 263.2606, Accuracy: 0.8073
Training loss (for one batch) at step 140: 275.5289, Accuracy: 0.8055
---- Training ----
Training loss: 245.3577
Training acc over epoch: 0.8051
---- Validation ----
Validation loss: 66.0141
Validation acc: 0.7281
Time taken: 51.68s
Start of epoch 17
Training loss (for one batch) at step 0: 274.8900, Accuracy: 0.7500
Training loss (for one batch) at step 10: 269.1225, Accuracy: 0.8100
Training loss (for one batch) at step 20: 269.4502, Accuracy: 0.8119
Training loss (for one batch) at step 30: 283.3691, Accuracy: 0.8023
Training loss (for one batch) at step 40: 247.5764, Accuracy: 0.8080
Training loss (for one batch) at step 50: 269.8650, Accuracy: 0.8102
Training loss (for one batch) at step 60: 267.3361, Accuracy: 0.8080
Training loss (for one batch) at step 70: 274.3218, Accuracy: 0.8094
Training loss (for one batch) at step 80: 287.8150, Accuracy: 0.8112
Training loss (for one batch) at step 90: 274.4793, Accuracy: 0.8093
Training loss (for one batch) at step 100: 290.6978, Accuracy: 0.8061
Training loss (for one batch) at step 110: 279.4243, Accuracy: 0.8057
Training loss (for one batch) at step 120: 264.9689, Accuracy: 0.8066
Training loss (for one batch) at step 130: 281.6041, Accuracy: 0.8083
Training loss (for one batch) at step 140: 284.7175, Accuracy: 0.8078
---- Training ----
Training loss: 248.8784
Training acc over epoch: 0.8076
---- Validation ----
Validation loss: 63.9478
Validation acc: 0.7491
Time taken: 70.26s
Start of epoch 18
Training loss (for one batch) at step 0: 272.9466, Accuracy: 0.8400
Training loss (for one batch) at step 10: 247.9762, Accuracy: 0.8391
Training loss (for one batch) at step 20: 252.5648, Accuracy: 0.8267
Training loss (for one batch) at step 30: 263.1140, Accuracy: 0.8213
Training loss (for one batch) at step 40: 292.3441, Accuracy: 0.8198
Training loss (for one batch) at step 50: 291.2531, Accuracy: 0.8198
Training loss (for one batch) at step 60: 278.8389, Accuracy: 0.8208
Training loss (for one batch) at step 70: 285.0758, Accuracy: 0.8203
Training loss (for one batch) at step 80: 271.2791, Accuracy: 0.8178
Training loss (for one batch) at step 90: 291.3209, Accuracy: 0.8165
Training loss (for one batch) at step 100: 266.3471, Accuracy: 0.8134
Training loss (for one batch) at step 110: 269.3744, Accuracy: 0.8148
Training loss (for one batch) at step 120: 274.5425, Accuracy: 0.8140
Training loss (for one batch) at step 130: 279.9560, Accuracy: 0.8140
Training loss (for one batch) at step 140: 277.9289, Accuracy: 0.8141
---- Training ----
Training loss: 233.6852
Training acc over epoch: 0.8138
---- Validation ----
Validation loss: 68.4359
Validation acc: 0.7389
Time taken: 52.34s
Start of epoch 19
Training loss (for one batch) at step 0: 279.3813, Accuracy: 0.8000
Training loss (for one batch) at step 10: 278.1928, Accuracy: 0.8009
Training loss (for one batch) at step 20: 302.7805, Accuracy: 0.8224
Training loss (for one batch) at step 30: 271.0864, Accuracy: 0.8155
Training loss (for one batch) at step 40: 264.1951, Accuracy: 0.8151
Training loss (for one batch) at step 50: 263.7999, Accuracy: 0.8196
Training loss (for one batch) at step 60: 275.9771, Accuracy: 0.8241
Training loss (for one batch) at step 70: 275.3202, Accuracy: 0.8207
Training loss (for one batch) at step 80: 250.8969, Accuracy: 0.8185
Training loss (for one batch) at step 90: 280.7341, Accuracy: 0.8185
Training loss (for one batch) at step 100: 272.0509, Accuracy: 0.8172
Training loss (for one batch) at step 110: 245.7936, Accuracy: 0.8185
Training loss (for one batch) at step 120: 274.5477, Accuracy: 0.8171
Training loss (for one batch) at step 130: 257.8491, Accuracy: 0.8163
Training loss (for one batch) at step 140: 251.0247, Accuracy: 0.8165
---- Training ----
Training loss: 232.0712
Training acc over epoch: 0.8174
---- Validation ----
Validation loss: 70.7217
Validation acc: 0.7544
Time taken: 71.62s
Start of epoch 20
Training loss (for one batch) at step 0: 276.1212, Accuracy: 0.7700
Training loss (for one batch) at step 10: 259.0184, Accuracy: 0.8055
Training loss (for one batch) at step 20: 275.1533, Accuracy: 0.8243
Training loss (for one batch) at step 30: 280.4181, Accuracy: 0.8206
Training loss (for one batch) at step 40: 274.4850, Accuracy: 0.8244
Training loss (for one batch) at step 50: 261.8997, Accuracy: 0.8276
Training loss (for one batch) at step 60: 279.0751, Accuracy: 0.8262
Training loss (for one batch) at step 70: 266.4477, Accuracy: 0.8258
Training loss (for one batch) at step 80: 246.7224, Accuracy: 0.8272
Training loss (for one batch) at step 90: 263.9696, Accuracy: 0.8237
Training loss (for one batch) at step 100: 267.0291, Accuracy: 0.8229
Training loss (for one batch) at step 110: 250.1994, Accuracy: 0.8232
Training loss (for one batch) at step 120: 272.1264, Accuracy: 0.8221
Training loss (for one batch) at step 130: 269.3633, Accuracy: 0.8219
Training loss (for one batch) at step 140: 274.8252, Accuracy: 0.8206
---- Training ----
Training loss: 237.5782
Training acc over epoch: 0.8205
---- Validation ----
Validation loss: 74.7589
Validation acc: 0.7453
Time taken: 45.99s
Start of epoch 21
Training loss (for one batch) at step 0: 257.6081, Accuracy: 0.8300
Training loss (for one batch) at step 10: 266.4244, Accuracy: 0.8164
Training loss (for one batch) at step 20: 248.3438, Accuracy: 0.8238
Training loss (for one batch) at step 30: 249.7629, Accuracy: 0.8219
Training loss (for one batch) at step 40: 254.5240, Accuracy: 0.8222
Training loss (for one batch) at step 50: 275.6307, Accuracy: 0.8267
Training loss (for one batch) at step 60: 268.2240, Accuracy: 0.8293
Training loss (for one batch) at step 70: 283.9367, Accuracy: 0.8265
Training loss (for one batch) at step 80: 260.9476, Accuracy: 0.8253
Training loss (for one batch) at step 90: 269.2513, Accuracy: 0.8242
Training loss (for one batch) at step 100: 260.6063, Accuracy: 0.8224
Training loss (for one batch) at step 110: 256.6330, Accuracy: 0.8234
Training loss (for one batch) at step 120: 262.2977, Accuracy: 0.8239
Training loss (for one batch) at step 130: 258.6865, Accuracy: 0.8222
Training loss (for one batch) at step 140: 260.8006, Accuracy: 0.8216
---- Training ----
Training loss: 217.3291
Training acc over epoch: 0.8223
---- Validation ----
Validation loss: 72.3949
Validation acc: 0.7421
Time taken: 71.40s
Start of epoch 22
Training loss (for one batch) at step 0: 276.4825, Accuracy: 0.7900
Training loss (for one batch) at step 10: 259.5473, Accuracy: 0.8209
Training loss (for one batch) at step 20: 267.1228, Accuracy: 0.8214
Training loss (for one batch) at step 30: 240.8865, Accuracy: 0.8232
Training loss (for one batch) at step 40: 254.7351, Accuracy: 0.8259
Training loss (for one batch) at step 50: 267.2617, Accuracy: 0.8292
Training loss (for one batch) at step 60: 258.9462, Accuracy: 0.8302
Training loss (for one batch) at step 70: 272.0041, Accuracy: 0.8315
Training loss (for one batch) at step 80: 246.8758, Accuracy: 0.8285
Training loss (for one batch) at step 90: 258.6791, Accuracy: 0.8259
Training loss (for one batch) at step 100: 265.4581, Accuracy: 0.8258
Training loss (for one batch) at step 110: 270.8426, Accuracy: 0.8250
Training loss (for one batch) at step 120: 280.7317, Accuracy: 0.8236
Training loss (for one batch) at step 130: 257.4622, Accuracy: 0.8237
Training loss (for one batch) at step 140: 270.6691, Accuracy: 0.8224
---- Training ----
Training loss: 232.1438
Training acc over epoch: 0.8224
---- Validation ----
Validation loss: 61.9496
Validation acc: 0.7539
Time taken: 46.75s
Start of epoch 23
Training loss (for one batch) at step 0: 265.7611, Accuracy: 0.7700
Training loss (for one batch) at step 10: 242.0949, Accuracy: 0.8264
Training loss (for one batch) at step 20: 265.8304, Accuracy: 0.8333
Training loss (for one batch) at step 30: 270.2719, Accuracy: 0.8297
Training loss (for one batch) at step 40: 265.8326, Accuracy: 0.8305
Training loss (for one batch) at step 50: 251.3139, Accuracy: 0.8367
Training loss (for one batch) at step 60: 248.8955, Accuracy: 0.8321
Training loss (for one batch) at step 70: 243.8008, Accuracy: 0.8301
Training loss (for one batch) at step 80: 270.3539, Accuracy: 0.8284
Training loss (for one batch) at step 90: 239.1412, Accuracy: 0.8280
Training loss (for one batch) at step 100: 270.8295, Accuracy: 0.8265
Training loss (for one batch) at step 110: 251.3218, Accuracy: 0.8264
Training loss (for one batch) at step 120: 236.0970, Accuracy: 0.8281
Training loss (for one batch) at step 130: 244.9296, Accuracy: 0.8287
Training loss (for one batch) at step 140: 267.5606, Accuracy: 0.8278
---- Training ----
Training loss: 227.3950
Training acc over epoch: 0.8268
---- Validation ----
Validation loss: 69.4867
Validation acc: 0.7480
Time taken: 69.38s
Start of epoch 24
Training loss (for one batch) at step 0: 272.0812, Accuracy: 0.8000
Training loss (for one batch) at step 10: 245.1203, Accuracy: 0.8282
Training loss (for one batch) at step 20: 243.8790, Accuracy: 0.8271
Training loss (for one batch) at step 30: 263.6561, Accuracy: 0.8210
Training loss (for one batch) at step 40: 246.8000, Accuracy: 0.8251
Training loss (for one batch) at step 50: 242.7812, Accuracy: 0.8343
Training loss (for one batch) at step 60: 255.8884, Accuracy: 0.8334
Training loss (for one batch) at step 70: 248.4144, Accuracy: 0.8314
Training loss (for one batch) at step 80: 246.6351, Accuracy: 0.8275
Training loss (for one batch) at step 90: 271.6547, Accuracy: 0.8265
Training loss (for one batch) at step 100: 259.4817, Accuracy: 0.8273
Training loss (for one batch) at step 110: 240.5323, Accuracy: 0.8279
Training loss (for one batch) at step 120: 254.9436, Accuracy: 0.8278
Training loss (for one batch) at step 130: 238.9370, Accuracy: 0.8271
Training loss (for one batch) at step 140: 258.3929, Accuracy: 0.8269
---- Training ----
Training loss: 228.5497
Training acc over epoch: 0.8266
---- Validation ----
Validation loss: 74.4236
Validation acc: 0.7168
Time taken: 49.24s
Start of epoch 25
Training loss (for one batch) at step 0: 256.0342, Accuracy: 0.8100
Training loss (for one batch) at step 10: 254.4592, Accuracy: 0.8191
Training loss (for one batch) at step 20: 251.3453, Accuracy: 0.8190
Training loss (for one batch) at step 30: 257.1556, Accuracy: 0.8277
Training loss (for one batch) at step 40: 247.4315, Accuracy: 0.8320
Training loss (for one batch) at step 50: 241.7233, Accuracy: 0.8341
Training loss (for one batch) at step 60: 243.6523, Accuracy: 0.8359
Training loss (for one batch) at step 70: 264.5139, Accuracy: 0.8358
Training loss (for one batch) at step 80: 249.9616, Accuracy: 0.8356
Training loss (for one batch) at step 90: 252.9820, Accuracy: 0.8345
Training loss (for one batch) at step 100: 261.4125, Accuracy: 0.8334
Training loss (for one batch) at step 110: 247.1939, Accuracy: 0.8338
Training loss (for one batch) at step 120: 252.2254, Accuracy: 0.8340
Training loss (for one batch) at step 130: 245.3759, Accuracy: 0.8344
Training loss (for one batch) at step 140: 256.6851, Accuracy: 0.8345
---- Training ----
Training loss: 212.4499
Training acc over epoch: 0.8330
---- Validation ----
Validation loss: 69.2358
Validation acc: 0.7421
Time taken: 69.31s
Start of epoch 26
Training loss (for one batch) at step 0: 246.8632, Accuracy: 0.8500
Training loss (for one batch) at step 10: 247.7753, Accuracy: 0.8464
Training loss (for one batch) at step 20: 258.1092, Accuracy: 0.8419
Training loss (for one batch) at step 30: 233.8673, Accuracy: 0.8416
Training loss (for one batch) at step 40: 235.2881, Accuracy: 0.8449
Training loss (for one batch) at step 50: 250.0424, Accuracy: 0.8420
Training loss (for one batch) at step 60: 242.9477, Accuracy: 0.8398
Training loss (for one batch) at step 70: 253.6862, Accuracy: 0.8415
Training loss (for one batch) at step 80: 256.7579, Accuracy: 0.8391
Training loss (for one batch) at step 90: 230.7756, Accuracy: 0.8390
Training loss (for one batch) at step 100: 236.8679, Accuracy: 0.8377
Training loss (for one batch) at step 110: 264.4499, Accuracy: 0.8385
Training loss (for one batch) at step 120: 246.7062, Accuracy: 0.8389
Training loss (for one batch) at step 130: 256.0327, Accuracy: 0.8376
Training loss (for one batch) at step 140: 256.4574, Accuracy: 0.8365
---- Training ----
Training loss: 215.5723
Training acc over epoch: 0.8360
---- Validation ----
Validation loss: 76.6262
Validation acc: 0.7426
Time taken: 50.23s
Start of epoch 27
Training loss (for one batch) at step 0: 277.0906, Accuracy: 0.7700
Training loss (for one batch) at step 10: 262.1761, Accuracy: 0.8445
Training loss (for one batch) at step 20: 246.4196, Accuracy: 0.8471
Training loss (for one batch) at step 30: 238.4443, Accuracy: 0.8442
Training loss (for one batch) at step 40: 238.5031, Accuracy: 0.8478
Training loss (for one batch) at step 50: 252.5024, Accuracy: 0.8463
Training loss (for one batch) at step 60: 224.9248, Accuracy: 0.8487
Training loss (for one batch) at step 70: 246.4198, Accuracy: 0.8485
Training loss (for one batch) at step 80: 241.9708, Accuracy: 0.8452
Training loss (for one batch) at step 90: 254.1552, Accuracy: 0.8438
Training loss (for one batch) at step 100: 256.2594, Accuracy: 0.8432
Training loss (for one batch) at step 110: 246.4868, Accuracy: 0.8436
Training loss (for one batch) at step 120: 244.1653, Accuracy: 0.8432
Training loss (for one batch) at step 130: 290.4219, Accuracy: 0.8408
Training loss (for one batch) at step 140: 251.2270, Accuracy: 0.8408
---- Training ----
Training loss: 231.0719
Training acc over epoch: 0.8397
---- Validation ----
Validation loss: 86.7389
Validation acc: 0.7230
Time taken: 65.27s
Start of epoch 28
Training loss (for one batch) at step 0: 255.8387, Accuracy: 0.8200
Training loss (for one batch) at step 10: 225.8213, Accuracy: 0.8418
Training loss (for one batch) at step 20: 218.8719, Accuracy: 0.8386
Training loss (for one batch) at step 30: 241.3032, Accuracy: 0.8387
Training loss (for one batch) at step 40: 244.0343, Accuracy: 0.8415
Training loss (for one batch) at step 50: 256.6502, Accuracy: 0.8439
Training loss (for one batch) at step 60: 236.6000, Accuracy: 0.8451
Training loss (for one batch) at step 70: 241.3197, Accuracy: 0.8461
Training loss (for one batch) at step 80: 251.1613, Accuracy: 0.8435
Training loss (for one batch) at step 90: 244.2081, Accuracy: 0.8423
Training loss (for one batch) at step 100: 236.1856, Accuracy: 0.8434
Training loss (for one batch) at step 110: 223.7601, Accuracy: 0.8432
Training loss (for one batch) at step 120: 246.5681, Accuracy: 0.8431
Training loss (for one batch) at step 130: 231.0123, Accuracy: 0.8437
Training loss (for one batch) at step 140: 250.5249, Accuracy: 0.8416
---- Training ----
Training loss: 224.3437
Training acc over epoch: 0.8407
---- Validation ----
Validation loss: 74.3685
Validation acc: 0.7410
Time taken: 51.54s
Start of epoch 29
Training loss (for one batch) at step 0: 241.6153, Accuracy: 0.8200
Training loss (for one batch) at step 10: 253.2788, Accuracy: 0.8418
Training loss (for one batch) at step 20: 225.0560, Accuracy: 0.8467
Training loss (for one batch) at step 30: 227.9296, Accuracy: 0.8439
Training loss (for one batch) at step 40: 249.8547, Accuracy: 0.8427
Training loss (for one batch) at step 50: 223.9567, Accuracy: 0.8496
Training loss (for one batch) at step 60: 234.1541, Accuracy: 0.8498
Training loss (for one batch) at step 70: 241.9895, Accuracy: 0.8480
Training loss (for one batch) at step 80: 244.4040, Accuracy: 0.8448
Training loss (for one batch) at step 90: 239.4972, Accuracy: 0.8452
Training loss (for one batch) at step 100: 244.7682, Accuracy: 0.8441
Training loss (for one batch) at step 110: 250.7487, Accuracy: 0.8435
Training loss (for one batch) at step 120: 234.4897, Accuracy: 0.8438
Training loss (for one batch) at step 130: 243.3691, Accuracy: 0.8431
Training loss (for one batch) at step 140: 228.3394, Accuracy: 0.8443
---- Training ----
Training loss: 217.6007
Training acc over epoch: 0.8438
---- Validation ----
Validation loss: 79.2367
Validation acc: 0.7281
Time taken: 63.30s
Start of epoch 30
Training loss (for one batch) at step 0: 259.1945, Accuracy: 0.7800
Training loss (for one batch) at step 10: 223.3891, Accuracy: 0.8336
Training loss (for one batch) at step 20: 230.8128, Accuracy: 0.8476
Training loss (for one batch) at step 30: 231.3328, Accuracy: 0.8490
Training loss (for one batch) at step 40: 240.7964, Accuracy: 0.8473
Training loss (for one batch) at step 50: 229.8090, Accuracy: 0.8506
Training loss (for one batch) at step 60: 233.2526, Accuracy: 0.8497
Training loss (for one batch) at step 70: 236.0879, Accuracy: 0.8506
Training loss (for one batch) at step 80: 241.9864, Accuracy: 0.8510
Training loss (for one batch) at step 90: 236.8143, Accuracy: 0.8484
Training loss (for one batch) at step 100: 247.7609, Accuracy: 0.8485
Training loss (for one batch) at step 110: 226.1624, Accuracy: 0.8487
Training loss (for one batch) at step 120: 242.7240, Accuracy: 0.8496
Training loss (for one batch) at step 130: 246.0476, Accuracy: 0.8488
Training loss (for one batch) at step 140: 236.3339, Accuracy: 0.8482
---- Training ----
Training loss: 225.1215
Training acc over epoch: 0.8481
---- Validation ----
Validation loss: 75.7910
Validation acc: 0.7286
Time taken: 53.81s
Start of epoch 31
Training loss (for one batch) at step 0: 244.4957, Accuracy: 0.8900
Training loss (for one batch) at step 10: 234.6986, Accuracy: 0.8527
Training loss (for one batch) at step 20: 238.7920, Accuracy: 0.8452
Training loss (for one batch) at step 30: 245.8981, Accuracy: 0.8455
Training loss (for one batch) at step 40: 257.8392, Accuracy: 0.8459
Training loss (for one batch) at step 50: 247.9574, Accuracy: 0.8490
Training loss (for one batch) at step 60: 254.8295, Accuracy: 0.8505
Training loss (for one batch) at step 70: 232.8447, Accuracy: 0.8479
Training loss (for one batch) at step 80: 237.8924, Accuracy: 0.8427
Training loss (for one batch) at step 90: 240.3538, Accuracy: 0.8413
Training loss (for one batch) at step 100: 229.9748, Accuracy: 0.8424
Training loss (for one batch) at step 110: 227.2185, Accuracy: 0.8427
Training loss (for one batch) at step 120: 209.5623, Accuracy: 0.8431
Training loss (for one batch) at step 130: 220.3897, Accuracy: 0.8426
Training loss (for one batch) at step 140: 227.3054, Accuracy: 0.8421
---- Training ----
Training loss: 203.9045
Training acc over epoch: 0.8416
---- Validation ----
Validation loss: 69.1030
Validation acc: 0.7227
Time taken: 60.09s
Start of epoch 32
Training loss (for one batch) at step 0: 213.5802, Accuracy: 0.9200
Training loss (for one batch) at step 10: 235.1485, Accuracy: 0.8518
Training loss (for one batch) at step 20: 251.9877, Accuracy: 0.8538
Training loss (for one batch) at step 30: 232.8600, Accuracy: 0.8510
Training loss (for one batch) at step 40: 228.4192, Accuracy: 0.8510
Training loss (for one batch) at step 50: 208.7157, Accuracy: 0.8575
Training loss (for one batch) at step 60: 220.2924, Accuracy: 0.8582
Training loss (for one batch) at step 70: 243.8316, Accuracy: 0.8537
Training loss (for one batch) at step 80: 243.9522, Accuracy: 0.8507
Training loss (for one batch) at step 90: 230.8679, Accuracy: 0.8498
Training loss (for one batch) at step 100: 246.1116, Accuracy: 0.8501
Training loss (for one batch) at step 110: 234.1390, Accuracy: 0.8493
Training loss (for one batch) at step 120: 246.8664, Accuracy: 0.8498
Training loss (for one batch) at step 130: 263.2541, Accuracy: 0.8500
Training loss (for one batch) at step 140: 239.6235, Accuracy: 0.8507
---- Training ----
Training loss: 213.3511
Training acc over epoch: 0.8503
---- Validation ----
Validation loss: 70.8767
Validation acc: 0.7289
Time taken: 54.89s
Start of epoch 33
Training loss (for one batch) at step 0: 255.6301, Accuracy: 0.8300
Training loss (for one batch) at step 10: 228.9266, Accuracy: 0.8482
Training loss (for one batch) at step 20: 227.9070, Accuracy: 0.8538
Training loss (for one batch) at step 30: 231.9559, Accuracy: 0.8500
Training loss (for one batch) at step 40: 226.0088, Accuracy: 0.8510
Training loss (for one batch) at step 50: 256.7722, Accuracy: 0.8527
Training loss (for one batch) at step 60: 214.6650, Accuracy: 0.8552
Training loss (for one batch) at step 70: 237.6623, Accuracy: 0.8535
Training loss (for one batch) at step 80: 231.9537, Accuracy: 0.8517
Training loss (for one batch) at step 90: 230.2535, Accuracy: 0.8512
Training loss (for one batch) at step 100: 245.0045, Accuracy: 0.8488
Training loss (for one batch) at step 110: 237.9336, Accuracy: 0.8498
Training loss (for one batch) at step 120: 222.8438, Accuracy: 0.8509
Training loss (for one batch) at step 130: 240.9932, Accuracy: 0.8492
Training loss (for one batch) at step 140: 230.7503, Accuracy: 0.8492
---- Training ----
Training loss: 196.4485
Training acc over epoch: 0.8500
---- Validation ----
Validation loss: 84.7091
Validation acc: 0.7316
Time taken: 59.51s
Start of epoch 34
Training loss (for one batch) at step 0: 216.3004, Accuracy: 0.8600
Training loss (for one batch) at step 10: 246.1683, Accuracy: 0.8527
Training loss (for one batch) at step 20: 217.2094, Accuracy: 0.8524
Training loss (for one batch) at step 30: 246.5513, Accuracy: 0.8568
Training loss (for one batch) at step 40: 242.6228, Accuracy: 0.8532
Training loss (for one batch) at step 50: 218.4184, Accuracy: 0.8553
Training loss (for one batch) at step 60: 208.4032, Accuracy: 0.8564
Training loss (for one batch) at step 70: 224.3983, Accuracy: 0.8570
Training loss (for one batch) at step 80: 226.0243, Accuracy: 0.8540
Training loss (for one batch) at step 90: 240.9213, Accuracy: 0.8508
Training loss (for one batch) at step 100: 219.7956, Accuracy: 0.8511
Training loss (for one batch) at step 110: 218.1034, Accuracy: 0.8530
Training loss (for one batch) at step 120: 229.8796, Accuracy: 0.8515
Training loss (for one batch) at step 130: 218.8590, Accuracy: 0.8527
Training loss (for one batch) at step 140: 222.4816, Accuracy: 0.8513
---- Training ----
Training loss: 192.9897
Training acc over epoch: 0.8514
---- Validation ----
Validation loss: 70.8462
Validation acc: 0.7431
Time taken: 57.13s
Start of epoch 35
Training loss (for one batch) at step 0: 267.2434, Accuracy: 0.7400
Training loss (for one batch) at step 10: 218.1785, Accuracy: 0.8482
Training loss (for one batch) at step 20: 214.4615, Accuracy: 0.8638
Training loss (for one batch) at step 30: 224.0377, Accuracy: 0.8581
Training loss (for one batch) at step 40: 237.7624, Accuracy: 0.8568
Training loss (for one batch) at step 50: 228.3045, Accuracy: 0.8590
Training loss (for one batch) at step 60: 217.6089, Accuracy: 0.8570
Training loss (for one batch) at step 70: 212.1713, Accuracy: 0.8597
Training loss (for one batch) at step 80: 230.1659, Accuracy: 0.8573
Training loss (for one batch) at step 90: 254.7918, Accuracy: 0.8527
Training loss (for one batch) at step 100: 224.2519, Accuracy: 0.8527
Training loss (for one batch) at step 110: 233.6137, Accuracy: 0.8538
Training loss (for one batch) at step 120: 226.4637, Accuracy: 0.8555
Training loss (for one batch) at step 130: 221.4399, Accuracy: 0.8547
Training loss (for one batch) at step 140: 233.4846, Accuracy: 0.8533
---- Training ----
Training loss: 183.0642
Training acc over epoch: 0.8522
---- Validation ----
Validation loss: 70.7151
Validation acc: 0.7335
Time taken: 56.19s
Start of epoch 36
Training loss (for one batch) at step 0: 233.8235, Accuracy: 0.8500
Training loss (for one batch) at step 10: 212.6081, Accuracy: 0.8582
Training loss (for one batch) at step 20: 213.5287, Accuracy: 0.8533
Training loss (for one batch) at step 30: 227.9007, Accuracy: 0.8613
Training loss (for one batch) at step 40: 228.0869, Accuracy: 0.8590
Training loss (for one batch) at step 50: 220.6433, Accuracy: 0.8598
Training loss (for one batch) at step 60: 217.8183, Accuracy: 0.8620
Training loss (for one batch) at step 70: 243.0871, Accuracy: 0.8600
Training loss (for one batch) at step 80: 247.3833, Accuracy: 0.8557
Training loss (for one batch) at step 90: 251.6672, Accuracy: 0.8543
Training loss (for one batch) at step 100: 218.7553, Accuracy: 0.8524
Training loss (for one batch) at step 110: 220.2776, Accuracy: 0.8547
Training loss (for one batch) at step 120: 233.2729, Accuracy: 0.8550
Training loss (for one batch) at step 130: 231.0576, Accuracy: 0.8534
Training loss (for one batch) at step 140: 217.3413, Accuracy: 0.8533
---- Training ----
Training loss: 209.7606
Training acc over epoch: 0.8531
---- Validation ----
Validation loss: 85.5308
Validation acc: 0.7343
Time taken: 61.82s
Start of epoch 37
Training loss (for one batch) at step 0: 217.9160, Accuracy: 0.8500
Training loss (for one batch) at step 10: 219.3918, Accuracy: 0.8445
Training loss (for one batch) at step 20: 231.2528, Accuracy: 0.8562
Training loss (for one batch) at step 30: 202.6178, Accuracy: 0.8561
Training loss (for one batch) at step 40: 225.8084, Accuracy: 0.8566
Training loss (for one batch) at step 50: 213.6163, Accuracy: 0.8580
Training loss (for one batch) at step 60: 220.8907, Accuracy: 0.8574
Training loss (for one batch) at step 70: 222.5047, Accuracy: 0.8568
Training loss (for one batch) at step 80: 227.2559, Accuracy: 0.8560
Training loss (for one batch) at step 90: 223.6768, Accuracy: 0.8535
Training loss (for one batch) at step 100: 241.9527, Accuracy: 0.8538
Training loss (for one batch) at step 110: 223.0481, Accuracy: 0.8544
Training loss (for one batch) at step 120: 216.9584, Accuracy: 0.8535
Training loss (for one batch) at step 130: 241.9727, Accuracy: 0.8534
Training loss (for one batch) at step 140: 229.2938, Accuracy: 0.8535
---- Training ----
Training loss: 193.3904
Training acc over epoch: 0.8546
---- Validation ----
Validation loss: 66.6012
Validation acc: 0.7273
Time taken: 55.07s
Start of epoch 38
Training loss (for one batch) at step 0: 215.4482, Accuracy: 0.8800
Training loss (for one batch) at step 10: 227.6439, Accuracy: 0.8245
Training loss (for one batch) at step 20: 202.9946, Accuracy: 0.8467
Training loss (for one batch) at step 30: 224.3863, Accuracy: 0.8552
Training loss (for one batch) at step 40: 215.3178, Accuracy: 0.8593
Training loss (for one batch) at step 50: 227.8319, Accuracy: 0.8625
Training loss (for one batch) at step 60: 213.2877, Accuracy: 0.8616
Training loss (for one batch) at step 70: 213.3867, Accuracy: 0.8614
Training loss (for one batch) at step 80: 227.9424, Accuracy: 0.8588
Training loss (for one batch) at step 90: 232.0980, Accuracy: 0.8586
Training loss (for one batch) at step 100: 220.7150, Accuracy: 0.8580
Training loss (for one batch) at step 110: 195.0101, Accuracy: 0.8592
Training loss (for one batch) at step 120: 229.7917, Accuracy: 0.8588
Training loss (for one batch) at step 130: 243.2842, Accuracy: 0.8591
Training loss (for one batch) at step 140: 210.2522, Accuracy: 0.8584
---- Training ----
Training loss: 197.0641
Training acc over epoch: 0.8588
---- Validation ----
Validation loss: 65.6224
Validation acc: 0.7429
Time taken: 63.90s
Start of epoch 39
Training loss (for one batch) at step 0: 238.4746, Accuracy: 0.8000
Training loss (for one batch) at step 10: 230.3143, Accuracy: 0.8573
Training loss (for one batch) at step 20: 223.6493, Accuracy: 0.8495
Training loss (for one batch) at step 30: 222.8586, Accuracy: 0.8513
Training loss (for one batch) at step 40: 220.3348, Accuracy: 0.8563
Training loss (for one batch) at step 50: 228.5116, Accuracy: 0.8576
Training loss (for one batch) at step 60: 207.0388, Accuracy: 0.8595
Training loss (for one batch) at step 70: 241.3827, Accuracy: 0.8580
Training loss (for one batch) at step 80: 226.1390, Accuracy: 0.8557
Training loss (for one batch) at step 90: 214.9415, Accuracy: 0.8557
Training loss (for one batch) at step 100: 209.0433, Accuracy: 0.8556
Training loss (for one batch) at step 110: 221.9224, Accuracy: 0.8591
Training loss (for one batch) at step 120: 203.4293, Accuracy: 0.8579
Training loss (for one batch) at step 130: 210.0576, Accuracy: 0.8567
Training loss (for one batch) at step 140: 223.2287, Accuracy: 0.8570
---- Training ----
Training loss: 216.0514
Training acc over epoch: 0.8573
---- Validation ----
Validation loss: 75.0526
Validation acc: 0.7321
Time taken: 54.03s
Start of epoch 40
Training loss (for one batch) at step 0: 210.0832, Accuracy: 0.8600
Training loss (for one batch) at step 10: 213.5945, Accuracy: 0.8600
Training loss (for one batch) at step 20: 215.8032, Accuracy: 0.8643
Training loss (for one batch) at step 30: 222.8817, Accuracy: 0.8584
Training loss (for one batch) at step 40: 222.6540, Accuracy: 0.8590
Training loss (for one batch) at step 50: 212.5570, Accuracy: 0.8620
Training loss (for one batch) at step 60: 241.9197, Accuracy: 0.8641
Training loss (for one batch) at step 70: 242.6893, Accuracy: 0.8601
Training loss (for one batch) at step 80: 214.0489, Accuracy: 0.8596
Training loss (for one batch) at step 90: 215.5048, Accuracy: 0.8586
Training loss (for one batch) at step 100: 190.9990, Accuracy: 0.8588
Training loss (for one batch) at step 110: 227.5381, Accuracy: 0.8605
Training loss (for one batch) at step 120: 211.4626, Accuracy: 0.8589
Training loss (for one batch) at step 130: 225.6810, Accuracy: 0.8586
Training loss (for one batch) at step 140: 208.4209, Accuracy: 0.8585
---- Training ----
Training loss: 211.6571
Training acc over epoch: 0.8587
---- Validation ----
Validation loss: 75.7190
Validation acc: 0.7405
Time taken: 67.00s
Start of epoch 41
Training loss (for one batch) at step 0: 224.8729, Accuracy: 0.8200
Training loss (for one batch) at step 10: 207.7277, Accuracy: 0.8518
Training loss (for one batch) at step 20: 215.5733, Accuracy: 0.8624
Training loss (for one batch) at step 30: 207.2203, Accuracy: 0.8619
Training loss (for one batch) at step 40: 209.2862, Accuracy: 0.8641
Training loss (for one batch) at step 50: 207.7836, Accuracy: 0.8692
Training loss (for one batch) at step 60: 218.2404, Accuracy: 0.8661
Training loss (for one batch) at step 70: 207.7049, Accuracy: 0.8642
Training loss (for one batch) at step 80: 236.5039, Accuracy: 0.8621
Training loss (for one batch) at step 90: 208.1163, Accuracy: 0.8582
Training loss (for one batch) at step 100: 231.4270, Accuracy: 0.8556
Training loss (for one batch) at step 110: 196.9257, Accuracy: 0.8561
Training loss (for one batch) at step 120: 221.2243, Accuracy: 0.8567
Training loss (for one batch) at step 130: 206.7386, Accuracy: 0.8581
Training loss (for one batch) at step 140: 213.0825, Accuracy: 0.8574
---- Training ----
Training loss: 202.3738
Training acc over epoch: 0.8567
---- Validation ----
Validation loss: 81.5544
Validation acc: 0.7319
Time taken: 52.42s
Start of epoch 42
Training loss (for one batch) at step 0: 191.8417, Accuracy: 0.8900
Training loss (for one batch) at step 10: 215.4447, Accuracy: 0.8700
Training loss (for one batch) at step 20: 264.2191, Accuracy: 0.8714
Training loss (for one batch) at step 30: 229.9935, Accuracy: 0.8684
Training loss (for one batch) at step 40: 219.8963, Accuracy: 0.8651
Training loss (for one batch) at step 50: 211.9828, Accuracy: 0.8676
Training loss (for one batch) at step 60: 222.1944, Accuracy: 0.8670
Training loss (for one batch) at step 70: 201.8228, Accuracy: 0.8665
Training loss (for one batch) at step 80: 223.0892, Accuracy: 0.8635
Training loss (for one batch) at step 90: 195.3042, Accuracy: 0.8632
Training loss (for one batch) at step 100: 212.5520, Accuracy: 0.8624
Training loss (for one batch) at step 110: 220.4249, Accuracy: 0.8633
Training loss (for one batch) at step 120: 214.3669, Accuracy: 0.8631
Training loss (for one batch) at step 130: 240.6119, Accuracy: 0.8624
Training loss (for one batch) at step 140: 211.5039, Accuracy: 0.8625
---- Training ----
Training loss: 207.9323
Training acc over epoch: 0.8626
---- Validation ----
Validation loss: 48.7808
Validation acc: 0.7230
Time taken: 68.50s
Start of epoch 43
Training loss (for one batch) at step 0: 239.1552, Accuracy: 0.8400
Training loss (for one batch) at step 10: 215.0515, Accuracy: 0.8645
Training loss (for one batch) at step 20: 204.8672, Accuracy: 0.8681
Training loss (for one batch) at step 30: 210.7895, Accuracy: 0.8652
Training loss (for one batch) at step 40: 201.3899, Accuracy: 0.8688
Training loss (for one batch) at step 50: 204.3907, Accuracy: 0.8692
Training loss (for one batch) at step 60: 213.1714, Accuracy: 0.8669
Training loss (for one batch) at step 70: 200.3799, Accuracy: 0.8662
Training loss (for one batch) at step 80: 212.8331, Accuracy: 0.8642
Training loss (for one batch) at step 90: 205.2450, Accuracy: 0.8630
Training loss (for one batch) at step 100: 199.6923, Accuracy: 0.8632
Training loss (for one batch) at step 110: 227.0561, Accuracy: 0.8625
Training loss (for one batch) at step 120: 198.3493, Accuracy: 0.8626
Training loss (for one batch) at step 130: 209.1645, Accuracy: 0.8629
Training loss (for one batch) at step 140: 231.2700, Accuracy: 0.8618
---- Training ----
Training loss: 180.6710
Training acc over epoch: 0.8624
---- Validation ----
Validation loss: 84.7197
Validation acc: 0.7410
Time taken: 51.33s
Start of epoch 44
Training loss (for one batch) at step 0: 221.9289, Accuracy: 0.8400
Training loss (for one batch) at step 10: 187.0391, Accuracy: 0.8582
Training loss (for one batch) at step 20: 204.9850, Accuracy: 0.8629
Training loss (for one batch) at step 30: 204.6988, Accuracy: 0.8635
Training loss (for one batch) at step 40: 209.1958, Accuracy: 0.8649
Training loss (for one batch) at step 50: 222.0705, Accuracy: 0.8663
Training loss (for one batch) at step 60: 189.1868, Accuracy: 0.8667
Training loss (for one batch) at step 70: 211.6730, Accuracy: 0.8668
Training loss (for one batch) at step 80: 234.1201, Accuracy: 0.8643
Training loss (for one batch) at step 90: 225.8559, Accuracy: 0.8618
Training loss (for one batch) at step 100: 208.1099, Accuracy: 0.8621
Training loss (for one batch) at step 110: 192.4733, Accuracy: 0.8627
Training loss (for one batch) at step 120: 207.3810, Accuracy: 0.8622
Training loss (for one batch) at step 130: 231.5769, Accuracy: 0.8622
Training loss (for one batch) at step 140: 209.8936, Accuracy: 0.8623
---- Training ----
Training loss: 171.4940
Training acc over epoch: 0.8622
---- Validation ----
Validation loss: 77.5996
Validation acc: 0.7316
Time taken: 66.09s
Start of epoch 45
Training loss (for one batch) at step 0: 215.6532, Accuracy: 0.7500
Training loss (for one batch) at step 10: 196.9977, Accuracy: 0.8509
Training loss (for one batch) at step 20: 190.9654, Accuracy: 0.8643
Training loss (for one batch) at step 30: 208.2529, Accuracy: 0.8613
Training loss (for one batch) at step 40: 215.6306, Accuracy: 0.8620
Training loss (for one batch) at step 50: 208.1503, Accuracy: 0.8604
Training loss (for one batch) at step 60: 198.4570, Accuracy: 0.8575
Training loss (for one batch) at step 70: 217.2853, Accuracy: 0.8576
Training loss (for one batch) at step 80: 207.6673, Accuracy: 0.8569
Training loss (for one batch) at step 90: 208.2173, Accuracy: 0.8569
Training loss (for one batch) at step 100: 238.3291, Accuracy: 0.8555
Training loss (for one batch) at step 110: 198.5829, Accuracy: 0.8575
Training loss (for one batch) at step 120: 228.1394, Accuracy: 0.8591
Training loss (for one batch) at step 130: 231.0985, Accuracy: 0.8579
Training loss (for one batch) at step 140: 214.2993, Accuracy: 0.8578
---- Training ----
Training loss: 195.8482
Training acc over epoch: 0.8588
---- Validation ----
Validation loss: 70.9614
Validation acc: 0.7378
Time taken: 54.83s
Start of epoch 46
Training loss (for one batch) at step 0: 215.0208, Accuracy: 0.8100
Training loss (for one batch) at step 10: 211.5710, Accuracy: 0.8691
Training loss (for one batch) at step 20: 215.3608, Accuracy: 0.8652
Training loss (for one batch) at step 30: 235.3496, Accuracy: 0.8694
Training loss (for one batch) at step 40: 199.7101, Accuracy: 0.8680
Training loss (for one batch) at step 50: 203.6368, Accuracy: 0.8702
Training loss (for one batch) at step 60: 202.3418, Accuracy: 0.8648
Training loss (for one batch) at step 70: 227.0818, Accuracy: 0.8663
Training loss (for one batch) at step 80: 206.1741, Accuracy: 0.8651
Training loss (for one batch) at step 90: 201.0674, Accuracy: 0.8646
Training loss (for one batch) at step 100: 207.4842, Accuracy: 0.8646
Training loss (for one batch) at step 110: 225.7283, Accuracy: 0.8659
Training loss (for one batch) at step 120: 199.6247, Accuracy: 0.8657
Training loss (for one batch) at step 130: 218.7323, Accuracy: 0.8653
Training loss (for one batch) at step 140: 203.7116, Accuracy: 0.8640
---- Training ----
Training loss: 194.4820
Training acc over epoch: 0.8642
---- Validation ----
Validation loss: 70.2130
Validation acc: 0.7308
Time taken: 63.98s
Start of epoch 47
Training loss (for one batch) at step 0: 230.8803, Accuracy: 0.8300
Training loss (for one batch) at step 10: 197.1341, Accuracy: 0.8773
Training loss (for one batch) at step 20: 196.8264, Accuracy: 0.8771
Training loss (for one batch) at step 30: 205.3613, Accuracy: 0.8735
Training loss (for one batch) at step 40: 209.0168, Accuracy: 0.8761
Training loss (for one batch) at step 50: 220.6655, Accuracy: 0.8743
Training loss (for one batch) at step 60: 205.7396, Accuracy: 0.8716
Training loss (for one batch) at step 70: 211.4763, Accuracy: 0.8693
Training loss (for one batch) at step 80: 220.8356, Accuracy: 0.8674
Training loss (for one batch) at step 90: 209.3886, Accuracy: 0.8638
Training loss (for one batch) at step 100: 224.1730, Accuracy: 0.8638
Training loss (for one batch) at step 110: 198.9174, Accuracy: 0.8658
Training loss (for one batch) at step 120: 224.3681, Accuracy: 0.8648
Training loss (for one batch) at step 130: 220.3005, Accuracy: 0.8661
Training loss (for one batch) at step 140: 230.1022, Accuracy: 0.8649
---- Training ----
Training loss: 175.7946
Training acc over epoch: 0.8645
---- Validation ----
Validation loss: 51.7637
Validation acc: 0.7316
Time taken: 54.96s
Start of epoch 48
Training loss (for one batch) at step 0: 215.2122, Accuracy: 0.8300
Training loss (for one batch) at step 10: 190.4184, Accuracy: 0.8682
Training loss (for one batch) at step 20: 198.8093, Accuracy: 0.8695
Training loss (for one batch) at step 30: 206.8180, Accuracy: 0.8690
Training loss (for one batch) at step 40: 221.2284, Accuracy: 0.8698
Training loss (for one batch) at step 50: 224.2672, Accuracy: 0.8702
Training loss (for one batch) at step 60: 208.2581, Accuracy: 0.8723
Training loss (for one batch) at step 70: 208.7904, Accuracy: 0.8703
Training loss (for one batch) at step 80: 210.7225, Accuracy: 0.8693
Training loss (for one batch) at step 90: 200.7899, Accuracy: 0.8679
Training loss (for one batch) at step 100: 226.8988, Accuracy: 0.8659
Training loss (for one batch) at step 110: 198.1590, Accuracy: 0.8678
Training loss (for one batch) at step 120: 202.0061, Accuracy: 0.8683
Training loss (for one batch) at step 130: 206.7579, Accuracy: 0.8684
Training loss (for one batch) at step 140: 199.1331, Accuracy: 0.8680
---- Training ----
Training loss: 187.1591
Training acc over epoch: 0.8676
---- Validation ----
Validation loss: 81.6912
Validation acc: 0.7308
Time taken: 94.50s
Start of epoch 49
Training loss (for one batch) at step 0: 219.1730, Accuracy: 0.9000
Training loss (for one batch) at step 10: 208.4199, Accuracy: 0.8700
Training loss (for one batch) at step 20: 200.9400, Accuracy: 0.8786
Training loss (for one batch) at step 30: 214.1836, Accuracy: 0.8700
Training loss (for one batch) at step 40: 210.3602, Accuracy: 0.8744
Training loss (for one batch) at step 50: 207.3212, Accuracy: 0.8749
Training loss (for one batch) at step 60: 190.6501, Accuracy: 0.8748
Training loss (for one batch) at step 70: 209.0744, Accuracy: 0.8724
Training loss (for one batch) at step 80: 203.5255, Accuracy: 0.8690
Training loss (for one batch) at step 90: 212.0255, Accuracy: 0.8696
Training loss (for one batch) at step 100: 192.4191, Accuracy: 0.8690
Training loss (for one batch) at step 110: 203.2726, Accuracy: 0.8701
Training loss (for one batch) at step 120: 198.3244, Accuracy: 0.8698
Training loss (for one batch) at step 130: 206.4267, Accuracy: 0.8685
Training loss (for one batch) at step 140: 200.6184, Accuracy: 0.8671
---- Training ----
Training loss: 173.5077
Training acc over epoch: 0.8671
---- Validation ----
Validation loss: 88.4884
Validation acc: 0.7386
Time taken: 57.19s
===== Q: 0.0001
Validation acc: 0.7504
Validation AUC: 0.7482
Validation Balanced_ACC: 0.4844
Validation MI: 0.1390
Validation Normalized MI: 0.2076
Validation Adjusted MI: 0.2076
Validation aUc_Sklearn: 0.8305
Start of epoch 0
Training loss (for one batch) at step 0: 568.2831, Accuracy: 0.4900
Training loss (for one batch) at step 10: 437.3378, Accuracy: 0.5036
Training loss (for one batch) at step 20: 451.9300, Accuracy: 0.5438
Training loss (for one batch) at step 30: 486.5266, Accuracy: 0.5500
Training loss (for one batch) at step 40: 422.0882, Accuracy: 0.5534
Training loss (for one batch) at step 50: 466.7481, Accuracy: 0.5537
Training loss (for one batch) at step 60: 473.6014, Accuracy: 0.5564
Training loss (for one batch) at step 70: 416.8942, Accuracy: 0.5566
Training loss (for one batch) at step 80: 466.3800, Accuracy: 0.5563
Training loss (for one batch) at step 90: 472.3071, Accuracy: 0.5563
Training loss (for one batch) at step 100: 436.4700, Accuracy: 0.5597
Training loss (for one batch) at step 110: 406.7233, Accuracy: 0.5622
Training loss (for one batch) at step 120: 441.3499, Accuracy: 0.5639
Training loss (for one batch) at step 130: 458.7116, Accuracy: 0.5653
Training loss (for one batch) at step 140: 408.3489, Accuracy: 0.5642
---- Training ----
Training loss: 412.6354
Training acc over epoch: 0.5650
---- Validation ----
Validation loss: 111.9293
Validation acc: 0.5134
Time taken: 68.80s
Start of epoch 1
Training loss (for one batch) at step 0: 438.6027, Accuracy: 0.6200
Training loss (for one batch) at step 10: 412.2669, Accuracy: 0.6282
Training loss (for one batch) at step 20: 388.8326, Accuracy: 0.6171
Training loss (for one batch) at step 30: 438.9496, Accuracy: 0.6174
Training loss (for one batch) at step 40: 429.8932, Accuracy: 0.6178
Training loss (for one batch) at step 50: 406.5327, Accuracy: 0.6192
Training loss (for one batch) at step 60: 388.8469, Accuracy: 0.6185
Training loss (for one batch) at step 70: 400.0368, Accuracy: 0.6165
Training loss (for one batch) at step 80: 390.3627, Accuracy: 0.6175
Training loss (for one batch) at step 90: 415.1183, Accuracy: 0.6169
Training loss (for one batch) at step 100: 447.9932, Accuracy: 0.6175
Training loss (for one batch) at step 110: 377.0223, Accuracy: 0.6171
Training loss (for one batch) at step 120: 391.4781, Accuracy: 0.6158
Training loss (for one batch) at step 130: 407.4236, Accuracy: 0.6175
Training loss (for one batch) at step 140: 404.4418, Accuracy: 0.6172
---- Training ----
Training loss: 332.7355
Training acc over epoch: 0.6173
---- Validation ----
Validation loss: 116.7803
Validation acc: 0.5215
Time taken: 59.94s
Start of epoch 2
Training loss (for one batch) at step 0: 370.5181, Accuracy: 0.6300
Training loss (for one batch) at step 10: 383.8105, Accuracy: 0.6445
Training loss (for one batch) at step 20: 379.3936, Accuracy: 0.6414
Training loss (for one batch) at step 30: 357.8014, Accuracy: 0.6371
Training loss (for one batch) at step 40: 414.6072, Accuracy: 0.6439
Training loss (for one batch) at step 50: 414.7042, Accuracy: 0.6422
Training loss (for one batch) at step 60: 391.1946, Accuracy: 0.6387
Training loss (for one batch) at step 70: 405.6899, Accuracy: 0.6372
Training loss (for one batch) at step 80: 387.7918, Accuracy: 0.6375
Training loss (for one batch) at step 90: 381.6738, Accuracy: 0.6377
Training loss (for one batch) at step 100: 368.9249, Accuracy: 0.6350
Training loss (for one batch) at step 110: 357.1618, Accuracy: 0.6352
Training loss (for one batch) at step 120: 393.7811, Accuracy: 0.6361
Training loss (for one batch) at step 130: 403.5355, Accuracy: 0.6375
Training loss (for one batch) at step 140: 364.4939, Accuracy: 0.6384
---- Training ----
Training loss: 339.4365
Training acc over epoch: 0.6384
---- Validation ----
Validation loss: 72.5300
Validation acc: 0.6577
Time taken: 61.15s
Start of epoch 3
Training loss (for one batch) at step 0: 379.4445, Accuracy: 0.6400
Training loss (for one batch) at step 10: 372.8696, Accuracy: 0.6518
Training loss (for one batch) at step 20: 374.9841, Accuracy: 0.6519
Training loss (for one batch) at step 30: 347.1618, Accuracy: 0.6490
Training loss (for one batch) at step 40: 351.9590, Accuracy: 0.6551
Training loss (for one batch) at step 50: 358.9036, Accuracy: 0.6537
Training loss (for one batch) at step 60: 354.5307, Accuracy: 0.6513
Training loss (for one batch) at step 70: 336.0504, Accuracy: 0.6523
Training loss (for one batch) at step 80: 349.8038, Accuracy: 0.6494
Training loss (for one batch) at step 90: 353.6759, Accuracy: 0.6518
Training loss (for one batch) at step 100: 350.6976, Accuracy: 0.6525
Training loss (for one batch) at step 110: 364.6344, Accuracy: 0.6547
Training loss (for one batch) at step 120: 352.3576, Accuracy: 0.6553
Training loss (for one batch) at step 130: 349.2221, Accuracy: 0.6547
Training loss (for one batch) at step 140: 347.7408, Accuracy: 0.6541
---- Training ----
Training loss: 351.7236
Training acc over epoch: 0.6546
---- Validation ----
Validation loss: 68.7718
Validation acc: 0.6983
Time taken: 66.51s
Start of epoch 4
Training loss (for one batch) at step 0: 344.7408, Accuracy: 0.7200
Training loss (for one batch) at step 10: 364.4579, Accuracy: 0.6673
Training loss (for one batch) at step 20: 387.5877, Accuracy: 0.6681
Training loss (for one batch) at step 30: 386.8831, Accuracy: 0.6610
Training loss (for one batch) at step 40: 378.7969, Accuracy: 0.6556
Training loss (for one batch) at step 50: 369.3926, Accuracy: 0.6575
Training loss (for one batch) at step 60: 345.5716, Accuracy: 0.6598
Training loss (for one batch) at step 70: 338.2794, Accuracy: 0.6615
Training loss (for one batch) at step 80: 348.9991, Accuracy: 0.6625
Training loss (for one batch) at step 90: 353.6831, Accuracy: 0.6634
Training loss (for one batch) at step 100: 349.4561, Accuracy: 0.6627
Training loss (for one batch) at step 110: 375.4094, Accuracy: 0.6624
Training loss (for one batch) at step 120: 362.9272, Accuracy: 0.6609
Training loss (for one batch) at step 130: 357.0318, Accuracy: 0.6622
Training loss (for one batch) at step 140: 329.0805, Accuracy: 0.6644
---- Training ----
Training loss: 308.0832
Training acc over epoch: 0.6636
---- Validation ----
Validation loss: 75.3626
Validation acc: 0.7058
Time taken: 62.33s
Start of epoch 5
Training loss (for one batch) at step 0: 335.2616, Accuracy: 0.7400
Training loss (for one batch) at step 10: 346.1322, Accuracy: 0.6691
Training loss (for one batch) at step 20: 355.4135, Accuracy: 0.6771
Training loss (for one batch) at step 30: 351.8374, Accuracy: 0.6797
Training loss (for one batch) at step 40: 351.6768, Accuracy: 0.6741
Training loss (for one batch) at step 50: 356.8464, Accuracy: 0.6780
Training loss (for one batch) at step 60: 325.4394, Accuracy: 0.6805
Training loss (for one batch) at step 70: 350.4631, Accuracy: 0.6832
Training loss (for one batch) at step 80: 337.1670, Accuracy: 0.6804
Training loss (for one batch) at step 90: 316.0183, Accuracy: 0.6793
Training loss (for one batch) at step 100: 336.9574, Accuracy: 0.6781
Training loss (for one batch) at step 110: 328.0048, Accuracy: 0.6787
Training loss (for one batch) at step 120: 332.7192, Accuracy: 0.6779
Training loss (for one batch) at step 130: 345.3898, Accuracy: 0.6786
Training loss (for one batch) at step 140: 373.0555, Accuracy: 0.6795
---- Training ----
Training loss: 313.9550
Training acc over epoch: 0.6789
---- Validation ----
Validation loss: 79.0183
Validation acc: 0.6929
Time taken: 61.90s
Start of epoch 6
Training loss (for one batch) at step 0: 348.7469, Accuracy: 0.6800
Training loss (for one batch) at step 10: 344.4017, Accuracy: 0.6973
Training loss (for one batch) at step 20: 350.7155, Accuracy: 0.6924
Training loss (for one batch) at step 30: 335.3892, Accuracy: 0.6939
Training loss (for one batch) at step 40: 321.6350, Accuracy: 0.6951
Training loss (for one batch) at step 50: 332.8470, Accuracy: 0.6961
Training loss (for one batch) at step 60: 323.5509, Accuracy: 0.7005
Training loss (for one batch) at step 70: 350.9400, Accuracy: 0.7015
Training loss (for one batch) at step 80: 341.6899, Accuracy: 0.7002
Training loss (for one batch) at step 90: 320.9451, Accuracy: 0.7012
Training loss (for one batch) at step 100: 344.4136, Accuracy: 0.6985
Training loss (for one batch) at step 110: 343.0099, Accuracy: 0.7003
Training loss (for one batch) at step 120: 342.8688, Accuracy: 0.6997
Training loss (for one batch) at step 130: 321.8667, Accuracy: 0.7015
Training loss (for one batch) at step 140: 322.1480, Accuracy: 0.7021
---- Training ----
Training loss: 314.0279
Training acc over epoch: 0.7001
---- Validation ----
Validation loss: 75.6584
Validation acc: 0.7037
Time taken: 61.02s
Start of epoch 7
Training loss (for one batch) at step 0: 313.5968, Accuracy: 0.6900
Training loss (for one batch) at step 10: 322.5370, Accuracy: 0.7191
Training loss (for one batch) at step 20: 341.3511, Accuracy: 0.7205
Training loss (for one batch) at step 30: 335.2746, Accuracy: 0.7203
Training loss (for one batch) at step 40: 301.4094, Accuracy: 0.7195
Training loss (for one batch) at step 50: 323.3734, Accuracy: 0.7180
Training loss (for one batch) at step 60: 331.7534, Accuracy: 0.7144
Training loss (for one batch) at step 70: 355.3971, Accuracy: 0.7148
Training loss (for one batch) at step 80: 336.0733, Accuracy: 0.7119
Training loss (for one batch) at step 90: 319.7128, Accuracy: 0.7116
Training loss (for one batch) at step 100: 326.7942, Accuracy: 0.7129
Training loss (for one batch) at step 110: 329.0853, Accuracy: 0.7158
Training loss (for one batch) at step 120: 326.3042, Accuracy: 0.7138
Training loss (for one batch) at step 130: 305.1540, Accuracy: 0.7144
Training loss (for one batch) at step 140: 329.6453, Accuracy: 0.7143
---- Training ----
Training loss: 273.9512
Training acc over epoch: 0.7147
---- Validation ----
Validation loss: 72.1647
Validation acc: 0.7071
Time taken: 65.00s
Start of epoch 8
Training loss (for one batch) at step 0: 309.6161, Accuracy: 0.7600
Training loss (for one batch) at step 10: 313.6886, Accuracy: 0.7436
Training loss (for one batch) at step 20: 322.5263, Accuracy: 0.7362
Training loss (for one batch) at step 30: 329.0206, Accuracy: 0.7242
Training loss (for one batch) at step 40: 315.5564, Accuracy: 0.7276
Training loss (for one batch) at step 50: 311.4828, Accuracy: 0.7278
Training loss (for one batch) at step 60: 319.9177, Accuracy: 0.7270
Training loss (for one batch) at step 70: 317.6866, Accuracy: 0.7277
Training loss (for one batch) at step 80: 320.6369, Accuracy: 0.7228
Training loss (for one batch) at step 90: 327.1533, Accuracy: 0.7236
Training loss (for one batch) at step 100: 322.1693, Accuracy: 0.7230
Training loss (for one batch) at step 110: 328.1114, Accuracy: 0.7219
Training loss (for one batch) at step 120: 321.6311, Accuracy: 0.7224
Training loss (for one batch) at step 130: 297.9505, Accuracy: 0.7240
Training loss (for one batch) at step 140: 335.2967, Accuracy: 0.7254
---- Training ----
Training loss: 286.4736
Training acc over epoch: 0.7260
---- Validation ----
Validation loss: 62.7328
Validation acc: 0.7157
Time taken: 61.32s
Start of epoch 9
Training loss (for one batch) at step 0: 318.9900, Accuracy: 0.7700
Training loss (for one batch) at step 10: 302.2181, Accuracy: 0.7309
Training loss (for one batch) at step 20: 325.4204, Accuracy: 0.7357
Training loss (for one batch) at step 30: 309.2781, Accuracy: 0.7358
Training loss (for one batch) at step 40: 298.2429, Accuracy: 0.7307
Training loss (for one batch) at step 50: 302.3362, Accuracy: 0.7345
Training loss (for one batch) at step 60: 299.2465, Accuracy: 0.7364
Training loss (for one batch) at step 70: 321.8457, Accuracy: 0.7370
Training loss (for one batch) at step 80: 316.0764, Accuracy: 0.7367
Training loss (for one batch) at step 90: 341.8124, Accuracy: 0.7335
Training loss (for one batch) at step 100: 308.9229, Accuracy: 0.7327
Training loss (for one batch) at step 110: 303.6467, Accuracy: 0.7336
Training loss (for one batch) at step 120: 299.8018, Accuracy: 0.7350
Training loss (for one batch) at step 130: 300.6668, Accuracy: 0.7373
Training loss (for one batch) at step 140: 315.6631, Accuracy: 0.7362
---- Training ----
Training loss: 278.0719
Training acc over epoch: 0.7366
---- Validation ----
Validation loss: 61.5623
Validation acc: 0.7198
Time taken: 64.97s
Start of epoch 10
Training loss (for one batch) at step 0: 326.9460, Accuracy: 0.7400
Training loss (for one batch) at step 10: 320.6804, Accuracy: 0.7473
Training loss (for one batch) at step 20: 300.0578, Accuracy: 0.7443
Training loss (for one batch) at step 30: 314.2905, Accuracy: 0.7397
Training loss (for one batch) at step 40: 293.0572, Accuracy: 0.7407
Training loss (for one batch) at step 50: 323.0417, Accuracy: 0.7422
Training loss (for one batch) at step 60: 299.3143, Accuracy: 0.7448
Training loss (for one batch) at step 70: 303.9180, Accuracy: 0.7432
Training loss (for one batch) at step 80: 316.0914, Accuracy: 0.7430
Training loss (for one batch) at step 90: 321.6951, Accuracy: 0.7421
Training loss (for one batch) at step 100: 301.6747, Accuracy: 0.7414
Training loss (for one batch) at step 110: 309.5676, Accuracy: 0.7418
Training loss (for one batch) at step 120: 322.7063, Accuracy: 0.7419
Training loss (for one batch) at step 130: 298.5293, Accuracy: 0.7414
Training loss (for one batch) at step 140: 291.1007, Accuracy: 0.7416
---- Training ----
Training loss: 274.7017
Training acc over epoch: 0.7414
---- Validation ----
Validation loss: 80.6227
Validation acc: 0.7303
Time taken: 60.04s
Start of epoch 11
Training loss (for one batch) at step 0: 304.2397, Accuracy: 0.6900
Training loss (for one batch) at step 10: 293.8421, Accuracy: 0.7509
Training loss (for one batch) at step 20: 319.0310, Accuracy: 0.7605
Training loss (for one batch) at step 30: 312.1626, Accuracy: 0.7494
Training loss (for one batch) at step 40: 312.9362, Accuracy: 0.7483
Training loss (for one batch) at step 50: 297.4744, Accuracy: 0.7539
Training loss (for one batch) at step 60: 285.4249, Accuracy: 0.7523
Training loss (for one batch) at step 70: 311.0377, Accuracy: 0.7530
Training loss (for one batch) at step 80: 302.7659, Accuracy: 0.7509
Training loss (for one batch) at step 90: 289.6555, Accuracy: 0.7499
Training loss (for one batch) at step 100: 308.4319, Accuracy: 0.7509
Training loss (for one batch) at step 110: 292.1535, Accuracy: 0.7523
Training loss (for one batch) at step 120: 304.6355, Accuracy: 0.7506
Training loss (for one batch) at step 130: 320.2207, Accuracy: 0.7511
Training loss (for one batch) at step 140: 321.7360, Accuracy: 0.7532
---- Training ----
Training loss: 272.6035
Training acc over epoch: 0.7538
---- Validation ----
Validation loss: 65.0247
Validation acc: 0.7362
Time taken: 63.73s
Start of epoch 12
Training loss (for one batch) at step 0: 327.2354, Accuracy: 0.7400
Training loss (for one batch) at step 10: 288.0884, Accuracy: 0.7518
Training loss (for one batch) at step 20: 294.4939, Accuracy: 0.7643
Training loss (for one batch) at step 30: 301.4861, Accuracy: 0.7587
Training loss (for one batch) at step 40: 299.5453, Accuracy: 0.7571
Training loss (for one batch) at step 50: 308.1908, Accuracy: 0.7578
Training loss (for one batch) at step 60: 309.6224, Accuracy: 0.7590
Training loss (for one batch) at step 70: 313.6192, Accuracy: 0.7580
Training loss (for one batch) at step 80: 293.9702, Accuracy: 0.7593
Training loss (for one batch) at step 90: 316.3510, Accuracy: 0.7570
Training loss (for one batch) at step 100: 308.5866, Accuracy: 0.7583
Training loss (for one batch) at step 110: 292.3713, Accuracy: 0.7589
Training loss (for one batch) at step 120: 317.2153, Accuracy: 0.7588
Training loss (for one batch) at step 130: 302.3272, Accuracy: 0.7579
Training loss (for one batch) at step 140: 302.9681, Accuracy: 0.7591
---- Training ----
Training loss: 278.3014
Training acc over epoch: 0.7599
---- Validation ----
Validation loss: 74.6865
Validation acc: 0.7101
Time taken: 60.17s
Start of epoch 13
Training loss (for one batch) at step 0: 305.9536, Accuracy: 0.7800
Training loss (for one batch) at step 10: 292.5565, Accuracy: 0.7627
Training loss (for one batch) at step 20: 310.2082, Accuracy: 0.7733
Training loss (for one batch) at step 30: 291.7585, Accuracy: 0.7703
Training loss (for one batch) at step 40: 305.9544, Accuracy: 0.7698
Training loss (for one batch) at step 50: 279.9608, Accuracy: 0.7739
Training loss (for one batch) at step 60: 283.8403, Accuracy: 0.7725
Training loss (for one batch) at step 70: 294.2223, Accuracy: 0.7721
Training loss (for one batch) at step 80: 293.4728, Accuracy: 0.7736
Training loss (for one batch) at step 90: 292.2520, Accuracy: 0.7719
Training loss (for one batch) at step 100: 286.7873, Accuracy: 0.7711
Training loss (for one batch) at step 110: 290.5057, Accuracy: 0.7738
Training loss (for one batch) at step 120: 298.0457, Accuracy: 0.7723
Training loss (for one batch) at step 130: 290.2370, Accuracy: 0.7704
Training loss (for one batch) at step 140: 311.6062, Accuracy: 0.7718
---- Training ----
Training loss: 274.8362
Training acc over epoch: 0.7710
---- Validation ----
Validation loss: 75.4801
Validation acc: 0.7410
Time taken: 64.29s
Start of epoch 14
Training loss (for one batch) at step 0: 291.9652, Accuracy: 0.7700
Training loss (for one batch) at step 10: 294.9038, Accuracy: 0.7945
Training loss (for one batch) at step 20: 317.3183, Accuracy: 0.7829
Training loss (for one batch) at step 30: 302.1375, Accuracy: 0.7752
Training loss (for one batch) at step 40: 311.7945, Accuracy: 0.7761
Training loss (for one batch) at step 50: 289.2685, Accuracy: 0.7776
Training loss (for one batch) at step 60: 291.3695, Accuracy: 0.7733
Training loss (for one batch) at step 70: 296.4028, Accuracy: 0.7731
Training loss (for one batch) at step 80: 284.3185, Accuracy: 0.7760
Training loss (for one batch) at step 90: 299.1676, Accuracy: 0.7749
Training loss (for one batch) at step 100: 298.5912, Accuracy: 0.7729
Training loss (for one batch) at step 110: 303.7232, Accuracy: 0.7734
Training loss (for one batch) at step 120: 288.7953, Accuracy: 0.7727
Training loss (for one batch) at step 130: 307.7017, Accuracy: 0.7729
Training loss (for one batch) at step 140: 296.4705, Accuracy: 0.7743
---- Training ----
Training loss: 255.1895
Training acc over epoch: 0.7753
---- Validation ----
Validation loss: 70.6533
Validation acc: 0.7340
Time taken: 61.16s
Start of epoch 15
Training loss (for one batch) at step 0: 293.9965, Accuracy: 0.8000
Training loss (for one batch) at step 10: 294.9586, Accuracy: 0.7655
Training loss (for one batch) at step 20: 289.5933, Accuracy: 0.7752
Training loss (for one batch) at step 30: 282.5462, Accuracy: 0.7771
Training loss (for one batch) at step 40: 302.3198, Accuracy: 0.7702
Training loss (for one batch) at step 50: 278.3625, Accuracy: 0.7765
Training loss (for one batch) at step 60: 293.8387, Accuracy: 0.7785
Training loss (for one batch) at step 70: 297.6463, Accuracy: 0.7769
Training loss (for one batch) at step 80: 297.8519, Accuracy: 0.7781
Training loss (for one batch) at step 90: 304.5286, Accuracy: 0.7777
Training loss (for one batch) at step 100: 300.4453, Accuracy: 0.7790
Training loss (for one batch) at step 110: 281.3574, Accuracy: 0.7803
Training loss (for one batch) at step 120: 287.2178, Accuracy: 0.7799
Training loss (for one batch) at step 130: 289.8464, Accuracy: 0.7795
Training loss (for one batch) at step 140: 292.2887, Accuracy: 0.7811
---- Training ----
Training loss: 252.5204
Training acc over epoch: 0.7803
---- Validation ----
Validation loss: 69.4193
Validation acc: 0.7324
Time taken: 64.52s
Start of epoch 16
Training loss (for one batch) at step 0: 295.8695, Accuracy: 0.7600
Training loss (for one batch) at step 10: 286.6907, Accuracy: 0.8009
Training loss (for one batch) at step 20: 286.0630, Accuracy: 0.7948
Training loss (for one batch) at step 30: 299.8842, Accuracy: 0.7868
Training loss (for one batch) at step 40: 297.5775, Accuracy: 0.7861
Training loss (for one batch) at step 50: 274.2227, Accuracy: 0.7904
Training loss (for one batch) at step 60: 298.6975, Accuracy: 0.7890
Training loss (for one batch) at step 70: 278.7169, Accuracy: 0.7906
Training loss (for one batch) at step 80: 282.7876, Accuracy: 0.7904
Training loss (for one batch) at step 90: 279.3397, Accuracy: 0.7899
Training loss (for one batch) at step 100: 281.4591, Accuracy: 0.7869
Training loss (for one batch) at step 110: 285.3374, Accuracy: 0.7877
Training loss (for one batch) at step 120: 286.2715, Accuracy: 0.7865
Training loss (for one batch) at step 130: 296.9460, Accuracy: 0.7868
Training loss (for one batch) at step 140: 298.2442, Accuracy: 0.7862
---- Training ----
Training loss: 243.2220
Training acc over epoch: 0.7866
---- Validation ----
Validation loss: 69.1680
Validation acc: 0.7501
Time taken: 68.49s
Start of epoch 17
Training loss (for one batch) at step 0: 287.6788, Accuracy: 0.7700
Training loss (for one batch) at step 10: 300.5267, Accuracy: 0.7909
Training loss (for one batch) at step 20: 277.8903, Accuracy: 0.7876
Training loss (for one batch) at step 30: 282.7995, Accuracy: 0.7800
Training loss (for one batch) at step 40: 279.7767, Accuracy: 0.7834
Training loss (for one batch) at step 50: 278.8689, Accuracy: 0.7892
Training loss (for one batch) at step 60: 278.0653, Accuracy: 0.7903
Training loss (for one batch) at step 70: 286.2449, Accuracy: 0.7900
Training loss (for one batch) at step 80: 293.3292, Accuracy: 0.7891
Training loss (for one batch) at step 90: 264.5290, Accuracy: 0.7878
Training loss (for one batch) at step 100: 276.8055, Accuracy: 0.7860
Training loss (for one batch) at step 110: 287.2839, Accuracy: 0.7859
Training loss (for one batch) at step 120: 309.3788, Accuracy: 0.7856
Training loss (for one batch) at step 130: 290.3387, Accuracy: 0.7865
Training loss (for one batch) at step 140: 288.9032, Accuracy: 0.7864
---- Training ----
Training loss: 265.9966
Training acc over epoch: 0.7867
---- Validation ----
Validation loss: 67.6321
Validation acc: 0.7397
Time taken: 65.69s
Start of epoch 18
Training loss (for one batch) at step 0: 287.2849, Accuracy: 0.7500
Training loss (for one batch) at step 10: 279.0876, Accuracy: 0.7936
Training loss (for one batch) at step 20: 283.4956, Accuracy: 0.7910
Training loss (for one batch) at step 30: 282.6040, Accuracy: 0.7823
Training loss (for one batch) at step 40: 297.4592, Accuracy: 0.7846
Training loss (for one batch) at step 50: 256.6135, Accuracy: 0.7927
Training loss (for one batch) at step 60: 285.0305, Accuracy: 0.7933
Training loss (for one batch) at step 70: 257.3356, Accuracy: 0.7949
Training loss (for one batch) at step 80: 273.1933, Accuracy: 0.7951
Training loss (for one batch) at step 90: 292.5005, Accuracy: 0.7940
Training loss (for one batch) at step 100: 285.5302, Accuracy: 0.7932
Training loss (for one batch) at step 110: 287.1104, Accuracy: 0.7944
Training loss (for one batch) at step 120: 270.4535, Accuracy: 0.7943
Training loss (for one batch) at step 130: 288.0115, Accuracy: 0.7952
Training loss (for one batch) at step 140: 265.4819, Accuracy: 0.7957
---- Training ----
Training loss: 263.6748
Training acc over epoch: 0.7943
---- Validation ----
Validation loss: 71.4707
Validation acc: 0.7254
Time taken: 63.41s
Start of epoch 19
Training loss (for one batch) at step 0: 276.2089, Accuracy: 0.8400
Training loss (for one batch) at step 10: 266.6432, Accuracy: 0.7982
Training loss (for one batch) at step 20: 269.7543, Accuracy: 0.7990
Training loss (for one batch) at step 30: 287.0071, Accuracy: 0.7913
Training loss (for one batch) at step 40: 268.1562, Accuracy: 0.7961
Training loss (for one batch) at step 50: 283.4767, Accuracy: 0.7975
Training loss (for one batch) at step 60: 270.7737, Accuracy: 0.7934
Training loss (for one batch) at step 70: 280.0797, Accuracy: 0.7961
Training loss (for one batch) at step 80: 273.4070, Accuracy: 0.7931
Training loss (for one batch) at step 90: 281.4162, Accuracy: 0.7935
Training loss (for one batch) at step 100: 265.3513, Accuracy: 0.7934
Training loss (for one batch) at step 110: 285.8152, Accuracy: 0.7941
Training loss (for one batch) at step 120: 269.8726, Accuracy: 0.7954
Training loss (for one batch) at step 130: 295.2633, Accuracy: 0.7947
Training loss (for one batch) at step 140: 290.7611, Accuracy: 0.7940
---- Training ----
Training loss: 234.0087
Training acc over epoch: 0.7940
---- Validation ----
Validation loss: 68.8456
Validation acc: 0.7491
Time taken: 65.90s
Start of epoch 20
Training loss (for one batch) at step 0: 264.8388, Accuracy: 0.8000
Training loss (for one batch) at step 10: 263.9401, Accuracy: 0.8100
Training loss (for one batch) at step 20: 287.3592, Accuracy: 0.8052
Training loss (for one batch) at step 30: 288.9638, Accuracy: 0.8068
Training loss (for one batch) at step 40: 294.9315, Accuracy: 0.8007
Training loss (for one batch) at step 50: 270.0282, Accuracy: 0.7992
Training loss (for one batch) at step 60: 269.6413, Accuracy: 0.8007
Training loss (for one batch) at step 70: 270.5291, Accuracy: 0.7996
Training loss (for one batch) at step 80: 300.7979, Accuracy: 0.7956
Training loss (for one batch) at step 90: 280.7478, Accuracy: 0.7943
Training loss (for one batch) at step 100: 267.3016, Accuracy: 0.7956
Training loss (for one batch) at step 110: 262.2457, Accuracy: 0.7960
Training loss (for one batch) at step 120: 271.8864, Accuracy: 0.7983
Training loss (for one batch) at step 130: 279.3852, Accuracy: 0.7962
Training loss (for one batch) at step 140: 257.5609, Accuracy: 0.7972
---- Training ----
Training loss: 232.2364
Training acc over epoch: 0.7976
---- Validation ----
Validation loss: 63.3904
Validation acc: 0.7311
Time taken: 60.80s
Start of epoch 21
Training loss (for one batch) at step 0: 275.0459, Accuracy: 0.8300
Training loss (for one batch) at step 10: 266.6565, Accuracy: 0.8236
Training loss (for one batch) at step 20: 275.9676, Accuracy: 0.8095
Training loss (for one batch) at step 30: 275.9047, Accuracy: 0.8061
Training loss (for one batch) at step 40: 274.3126, Accuracy: 0.8041
Training loss (for one batch) at step 50: 259.8558, Accuracy: 0.8049
Training loss (for one batch) at step 60: 258.5053, Accuracy: 0.8105
Training loss (for one batch) at step 70: 264.7143, Accuracy: 0.8099
Training loss (for one batch) at step 80: 281.0883, Accuracy: 0.8072
Training loss (for one batch) at step 90: 256.6648, Accuracy: 0.8090
Training loss (for one batch) at step 100: 270.2168, Accuracy: 0.8075
Training loss (for one batch) at step 110: 250.5022, Accuracy: 0.8107
Training loss (for one batch) at step 120: 270.2244, Accuracy: 0.8088
Training loss (for one batch) at step 130: 265.8899, Accuracy: 0.8086
Training loss (for one batch) at step 140: 270.2143, Accuracy: 0.8083
---- Training ----
Training loss: 245.8899
Training acc over epoch: 0.8082
---- Validation ----
Validation loss: 65.7369
Validation acc: 0.7456
Time taken: 64.39s
Start of epoch 22
Training loss (for one batch) at step 0: 276.1351, Accuracy: 0.7500
Training loss (for one batch) at step 10: 291.4082, Accuracy: 0.7955
Training loss (for one batch) at step 20: 268.5733, Accuracy: 0.8090
Training loss (for one batch) at step 30: 254.4498, Accuracy: 0.8129
Training loss (for one batch) at step 40: 265.7195, Accuracy: 0.8102
Training loss (for one batch) at step 50: 266.6225, Accuracy: 0.8092
Training loss (for one batch) at step 60: 263.0423, Accuracy: 0.8105
Training loss (for one batch) at step 70: 262.4725, Accuracy: 0.8101
Training loss (for one batch) at step 80: 262.0826, Accuracy: 0.8059
Training loss (for one batch) at step 90: 261.9514, Accuracy: 0.8082
Training loss (for one batch) at step 100: 272.9232, Accuracy: 0.8061
Training loss (for one batch) at step 110: 261.0333, Accuracy: 0.8063
Training loss (for one batch) at step 120: 278.7687, Accuracy: 0.8080
Training loss (for one batch) at step 130: 263.6119, Accuracy: 0.8079
Training loss (for one batch) at step 140: 251.8516, Accuracy: 0.8079
---- Training ----
Training loss: 225.3245
Training acc over epoch: 0.8082
---- Validation ----
Validation loss: 86.2905
Validation acc: 0.7324
Time taken: 59.72s
Start of epoch 23
Training loss (for one batch) at step 0: 256.5095, Accuracy: 0.8300
Training loss (for one batch) at step 10: 272.0696, Accuracy: 0.8309
Training loss (for one batch) at step 20: 270.7391, Accuracy: 0.8281
Training loss (for one batch) at step 30: 269.6561, Accuracy: 0.8200
Training loss (for one batch) at step 40: 236.1044, Accuracy: 0.8166
Training loss (for one batch) at step 50: 260.8615, Accuracy: 0.8173
Training loss (for one batch) at step 60: 271.0296, Accuracy: 0.8166
Training loss (for one batch) at step 70: 253.4761, Accuracy: 0.8203
Training loss (for one batch) at step 80: 272.0312, Accuracy: 0.8153
Training loss (for one batch) at step 90: 292.8952, Accuracy: 0.8127
Training loss (for one batch) at step 100: 248.8977, Accuracy: 0.8121
Training loss (for one batch) at step 110: 267.0077, Accuracy: 0.8134
Training loss (for one batch) at step 120: 260.1814, Accuracy: 0.8131
Training loss (for one batch) at step 130: 271.7419, Accuracy: 0.8122
Training loss (for one batch) at step 140: 262.2242, Accuracy: 0.8125
---- Training ----
Training loss: 235.3381
Training acc over epoch: 0.8125
---- Validation ----
Validation loss: 72.8500
Validation acc: 0.7372
Time taken: 64.07s
Start of epoch 24
Training loss (for one batch) at step 0: 267.8811, Accuracy: 0.8300
Training loss (for one batch) at step 10: 247.8776, Accuracy: 0.8091
Training loss (for one batch) at step 20: 267.8331, Accuracy: 0.8133
Training loss (for one batch) at step 30: 261.3820, Accuracy: 0.8061
Training loss (for one batch) at step 40: 270.7435, Accuracy: 0.8115
Training loss (for one batch) at step 50: 251.7231, Accuracy: 0.8116
Training loss (for one batch) at step 60: 256.8279, Accuracy: 0.8072
Training loss (for one batch) at step 70: 245.6734, Accuracy: 0.8106
Training loss (for one batch) at step 80: 268.0728, Accuracy: 0.8102
Training loss (for one batch) at step 90: 287.5371, Accuracy: 0.8102
Training loss (for one batch) at step 100: 261.7558, Accuracy: 0.8102
Training loss (for one batch) at step 110: 266.9334, Accuracy: 0.8087
Training loss (for one batch) at step 120: 249.5676, Accuracy: 0.8102
Training loss (for one batch) at step 130: 251.9732, Accuracy: 0.8105
Training loss (for one batch) at step 140: 249.8629, Accuracy: 0.8098
---- Training ----
Training loss: 233.2699
Training acc over epoch: 0.8099
---- Validation ----
Validation loss: 76.3288
Validation acc: 0.7367
Time taken: 60.78s
Start of epoch 25
Training loss (for one batch) at step 0: 249.4137, Accuracy: 0.8200
Training loss (for one batch) at step 10: 253.2252, Accuracy: 0.8291
Training loss (for one batch) at step 20: 255.6759, Accuracy: 0.8224
Training loss (for one batch) at step 30: 242.6517, Accuracy: 0.8165
Training loss (for one batch) at step 40: 253.5451, Accuracy: 0.8210
Training loss (for one batch) at step 50: 248.8409, Accuracy: 0.8206
Training loss (for one batch) at step 60: 252.3577, Accuracy: 0.8197
Training loss (for one batch) at step 70: 276.0668, Accuracy: 0.8177
Training loss (for one batch) at step 80: 263.3936, Accuracy: 0.8159
Training loss (for one batch) at step 90: 242.4599, Accuracy: 0.8176
Training loss (for one batch) at step 100: 249.0356, Accuracy: 0.8172
Training loss (for one batch) at step 110: 259.4251, Accuracy: 0.8173
Training loss (for one batch) at step 120: 243.3459, Accuracy: 0.8175
Training loss (for one batch) at step 130: 290.0939, Accuracy: 0.8167
Training loss (for one batch) at step 140: 271.7534, Accuracy: 0.8184
---- Training ----
Training loss: 234.4030
Training acc over epoch: 0.8180
---- Validation ----
Validation loss: 83.1037
Validation acc: 0.7426
Time taken: 64.01s
Start of epoch 26
Training loss (for one batch) at step 0: 260.2676, Accuracy: 0.7700
Training loss (for one batch) at step 10: 245.9767, Accuracy: 0.8227
Training loss (for one batch) at step 20: 247.8090, Accuracy: 0.8233
Training loss (for one batch) at step 30: 270.2077, Accuracy: 0.8135
Training loss (for one batch) at step 40: 256.2837, Accuracy: 0.8163
Training loss (for one batch) at step 50: 249.5801, Accuracy: 0.8210
Training loss (for one batch) at step 60: 256.0130, Accuracy: 0.8200
Training loss (for one batch) at step 70: 244.1199, Accuracy: 0.8213
Training loss (for one batch) at step 80: 264.7918, Accuracy: 0.8183
Training loss (for one batch) at step 90: 269.7300, Accuracy: 0.8166
Training loss (for one batch) at step 100: 240.6131, Accuracy: 0.8177
Training loss (for one batch) at step 110: 243.9978, Accuracy: 0.8193
Training loss (for one batch) at step 120: 254.1115, Accuracy: 0.8183
Training loss (for one batch) at step 130: 273.2556, Accuracy: 0.8187
Training loss (for one batch) at step 140: 259.5561, Accuracy: 0.8187
---- Training ----
Training loss: 222.6881
Training acc over epoch: 0.8180
---- Validation ----
Validation loss: 74.0773
Validation acc: 0.7448
Time taken: 60.56s
Start of epoch 27
Training loss (for one batch) at step 0: 232.2290, Accuracy: 0.8600
Training loss (for one batch) at step 10: 261.5714, Accuracy: 0.8245
Training loss (for one batch) at step 20: 243.0491, Accuracy: 0.8248
Training loss (for one batch) at step 30: 251.9718, Accuracy: 0.8284
Training loss (for one batch) at step 40: 249.7420, Accuracy: 0.8271
Training loss (for one batch) at step 50: 238.1377, Accuracy: 0.8284
Training loss (for one batch) at step 60: 230.1080, Accuracy: 0.8289
Training loss (for one batch) at step 70: 252.6935, Accuracy: 0.8300
Training loss (for one batch) at step 80: 267.6059, Accuracy: 0.8259
Training loss (for one batch) at step 90: 237.7122, Accuracy: 0.8259
Training loss (for one batch) at step 100: 265.2425, Accuracy: 0.8252
Training loss (for one batch) at step 110: 238.5276, Accuracy: 0.8252
Training loss (for one batch) at step 120: 249.6782, Accuracy: 0.8255
Training loss (for one batch) at step 130: 272.7804, Accuracy: 0.8250
Training loss (for one batch) at step 140: 250.0237, Accuracy: 0.8245
---- Training ----
Training loss: 220.8620
Training acc over epoch: 0.8251
---- Validation ----
Validation loss: 69.8786
Validation acc: 0.7370
Time taken: 65.37s
Start of epoch 28
Training loss (for one batch) at step 0: 267.0614, Accuracy: 0.7500
Training loss (for one batch) at step 10: 231.5706, Accuracy: 0.8309
Training loss (for one batch) at step 20: 251.8730, Accuracy: 0.8248
Training loss (for one batch) at step 30: 263.0667, Accuracy: 0.8184
Training loss (for one batch) at step 40: 231.5031, Accuracy: 0.8266
Training loss (for one batch) at step 50: 245.4556, Accuracy: 0.8271
Training loss (for one batch) at step 60: 251.4777, Accuracy: 0.8257
Training loss (for one batch) at step 70: 264.3494, Accuracy: 0.8246
Training loss (for one batch) at step 80: 244.8515, Accuracy: 0.8247
Training loss (for one batch) at step 90: 227.7712, Accuracy: 0.8247
Training loss (for one batch) at step 100: 259.4113, Accuracy: 0.8233
Training loss (for one batch) at step 110: 238.0053, Accuracy: 0.8258
Training loss (for one batch) at step 120: 249.3116, Accuracy: 0.8261
Training loss (for one batch) at step 130: 262.4524, Accuracy: 0.8253
Training loss (for one batch) at step 140: 243.7264, Accuracy: 0.8235
---- Training ----
Training loss: 213.0100
Training acc over epoch: 0.8242
---- Validation ----
Validation loss: 72.2446
Validation acc: 0.7458
Time taken: 61.85s
Start of epoch 29
Training loss (for one batch) at step 0: 238.2764, Accuracy: 0.8600
Training loss (for one batch) at step 10: 239.8331, Accuracy: 0.8291
Training loss (for one batch) at step 20: 265.0829, Accuracy: 0.8281
Training loss (for one batch) at step 30: 246.6732, Accuracy: 0.8281
Training loss (for one batch) at step 40: 237.7912, Accuracy: 0.8288
Training loss (for one batch) at step 50: 256.3950, Accuracy: 0.8304
Training loss (for one batch) at step 60: 260.2420, Accuracy: 0.8321
Training loss (for one batch) at step 70: 257.2775, Accuracy: 0.8299
Training loss (for one batch) at step 80: 287.9333, Accuracy: 0.8256
Training loss (for one batch) at step 90: 240.2014, Accuracy: 0.8267
Training loss (for one batch) at step 100: 240.8687, Accuracy: 0.8263
Training loss (for one batch) at step 110: 255.1096, Accuracy: 0.8275
Training loss (for one batch) at step 120: 255.1806, Accuracy: 0.8279
Training loss (for one batch) at step 130: 272.5242, Accuracy: 0.8274
Training loss (for one batch) at step 140: 251.8438, Accuracy: 0.8293
---- Training ----
Training loss: 225.0919
Training acc over epoch: 0.8287
---- Validation ----
Validation loss: 79.6905
Validation acc: 0.7289
Time taken: 64.75s
Start of epoch 30
Training loss (for one batch) at step 0: 247.9711, Accuracy: 0.8100
Training loss (for one batch) at step 10: 259.7964, Accuracy: 0.8427
Training loss (for one batch) at step 20: 244.8900, Accuracy: 0.8357
Training loss (for one batch) at step 30: 254.7878, Accuracy: 0.8277
Training loss (for one batch) at step 40: 215.8607, Accuracy: 0.8285
Training loss (for one batch) at step 50: 225.9349, Accuracy: 0.8345
Training loss (for one batch) at step 60: 248.0441, Accuracy: 0.8339
Training loss (for one batch) at step 70: 246.8478, Accuracy: 0.8318
Training loss (for one batch) at step 80: 244.3359, Accuracy: 0.8304
Training loss (for one batch) at step 90: 250.8904, Accuracy: 0.8297
Training loss (for one batch) at step 100: 247.1734, Accuracy: 0.8290
Training loss (for one batch) at step 110: 267.8675, Accuracy: 0.8288
Training loss (for one batch) at step 120: 249.0680, Accuracy: 0.8289
Training loss (for one batch) at step 130: 265.7591, Accuracy: 0.8283
Training loss (for one batch) at step 140: 241.6664, Accuracy: 0.8278
---- Training ----
Training loss: 211.0540
Training acc over epoch: 0.8275
---- Validation ----
Validation loss: 63.9111
Validation acc: 0.7421
Time taken: 60.18s
Start of epoch 31
Training loss (for one batch) at step 0: 233.4685, Accuracy: 0.8200
Training loss (for one batch) at step 10: 228.6099, Accuracy: 0.8373
Training loss (for one batch) at step 20: 243.1700, Accuracy: 0.8357
Training loss (for one batch) at step 30: 263.4500, Accuracy: 0.8329
Training loss (for one batch) at step 40: 246.0653, Accuracy: 0.8354
Training loss (for one batch) at step 50: 235.3599, Accuracy: 0.8392
Training loss (for one batch) at step 60: 213.6123, Accuracy: 0.8364
Training loss (for one batch) at step 70: 262.0391, Accuracy: 0.8376
Training loss (for one batch) at step 80: 235.3920, Accuracy: 0.8375
Training loss (for one batch) at step 90: 249.2869, Accuracy: 0.8354
Training loss (for one batch) at step 100: 255.4146, Accuracy: 0.8357
Training loss (for one batch) at step 110: 222.9192, Accuracy: 0.8350
Training loss (for one batch) at step 120: 241.1500, Accuracy: 0.8342
Training loss (for one batch) at step 130: 265.1850, Accuracy: 0.8338
Training loss (for one batch) at step 140: 256.8575, Accuracy: 0.8333
---- Training ----
Training loss: 210.4998
Training acc over epoch: 0.8336
---- Validation ----
Validation loss: 65.5111
Validation acc: 0.7413
Time taken: 65.43s
Start of epoch 32
Training loss (for one batch) at step 0: 250.3142, Accuracy: 0.7600
Training loss (for one batch) at step 10: 222.0178, Accuracy: 0.8336
Training loss (for one batch) at step 20: 246.6547, Accuracy: 0.8243
Training loss (for one batch) at step 30: 237.3751, Accuracy: 0.8297
Training loss (for one batch) at step 40: 237.4019, Accuracy: 0.8320
Training loss (for one batch) at step 50: 218.1908, Accuracy: 0.8365
Training loss (for one batch) at step 60: 224.8291, Accuracy: 0.8372
Training loss (for one batch) at step 70: 261.4745, Accuracy: 0.8335
Training loss (for one batch) at step 80: 249.2448, Accuracy: 0.8302
Training loss (for one batch) at step 90: 232.7803, Accuracy: 0.8319
Training loss (for one batch) at step 100: 241.0771, Accuracy: 0.8318
Training loss (for one batch) at step 110: 224.0091, Accuracy: 0.8326
Training loss (for one batch) at step 120: 240.4555, Accuracy: 0.8325
Training loss (for one batch) at step 130: 262.9789, Accuracy: 0.8317
Training loss (for one batch) at step 140: 237.9560, Accuracy: 0.8300
---- Training ----
Training loss: 223.2001
Training acc over epoch: 0.8296
---- Validation ----
Validation loss: 80.8375
Validation acc: 0.7448
Time taken: 62.10s
Start of epoch 33
Training loss (for one batch) at step 0: 236.3316, Accuracy: 0.8900
Training loss (for one batch) at step 10: 217.8042, Accuracy: 0.8427
Training loss (for one batch) at step 20: 246.2785, Accuracy: 0.8357
Training loss (for one batch) at step 30: 243.9414, Accuracy: 0.8303
Training loss (for one batch) at step 40: 222.4774, Accuracy: 0.8305
Training loss (for one batch) at step 50: 227.3203, Accuracy: 0.8343
Training loss (for one batch) at step 60: 231.0095, Accuracy: 0.8366
Training loss (for one batch) at step 70: 247.9220, Accuracy: 0.8355
Training loss (for one batch) at step 80: 236.7035, Accuracy: 0.8337
Training loss (for one batch) at step 90: 250.3729, Accuracy: 0.8349
Training loss (for one batch) at step 100: 225.0669, Accuracy: 0.8344
Training loss (for one batch) at step 110: 246.6193, Accuracy: 0.8363
Training loss (for one batch) at step 120: 240.5588, Accuracy: 0.8339
Training loss (for one batch) at step 130: 247.9467, Accuracy: 0.8331
Training loss (for one batch) at step 140: 243.9002, Accuracy: 0.8318
---- Training ----
Training loss: 209.8524
Training acc over epoch: 0.8328
---- Validation ----
Validation loss: 76.5350
Validation acc: 0.7294
Time taken: 66.55s
Start of epoch 34
Training loss (for one batch) at step 0: 234.2189, Accuracy: 0.8300
Training loss (for one batch) at step 10: 248.5501, Accuracy: 0.8427
Training loss (for one batch) at step 20: 231.7302, Accuracy: 0.8486
Training loss (for one batch) at step 30: 255.7598, Accuracy: 0.8426
Training loss (for one batch) at step 40: 227.9252, Accuracy: 0.8434
Training loss (for one batch) at step 50: 221.3380, Accuracy: 0.8441
Training loss (for one batch) at step 60: 229.0824, Accuracy: 0.8420
Training loss (for one batch) at step 70: 210.9901, Accuracy: 0.8401
Training loss (for one batch) at step 80: 252.5154, Accuracy: 0.8373
Training loss (for one batch) at step 90: 247.1678, Accuracy: 0.8359
Training loss (for one batch) at step 100: 229.2041, Accuracy: 0.8360
Training loss (for one batch) at step 110: 218.3767, Accuracy: 0.8359
Training loss (for one batch) at step 120: 235.7198, Accuracy: 0.8360
Training loss (for one batch) at step 130: 217.2781, Accuracy: 0.8360
Training loss (for one batch) at step 140: 231.0072, Accuracy: 0.8357
---- Training ----
Training loss: 214.5541
Training acc over epoch: 0.8354
---- Validation ----
Validation loss: 81.2397
Validation acc: 0.7348
Time taken: 61.80s
Start of epoch 35
Training loss (for one batch) at step 0: 238.7279, Accuracy: 0.8600
Training loss (for one batch) at step 10: 235.9412, Accuracy: 0.8573
Training loss (for one batch) at step 20: 231.8674, Accuracy: 0.8476
Training loss (for one batch) at step 30: 222.4182, Accuracy: 0.8394
Training loss (for one batch) at step 40: 226.7190, Accuracy: 0.8393
Training loss (for one batch) at step 50: 239.5313, Accuracy: 0.8408
Training loss (for one batch) at step 60: 251.8678, Accuracy: 0.8403
Training loss (for one batch) at step 70: 218.2096, Accuracy: 0.8399
Training loss (for one batch) at step 80: 246.0926, Accuracy: 0.8396
Training loss (for one batch) at step 90: 238.2421, Accuracy: 0.8378
Training loss (for one batch) at step 100: 231.9322, Accuracy: 0.8342
Training loss (for one batch) at step 110: 239.9175, Accuracy: 0.8359
Training loss (for one batch) at step 120: 233.6303, Accuracy: 0.8355
Training loss (for one batch) at step 130: 246.1391, Accuracy: 0.8343
Training loss (for one batch) at step 140: 235.0736, Accuracy: 0.8357
---- Training ----
Training loss: 211.4016
Training acc over epoch: 0.8352
---- Validation ----
Validation loss: 76.6281
Validation acc: 0.7308
Time taken: 65.35s
Start of epoch 36
Training loss (for one batch) at step 0: 238.6932, Accuracy: 0.8500
Training loss (for one batch) at step 10: 218.2681, Accuracy: 0.8500
Training loss (for one batch) at step 20: 227.9955, Accuracy: 0.8467
Training loss (for one batch) at step 30: 255.4380, Accuracy: 0.8400
Training loss (for one batch) at step 40: 237.3891, Accuracy: 0.8432
Training loss (for one batch) at step 50: 219.0266, Accuracy: 0.8465
Training loss (for one batch) at step 60: 227.9399, Accuracy: 0.8479
Training loss (for one batch) at step 70: 229.6174, Accuracy: 0.8437
Training loss (for one batch) at step 80: 234.7866, Accuracy: 0.8388
Training loss (for one batch) at step 90: 245.9583, Accuracy: 0.8392
Training loss (for one batch) at step 100: 230.3555, Accuracy: 0.8377
Training loss (for one batch) at step 110: 231.9246, Accuracy: 0.8375
Training loss (for one batch) at step 120: 226.8226, Accuracy: 0.8384
Training loss (for one batch) at step 130: 244.9249, Accuracy: 0.8385
Training loss (for one batch) at step 140: 228.1391, Accuracy: 0.8378
---- Training ----
Training loss: 213.8731
Training acc over epoch: 0.8383
---- Validation ----
Validation loss: 65.4396
Validation acc: 0.7332
Time taken: 65.13s
Start of epoch 37
Training loss (for one batch) at step 0: 214.7086, Accuracy: 0.8400
Training loss (for one batch) at step 10: 204.4968, Accuracy: 0.8445
Training loss (for one batch) at step 20: 220.8909, Accuracy: 0.8338
Training loss (for one batch) at step 30: 232.6674, Accuracy: 0.8387
Training loss (for one batch) at step 40: 221.8589, Accuracy: 0.8422
Training loss (for one batch) at step 50: 214.9378, Accuracy: 0.8427
Training loss (for one batch) at step 60: 243.7677, Accuracy: 0.8411
Training loss (for one batch) at step 70: 241.7416, Accuracy: 0.8373
Training loss (for one batch) at step 80: 225.1516, Accuracy: 0.8360
Training loss (for one batch) at step 90: 229.0391, Accuracy: 0.8358
Training loss (for one batch) at step 100: 229.2784, Accuracy: 0.8362
Training loss (for one batch) at step 110: 229.3720, Accuracy: 0.8377
Training loss (for one batch) at step 120: 215.9642, Accuracy: 0.8377
Training loss (for one batch) at step 130: 239.4437, Accuracy: 0.8379
Training loss (for one batch) at step 140: 234.9407, Accuracy: 0.8377
---- Training ----
Training loss: 205.1173
Training acc over epoch: 0.8370
---- Validation ----
Validation loss: 68.3316
Validation acc: 0.7364
Time taken: 66.56s
Start of epoch 38
Training loss (for one batch) at step 0: 213.8399, Accuracy: 0.9000
Training loss (for one batch) at step 10: 231.6600, Accuracy: 0.8418
Training loss (for one batch) at step 20: 231.5735, Accuracy: 0.8490
Training loss (for one batch) at step 30: 230.3202, Accuracy: 0.8397
Training loss (for one batch) at step 40: 216.5244, Accuracy: 0.8415
Training loss (for one batch) at step 50: 241.8613, Accuracy: 0.8433
Training loss (for one batch) at step 60: 235.0100, Accuracy: 0.8464
Training loss (for one batch) at step 70: 240.9852, Accuracy: 0.8477
Training loss (for one batch) at step 80: 211.2136, Accuracy: 0.8452
Training loss (for one batch) at step 90: 223.6532, Accuracy: 0.8419
Training loss (for one batch) at step 100: 228.4670, Accuracy: 0.8418
Training loss (for one batch) at step 110: 235.2991, Accuracy: 0.8431
Training loss (for one batch) at step 120: 229.2585, Accuracy: 0.8431
Training loss (for one batch) at step 130: 221.5671, Accuracy: 0.8432
Training loss (for one batch) at step 140: 227.6562, Accuracy: 0.8422
---- Training ----
Training loss: 189.6470
Training acc over epoch: 0.8432
---- Validation ----
Validation loss: 64.1676
Validation acc: 0.7370
Time taken: 68.39s
Start of epoch 39
Training loss (for one batch) at step 0: 237.1278, Accuracy: 0.8500
Training loss (for one batch) at step 10: 205.4793, Accuracy: 0.8436
Training loss (for one batch) at step 20: 268.2793, Accuracy: 0.8390
Training loss (for one batch) at step 30: 235.5998, Accuracy: 0.8387
Training loss (for one batch) at step 40: 208.8513, Accuracy: 0.8420
Training loss (for one batch) at step 50: 230.6293, Accuracy: 0.8443
Training loss (for one batch) at step 60: 235.8383, Accuracy: 0.8443
Training loss (for one batch) at step 70: 239.6381, Accuracy: 0.8435
Training loss (for one batch) at step 80: 234.1430, Accuracy: 0.8443
Training loss (for one batch) at step 90: 220.4178, Accuracy: 0.8425
Training loss (for one batch) at step 100: 225.1412, Accuracy: 0.8433
Training loss (for one batch) at step 110: 223.0596, Accuracy: 0.8439
Training loss (for one batch) at step 120: 252.3037, Accuracy: 0.8436
Training loss (for one batch) at step 130: 243.3176, Accuracy: 0.8437
Training loss (for one batch) at step 140: 227.3838, Accuracy: 0.8435
---- Training ----
Training loss: 178.0078
Training acc over epoch: 0.8437
---- Validation ----
Validation loss: 88.5662
Validation acc: 0.7273
Time taken: 66.35s
Start of epoch 40
Training loss (for one batch) at step 0: 216.4188, Accuracy: 0.8100
Training loss (for one batch) at step 10: 214.0938, Accuracy: 0.8527
Training loss (for one batch) at step 20: 235.9726, Accuracy: 0.8457
Training loss (for one batch) at step 30: 227.4235, Accuracy: 0.8471
Training loss (for one batch) at step 40: 226.2378, Accuracy: 0.8502
Training loss (for one batch) at step 50: 190.1640, Accuracy: 0.8518
Training loss (for one batch) at step 60: 205.0281, Accuracy: 0.8502
Training loss (for one batch) at step 70: 227.6306, Accuracy: 0.8501
Training loss (for one batch) at step 80: 226.0949, Accuracy: 0.8462
Training loss (for one batch) at step 90: 216.7089, Accuracy: 0.8454
Training loss (for one batch) at step 100: 219.3318, Accuracy: 0.8446
Training loss (for one batch) at step 110: 219.0894, Accuracy: 0.8453
Training loss (for one batch) at step 120: 235.9011, Accuracy: 0.8457
Training loss (for one batch) at step 130: 213.5753, Accuracy: 0.8456
Training loss (for one batch) at step 140: 226.0894, Accuracy: 0.8440
---- Training ----
Training loss: 195.6663
Training acc over epoch: 0.8442
---- Validation ----
Validation loss: 76.6389
Validation acc: 0.7303
Time taken: 61.15s
Start of epoch 41
Training loss (for one batch) at step 0: 225.0316, Accuracy: 0.8700
Training loss (for one batch) at step 10: 216.7598, Accuracy: 0.8373
Training loss (for one batch) at step 20: 217.8878, Accuracy: 0.8543
Training loss (for one batch) at step 30: 213.2420, Accuracy: 0.8561
Training loss (for one batch) at step 40: 201.4889, Accuracy: 0.8546
Training loss (for one batch) at step 50: 201.1270, Accuracy: 0.8539
Training loss (for one batch) at step 60: 205.9330, Accuracy: 0.8523
Training loss (for one batch) at step 70: 231.7452, Accuracy: 0.8521
Training loss (for one batch) at step 80: 234.2701, Accuracy: 0.8473
Training loss (for one batch) at step 90: 218.1704, Accuracy: 0.8457
Training loss (for one batch) at step 100: 212.5427, Accuracy: 0.8444
Training loss (for one batch) at step 110: 204.7717, Accuracy: 0.8440
Training loss (for one batch) at step 120: 225.0644, Accuracy: 0.8450
Training loss (for one batch) at step 130: 222.9135, Accuracy: 0.8444
Training loss (for one batch) at step 140: 250.2830, Accuracy: 0.8429
---- Training ----
Training loss: 188.5053
Training acc over epoch: 0.8429
---- Validation ----
Validation loss: 67.1551
Validation acc: 0.7383
Time taken: 65.77s
Start of epoch 42
Training loss (for one batch) at step 0: 225.8636, Accuracy: 0.8400
Training loss (for one batch) at step 10: 216.1869, Accuracy: 0.8482
Training loss (for one batch) at step 20: 219.3552, Accuracy: 0.8457
Training loss (for one batch) at step 30: 225.4915, Accuracy: 0.8474
Training loss (for one batch) at step 40: 223.6529, Accuracy: 0.8490
Training loss (for one batch) at step 50: 215.7772, Accuracy: 0.8478
Training loss (for one batch) at step 60: 210.5129, Accuracy: 0.8469
Training loss (for one batch) at step 70: 239.4937, Accuracy: 0.8473
Training loss (for one batch) at step 80: 226.0515, Accuracy: 0.8456
Training loss (for one batch) at step 90: 209.4474, Accuracy: 0.8445
Training loss (for one batch) at step 100: 197.7493, Accuracy: 0.8435
Training loss (for one batch) at step 110: 211.2419, Accuracy: 0.8426
Training loss (for one batch) at step 120: 214.8913, Accuracy: 0.8431
Training loss (for one batch) at step 130: 236.1486, Accuracy: 0.8424
Training loss (for one batch) at step 140: 237.5836, Accuracy: 0.8432
---- Training ----
Training loss: 181.0783
Training acc over epoch: 0.8434
---- Validation ----
Validation loss: 80.9885
Validation acc: 0.7402
Time taken: 60.94s
Start of epoch 43
Training loss (for one batch) at step 0: 202.6954, Accuracy: 0.8900
Training loss (for one batch) at step 10: 216.0591, Accuracy: 0.8718
Training loss (for one batch) at step 20: 225.6759, Accuracy: 0.8562
Training loss (for one batch) at step 30: 213.2162, Accuracy: 0.8558
Training loss (for one batch) at step 40: 212.0256, Accuracy: 0.8546
Training loss (for one batch) at step 50: 206.8369, Accuracy: 0.8549
Training loss (for one batch) at step 60: 226.9923, Accuracy: 0.8541
Training loss (for one batch) at step 70: 204.6375, Accuracy: 0.8530
Training loss (for one batch) at step 80: 216.5882, Accuracy: 0.8506
Training loss (for one batch) at step 90: 216.0438, Accuracy: 0.8479
Training loss (for one batch) at step 100: 223.0962, Accuracy: 0.8457
Training loss (for one batch) at step 110: 233.1980, Accuracy: 0.8461
Training loss (for one batch) at step 120: 227.0358, Accuracy: 0.8461
Training loss (for one batch) at step 130: 219.4743, Accuracy: 0.8469
Training loss (for one batch) at step 140: 245.4401, Accuracy: 0.8459
---- Training ----
Training loss: 199.1525
Training acc over epoch: 0.8458
---- Validation ----
Validation loss: 89.1501
Validation acc: 0.7327
Time taken: 67.79s
Start of epoch 44
Training loss (for one batch) at step 0: 229.6684, Accuracy: 0.8500
Training loss (for one batch) at step 10: 217.3118, Accuracy: 0.8436
Training loss (for one batch) at step 20: 213.0234, Accuracy: 0.8471
Training loss (for one batch) at step 30: 201.2134, Accuracy: 0.8510
Training loss (for one batch) at step 40: 225.7282, Accuracy: 0.8488
Training loss (for one batch) at step 50: 208.7262, Accuracy: 0.8461
Training loss (for one batch) at step 60: 222.6118, Accuracy: 0.8456
Training loss (for one batch) at step 70: 245.3731, Accuracy: 0.8435
Training loss (for one batch) at step 80: 217.2943, Accuracy: 0.8422
Training loss (for one batch) at step 90: 228.1328, Accuracy: 0.8423
Training loss (for one batch) at step 100: 225.1113, Accuracy: 0.8430
Training loss (for one batch) at step 110: 205.0457, Accuracy: 0.8435
Training loss (for one batch) at step 120: 199.2538, Accuracy: 0.8432
Training loss (for one batch) at step 130: 219.1492, Accuracy: 0.8434
Training loss (for one batch) at step 140: 215.4514, Accuracy: 0.8431
---- Training ----
Training loss: 187.3772
Training acc over epoch: 0.8438
---- Validation ----
Validation loss: 81.6574
Validation acc: 0.7346
Time taken: 61.85s
Start of epoch 45
Training loss (for one batch) at step 0: 203.3833, Accuracy: 0.8800
Training loss (for one batch) at step 10: 219.9881, Accuracy: 0.8618
Training loss (for one batch) at step 20: 194.4039, Accuracy: 0.8581
Training loss (for one batch) at step 30: 220.7786, Accuracy: 0.8568
Training loss (for one batch) at step 40: 208.8696, Accuracy: 0.8559
Training loss (for one batch) at step 50: 210.0050, Accuracy: 0.8547
Training loss (for one batch) at step 60: 202.9584, Accuracy: 0.8520
Training loss (for one batch) at step 70: 234.5377, Accuracy: 0.8524
Training loss (for one batch) at step 80: 231.1420, Accuracy: 0.8506
Training loss (for one batch) at step 90: 214.6229, Accuracy: 0.8487
Training loss (for one batch) at step 100: 217.6947, Accuracy: 0.8489
Training loss (for one batch) at step 110: 202.5582, Accuracy: 0.8490
Training loss (for one batch) at step 120: 214.1706, Accuracy: 0.8498
Training loss (for one batch) at step 130: 224.6547, Accuracy: 0.8495
Training loss (for one batch) at step 140: 212.0076, Accuracy: 0.8497
---- Training ----
Training loss: 183.5596
Training acc over epoch: 0.8488
---- Validation ----
Validation loss: 65.4504
Validation acc: 0.7294
Time taken: 66.87s
Start of epoch 46
Training loss (for one batch) at step 0: 228.8866, Accuracy: 0.8400
Training loss (for one batch) at step 10: 207.2851, Accuracy: 0.8427
Training loss (for one batch) at step 20: 205.6905, Accuracy: 0.8476
Training loss (for one batch) at step 30: 216.9939, Accuracy: 0.8458
Training loss (for one batch) at step 40: 215.1180, Accuracy: 0.8405
Training loss (for one batch) at step 50: 233.5157, Accuracy: 0.8418
Training loss (for one batch) at step 60: 193.3800, Accuracy: 0.8446
Training loss (for one batch) at step 70: 197.5421, Accuracy: 0.8446
Training loss (for one batch) at step 80: 201.9869, Accuracy: 0.8458
Training loss (for one batch) at step 90: 223.5165, Accuracy: 0.8478
Training loss (for one batch) at step 100: 219.3352, Accuracy: 0.8468
Training loss (for one batch) at step 110: 217.6826, Accuracy: 0.8470
Training loss (for one batch) at step 120: 217.2001, Accuracy: 0.8458
Training loss (for one batch) at step 130: 212.4535, Accuracy: 0.8465
Training loss (for one batch) at step 140: 207.2098, Accuracy: 0.8472
---- Training ----
Training loss: 193.4420
Training acc over epoch: 0.8466
---- Validation ----
Validation loss: 73.7934
Validation acc: 0.7329
Time taken: 62.83s
Start of epoch 47
Training loss (for one batch) at step 0: 228.5602, Accuracy: 0.7900
Training loss (for one batch) at step 10: 191.1613, Accuracy: 0.8427
Training loss (for one batch) at step 20: 207.2621, Accuracy: 0.8538
Training loss (for one batch) at step 30: 216.3276, Accuracy: 0.8445
Training loss (for one batch) at step 40: 202.8960, Accuracy: 0.8459
Training loss (for one batch) at step 50: 222.4717, Accuracy: 0.8467
Training loss (for one batch) at step 60: 200.9142, Accuracy: 0.8470
Training loss (for one batch) at step 70: 215.5580, Accuracy: 0.8466
Training loss (for one batch) at step 80: 224.5522, Accuracy: 0.8464
Training loss (for one batch) at step 90: 220.8380, Accuracy: 0.8440
Training loss (for one batch) at step 100: 211.4236, Accuracy: 0.8419
Training loss (for one batch) at step 110: 201.1813, Accuracy: 0.8438
Training loss (for one batch) at step 120: 211.5011, Accuracy: 0.8439
Training loss (for one batch) at step 130: 201.7614, Accuracy: 0.8450
Training loss (for one batch) at step 140: 204.2637, Accuracy: 0.8460
---- Training ----
Training loss: 191.6983
Training acc over epoch: 0.8453
---- Validation ----
Validation loss: 67.1076
Validation acc: 0.7372
Time taken: 68.00s
Start of epoch 48
Training loss (for one batch) at step 0: 191.9689, Accuracy: 0.8400
Training loss (for one batch) at step 10: 211.0682, Accuracy: 0.8355
Training loss (for one batch) at step 20: 179.6260, Accuracy: 0.8419
Training loss (for one batch) at step 30: 201.9257, Accuracy: 0.8477
Training loss (for one batch) at step 40: 210.9674, Accuracy: 0.8478
Training loss (for one batch) at step 50: 214.6008, Accuracy: 0.8508
Training loss (for one batch) at step 60: 209.1919, Accuracy: 0.8515
Training loss (for one batch) at step 70: 236.3939, Accuracy: 0.8468
Training loss (for one batch) at step 80: 208.3466, Accuracy: 0.8473
Training loss (for one batch) at step 90: 227.0635, Accuracy: 0.8444
Training loss (for one batch) at step 100: 184.3079, Accuracy: 0.8451
Training loss (for one batch) at step 110: 222.1685, Accuracy: 0.8447
Training loss (for one batch) at step 120: 200.5127, Accuracy: 0.8456
Training loss (for one batch) at step 130: 222.7606, Accuracy: 0.8447
Training loss (for one batch) at step 140: 231.4540, Accuracy: 0.8463
---- Training ----
Training loss: 185.7916
Training acc over epoch: 0.8458
---- Validation ----
Validation loss: 60.5733
Validation acc: 0.7332
Time taken: 63.38s
Start of epoch 49
Training loss (for one batch) at step 0: 205.8401, Accuracy: 0.8700
Training loss (for one batch) at step 10: 207.2703, Accuracy: 0.8645
Training loss (for one batch) at step 20: 210.0096, Accuracy: 0.8590
Training loss (for one batch) at step 30: 195.4416, Accuracy: 0.8606
Training loss (for one batch) at step 40: 210.9266, Accuracy: 0.8585
Training loss (for one batch) at step 50: 189.7214, Accuracy: 0.8608
Training loss (for one batch) at step 60: 196.2239, Accuracy: 0.8605
Training loss (for one batch) at step 70: 227.3105, Accuracy: 0.8558
Training loss (for one batch) at step 80: 218.6985, Accuracy: 0.8547
Training loss (for one batch) at step 90: 217.8864, Accuracy: 0.8532
Training loss (for one batch) at step 100: 220.0170, Accuracy: 0.8532
Training loss (for one batch) at step 110: 195.9427, Accuracy: 0.8528
Training loss (for one batch) at step 120: 191.4595, Accuracy: 0.8535
Training loss (for one batch) at step 130: 212.8763, Accuracy: 0.8527
Training loss (for one batch) at step 140: 230.2480, Accuracy: 0.8519
---- Training ----
Training loss: 174.7856
Training acc over epoch: 0.8526
---- Validation ----
Validation loss: 84.9808
Validation acc: 0.7311
Time taken: 65.71s
===== Q: 0.0001
Validation acc: 0.7326
Validation AUC: 0.7294
Validation Balanced_ACC: 0.4802
Validation MI: 0.1378
Validation Normalized MI: 0.2061
Validation Adjusted MI: 0.2061
Validation aUc_Sklearn: 0.8307
Start of epoch 0
2023-02-15 00:54:50.466304: I tensorflow/core/kernels/data/shuffle_dataset_op.cc:175] Filling up shuffle buffer (this may take a while): 1 of 1024
2023-02-15 00:54:53.994200: I tensorflow/core/kernels/data/shuffle_dataset_op.cc:228] Shuffle buffer filled.
Training loss (for one batch) at step 0: 545.6800, Accuracy: 0.5100
Training loss (for one batch) at step 10: 495.1459, Accuracy: 0.5127
Training loss (for one batch) at step 20: 490.8329, Accuracy: 0.5248
Training loss (for one batch) at step 30: 425.7359, Accuracy: 0.5358
Training loss (for one batch) at step 40: 438.1086, Accuracy: 0.5522
Training loss (for one batch) at step 50: 437.8628, Accuracy: 0.5557
Training loss (for one batch) at step 60: 432.2560, Accuracy: 0.5548
Training loss (for one batch) at step 70: 464.4137, Accuracy: 0.5606
Training loss (for one batch) at step 80: 417.0397, Accuracy: 0.5611
Training loss (for one batch) at step 90: 430.1098, Accuracy: 0.5642
Training loss (for one batch) at step 100: 430.7260, Accuracy: 0.5689
Training loss (for one batch) at step 110: 410.6523, Accuracy: 0.5713
Training loss (for one batch) at step 120: 439.1637, Accuracy: 0.5726
Training loss (for one batch) at step 130: 417.6356, Accuracy: 0.5734
Training loss (for one batch) at step 140: 450.7852, Accuracy: 0.5729
---- Training ----
Training loss: 393.9187
Training acc over epoch: 0.5727
---- Validation ----
Validation loss: 85.3938
Validation acc: 0.5134
Time taken: 80.13s
Start of epoch 1
Training loss (for one batch) at step 0: 407.4026, Accuracy: 0.6200
Training loss (for one batch) at step 10: 361.6945, Accuracy: 0.6200
Training loss (for one batch) at step 20: 424.7666, Accuracy: 0.6100
Training loss (for one batch) at step 30: 444.8634, Accuracy: 0.6042
Training loss (for one batch) at step 40: 393.1862, Accuracy: 0.6000
Training loss (for one batch) at step 50: 391.3019, Accuracy: 0.5990
Training loss (for one batch) at step 60: 424.5691, Accuracy: 0.5989
Training loss (for one batch) at step 70: 389.5775, Accuracy: 0.5979
Training loss (for one batch) at step 80: 360.7109, Accuracy: 0.6037
Training loss (for one batch) at step 90: 367.6332, Accuracy: 0.6057
Training loss (for one batch) at step 100: 379.7974, Accuracy: 0.6047
Training loss (for one batch) at step 110: 390.2852, Accuracy: 0.6052
Training loss (for one batch) at step 120: 379.6125, Accuracy: 0.6050
Training loss (for one batch) at step 130: 391.9070, Accuracy: 0.6064
Training loss (for one batch) at step 140: 383.6197, Accuracy: 0.6074
---- Training ----
Training loss: 322.8352
Training acc over epoch: 0.6075
---- Validation ----
Validation loss: 80.8852
Validation acc: 0.5355
Time taken: 58.11s
Start of epoch 2
Training loss (for one batch) at step 0: 377.7250, Accuracy: 0.6100
Training loss (for one batch) at step 10: 364.0365, Accuracy: 0.6291
Training loss (for one batch) at step 20: 370.7310, Accuracy: 0.6314
Training loss (for one batch) at step 30: 385.7524, Accuracy: 0.6313
Training loss (for one batch) at step 40: 373.0836, Accuracy: 0.6249
Training loss (for one batch) at step 50: 359.8702, Accuracy: 0.6239
Training loss (for one batch) at step 60: 367.8431, Accuracy: 0.6266
Training loss (for one batch) at step 70: 372.2193, Accuracy: 0.6263
Training loss (for one batch) at step 80: 367.2703, Accuracy: 0.6280
Training loss (for one batch) at step 90: 355.3396, Accuracy: 0.6292
Training loss (for one batch) at step 100: 349.9782, Accuracy: 0.6321
Training loss (for one batch) at step 110: 344.6563, Accuracy: 0.6338
Training loss (for one batch) at step 120: 373.5445, Accuracy: 0.6346
Training loss (for one batch) at step 130: 375.1917, Accuracy: 0.6341
Training loss (for one batch) at step 140: 355.6011, Accuracy: 0.6342
---- Training ----
Training loss: 309.4636
Training acc over epoch: 0.6351
---- Validation ----
Validation loss: 73.6118
Validation acc: 0.6644
Time taken: 70.40s
Start of epoch 3
Training loss (for one batch) at step 0: 358.9247, Accuracy: 0.6500
Training loss (for one batch) at step 10: 391.4939, Accuracy: 0.6555
Training loss (for one batch) at step 20: 351.8127, Accuracy: 0.6452
Training loss (for one batch) at step 30: 335.7517, Accuracy: 0.6510
Training loss (for one batch) at step 40: 354.4075, Accuracy: 0.6495
Training loss (for one batch) at step 50: 364.5615, Accuracy: 0.6533
Training loss (for one batch) at step 60: 349.4557, Accuracy: 0.6507
Training loss (for one batch) at step 70: 345.6068, Accuracy: 0.6545
Training loss (for one batch) at step 80: 350.7102, Accuracy: 0.6505
Training loss (for one batch) at step 90: 379.4396, Accuracy: 0.6497
Training loss (for one batch) at step 100: 341.4798, Accuracy: 0.6505
Training loss (for one batch) at step 110: 353.7067, Accuracy: 0.6526
Training loss (for one batch) at step 120: 338.0478, Accuracy: 0.6545
Training loss (for one batch) at step 130: 363.7738, Accuracy: 0.6540
Training loss (for one batch) at step 140: 339.6879, Accuracy: 0.6543
---- Training ----
Training loss: 295.2355
Training acc over epoch: 0.6537
---- Validation ----
Validation loss: 72.7792
Validation acc: 0.6851
Time taken: 41.02s
Start of epoch 4
Training loss (for one batch) at step 0: 345.4147, Accuracy: 0.7100
Training loss (for one batch) at step 10: 331.6015, Accuracy: 0.6600
Training loss (for one batch) at step 20: 351.4102, Accuracy: 0.6595
Training loss (for one batch) at step 30: 349.5045, Accuracy: 0.6632
Training loss (for one batch) at step 40: 343.8345, Accuracy: 0.6646
Training loss (for one batch) at step 50: 331.9045, Accuracy: 0.6696
Training loss (for one batch) at step 60: 340.4880, Accuracy: 0.6695
Training loss (for one batch) at step 70: 364.7793, Accuracy: 0.6673
Training loss (for one batch) at step 80: 353.3133, Accuracy: 0.6657
Training loss (for one batch) at step 90: 366.5410, Accuracy: 0.6663
Training loss (for one batch) at step 100: 342.2521, Accuracy: 0.6650
Training loss (for one batch) at step 110: 340.5629, Accuracy: 0.6651
Training loss (for one batch) at step 120: 363.6970, Accuracy: 0.6672
Training loss (for one batch) at step 130: 336.9555, Accuracy: 0.6679
Training loss (for one batch) at step 140: 341.2458, Accuracy: 0.6669
---- Training ----
Training loss: 287.5524
Training acc over epoch: 0.6682
---- Validation ----
Validation loss: 62.8962
Validation acc: 0.6940
Time taken: 68.05s
Start of epoch 5
Training loss (for one batch) at step 0: 333.6786, Accuracy: 0.7200
Training loss (for one batch) at step 10: 342.5051, Accuracy: 0.6645
Training loss (for one batch) at step 20: 331.6709, Accuracy: 0.6819
Training loss (for one batch) at step 30: 324.5829, Accuracy: 0.6816
Training loss (for one batch) at step 40: 357.9179, Accuracy: 0.6827
Training loss (for one batch) at step 50: 337.0055, Accuracy: 0.6837
Training loss (for one batch) at step 60: 316.3197, Accuracy: 0.6846
Training loss (for one batch) at step 70: 324.5354, Accuracy: 0.6849
Training loss (for one batch) at step 80: 330.8062, Accuracy: 0.6822
Training loss (for one batch) at step 90: 306.9734, Accuracy: 0.6824
Training loss (for one batch) at step 100: 330.3593, Accuracy: 0.6828
Training loss (for one batch) at step 110: 323.9038, Accuracy: 0.6821
Training loss (for one batch) at step 120: 333.3981, Accuracy: 0.6826
Training loss (for one batch) at step 130: 305.9286, Accuracy: 0.6836
Training loss (for one batch) at step 140: 350.6210, Accuracy: 0.6835
---- Training ----
Training loss: 295.3501
Training acc over epoch: 0.6842
---- Validation ----
Validation loss: 71.8306
Validation acc: 0.6937
Time taken: 40.72s
Start of epoch 6
Training loss (for one batch) at step 0: 350.2249, Accuracy: 0.6800
Training loss (for one batch) at step 10: 328.0069, Accuracy: 0.6964
Training loss (for one batch) at step 20: 334.0849, Accuracy: 0.6933
Training loss (for one batch) at step 30: 345.8398, Accuracy: 0.6948
Training loss (for one batch) at step 40: 319.3948, Accuracy: 0.6980
Training loss (for one batch) at step 50: 325.6580, Accuracy: 0.7014
Training loss (for one batch) at step 60: 327.7035, Accuracy: 0.7036
Training loss (for one batch) at step 70: 334.6507, Accuracy: 0.6999
Training loss (for one batch) at step 80: 319.2107, Accuracy: 0.6986
Training loss (for one batch) at step 90: 338.7266, Accuracy: 0.6985
Training loss (for one batch) at step 100: 324.2195, Accuracy: 0.6979
Training loss (for one batch) at step 110: 312.8528, Accuracy: 0.7004
Training loss (for one batch) at step 120: 340.3146, Accuracy: 0.7011
Training loss (for one batch) at step 130: 317.3104, Accuracy: 0.6988
Training loss (for one batch) at step 140: 323.9902, Accuracy: 0.6987
---- Training ----
Training loss: 281.3472
Training acc over epoch: 0.6991
---- Validation ----
Validation loss: 78.8091
Validation acc: 0.7071
Time taken: 67.51s
Start of epoch 7
Training loss (for one batch) at step 0: 327.6902, Accuracy: 0.6500
Training loss (for one batch) at step 10: 326.8561, Accuracy: 0.7045
Training loss (for one batch) at step 20: 335.8393, Accuracy: 0.7038
Training loss (for one batch) at step 30: 317.1183, Accuracy: 0.7061
Training loss (for one batch) at step 40: 305.7310, Accuracy: 0.7068
Training loss (for one batch) at step 50: 323.1814, Accuracy: 0.7075
Training loss (for one batch) at step 60: 345.1830, Accuracy: 0.7098
Training loss (for one batch) at step 70: 336.6928, Accuracy: 0.7108
Training loss (for one batch) at step 80: 294.5559, Accuracy: 0.7114
Training loss (for one batch) at step 90: 320.8440, Accuracy: 0.7100
Training loss (for one batch) at step 100: 300.9401, Accuracy: 0.7077
Training loss (for one batch) at step 110: 310.1031, Accuracy: 0.7108
Training loss (for one batch) at step 120: 310.9427, Accuracy: 0.7119
Training loss (for one batch) at step 130: 292.9605, Accuracy: 0.7106
Training loss (for one batch) at step 140: 325.5179, Accuracy: 0.7104
---- Training ----
Training loss: 272.8610
Training acc over epoch: 0.7123
---- Validation ----
Validation loss: 67.2095
Validation acc: 0.7246
Time taken: 39.73s
Start of epoch 8
Training loss (for one batch) at step 0: 307.7834, Accuracy: 0.8000
Training loss (for one batch) at step 10: 320.1678, Accuracy: 0.7409
Training loss (for one batch) at step 20: 340.8282, Accuracy: 0.7329
Training loss (for one batch) at step 30: 311.2838, Accuracy: 0.7310
Training loss (for one batch) at step 40: 315.8815, Accuracy: 0.7324
Training loss (for one batch) at step 50: 336.0710, Accuracy: 0.7331
Training loss (for one batch) at step 60: 329.4042, Accuracy: 0.7344
Training loss (for one batch) at step 70: 317.3655, Accuracy: 0.7301
Training loss (for one batch) at step 80: 310.3604, Accuracy: 0.7277
Training loss (for one batch) at step 90: 313.7139, Accuracy: 0.7267
Training loss (for one batch) at step 100: 305.7078, Accuracy: 0.7263
Training loss (for one batch) at step 110: 315.9340, Accuracy: 0.7285
Training loss (for one batch) at step 120: 309.2935, Accuracy: 0.7294
Training loss (for one batch) at step 130: 317.5342, Accuracy: 0.7292
Training loss (for one batch) at step 140: 292.7810, Accuracy: 0.7278
---- Training ----
Training loss: 299.0878
Training acc over epoch: 0.7276
---- Validation ----
Validation loss: 70.1739
Validation acc: 0.7053
Time taken: 68.22s
Start of epoch 9
Training loss (for one batch) at step 0: 316.2713, Accuracy: 0.6500
Training loss (for one batch) at step 10: 292.0122, Accuracy: 0.7355
Training loss (for one batch) at step 20: 302.4611, Accuracy: 0.7329
Training loss (for one batch) at step 30: 321.5370, Accuracy: 0.7371
Training loss (for one batch) at step 40: 313.6048, Accuracy: 0.7349
Training loss (for one batch) at step 50: 316.6800, Accuracy: 0.7400
Training loss (for one batch) at step 60: 308.0285, Accuracy: 0.7398
Training loss (for one batch) at step 70: 312.4781, Accuracy: 0.7385
Training loss (for one batch) at step 80: 312.4885, Accuracy: 0.7380
Training loss (for one batch) at step 90: 310.7767, Accuracy: 0.7341
Training loss (for one batch) at step 100: 309.3721, Accuracy: 0.7337
Training loss (for one batch) at step 110: 307.0711, Accuracy: 0.7332
Training loss (for one batch) at step 120: 322.7055, Accuracy: 0.7352
Training loss (for one batch) at step 130: 318.4422, Accuracy: 0.7346
Training loss (for one batch) at step 140: 318.7543, Accuracy: 0.7360
---- Training ----
Training loss: 278.0220
Training acc over epoch: 0.7368
---- Validation ----
Validation loss: 65.0744
Validation acc: 0.7214
Time taken: 39.97s
Start of epoch 10
Training loss (for one batch) at step 0: 311.4355, Accuracy: 0.7400
Training loss (for one batch) at step 10: 298.1284, Accuracy: 0.7291
Training loss (for one batch) at step 20: 303.1657, Accuracy: 0.7438
Training loss (for one batch) at step 30: 297.2698, Accuracy: 0.7397
Training loss (for one batch) at step 40: 325.3689, Accuracy: 0.7400
Training loss (for one batch) at step 50: 298.9141, Accuracy: 0.7435
Training loss (for one batch) at step 60: 289.5235, Accuracy: 0.7474
Training loss (for one batch) at step 70: 313.0559, Accuracy: 0.7459
Training loss (for one batch) at step 80: 313.6139, Accuracy: 0.7441
Training loss (for one batch) at step 90: 296.2527, Accuracy: 0.7447
Training loss (for one batch) at step 100: 295.1881, Accuracy: 0.7439
Training loss (for one batch) at step 110: 296.0001, Accuracy: 0.7438
Training loss (for one batch) at step 120: 313.2403, Accuracy: 0.7434
Training loss (for one batch) at step 130: 296.1858, Accuracy: 0.7454
Training loss (for one batch) at step 140: 305.0052, Accuracy: 0.7449
---- Training ----
Training loss: 255.2052
Training acc over epoch: 0.7458
---- Validation ----
Validation loss: 68.8803
Validation acc: 0.7286
Time taken: 66.28s
Start of epoch 11
Training loss (for one batch) at step 0: 303.8528, Accuracy: 0.7800
Training loss (for one batch) at step 10: 284.4423, Accuracy: 0.7600
Training loss (for one batch) at step 20: 291.6891, Accuracy: 0.7600
Training loss (for one batch) at step 30: 292.7007, Accuracy: 0.7603
Training loss (for one batch) at step 40: 317.7006, Accuracy: 0.7607
Training loss (for one batch) at step 50: 312.8318, Accuracy: 0.7624
Training loss (for one batch) at step 60: 294.3421, Accuracy: 0.7651
Training loss (for one batch) at step 70: 304.1695, Accuracy: 0.7638
Training loss (for one batch) at step 80: 305.7187, Accuracy: 0.7622
Training loss (for one batch) at step 90: 300.9535, Accuracy: 0.7597
Training loss (for one batch) at step 100: 302.1191, Accuracy: 0.7573
Training loss (for one batch) at step 110: 293.2072, Accuracy: 0.7578
Training loss (for one batch) at step 120: 303.5265, Accuracy: 0.7569
Training loss (for one batch) at step 130: 304.8642, Accuracy: 0.7579
Training loss (for one batch) at step 140: 301.4506, Accuracy: 0.7571
---- Training ----
Training loss: 273.7491
Training acc over epoch: 0.7581
---- Validation ----
Validation loss: 62.9857
Validation acc: 0.7303
Time taken: 38.79s
Start of epoch 12
Training loss (for one batch) at step 0: 313.0536, Accuracy: 0.7700
Training loss (for one batch) at step 10: 283.7522, Accuracy: 0.7682
Training loss (for one batch) at step 20: 296.0261, Accuracy: 0.7667
Training loss (for one batch) at step 30: 301.6374, Accuracy: 0.7690
Training loss (for one batch) at step 40: 306.0946, Accuracy: 0.7663
Training loss (for one batch) at step 50: 297.3725, Accuracy: 0.7663
Training loss (for one batch) at step 60: 301.3681, Accuracy: 0.7680
Training loss (for one batch) at step 70: 299.6974, Accuracy: 0.7659
Training loss (for one batch) at step 80: 280.9861, Accuracy: 0.7658
Training loss (for one batch) at step 90: 288.2669, Accuracy: 0.7671
Training loss (for one batch) at step 100: 301.1797, Accuracy: 0.7651
Training loss (for one batch) at step 110: 289.5865, Accuracy: 0.7665
Training loss (for one batch) at step 120: 293.1295, Accuracy: 0.7655
Training loss (for one batch) at step 130: 295.6617, Accuracy: 0.7651
Training loss (for one batch) at step 140: 312.6779, Accuracy: 0.7648
---- Training ----
Training loss: 265.8235
Training acc over epoch: 0.7648
---- Validation ----
Validation loss: 78.2470
Validation acc: 0.7292
Time taken: 67.38s
Start of epoch 13
Training loss (for one batch) at step 0: 294.2739, Accuracy: 0.7500
Training loss (for one batch) at step 10: 294.1396, Accuracy: 0.7682
Training loss (for one batch) at step 20: 297.5699, Accuracy: 0.7652
Training loss (for one batch) at step 30: 299.1764, Accuracy: 0.7587
Training loss (for one batch) at step 40: 284.9442, Accuracy: 0.7615
Training loss (for one batch) at step 50: 307.8511, Accuracy: 0.7659
Training loss (for one batch) at step 60: 292.0417, Accuracy: 0.7703
Training loss (for one batch) at step 70: 288.8527, Accuracy: 0.7711
Training loss (for one batch) at step 80: 291.9001, Accuracy: 0.7701
Training loss (for one batch) at step 90: 303.5839, Accuracy: 0.7680
Training loss (for one batch) at step 100: 277.9055, Accuracy: 0.7695
Training loss (for one batch) at step 110: 297.5024, Accuracy: 0.7702
Training loss (for one batch) at step 120: 290.7953, Accuracy: 0.7702
Training loss (for one batch) at step 130: 274.9261, Accuracy: 0.7692
Training loss (for one batch) at step 140: 272.5548, Accuracy: 0.7693
---- Training ----
Training loss: 253.3904
Training acc over epoch: 0.7693
---- Validation ----
Validation loss: 68.7857
Validation acc: 0.7321
Time taken: 39.12s
Start of epoch 14
Training loss (for one batch) at step 0: 287.0439, Accuracy: 0.8300
Training loss (for one batch) at step 10: 278.1053, Accuracy: 0.8000
Training loss (for one batch) at step 20: 294.4882, Accuracy: 0.7900
Training loss (for one batch) at step 30: 300.0937, Accuracy: 0.7845
Training loss (for one batch) at step 40: 278.7210, Accuracy: 0.7844
Training loss (for one batch) at step 50: 272.2222, Accuracy: 0.7910
Training loss (for one batch) at step 60: 296.9243, Accuracy: 0.7889
Training loss (for one batch) at step 70: 305.9151, Accuracy: 0.7852
Training loss (for one batch) at step 80: 278.8784, Accuracy: 0.7812
Training loss (for one batch) at step 90: 296.2625, Accuracy: 0.7795
Training loss (for one batch) at step 100: 286.6433, Accuracy: 0.7796
Training loss (for one batch) at step 110: 290.6607, Accuracy: 0.7797
Training loss (for one batch) at step 120: 289.5623, Accuracy: 0.7788
Training loss (for one batch) at step 130: 283.5416, Accuracy: 0.7779
Training loss (for one batch) at step 140: 283.1549, Accuracy: 0.7788
---- Training ----
Training loss: 266.6304
Training acc over epoch: 0.7774
---- Validation ----
Validation loss: 73.6558
Validation acc: 0.7423
Time taken: 66.32s
Start of epoch 15
Training loss (for one batch) at step 0: 271.9902, Accuracy: 0.8200
Training loss (for one batch) at step 10: 269.7357, Accuracy: 0.7745
Training loss (for one batch) at step 20: 282.8263, Accuracy: 0.7795
Training loss (for one batch) at step 30: 298.6555, Accuracy: 0.7761
Training loss (for one batch) at step 40: 280.9085, Accuracy: 0.7778
Training loss (for one batch) at step 50: 288.6775, Accuracy: 0.7855
Training loss (for one batch) at step 60: 275.0435, Accuracy: 0.7867
Training loss (for one batch) at step 70: 286.2972, Accuracy: 0.7849
Training loss (for one batch) at step 80: 280.3688, Accuracy: 0.7860
Training loss (for one batch) at step 90: 273.1516, Accuracy: 0.7844
Training loss (for one batch) at step 100: 290.1310, Accuracy: 0.7821
Training loss (for one batch) at step 110: 264.9464, Accuracy: 0.7820
Training loss (for one batch) at step 120: 295.4944, Accuracy: 0.7813
Training loss (for one batch) at step 130: 286.2273, Accuracy: 0.7796
Training loss (for one batch) at step 140: 297.2076, Accuracy: 0.7797
---- Training ----
Training loss: 259.1566
Training acc over epoch: 0.7787
---- Validation ----
Validation loss: 80.6519
Validation acc: 0.7364
Time taken: 38.93s
Start of epoch 16
Training loss (for one batch) at step 0: 293.9340, Accuracy: 0.8000
Training loss (for one batch) at step 10: 285.1624, Accuracy: 0.7800
Training loss (for one batch) at step 20: 286.4086, Accuracy: 0.7862
Training loss (for one batch) at step 30: 276.6572, Accuracy: 0.7942
Training loss (for one batch) at step 40: 278.9475, Accuracy: 0.7932
Training loss (for one batch) at step 50: 281.2241, Accuracy: 0.7986
Training loss (for one batch) at step 60: 292.4484, Accuracy: 0.7962
Training loss (for one batch) at step 70: 282.9411, Accuracy: 0.7945
Training loss (for one batch) at step 80: 287.3175, Accuracy: 0.7922
Training loss (for one batch) at step 90: 299.9920, Accuracy: 0.7909
Training loss (for one batch) at step 100: 289.0371, Accuracy: 0.7898
Training loss (for one batch) at step 110: 272.8687, Accuracy: 0.7903
Training loss (for one batch) at step 120: 282.2369, Accuracy: 0.7911
Training loss (for one batch) at step 130: 269.1110, Accuracy: 0.7907
Training loss (for one batch) at step 140: 301.0891, Accuracy: 0.7901
---- Training ----
Training loss: 242.8848
Training acc over epoch: 0.7902
---- Validation ----
Validation loss: 70.5960
Validation acc: 0.7472
Time taken: 65.75s
Start of epoch 17
Training loss (for one batch) at step 0: 277.0214, Accuracy: 0.7800
Training loss (for one batch) at step 10: 282.9412, Accuracy: 0.8136
Training loss (for one batch) at step 20: 271.3755, Accuracy: 0.7905
Training loss (for one batch) at step 30: 282.5807, Accuracy: 0.7890
Training loss (for one batch) at step 40: 285.6729, Accuracy: 0.7907
Training loss (for one batch) at step 50: 287.7717, Accuracy: 0.7927
Training loss (for one batch) at step 60: 283.2824, Accuracy: 0.7959
Training loss (for one batch) at step 70: 302.6453, Accuracy: 0.7954
Training loss (for one batch) at step 80: 277.1131, Accuracy: 0.7938
Training loss (for one batch) at step 90: 272.5903, Accuracy: 0.7919
Training loss (for one batch) at step 100: 294.7510, Accuracy: 0.7920
Training loss (for one batch) at step 110: 275.6088, Accuracy: 0.7915
Training loss (for one batch) at step 120: 288.3119, Accuracy: 0.7923
Training loss (for one batch) at step 130: 283.3275, Accuracy: 0.7929
Training loss (for one batch) at step 140: 278.0891, Accuracy: 0.7924
---- Training ----
Training loss: 258.0831
Training acc over epoch: 0.7915
---- Validation ----
Validation loss: 67.1297
Validation acc: 0.7163
Time taken: 38.01s
Start of epoch 18
Training loss (for one batch) at step 0: 267.0449, Accuracy: 0.8200
Training loss (for one batch) at step 10: 274.1760, Accuracy: 0.8073
Training loss (for one batch) at step 20: 260.7248, Accuracy: 0.7962
Training loss (for one batch) at step 30: 279.9407, Accuracy: 0.7987
Training loss (for one batch) at step 40: 281.8289, Accuracy: 0.7941
Training loss (for one batch) at step 50: 264.4958, Accuracy: 0.7996
Training loss (for one batch) at step 60: 258.0450, Accuracy: 0.7985
Training loss (for one batch) at step 70: 274.3603, Accuracy: 0.7979
Training loss (for one batch) at step 80: 277.2342, Accuracy: 0.7957
Training loss (for one batch) at step 90: 287.3586, Accuracy: 0.7932
Training loss (for one batch) at step 100: 269.6772, Accuracy: 0.7934
Training loss (for one batch) at step 110: 278.3929, Accuracy: 0.7946
Training loss (for one batch) at step 120: 257.8979, Accuracy: 0.7965
Training loss (for one batch) at step 130: 272.9823, Accuracy: 0.7952
Training loss (for one batch) at step 140: 282.2632, Accuracy: 0.7954
---- Training ----
Training loss: 269.7241
Training acc over epoch: 0.7953
---- Validation ----
Validation loss: 74.4474
Validation acc: 0.7292
Time taken: 66.02s
Start of epoch 19
Training loss (for one batch) at step 0: 281.8826, Accuracy: 0.7500
Training loss (for one batch) at step 10: 274.4663, Accuracy: 0.8173
Training loss (for one batch) at step 20: 268.7200, Accuracy: 0.8062
Training loss (for one batch) at step 30: 290.0397, Accuracy: 0.7965
Training loss (for one batch) at step 40: 255.9625, Accuracy: 0.7980
Training loss (for one batch) at step 50: 258.1797, Accuracy: 0.8075
Training loss (for one batch) at step 60: 281.1497, Accuracy: 0.8064
Training loss (for one batch) at step 70: 281.7207, Accuracy: 0.8072
Training loss (for one batch) at step 80: 262.9015, Accuracy: 0.8022
Training loss (for one batch) at step 90: 273.4463, Accuracy: 0.8014
Training loss (for one batch) at step 100: 270.0117, Accuracy: 0.7982
Training loss (for one batch) at step 110: 244.3165, Accuracy: 0.8010
Training loss (for one batch) at step 120: 270.0508, Accuracy: 0.7996
Training loss (for one batch) at step 130: 268.8502, Accuracy: 0.7988
Training loss (for one batch) at step 140: 260.0322, Accuracy: 0.7992
---- Training ----
Training loss: 269.8491
Training acc over epoch: 0.8003
---- Validation ----
Validation loss: 72.2253
Validation acc: 0.7166
Time taken: 39.84s
Start of epoch 20
Training loss (for one batch) at step 0: 283.3231, Accuracy: 0.7800
Training loss (for one batch) at step 10: 265.2091, Accuracy: 0.8118
Training loss (for one batch) at step 20: 262.2683, Accuracy: 0.8024
Training loss (for one batch) at step 30: 286.8758, Accuracy: 0.8055
Training loss (for one batch) at step 40: 262.9738, Accuracy: 0.8061
Training loss (for one batch) at step 50: 272.9333, Accuracy: 0.8080
Training loss (for one batch) at step 60: 272.6455, Accuracy: 0.8098
Training loss (for one batch) at step 70: 275.4010, Accuracy: 0.8077
Training loss (for one batch) at step 80: 261.3798, Accuracy: 0.8037
Training loss (for one batch) at step 90: 260.1746, Accuracy: 0.8040
Training loss (for one batch) at step 100: 284.5849, Accuracy: 0.8023
Training loss (for one batch) at step 110: 270.2447, Accuracy: 0.8016
Training loss (for one batch) at step 120: 283.0817, Accuracy: 0.8021
Training loss (for one batch) at step 130: 276.7613, Accuracy: 0.8011
Training loss (for one batch) at step 140: 271.3142, Accuracy: 0.8013
---- Training ----
Training loss: 228.9207
Training acc over epoch: 0.8018
---- Validation ----
Validation loss: 70.2356
Validation acc: 0.7496
Time taken: 66.29s
Start of epoch 21
Training loss (for one batch) at step 0: 261.5767, Accuracy: 0.8100
Training loss (for one batch) at step 10: 264.8340, Accuracy: 0.8155
Training loss (for one batch) at step 20: 284.0477, Accuracy: 0.8052
Training loss (for one batch) at step 30: 265.1070, Accuracy: 0.8048
Training loss (for one batch) at step 40: 266.2586, Accuracy: 0.8090
Training loss (for one batch) at step 50: 257.8621, Accuracy: 0.8092
Training loss (for one batch) at step 60: 259.6110, Accuracy: 0.8118
Training loss (for one batch) at step 70: 263.3839, Accuracy: 0.8085
Training loss (for one batch) at step 80: 270.4365, Accuracy: 0.8054
Training loss (for one batch) at step 90: 267.5217, Accuracy: 0.8042
Training loss (for one batch) at step 100: 262.6905, Accuracy: 0.8029
Training loss (for one batch) at step 110: 253.4150, Accuracy: 0.8039
Training loss (for one batch) at step 120: 276.2479, Accuracy: 0.8050
Training loss (for one batch) at step 130: 268.2697, Accuracy: 0.8044
Training loss (for one batch) at step 140: 265.4790, Accuracy: 0.8029
---- Training ----
Training loss: 245.3509
Training acc over epoch: 0.8025
---- Validation ----
Validation loss: 74.2644
Validation acc: 0.7466
Time taken: 38.38s
Start of epoch 22
Training loss (for one batch) at step 0: 252.2912, Accuracy: 0.8500
Training loss (for one batch) at step 10: 270.3866, Accuracy: 0.8291
Training loss (for one batch) at step 20: 265.6100, Accuracy: 0.8186
Training loss (for one batch) at step 30: 277.3905, Accuracy: 0.8158
Training loss (for one batch) at step 40: 239.8378, Accuracy: 0.8185
Training loss (for one batch) at step 50: 256.0900, Accuracy: 0.8200
Training loss (for one batch) at step 60: 252.8295, Accuracy: 0.8164
Training loss (for one batch) at step 70: 260.9091, Accuracy: 0.8155
Training loss (for one batch) at step 80: 262.5147, Accuracy: 0.8125
Training loss (for one batch) at step 90: 261.9024, Accuracy: 0.8127
Training loss (for one batch) at step 100: 268.8061, Accuracy: 0.8125
Training loss (for one batch) at step 110: 236.9369, Accuracy: 0.8129
Training loss (for one batch) at step 120: 272.2038, Accuracy: 0.8129
Training loss (for one batch) at step 130: 248.0572, Accuracy: 0.8128
Training loss (for one batch) at step 140: 262.9789, Accuracy: 0.8121
---- Training ----
Training loss: 221.2986
Training acc over epoch: 0.8123
---- Validation ----
Validation loss: 64.7940
Validation acc: 0.7354
Time taken: 65.56s
Start of epoch 23
Training loss (for one batch) at step 0: 261.0020, Accuracy: 0.7900
Training loss (for one batch) at step 10: 263.5709, Accuracy: 0.8109
Training loss (for one batch) at step 20: 258.6952, Accuracy: 0.8024
Training loss (for one batch) at step 30: 273.9492, Accuracy: 0.8052
Training loss (for one batch) at step 40: 255.9412, Accuracy: 0.8134
Training loss (for one batch) at step 50: 251.2284, Accuracy: 0.8127
Training loss (for one batch) at step 60: 256.5804, Accuracy: 0.8098
Training loss (for one batch) at step 70: 265.2065, Accuracy: 0.8099
Training loss (for one batch) at step 80: 307.5951, Accuracy: 0.8059
Training loss (for one batch) at step 90: 254.5491, Accuracy: 0.8027
Training loss (for one batch) at step 100: 265.7367, Accuracy: 0.8005
Training loss (for one batch) at step 110: 232.5560, Accuracy: 0.8006
Training loss (for one batch) at step 120: 259.8309, Accuracy: 0.8012
Training loss (for one batch) at step 130: 285.8835, Accuracy: 0.8030
Training loss (for one batch) at step 140: 286.5533, Accuracy: 0.8046
---- Training ----
Training loss: 225.5977
Training acc over epoch: 0.8053
---- Validation ----
Validation loss: 79.5885
Validation acc: 0.7230
Time taken: 37.99s
Start of epoch 24
Training loss (for one batch) at step 0: 253.1805, Accuracy: 0.8100
Training loss (for one batch) at step 10: 260.0168, Accuracy: 0.8091
Training loss (for one batch) at step 20: 259.2551, Accuracy: 0.8129
Training loss (for one batch) at step 30: 266.1500, Accuracy: 0.8103
Training loss (for one batch) at step 40: 248.7327, Accuracy: 0.8129
Training loss (for one batch) at step 50: 269.4868, Accuracy: 0.8161
Training loss (for one batch) at step 60: 244.5789, Accuracy: 0.8177
Training loss (for one batch) at step 70: 264.4471, Accuracy: 0.8159
Training loss (for one batch) at step 80: 254.6012, Accuracy: 0.8151
Training loss (for one batch) at step 90: 266.1863, Accuracy: 0.8122
Training loss (for one batch) at step 100: 265.5896, Accuracy: 0.8114
Training loss (for one batch) at step 110: 274.1886, Accuracy: 0.8105
Training loss (for one batch) at step 120: 246.1082, Accuracy: 0.8121
Training loss (for one batch) at step 130: 258.8216, Accuracy: 0.8113
Training loss (for one batch) at step 140: 265.9856, Accuracy: 0.8104
---- Training ----
Training loss: 228.0304
Training acc over epoch: 0.8108
---- Validation ----
Validation loss: 77.5830
Validation acc: 0.7233
Time taken: 65.46s
Start of epoch 25
Training loss (for one batch) at step 0: 260.5574, Accuracy: 0.7600
Training loss (for one batch) at step 10: 259.3604, Accuracy: 0.8100
Training loss (for one batch) at step 20: 242.0190, Accuracy: 0.8090
Training loss (for one batch) at step 30: 228.5586, Accuracy: 0.8171
Training loss (for one batch) at step 40: 242.7536, Accuracy: 0.8159
Training loss (for one batch) at step 50: 247.6724, Accuracy: 0.8210
Training loss (for one batch) at step 60: 248.3274, Accuracy: 0.8249
Training loss (for one batch) at step 70: 272.1266, Accuracy: 0.8224
Training loss (for one batch) at step 80: 263.2000, Accuracy: 0.8180
Training loss (for one batch) at step 90: 258.9083, Accuracy: 0.8173
Training loss (for one batch) at step 100: 258.8387, Accuracy: 0.8163
Training loss (for one batch) at step 110: 242.9895, Accuracy: 0.8145
Training loss (for one batch) at step 120: 273.8111, Accuracy: 0.8146
Training loss (for one batch) at step 130: 234.3223, Accuracy: 0.8140
Training loss (for one batch) at step 140: 267.3054, Accuracy: 0.8140
---- Training ----
Training loss: 239.2564
Training acc over epoch: 0.8135
---- Validation ----
Validation loss: 72.5420
Validation acc: 0.7214
Time taken: 38.26s
Start of epoch 26
Training loss (for one batch) at step 0: 252.9211, Accuracy: 0.7600
Training loss (for one batch) at step 10: 250.2838, Accuracy: 0.8109
Training loss (for one batch) at step 20: 255.2143, Accuracy: 0.8090
Training loss (for one batch) at step 30: 258.8029, Accuracy: 0.8106
Training loss (for one batch) at step 40: 231.0898, Accuracy: 0.8139
Training loss (for one batch) at step 50: 261.2003, Accuracy: 0.8165
Training loss (for one batch) at step 60: 253.1717, Accuracy: 0.8159
Training loss (for one batch) at step 70: 242.5367, Accuracy: 0.8162
Training loss (for one batch) at step 80: 256.9031, Accuracy: 0.8146
Training loss (for one batch) at step 90: 267.7472, Accuracy: 0.8135
Training loss (for one batch) at step 100: 264.5642, Accuracy: 0.8116
Training loss (for one batch) at step 110: 247.3413, Accuracy: 0.8132
Training loss (for one batch) at step 120: 266.5620, Accuracy: 0.8141
Training loss (for one batch) at step 130: 279.6115, Accuracy: 0.8153
Training loss (for one batch) at step 140: 226.5630, Accuracy: 0.8141
---- Training ----
Training loss: 226.3137
Training acc over epoch: 0.8135
---- Validation ----
Validation loss: 72.9198
Validation acc: 0.7206
Time taken: 65.04s
Start of epoch 27
Training loss (for one batch) at step 0: 275.8741, Accuracy: 0.7400
Training loss (for one batch) at step 10: 228.7854, Accuracy: 0.8155
Training loss (for one batch) at step 20: 251.8837, Accuracy: 0.8267
Training loss (for one batch) at step 30: 251.2052, Accuracy: 0.8245
Training loss (for one batch) at step 40: 250.4408, Accuracy: 0.8259
Training loss (for one batch) at step 50: 235.0696, Accuracy: 0.8267
Training loss (for one batch) at step 60: 251.1914, Accuracy: 0.8266
Training loss (for one batch) at step 70: 254.8408, Accuracy: 0.8249
Training loss (for one batch) at step 80: 242.4602, Accuracy: 0.8247
Training loss (for one batch) at step 90: 248.1003, Accuracy: 0.8237
Training loss (for one batch) at step 100: 257.4386, Accuracy: 0.8211
Training loss (for one batch) at step 110: 244.1203, Accuracy: 0.8192
Training loss (for one batch) at step 120: 247.9332, Accuracy: 0.8199
Training loss (for one batch) at step 130: 244.9213, Accuracy: 0.8196
Training loss (for one batch) at step 140: 256.4597, Accuracy: 0.8177
---- Training ----
Training loss: 204.4540
Training acc over epoch: 0.8175
---- Validation ----
Validation loss: 54.5762
Validation acc: 0.7380
Time taken: 37.56s
Start of epoch 28
Training loss (for one batch) at step 0: 263.3540, Accuracy: 0.7700
Training loss (for one batch) at step 10: 259.2111, Accuracy: 0.8136
Training loss (for one batch) at step 20: 246.8516, Accuracy: 0.8205
Training loss (for one batch) at step 30: 238.2427, Accuracy: 0.8190
Training loss (for one batch) at step 40: 216.3518, Accuracy: 0.8241
Training loss (for one batch) at step 50: 232.5271, Accuracy: 0.8275
Training loss (for one batch) at step 60: 257.6609, Accuracy: 0.8266
Training loss (for one batch) at step 70: 271.8972, Accuracy: 0.8232
Training loss (for one batch) at step 80: 242.0723, Accuracy: 0.8200
Training loss (for one batch) at step 90: 261.6023, Accuracy: 0.8181
Training loss (for one batch) at step 100: 244.6490, Accuracy: 0.8198
Training loss (for one batch) at step 110: 249.1104, Accuracy: 0.8188
Training loss (for one batch) at step 120: 231.2841, Accuracy: 0.8204
Training loss (for one batch) at step 130: 233.8869, Accuracy: 0.8173
Training loss (for one batch) at step 140: 244.7898, Accuracy: 0.8178
---- Training ----
Training loss: 225.4135
Training acc over epoch: 0.8180
---- Validation ----
Validation loss: 69.9166
Validation acc: 0.7364
Time taken: 65.86s
Start of epoch 29
Training loss (for one batch) at step 0: 240.2009, Accuracy: 0.8700
Training loss (for one batch) at step 10: 232.1888, Accuracy: 0.8300
Training loss (for one batch) at step 20: 256.5484, Accuracy: 0.8262
Training loss (for one batch) at step 30: 233.2800, Accuracy: 0.8226
Training loss (for one batch) at step 40: 245.3730, Accuracy: 0.8234
Training loss (for one batch) at step 50: 249.1418, Accuracy: 0.8261
Training loss (for one batch) at step 60: 231.9691, Accuracy: 0.8264
Training loss (for one batch) at step 70: 247.3479, Accuracy: 0.8234
Training loss (for one batch) at step 80: 251.1865, Accuracy: 0.8228
Training loss (for one batch) at step 90: 241.9819, Accuracy: 0.8235
Training loss (for one batch) at step 100: 233.8342, Accuracy: 0.8239
Training loss (for one batch) at step 110: 242.0074, Accuracy: 0.8239
Training loss (for one batch) at step 120: 220.5725, Accuracy: 0.8236
Training loss (for one batch) at step 130: 239.0746, Accuracy: 0.8226
Training loss (for one batch) at step 140: 240.2853, Accuracy: 0.8220
---- Training ----
Training loss: 216.5283
Training acc over epoch: 0.8231
---- Validation ----
Validation loss: 62.2217
Validation acc: 0.7335
Time taken: 38.02s
Start of epoch 30
Training loss (for one batch) at step 0: 241.6787, Accuracy: 0.8400
Training loss (for one batch) at step 10: 248.3027, Accuracy: 0.7973
Training loss (for one batch) at step 20: 234.7922, Accuracy: 0.8071
Training loss (for one batch) at step 30: 233.5747, Accuracy: 0.8187
Training loss (for one batch) at step 40: 230.8037, Accuracy: 0.8222
Training loss (for one batch) at step 50: 222.7518, Accuracy: 0.8237
Training loss (for one batch) at step 60: 231.8712, Accuracy: 0.8264
Training loss (for one batch) at step 70: 234.8821, Accuracy: 0.8254
Training loss (for one batch) at step 80: 218.3129, Accuracy: 0.8251
Training loss (for one batch) at step 90: 234.8781, Accuracy: 0.8244
Training loss (for one batch) at step 100: 227.4222, Accuracy: 0.8244
Training loss (for one batch) at step 110: 248.7650, Accuracy: 0.8238
Training loss (for one batch) at step 120: 260.3617, Accuracy: 0.8250
Training loss (for one batch) at step 130: 246.1893, Accuracy: 0.8244
Training loss (for one batch) at step 140: 228.9409, Accuracy: 0.8236
---- Training ----
Training loss: 244.2156
Training acc over epoch: 0.8225
---- Validation ----
Validation loss: 70.0685
Validation acc: 0.7300
Time taken: 64.26s
Start of epoch 31
Training loss (for one batch) at step 0: 243.0386, Accuracy: 0.7800
Training loss (for one batch) at step 10: 251.9852, Accuracy: 0.8209
Training loss (for one batch) at step 20: 257.1714, Accuracy: 0.8171
Training loss (for one batch) at step 30: 241.6427, Accuracy: 0.8171
Training loss (for one batch) at step 40: 239.5426, Accuracy: 0.8254
Training loss (for one batch) at step 50: 244.6678, Accuracy: 0.8243
Training loss (for one batch) at step 60: 229.5812, Accuracy: 0.8244
Training loss (for one batch) at step 70: 239.7380, Accuracy: 0.8245
Training loss (for one batch) at step 80: 247.2630, Accuracy: 0.8233
Training loss (for one batch) at step 90: 251.6795, Accuracy: 0.8225
Training loss (for one batch) at step 100: 243.9653, Accuracy: 0.8208
Training loss (for one batch) at step 110: 229.6485, Accuracy: 0.8215
Training loss (for one batch) at step 120: 252.0766, Accuracy: 0.8223
Training loss (for one batch) at step 130: 227.1138, Accuracy: 0.8201
Training loss (for one batch) at step 140: 229.0114, Accuracy: 0.8191
---- Training ----
Training loss: 224.5929
Training acc over epoch: 0.8197
---- Validation ----
Validation loss: 69.2726
Validation acc: 0.7211
Time taken: 37.62s
Start of epoch 32
Training loss (for one batch) at step 0: 271.6505, Accuracy: 0.7700
Training loss (for one batch) at step 10: 220.2006, Accuracy: 0.8391
Training loss (for one batch) at step 20: 245.9800, Accuracy: 0.8295
Training loss (for one batch) at step 30: 255.8607, Accuracy: 0.8290
Training loss (for one batch) at step 40: 240.1856, Accuracy: 0.8268
Training loss (for one batch) at step 50: 217.3540, Accuracy: 0.8314
Training loss (for one batch) at step 60: 232.1200, Accuracy: 0.8298
Training loss (for one batch) at step 70: 238.8280, Accuracy: 0.8270
Training loss (for one batch) at step 80: 234.9357, Accuracy: 0.8241
Training loss (for one batch) at step 90: 258.4650, Accuracy: 0.8246
Training loss (for one batch) at step 100: 249.7270, Accuracy: 0.8234
Training loss (for one batch) at step 110: 250.2388, Accuracy: 0.8226
Training loss (for one batch) at step 120: 228.7269, Accuracy: 0.8225
Training loss (for one batch) at step 130: 256.3141, Accuracy: 0.8225
Training loss (for one batch) at step 140: 224.9754, Accuracy: 0.8227
---- Training ----
Training loss: 223.0587
Training acc over epoch: 0.8223
---- Validation ----
Validation loss: 73.4559
Validation acc: 0.7217
Time taken: 65.59s
Start of epoch 33
Training loss (for one batch) at step 0: 220.2015, Accuracy: 0.8700
Training loss (for one batch) at step 10: 229.7063, Accuracy: 0.8427
Training loss (for one batch) at step 20: 235.6014, Accuracy: 0.8424
Training loss (for one batch) at step 30: 229.0674, Accuracy: 0.8365
Training loss (for one batch) at step 40: 239.4333, Accuracy: 0.8373
Training loss (for one batch) at step 50: 230.2215, Accuracy: 0.8394
Training loss (for one batch) at step 60: 231.3838, Accuracy: 0.8400
Training loss (for one batch) at step 70: 238.2717, Accuracy: 0.8386
Training loss (for one batch) at step 80: 246.0764, Accuracy: 0.8341
Training loss (for one batch) at step 90: 247.5864, Accuracy: 0.8334
Training loss (for one batch) at step 100: 233.3765, Accuracy: 0.8319
Training loss (for one batch) at step 110: 232.3233, Accuracy: 0.8328
Training loss (for one batch) at step 120: 245.4573, Accuracy: 0.8296
Training loss (for one batch) at step 130: 237.8051, Accuracy: 0.8289
Training loss (for one batch) at step 140: 237.6152, Accuracy: 0.8290
---- Training ----
Training loss: 229.0773
Training acc over epoch: 0.8287
---- Validation ----
Validation loss: 75.4221
Validation acc: 0.7225
Time taken: 38.13s
Start of epoch 34
Training loss (for one batch) at step 0: 248.8514, Accuracy: 0.7500
Training loss (for one batch) at step 10: 227.8739, Accuracy: 0.8282
Training loss (for one batch) at step 20: 251.3698, Accuracy: 0.8329
Training loss (for one batch) at step 30: 217.5492, Accuracy: 0.8348
Training loss (for one batch) at step 40: 233.7143, Accuracy: 0.8295
Training loss (for one batch) at step 50: 234.8620, Accuracy: 0.8308
Training loss (for one batch) at step 60: 246.6546, Accuracy: 0.8289
Training loss (for one batch) at step 70: 231.1816, Accuracy: 0.8300
Training loss (for one batch) at step 80: 243.9612, Accuracy: 0.8263
Training loss (for one batch) at step 90: 231.9786, Accuracy: 0.8273
Training loss (for one batch) at step 100: 229.2979, Accuracy: 0.8265
Training loss (for one batch) at step 110: 251.5372, Accuracy: 0.8267
Training loss (for one batch) at step 120: 225.8591, Accuracy: 0.8276
Training loss (for one batch) at step 130: 242.9096, Accuracy: 0.8262
Training loss (for one batch) at step 140: 233.1262, Accuracy: 0.8273
---- Training ----
Training loss: 196.5769
Training acc over epoch: 0.8265
---- Validation ----
Validation loss: 67.8755
Validation acc: 0.7429
Time taken: 72.57s
Start of epoch 35
Training loss (for one batch) at step 0: 251.6011, Accuracy: 0.7900
Training loss (for one batch) at step 10: 240.7859, Accuracy: 0.8309
Training loss (for one batch) at step 20: 228.1844, Accuracy: 0.8348
Training loss (for one batch) at step 30: 238.0390, Accuracy: 0.8277
Training loss (for one batch) at step 40: 219.0575, Accuracy: 0.8320
Training loss (for one batch) at step 50: 233.7171, Accuracy: 0.8341
Training loss (for one batch) at step 60: 219.2300, Accuracy: 0.8339
Training loss (for one batch) at step 70: 251.2744, Accuracy: 0.8296
Training loss (for one batch) at step 80: 233.7988, Accuracy: 0.8302
Training loss (for one batch) at step 90: 221.3503, Accuracy: 0.8278
Training loss (for one batch) at step 100: 220.4608, Accuracy: 0.8268
Training loss (for one batch) at step 110: 244.2804, Accuracy: 0.8273
Training loss (for one batch) at step 120: 213.3488, Accuracy: 0.8276
Training loss (for one batch) at step 130: 236.7100, Accuracy: 0.8267
Training loss (for one batch) at step 140: 229.8438, Accuracy: 0.8251
---- Training ----
Training loss: 208.8496
Training acc over epoch: 0.8257
---- Validation ----
Validation loss: 70.2737
Validation acc: 0.7125
Time taken: 38.47s
Start of epoch 36
Training loss (for one batch) at step 0: 228.1042, Accuracy: 0.8300
Training loss (for one batch) at step 10: 229.1110, Accuracy: 0.8091
Training loss (for one batch) at step 20: 230.0180, Accuracy: 0.8210
Training loss (for one batch) at step 30: 246.5198, Accuracy: 0.8261
Training loss (for one batch) at step 40: 221.1021, Accuracy: 0.8293
Training loss (for one batch) at step 50: 222.1232, Accuracy: 0.8324
Training loss (for one batch) at step 60: 219.4864, Accuracy: 0.8326
Training loss (for one batch) at step 70: 218.5452, Accuracy: 0.8311
Training loss (for one batch) at step 80: 234.2761, Accuracy: 0.8311
Training loss (for one batch) at step 90: 240.7597, Accuracy: 0.8291
Training loss (for one batch) at step 100: 242.9129, Accuracy: 0.8281
Training loss (for one batch) at step 110: 248.8243, Accuracy: 0.8270
Training loss (for one batch) at step 120: 225.5729, Accuracy: 0.8266
Training loss (for one batch) at step 130: 222.3100, Accuracy: 0.8266
Training loss (for one batch) at step 140: 240.8463, Accuracy: 0.8245
---- Training ----
Training loss: 192.1903
Training acc over epoch: 0.8245
---- Validation ----
Validation loss: 55.7181
Validation acc: 0.7208
Time taken: 65.86s
Start of epoch 37
Training loss (for one batch) at step 0: 235.8002, Accuracy: 0.8300
Training loss (for one batch) at step 10: 244.8454, Accuracy: 0.8464
Training loss (for one batch) at step 20: 213.9529, Accuracy: 0.8386
Training loss (for one batch) at step 30: 227.9485, Accuracy: 0.8297
Training loss (for one batch) at step 40: 221.9188, Accuracy: 0.8305
Training loss (for one batch) at step 50: 207.3791, Accuracy: 0.8312
Training loss (for one batch) at step 60: 226.3200, Accuracy: 0.8339
Training loss (for one batch) at step 70: 234.2954, Accuracy: 0.8301
Training loss (for one batch) at step 80: 246.6668, Accuracy: 0.8268
Training loss (for one batch) at step 90: 251.1544, Accuracy: 0.8268
Training loss (for one batch) at step 100: 232.6071, Accuracy: 0.8276
Training loss (for one batch) at step 110: 226.9445, Accuracy: 0.8266
Training loss (for one batch) at step 120: 234.8217, Accuracy: 0.8274
Training loss (for one batch) at step 130: 234.0620, Accuracy: 0.8269
Training loss (for one batch) at step 140: 225.7990, Accuracy: 0.8266
---- Training ----
Training loss: 193.7419
Training acc over epoch: 0.8263
---- Validation ----
Validation loss: 59.9010
Validation acc: 0.7233
Time taken: 38.45s
Start of epoch 38
Training loss (for one batch) at step 0: 229.3847, Accuracy: 0.7500
Training loss (for one batch) at step 10: 240.2244, Accuracy: 0.8236
Training loss (for one batch) at step 20: 216.0325, Accuracy: 0.8343
Training loss (for one batch) at step 30: 229.5385, Accuracy: 0.8306
Training loss (for one batch) at step 40: 206.5859, Accuracy: 0.8332
Training loss (for one batch) at step 50: 244.8557, Accuracy: 0.8380
Training loss (for one batch) at step 60: 212.8445, Accuracy: 0.8354
Training loss (for one batch) at step 70: 220.3538, Accuracy: 0.8349
Training loss (for one batch) at step 80: 219.2851, Accuracy: 0.8321
Training loss (for one batch) at step 90: 219.6088, Accuracy: 0.8310
Training loss (for one batch) at step 100: 208.5105, Accuracy: 0.8307
Training loss (for one batch) at step 110: 241.8223, Accuracy: 0.8301
Training loss (for one batch) at step 120: 227.5181, Accuracy: 0.8314
Training loss (for one batch) at step 130: 232.9738, Accuracy: 0.8319
Training loss (for one batch) at step 140: 231.8808, Accuracy: 0.8323
---- Training ----
Training loss: 197.7151
Training acc over epoch: 0.8326
---- Validation ----
Validation loss: 86.0767
Validation acc: 0.7163
Time taken: 95.50s
Start of epoch 39
Training loss (for one batch) at step 0: 261.9899, Accuracy: 0.7800
Training loss (for one batch) at step 10: 213.6380, Accuracy: 0.8345
Training loss (for one batch) at step 20: 242.8512, Accuracy: 0.8271
Training loss (for one batch) at step 30: 227.1693, Accuracy: 0.8277
Training loss (for one batch) at step 40: 241.1911, Accuracy: 0.8354
Training loss (for one batch) at step 50: 230.3850, Accuracy: 0.8353
Training loss (for one batch) at step 60: 205.4429, Accuracy: 0.8374
Training loss (for one batch) at step 70: 212.5689, Accuracy: 0.8390
Training loss (for one batch) at step 80: 231.4029, Accuracy: 0.8340
Training loss (for one batch) at step 90: 217.5745, Accuracy: 0.8336
Training loss (for one batch) at step 100: 225.2627, Accuracy: 0.8318
Training loss (for one batch) at step 110: 222.2124, Accuracy: 0.8329
Training loss (for one batch) at step 120: 242.0108, Accuracy: 0.8333
Training loss (for one batch) at step 130: 227.8949, Accuracy: 0.8335
Training loss (for one batch) at step 140: 242.5297, Accuracy: 0.8318
---- Training ----
Training loss: 199.7805
Training acc over epoch: 0.8324
---- Validation ----
Validation loss: 77.9836
Validation acc: 0.7166
Time taken: 37.42s
Start of epoch 40
Training loss (for one batch) at step 0: 213.9440, Accuracy: 0.8700
Training loss (for one batch) at step 10: 203.8948, Accuracy: 0.8427
Training loss (for one batch) at step 20: 199.8266, Accuracy: 0.8438
Training loss (for one batch) at step 30: 200.7892, Accuracy: 0.8384
Training loss (for one batch) at step 40: 208.9452, Accuracy: 0.8354
Training loss (for one batch) at step 50: 227.5830, Accuracy: 0.8375
Training loss (for one batch) at step 60: 215.2776, Accuracy: 0.8410
Training loss (for one batch) at step 70: 218.1238, Accuracy: 0.8393
Training loss (for one batch) at step 80: 224.6337, Accuracy: 0.8353
Training loss (for one batch) at step 90: 218.9991, Accuracy: 0.8326
Training loss (for one batch) at step 100: 216.6648, Accuracy: 0.8323
Training loss (for one batch) at step 110: 218.1716, Accuracy: 0.8345
Training loss (for one batch) at step 120: 212.0941, Accuracy: 0.8342
Training loss (for one batch) at step 130: 228.5353, Accuracy: 0.8332
Training loss (for one batch) at step 140: 233.0268, Accuracy: 0.8320
---- Training ----
Training loss: 183.2200
Training acc over epoch: 0.8320
---- Validation ----
Validation loss: 72.4891
Validation acc: 0.7222
Time taken: 66.00s
Start of epoch 41
Training loss (for one batch) at step 0: 212.2791, Accuracy: 0.8500
Training loss (for one batch) at step 10: 217.7918, Accuracy: 0.8364
Training loss (for one batch) at step 20: 222.9600, Accuracy: 0.8300
Training loss (for one batch) at step 30: 212.4073, Accuracy: 0.8358
Training loss (for one batch) at step 40: 209.3708, Accuracy: 0.8388
Training loss (for one batch) at step 50: 233.0493, Accuracy: 0.8371
Training loss (for one batch) at step 60: 228.9935, Accuracy: 0.8377
Training loss (for one batch) at step 70: 237.9861, Accuracy: 0.8359
Training loss (for one batch) at step 80: 218.7483, Accuracy: 0.8386
Training loss (for one batch) at step 90: 248.5565, Accuracy: 0.8334
Training loss (for one batch) at step 100: 227.8262, Accuracy: 0.8339
Training loss (for one batch) at step 110: 215.5889, Accuracy: 0.8346
Training loss (for one batch) at step 120: 205.8310, Accuracy: 0.8355
Training loss (for one batch) at step 130: 215.7945, Accuracy: 0.8350
Training loss (for one batch) at step 140: 235.8431, Accuracy: 0.8338
---- Training ----
Training loss: 197.8561
Training acc over epoch: 0.8336
---- Validation ----
Validation loss: 60.8140
Validation acc: 0.7184
Time taken: 37.70s
Start of epoch 42
Training loss (for one batch) at step 0: 214.6485, Accuracy: 0.8700
Training loss (for one batch) at step 10: 181.9791, Accuracy: 0.8427
Training loss (for one batch) at step 20: 224.0612, Accuracy: 0.8390
Training loss (for one batch) at step 30: 215.2913, Accuracy: 0.8368
Training loss (for one batch) at step 40: 222.3994, Accuracy: 0.8366
Training loss (for one batch) at step 50: 240.0242, Accuracy: 0.8388
Training loss (for one batch) at step 60: 211.2804, Accuracy: 0.8418
Training loss (for one batch) at step 70: 225.1783, Accuracy: 0.8415
Training loss (for one batch) at step 80: 224.0579, Accuracy: 0.8405
Training loss (for one batch) at step 90: 205.6766, Accuracy: 0.8401
Training loss (for one batch) at step 100: 225.3220, Accuracy: 0.8372
Training loss (for one batch) at step 110: 216.5259, Accuracy: 0.8369
Training loss (for one batch) at step 120: 197.8819, Accuracy: 0.8383
Training loss (for one batch) at step 130: 235.7797, Accuracy: 0.8370
Training loss (for one batch) at step 140: 233.7490, Accuracy: 0.8360
---- Training ----
Training loss: 200.7546
Training acc over epoch: 0.8361
---- Validation ----
Validation loss: 82.5835
Validation acc: 0.7168
Time taken: 65.14s
Start of epoch 43
Training loss (for one batch) at step 0: 229.3561, Accuracy: 0.8200
Training loss (for one batch) at step 10: 198.8737, Accuracy: 0.8336
Training loss (for one batch) at step 20: 217.4108, Accuracy: 0.8400
Training loss (for one batch) at step 30: 218.5291, Accuracy: 0.8403
Training loss (for one batch) at step 40: 232.2681, Accuracy: 0.8376
Training loss (for one batch) at step 50: 226.4752, Accuracy: 0.8378
Training loss (for one batch) at step 60: 209.1878, Accuracy: 0.8392
Training loss (for one batch) at step 70: 225.9359, Accuracy: 0.8359
Training loss (for one batch) at step 80: 209.7132, Accuracy: 0.8331
Training loss (for one batch) at step 90: 222.9254, Accuracy: 0.8331
Training loss (for one batch) at step 100: 203.8576, Accuracy: 0.8331
Training loss (for one batch) at step 110: 206.8454, Accuracy: 0.8332
Training loss (for one batch) at step 120: 227.8678, Accuracy: 0.8341
Training loss (for one batch) at step 130: 213.7478, Accuracy: 0.8331
Training loss (for one batch) at step 140: 228.0531, Accuracy: 0.8328
---- Training ----
Training loss: 189.8096
Training acc over epoch: 0.8325
---- Validation ----
Validation loss: 69.9581
Validation acc: 0.7332
Time taken: 38.10s
Start of epoch 44
Training loss (for one batch) at step 0: 200.8908, Accuracy: 0.8900
Training loss (for one batch) at step 10: 219.6393, Accuracy: 0.8445
Training loss (for one batch) at step 20: 207.8801, Accuracy: 0.8352
Training loss (for one batch) at step 30: 246.4021, Accuracy: 0.8284
Training loss (for one batch) at step 40: 209.6073, Accuracy: 0.8349
Training loss (for one batch) at step 50: 199.8738, Accuracy: 0.8406
Training loss (for one batch) at step 60: 228.0004, Accuracy: 0.8387
Training loss (for one batch) at step 70: 249.1757, Accuracy: 0.8386
Training loss (for one batch) at step 80: 224.6043, Accuracy: 0.8363
Training loss (for one batch) at step 90: 227.6601, Accuracy: 0.8351
Training loss (for one batch) at step 100: 213.1256, Accuracy: 0.8359
Training loss (for one batch) at step 110: 232.2686, Accuracy: 0.8362
Training loss (for one batch) at step 120: 223.9907, Accuracy: 0.8379
Training loss (for one batch) at step 130: 247.7912, Accuracy: 0.8376
Training loss (for one batch) at step 140: 246.0294, Accuracy: 0.8362
---- Training ----
Training loss: 187.7875
Training acc over epoch: 0.8366
---- Validation ----
Validation loss: 70.7350
Validation acc: 0.7184
Time taken: 65.56s
Start of epoch 45
Training loss (for one batch) at step 0: 217.7732, Accuracy: 0.8800
Training loss (for one batch) at step 10: 234.4266, Accuracy: 0.8482
Training loss (for one batch) at step 20: 203.2027, Accuracy: 0.8405
Training loss (for one batch) at step 30: 216.2579, Accuracy: 0.8400
Training loss (for one batch) at step 40: 207.5538, Accuracy: 0.8417
Training loss (for one batch) at step 50: 213.3472, Accuracy: 0.8425
Training loss (for one batch) at step 60: 215.3906, Accuracy: 0.8361
Training loss (for one batch) at step 70: 197.4723, Accuracy: 0.8377
Training loss (for one batch) at step 80: 210.1010, Accuracy: 0.8359
Training loss (for one batch) at step 90: 195.2778, Accuracy: 0.8360
Training loss (for one batch) at step 100: 222.4199, Accuracy: 0.8350
Training loss (for one batch) at step 110: 206.8688, Accuracy: 0.8364
Training loss (for one batch) at step 120: 211.1022, Accuracy: 0.8363
Training loss (for one batch) at step 130: 221.0536, Accuracy: 0.8362
Training loss (for one batch) at step 140: 212.9427, Accuracy: 0.8353
---- Training ----
Training loss: 191.5314
Training acc over epoch: 0.8352
---- Validation ----
Validation loss: 77.9508
Validation acc: 0.7195
Time taken: 37.59s
Start of epoch 46
Training loss (for one batch) at step 0: 223.6536, Accuracy: 0.8700
Training loss (for one batch) at step 10: 210.7723, Accuracy: 0.8345
Training loss (for one batch) at step 20: 213.7711, Accuracy: 0.8305
Training loss (for one batch) at step 30: 212.7658, Accuracy: 0.8377
Training loss (for one batch) at step 40: 241.1566, Accuracy: 0.8412
Training loss (for one batch) at step 50: 213.4596, Accuracy: 0.8488
Training loss (for one batch) at step 60: 209.3451, Accuracy: 0.8480
Training loss (for one batch) at step 70: 231.4615, Accuracy: 0.8437
Training loss (for one batch) at step 80: 219.8414, Accuracy: 0.8438
Training loss (for one batch) at step 90: 216.0857, Accuracy: 0.8416
Training loss (for one batch) at step 100: 211.8239, Accuracy: 0.8401
Training loss (for one batch) at step 110: 217.0179, Accuracy: 0.8408
Training loss (for one batch) at step 120: 220.0503, Accuracy: 0.8409
Training loss (for one batch) at step 130: 247.5967, Accuracy: 0.8407
Training loss (for one batch) at step 140: 211.8790, Accuracy: 0.8401
---- Training ----
Training loss: 170.4977
Training acc over epoch: 0.8393
---- Validation ----
Validation loss: 87.0815
Validation acc: 0.7243
Time taken: 74.55s
Start of epoch 47
Training loss (for one batch) at step 0: 233.3850, Accuracy: 0.7800
Training loss (for one batch) at step 10: 216.5984, Accuracy: 0.8355
Training loss (for one batch) at step 20: 205.7735, Accuracy: 0.8348
Training loss (for one batch) at step 30: 196.3538, Accuracy: 0.8368
Training loss (for one batch) at step 40: 225.3056, Accuracy: 0.8383
Training loss (for one batch) at step 50: 196.2516, Accuracy: 0.8418
Training loss (for one batch) at step 60: 218.9918, Accuracy: 0.8418
Training loss (for one batch) at step 70: 233.2673, Accuracy: 0.8394
Training loss (for one batch) at step 80: 255.6149, Accuracy: 0.8381
Training loss (for one batch) at step 90: 200.5958, Accuracy: 0.8381
Training loss (for one batch) at step 100: 216.1486, Accuracy: 0.8366
Training loss (for one batch) at step 110: 228.9607, Accuracy: 0.8368
Training loss (for one batch) at step 120: 253.8981, Accuracy: 0.8372
Training loss (for one batch) at step 130: 208.7144, Accuracy: 0.8373
Training loss (for one batch) at step 140: 209.7704, Accuracy: 0.8379
---- Training ----
Training loss: 192.3657
Training acc over epoch: 0.8382
---- Validation ----
Validation loss: 80.3164
Validation acc: 0.7184
Time taken: 40.68s
Start of epoch 48
Training loss (for one batch) at step 0: 217.6216, Accuracy: 0.8100
Training loss (for one batch) at step 10: 216.5425, Accuracy: 0.8327
Training loss (for one batch) at step 20: 237.0836, Accuracy: 0.8371
Training loss (for one batch) at step 30: 206.3041, Accuracy: 0.8410
Training loss (for one batch) at step 40: 204.0766, Accuracy: 0.8417
Training loss (for one batch) at step 50: 232.2241, Accuracy: 0.8424
Training loss (for one batch) at step 60: 217.6690, Accuracy: 0.8397
Training loss (for one batch) at step 70: 220.7417, Accuracy: 0.8397
Training loss (for one batch) at step 80: 199.8788, Accuracy: 0.8381
Training loss (for one batch) at step 90: 205.5216, Accuracy: 0.8364
Training loss (for one batch) at step 100: 207.1142, Accuracy: 0.8365
Training loss (for one batch) at step 110: 205.1105, Accuracy: 0.8393
Training loss (for one batch) at step 120: 217.7876, Accuracy: 0.8388
Training loss (for one batch) at step 130: 190.2369, Accuracy: 0.8390
Training loss (for one batch) at step 140: 216.8800, Accuracy: 0.8393
---- Training ----
Training loss: 198.1443
Training acc over epoch: 0.8385
---- Validation ----
Validation loss: 78.2435
Validation acc: 0.7200
Time taken: 63.97s
Start of epoch 49
Training loss (for one batch) at step 0: 230.1934, Accuracy: 0.7900
Training loss (for one batch) at step 10: 209.6478, Accuracy: 0.8291
Training loss (for one batch) at step 20: 213.9764, Accuracy: 0.8395
Training loss (for one batch) at step 30: 226.4637, Accuracy: 0.8345
Training loss (for one batch) at step 40: 190.1437, Accuracy: 0.8412
Training loss (for one batch) at step 50: 217.8707, Accuracy: 0.8443
Training loss (for one batch) at step 60: 220.8602, Accuracy: 0.8457
Training loss (for one batch) at step 70: 215.5857, Accuracy: 0.8424
Training loss (for one batch) at step 80: 199.4958, Accuracy: 0.8412
Training loss (for one batch) at step 90: 211.4374, Accuracy: 0.8405
Training loss (for one batch) at step 100: 196.8961, Accuracy: 0.8394
Training loss (for one batch) at step 110: 210.9171, Accuracy: 0.8386
Training loss (for one batch) at step 120: 227.3339, Accuracy: 0.8394
Training loss (for one batch) at step 130: 212.0144, Accuracy: 0.8392
Training loss (for one batch) at step 140: 240.9035, Accuracy: 0.8392
---- Training ----
Training loss: 181.9597
Training acc over epoch: 0.8384
---- Validation ----
Validation loss: 59.4744
Validation acc: 0.7214
Time taken: 38.05s
===== Q: 0.0001
Validation acc: 0.7324
Validation AUC: 0.7293
Validation Balanced_ACC: 0.4774
Validation MI: 0.1368
Validation Normalized MI: 0.2050
Validation Adjusted MI: 0.2050
Validation aUc_Sklearn: 0.8299
Start of epoch 0
Training loss (for one batch) at step 0: 479.5241, Accuracy: 0.5300
Training loss (for one batch) at step 10: 434.6303, Accuracy: 0.5500
Training loss (for one batch) at step 20: 480.0556, Accuracy: 0.5414
Training loss (for one batch) at step 30: 459.9586, Accuracy: 0.5371
Training loss (for one batch) at step 40: 437.3063, Accuracy: 0.5468
Training loss (for one batch) at step 50: 441.2097, Accuracy: 0.5506
Training loss (for one batch) at step 60: 437.8273, Accuracy: 0.5552
Training loss (for one batch) at step 70: 433.0073, Accuracy: 0.5603
Training loss (for one batch) at step 80: 409.3933, Accuracy: 0.5602
Training loss (for one batch) at step 90: 456.9695, Accuracy: 0.5609
Training loss (for one batch) at step 100: 397.3320, Accuracy: 0.5623
Training loss (for one batch) at step 110: 399.2796, Accuracy: 0.5634
Training loss (for one batch) at step 120: 425.2460, Accuracy: 0.5653
Training loss (for one batch) at step 130: 417.5290, Accuracy: 0.5678
Training loss (for one batch) at step 140: 421.9589, Accuracy: 0.5709
---- Training ----
Training loss: 339.3156
Training acc over epoch: 0.5713
---- Validation ----
Validation loss: 90.4750
Validation acc: 0.5134
Time taken: 75.59s
Start of epoch 1
Training loss (for one batch) at step 0: 395.2241, Accuracy: 0.6200
Training loss (for one batch) at step 10: 394.3415, Accuracy: 0.6445
Training loss (for one batch) at step 20: 391.3694, Accuracy: 0.6343
Training loss (for one batch) at step 30: 392.6755, Accuracy: 0.6239
Training loss (for one batch) at step 40: 366.5583, Accuracy: 0.6251
Training loss (for one batch) at step 50: 397.9012, Accuracy: 0.6169
Training loss (for one batch) at step 60: 401.4820, Accuracy: 0.6169
Training loss (for one batch) at step 70: 393.2722, Accuracy: 0.6154
Training loss (for one batch) at step 80: 382.2695, Accuracy: 0.6163
Training loss (for one batch) at step 90: 385.4666, Accuracy: 0.6156
Training loss (for one batch) at step 100: 393.0071, Accuracy: 0.6166
Training loss (for one batch) at step 110: 373.5232, Accuracy: 0.6183
Training loss (for one batch) at step 120: 442.0537, Accuracy: 0.6172
Training loss (for one batch) at step 130: 357.0536, Accuracy: 0.6172
Training loss (for one batch) at step 140: 363.0621, Accuracy: 0.6188
---- Training ----
Training loss: 321.2428
Training acc over epoch: 0.6196
---- Validation ----
Validation loss: 80.1559
Validation acc: 0.5247
Time taken: 48.99s
Start of epoch 2
Training loss (for one batch) at step 0: 355.7010, Accuracy: 0.6300
Training loss (for one batch) at step 10: 381.1049, Accuracy: 0.6491
Training loss (for one batch) at step 20: 367.4984, Accuracy: 0.6300
Training loss (for one batch) at step 30: 382.7675, Accuracy: 0.6319
Training loss (for one batch) at step 40: 359.8742, Accuracy: 0.6339
Training loss (for one batch) at step 50: 360.8224, Accuracy: 0.6300
Training loss (for one batch) at step 60: 383.7451, Accuracy: 0.6318
Training loss (for one batch) at step 70: 360.4893, Accuracy: 0.6303
Training loss (for one batch) at step 80: 382.1013, Accuracy: 0.6291
Training loss (for one batch) at step 90: 338.6798, Accuracy: 0.6301
Training loss (for one batch) at step 100: 363.4007, Accuracy: 0.6290
Training loss (for one batch) at step 110: 386.4888, Accuracy: 0.6300
Training loss (for one batch) at step 120: 366.4424, Accuracy: 0.6302
Training loss (for one batch) at step 130: 377.8067, Accuracy: 0.6327
Training loss (for one batch) at step 140: 354.8051, Accuracy: 0.6338
---- Training ----
Training loss: 331.3201
Training acc over epoch: 0.6327
---- Validation ----
Validation loss: 71.6832
Validation acc: 0.6690
Time taken: 65.55s
Start of epoch 3
Training loss (for one batch) at step 0: 359.6548, Accuracy: 0.6200
Training loss (for one batch) at step 10: 350.7328, Accuracy: 0.6564
Training loss (for one batch) at step 20: 360.8133, Accuracy: 0.6510
Training loss (for one batch) at step 30: 343.5828, Accuracy: 0.6587
Training loss (for one batch) at step 40: 363.3199, Accuracy: 0.6612
Training loss (for one batch) at step 50: 353.9315, Accuracy: 0.6592
Training loss (for one batch) at step 60: 357.2854, Accuracy: 0.6561
Training loss (for one batch) at step 70: 368.5715, Accuracy: 0.6577
Training loss (for one batch) at step 80: 377.2346, Accuracy: 0.6549
Training loss (for one batch) at step 90: 359.4851, Accuracy: 0.6524
Training loss (for one batch) at step 100: 352.9202, Accuracy: 0.6531
Training loss (for one batch) at step 110: 344.2189, Accuracy: 0.6514
Training loss (for one batch) at step 120: 340.2103, Accuracy: 0.6521
Training loss (for one batch) at step 130: 348.5594, Accuracy: 0.6524
Training loss (for one batch) at step 140: 340.0830, Accuracy: 0.6542
---- Training ----
Training loss: 310.6653
Training acc over epoch: 0.6555
---- Validation ----
Validation loss: 74.3464
Validation acc: 0.7133
Time taken: 39.16s
Start of epoch 4
Training loss (for one batch) at step 0: 351.3768, Accuracy: 0.6200
Training loss (for one batch) at step 10: 353.0665, Accuracy: 0.6636
Training loss (for one batch) at step 20: 354.0257, Accuracy: 0.6633
Training loss (for one batch) at step 30: 340.7885, Accuracy: 0.6616
Training loss (for one batch) at step 40: 336.6227, Accuracy: 0.6690
Training loss (for one batch) at step 50: 344.3452, Accuracy: 0.6669
Training loss (for one batch) at step 60: 342.9716, Accuracy: 0.6669
Training loss (for one batch) at step 70: 337.4590, Accuracy: 0.6700
Training loss (for one batch) at step 80: 350.5383, Accuracy: 0.6704
Training loss (for one batch) at step 90: 381.6825, Accuracy: 0.6695
Training loss (for one batch) at step 100: 315.9998, Accuracy: 0.6703
Training loss (for one batch) at step 110: 345.2491, Accuracy: 0.6732
Training loss (for one batch) at step 120: 334.2036, Accuracy: 0.6764
Training loss (for one batch) at step 130: 361.2713, Accuracy: 0.6763
Training loss (for one batch) at step 140: 347.9839, Accuracy: 0.6776
---- Training ----
Training loss: 308.9143
Training acc over epoch: 0.6781
---- Validation ----
Validation loss: 72.1825
Validation acc: 0.7023
Time taken: 65.45s
Start of epoch 5
Training loss (for one batch) at step 0: 349.2518, Accuracy: 0.7300
Training loss (for one batch) at step 10: 345.1181, Accuracy: 0.6882
Training loss (for one batch) at step 20: 339.7554, Accuracy: 0.6924
Training loss (for one batch) at step 30: 332.8319, Accuracy: 0.6929
Training loss (for one batch) at step 40: 337.7595, Accuracy: 0.6871
Training loss (for one batch) at step 50: 318.6205, Accuracy: 0.6920
Training loss (for one batch) at step 60: 329.6430, Accuracy: 0.6930
Training loss (for one batch) at step 70: 316.1043, Accuracy: 0.6900
Training loss (for one batch) at step 80: 338.8696, Accuracy: 0.6888
Training loss (for one batch) at step 90: 326.2845, Accuracy: 0.6897
Training loss (for one batch) at step 100: 319.5118, Accuracy: 0.6880
Training loss (for one batch) at step 110: 309.0623, Accuracy: 0.6910
Training loss (for one batch) at step 120: 362.3560, Accuracy: 0.6928
Training loss (for one batch) at step 130: 326.4955, Accuracy: 0.6944
Training loss (for one batch) at step 140: 327.5291, Accuracy: 0.6945
---- Training ----
Training loss: 309.4334
Training acc over epoch: 0.6939
---- Validation ----
Validation loss: 77.0379
Validation acc: 0.6814
Time taken: 38.38s
Start of epoch 6
Training loss (for one batch) at step 0: 314.7722, Accuracy: 0.6300
Training loss (for one batch) at step 10: 338.0643, Accuracy: 0.7155
Training loss (for one batch) at step 20: 329.1381, Accuracy: 0.7114
Training loss (for one batch) at step 30: 339.6018, Accuracy: 0.7145
Training loss (for one batch) at step 40: 309.0722, Accuracy: 0.7159
Training loss (for one batch) at step 50: 330.9476, Accuracy: 0.7175
Training loss (for one batch) at step 60: 327.5219, Accuracy: 0.7218
Training loss (for one batch) at step 70: 317.3719, Accuracy: 0.7159
Training loss (for one batch) at step 80: 315.6838, Accuracy: 0.7137
Training loss (for one batch) at step 90: 342.1706, Accuracy: 0.7136
Training loss (for one batch) at step 100: 316.2815, Accuracy: 0.7111
Training loss (for one batch) at step 110: 321.2736, Accuracy: 0.7090
Training loss (for one batch) at step 120: 315.9786, Accuracy: 0.7086
Training loss (for one batch) at step 130: 322.0745, Accuracy: 0.7102
Training loss (for one batch) at step 140: 315.4114, Accuracy: 0.7099
---- Training ----
Training loss: 294.9938
Training acc over epoch: 0.7089
---- Validation ----
Validation loss: 76.7858
Validation acc: 0.7163
Time taken: 66.39s
Start of epoch 7
Training loss (for one batch) at step 0: 334.0974, Accuracy: 0.7800
Training loss (for one batch) at step 10: 322.8423, Accuracy: 0.7373
Training loss (for one batch) at step 20: 331.6446, Accuracy: 0.7267
Training loss (for one batch) at step 30: 333.5007, Accuracy: 0.7229
Training loss (for one batch) at step 40: 317.0521, Accuracy: 0.7205
Training loss (for one batch) at step 50: 314.1695, Accuracy: 0.7190
Training loss (for one batch) at step 60: 297.0352, Accuracy: 0.7208
Training loss (for one batch) at step 70: 325.3018, Accuracy: 0.7234
Training loss (for one batch) at step 80: 328.4239, Accuracy: 0.7189
Training loss (for one batch) at step 90: 331.9845, Accuracy: 0.7180
Training loss (for one batch) at step 100: 324.1481, Accuracy: 0.7178
Training loss (for one batch) at step 110: 322.5584, Accuracy: 0.7200
Training loss (for one batch) at step 120: 297.7184, Accuracy: 0.7198
Training loss (for one batch) at step 130: 316.6651, Accuracy: 0.7201
Training loss (for one batch) at step 140: 328.7845, Accuracy: 0.7211
---- Training ----
Training loss: 273.9669
Training acc over epoch: 0.7219
---- Validation ----
Validation loss: 71.8118
Validation acc: 0.7359
Time taken: 38.33s
Start of epoch 8
Training loss (for one batch) at step 0: 308.8978, Accuracy: 0.6800
Training loss (for one batch) at step 10: 300.7010, Accuracy: 0.7464
Training loss (for one batch) at step 20: 307.9345, Accuracy: 0.7219
Training loss (for one batch) at step 30: 304.0897, Accuracy: 0.7248
Training loss (for one batch) at step 40: 300.8628, Accuracy: 0.7251
Training loss (for one batch) at step 50: 300.6904, Accuracy: 0.7298
Training loss (for one batch) at step 60: 333.6637, Accuracy: 0.7313
Training loss (for one batch) at step 70: 347.8326, Accuracy: 0.7297
Training loss (for one batch) at step 80: 307.9893, Accuracy: 0.7283
Training loss (for one batch) at step 90: 318.0943, Accuracy: 0.7307
Training loss (for one batch) at step 100: 318.7148, Accuracy: 0.7312
Training loss (for one batch) at step 110: 310.1903, Accuracy: 0.7339
Training loss (for one batch) at step 120: 328.8819, Accuracy: 0.7359
Training loss (for one batch) at step 130: 316.3365, Accuracy: 0.7353
Training loss (for one batch) at step 140: 301.1498, Accuracy: 0.7350
---- Training ----
Training loss: 285.0716
Training acc over epoch: 0.7349
---- Validation ----
Validation loss: 66.1246
Validation acc: 0.7176
Time taken: 65.71s
Start of epoch 9
Training loss (for one batch) at step 0: 300.6933, Accuracy: 0.7700
Training loss (for one batch) at step 10: 292.2977, Accuracy: 0.7382
Training loss (for one batch) at step 20: 311.1510, Accuracy: 0.7443
Training loss (for one batch) at step 30: 339.4004, Accuracy: 0.7371
Training loss (for one batch) at step 40: 307.5127, Accuracy: 0.7412
Training loss (for one batch) at step 50: 289.1755, Accuracy: 0.7473
Training loss (for one batch) at step 60: 311.5857, Accuracy: 0.7449
Training loss (for one batch) at step 70: 292.3828, Accuracy: 0.7445
Training loss (for one batch) at step 80: 318.7902, Accuracy: 0.7421
Training loss (for one batch) at step 90: 307.0590, Accuracy: 0.7423
Training loss (for one batch) at step 100: 321.7906, Accuracy: 0.7422
Training loss (for one batch) at step 110: 298.1156, Accuracy: 0.7444
Training loss (for one batch) at step 120: 307.2987, Accuracy: 0.7456
Training loss (for one batch) at step 130: 313.9139, Accuracy: 0.7471
Training loss (for one batch) at step 140: 297.6818, Accuracy: 0.7460
---- Training ----
Training loss: 268.8992
Training acc over epoch: 0.7460
---- Validation ----
Validation loss: 64.5709
Validation acc: 0.7380
Time taken: 39.95s
Start of epoch 10
Training loss (for one batch) at step 0: 301.4560, Accuracy: 0.8200
Training loss (for one batch) at step 10: 310.3856, Accuracy: 0.7527
Training loss (for one batch) at step 20: 313.4003, Accuracy: 0.7538
Training loss (for one batch) at step 30: 292.6399, Accuracy: 0.7513
Training loss (for one batch) at step 40: 304.9426, Accuracy: 0.7546
Training loss (for one batch) at step 50: 298.1487, Accuracy: 0.7582
Training loss (for one batch) at step 60: 290.6568, Accuracy: 0.7631
Training loss (for one batch) at step 70: 322.3645, Accuracy: 0.7617
Training loss (for one batch) at step 80: 298.8371, Accuracy: 0.7616
Training loss (for one batch) at step 90: 267.7661, Accuracy: 0.7601
Training loss (for one batch) at step 100: 294.8772, Accuracy: 0.7576
Training loss (for one batch) at step 110: 297.3646, Accuracy: 0.7577
Training loss (for one batch) at step 120: 298.3638, Accuracy: 0.7598
Training loss (for one batch) at step 130: 306.0370, Accuracy: 0.7589
Training loss (for one batch) at step 140: 302.5164, Accuracy: 0.7594
---- Training ----
Training loss: 269.8598
Training acc over epoch: 0.7599
---- Validation ----
Validation loss: 75.6907
Validation acc: 0.7230
Time taken: 64.81s
Start of epoch 11
Training loss (for one batch) at step 0: 299.8494, Accuracy: 0.7700
Training loss (for one batch) at step 10: 299.5659, Accuracy: 0.7455
Training loss (for one batch) at step 20: 286.4368, Accuracy: 0.7524
Training loss (for one batch) at step 30: 309.2848, Accuracy: 0.7587
Training loss (for one batch) at step 40: 305.0037, Accuracy: 0.7627
Training loss (for one batch) at step 50: 284.1733, Accuracy: 0.7643
Training loss (for one batch) at step 60: 286.8339, Accuracy: 0.7646
Training loss (for one batch) at step 70: 305.7822, Accuracy: 0.7648
Training loss (for one batch) at step 80: 301.5305, Accuracy: 0.7653
Training loss (for one batch) at step 90: 293.5418, Accuracy: 0.7658
Training loss (for one batch) at step 100: 308.7319, Accuracy: 0.7669
Training loss (for one batch) at step 110: 286.5334, Accuracy: 0.7658
Training loss (for one batch) at step 120: 294.0993, Accuracy: 0.7663
Training loss (for one batch) at step 130: 280.4850, Accuracy: 0.7656
Training loss (for one batch) at step 140: 304.8282, Accuracy: 0.7654
---- Training ----
Training loss: 276.5668
Training acc over epoch: 0.7659
---- Validation ----
Validation loss: 67.2533
Validation acc: 0.7254
Time taken: 37.87s
Start of epoch 12
Training loss (for one batch) at step 0: 286.7213, Accuracy: 0.7900
Training loss (for one batch) at step 10: 280.2058, Accuracy: 0.7727
Training loss (for one batch) at step 20: 281.2866, Accuracy: 0.7762
Training loss (for one batch) at step 30: 296.3166, Accuracy: 0.7794
Training loss (for one batch) at step 40: 290.6591, Accuracy: 0.7856
Training loss (for one batch) at step 50: 311.8596, Accuracy: 0.7814
Training loss (for one batch) at step 60: 276.2718, Accuracy: 0.7843
Training loss (for one batch) at step 70: 308.3509, Accuracy: 0.7793
Training loss (for one batch) at step 80: 293.8778, Accuracy: 0.7783
Training loss (for one batch) at step 90: 282.3287, Accuracy: 0.7744
Training loss (for one batch) at step 100: 302.3368, Accuracy: 0.7717
Training loss (for one batch) at step 110: 268.1348, Accuracy: 0.7724
Training loss (for one batch) at step 120: 307.6303, Accuracy: 0.7737
Training loss (for one batch) at step 130: 293.3424, Accuracy: 0.7727
Training loss (for one batch) at step 140: 286.4856, Accuracy: 0.7732
---- Training ----
Training loss: 256.0646
Training acc over epoch: 0.7740
---- Validation ----
Validation loss: 75.0658
Validation acc: 0.7182
Time taken: 65.20s
Start of epoch 13
Training loss (for one batch) at step 0: 289.9854, Accuracy: 0.8000
Training loss (for one batch) at step 10: 281.6703, Accuracy: 0.7864
Training loss (for one batch) at step 20: 286.9897, Accuracy: 0.7800
Training loss (for one batch) at step 30: 280.8076, Accuracy: 0.7784
Training loss (for one batch) at step 40: 298.1939, Accuracy: 0.7788
Training loss (for one batch) at step 50: 297.9860, Accuracy: 0.7847
Training loss (for one batch) at step 60: 288.4882, Accuracy: 0.7838
Training loss (for one batch) at step 70: 288.8973, Accuracy: 0.7837
Training loss (for one batch) at step 80: 288.4696, Accuracy: 0.7816
Training loss (for one batch) at step 90: 296.7395, Accuracy: 0.7829
Training loss (for one batch) at step 100: 277.1353, Accuracy: 0.7803
Training loss (for one batch) at step 110: 283.6693, Accuracy: 0.7816
Training loss (for one batch) at step 120: 282.2724, Accuracy: 0.7818
Training loss (for one batch) at step 130: 275.2750, Accuracy: 0.7821
Training loss (for one batch) at step 140: 278.1713, Accuracy: 0.7820
---- Training ----
Training loss: 254.8592
Training acc over epoch: 0.7818
---- Validation ----
Validation loss: 74.8755
Validation acc: 0.7431
Time taken: 39.00s
Start of epoch 14
Training loss (for one batch) at step 0: 279.8448, Accuracy: 0.7300
Training loss (for one batch) at step 10: 293.5523, Accuracy: 0.7845
Training loss (for one batch) at step 20: 286.2355, Accuracy: 0.7819
Training loss (for one batch) at step 30: 271.8284, Accuracy: 0.7861
Training loss (for one batch) at step 40: 275.5151, Accuracy: 0.7841
Training loss (for one batch) at step 50: 281.7759, Accuracy: 0.7847
Training loss (for one batch) at step 60: 289.5174, Accuracy: 0.7844
Training loss (for one batch) at step 70: 278.7155, Accuracy: 0.7831
Training loss (for one batch) at step 80: 290.5847, Accuracy: 0.7801
Training loss (for one batch) at step 90: 303.0349, Accuracy: 0.7823
Training loss (for one batch) at step 100: 289.9084, Accuracy: 0.7825
Training loss (for one batch) at step 110: 289.0851, Accuracy: 0.7811
Training loss (for one batch) at step 120: 296.3590, Accuracy: 0.7824
Training loss (for one batch) at step 130: 277.2641, Accuracy: 0.7834
Training loss (for one batch) at step 140: 307.4809, Accuracy: 0.7833
---- Training ----
Training loss: 253.9147
Training acc over epoch: 0.7835
---- Validation ----
Validation loss: 64.2057
Validation acc: 0.7397
Time taken: 66.00s
Start of epoch 15
Training loss (for one batch) at step 0: 282.5075, Accuracy: 0.7800
Training loss (for one batch) at step 10: 276.2860, Accuracy: 0.7964
Training loss (for one batch) at step 20: 289.4795, Accuracy: 0.7810
Training loss (for one batch) at step 30: 301.5453, Accuracy: 0.7887
Training loss (for one batch) at step 40: 289.0535, Accuracy: 0.7834
Training loss (for one batch) at step 50: 274.3402, Accuracy: 0.7839
Training loss (for one batch) at step 60: 283.3886, Accuracy: 0.7836
Training loss (for one batch) at step 70: 270.7652, Accuracy: 0.7851
Training loss (for one batch) at step 80: 278.5645, Accuracy: 0.7853
Training loss (for one batch) at step 90: 288.3759, Accuracy: 0.7820
Training loss (for one batch) at step 100: 287.1304, Accuracy: 0.7815
Training loss (for one batch) at step 110: 264.7650, Accuracy: 0.7827
Training loss (for one batch) at step 120: 283.5528, Accuracy: 0.7841
Training loss (for one batch) at step 130: 271.7293, Accuracy: 0.7849
Training loss (for one batch) at step 140: 287.8220, Accuracy: 0.7850
---- Training ----
Training loss: 252.3809
Training acc over epoch: 0.7859
---- Validation ----
Validation loss: 76.6500
Validation acc: 0.7523
Time taken: 38.30s
Start of epoch 16
Training loss (for one batch) at step 0: 279.6506, Accuracy: 0.8100
Training loss (for one batch) at step 10: 277.5917, Accuracy: 0.7964
Training loss (for one batch) at step 20: 291.1006, Accuracy: 0.8005
Training loss (for one batch) at step 30: 277.0858, Accuracy: 0.7948
Training loss (for one batch) at step 40: 263.8018, Accuracy: 0.7939
Training loss (for one batch) at step 50: 280.2572, Accuracy: 0.7978
Training loss (for one batch) at step 60: 269.4338, Accuracy: 0.7979
Training loss (for one batch) at step 70: 290.9586, Accuracy: 0.7999
Training loss (for one batch) at step 80: 286.7905, Accuracy: 0.7962
Training loss (for one batch) at step 90: 289.6342, Accuracy: 0.7960
Training loss (for one batch) at step 100: 256.5775, Accuracy: 0.7953
Training loss (for one batch) at step 110: 277.3252, Accuracy: 0.7961
Training loss (for one batch) at step 120: 269.9976, Accuracy: 0.7959
Training loss (for one batch) at step 130: 265.6883, Accuracy: 0.7965
Training loss (for one batch) at step 140: 273.1998, Accuracy: 0.7963
---- Training ----
Training loss: 246.7365
Training acc over epoch: 0.7960
---- Validation ----
Validation loss: 71.3381
Validation acc: 0.7327
Time taken: 67.18s
Start of epoch 17
Training loss (for one batch) at step 0: 274.3946, Accuracy: 0.8600
Training loss (for one batch) at step 10: 278.1289, Accuracy: 0.8091
Training loss (for one batch) at step 20: 280.4697, Accuracy: 0.8052
Training loss (for one batch) at step 30: 281.5846, Accuracy: 0.8100
Training loss (for one batch) at step 40: 255.4583, Accuracy: 0.8071
Training loss (for one batch) at step 50: 269.6871, Accuracy: 0.8053
Training loss (for one batch) at step 60: 269.4048, Accuracy: 0.8041
Training loss (for one batch) at step 70: 285.0975, Accuracy: 0.8004
Training loss (for one batch) at step 80: 285.1865, Accuracy: 0.7984
Training loss (for one batch) at step 90: 283.6025, Accuracy: 0.7960
Training loss (for one batch) at step 100: 273.6462, Accuracy: 0.7955
Training loss (for one batch) at step 110: 276.3793, Accuracy: 0.7970
Training loss (for one batch) at step 120: 281.1371, Accuracy: 0.7974
Training loss (for one batch) at step 130: 285.2191, Accuracy: 0.7975
Training loss (for one batch) at step 140: 271.5698, Accuracy: 0.7984
---- Training ----
Training loss: 234.5247
Training acc over epoch: 0.7975
---- Validation ----
Validation loss: 71.2114
Validation acc: 0.7380
Time taken: 39.28s
Start of epoch 18
Training loss (for one batch) at step 0: 276.9429, Accuracy: 0.8100
Training loss (for one batch) at step 10: 276.7552, Accuracy: 0.8018
Training loss (for one batch) at step 20: 294.2938, Accuracy: 0.8067
Training loss (for one batch) at step 30: 264.8023, Accuracy: 0.8116
Training loss (for one batch) at step 40: 289.8555, Accuracy: 0.8098
Training loss (for one batch) at step 50: 280.9545, Accuracy: 0.8125
Training loss (for one batch) at step 60: 267.7056, Accuracy: 0.8116
Training loss (for one batch) at step 70: 277.8397, Accuracy: 0.8100
Training loss (for one batch) at step 80: 286.8100, Accuracy: 0.8042
Training loss (for one batch) at step 90: 251.4415, Accuracy: 0.8042
Training loss (for one batch) at step 100: 268.6334, Accuracy: 0.8048
Training loss (for one batch) at step 110: 278.9490, Accuracy: 0.8036
Training loss (for one batch) at step 120: 274.5276, Accuracy: 0.8034
Training loss (for one batch) at step 130: 269.3804, Accuracy: 0.8034
Training loss (for one batch) at step 140: 293.4306, Accuracy: 0.8023
---- Training ----
Training loss: 247.7811
Training acc over epoch: 0.8031
---- Validation ----
Validation loss: 71.8373
Validation acc: 0.7281
Time taken: 67.82s
Start of epoch 19
Training loss (for one batch) at step 0: 272.2863, Accuracy: 0.8200
Training loss (for one batch) at step 10: 263.7847, Accuracy: 0.8182
Training loss (for one batch) at step 20: 280.6638, Accuracy: 0.8081
Training loss (for one batch) at step 30: 269.1390, Accuracy: 0.8113
Training loss (for one batch) at step 40: 252.1497, Accuracy: 0.8120
Training loss (for one batch) at step 50: 267.2648, Accuracy: 0.8129
Training loss (for one batch) at step 60: 266.2668, Accuracy: 0.8079
Training loss (for one batch) at step 70: 268.8604, Accuracy: 0.8055
Training loss (for one batch) at step 80: 272.5202, Accuracy: 0.8046
Training loss (for one batch) at step 90: 268.9320, Accuracy: 0.8049
Training loss (for one batch) at step 100: 265.1130, Accuracy: 0.8023
Training loss (for one batch) at step 110: 270.3014, Accuracy: 0.8028
Training loss (for one batch) at step 120: 276.8309, Accuracy: 0.8033
Training loss (for one batch) at step 130: 284.6234, Accuracy: 0.8034
Training loss (for one batch) at step 140: 280.2619, Accuracy: 0.8052
---- Training ----
Training loss: 231.6023
Training acc over epoch: 0.8049
---- Validation ----
Validation loss: 69.1793
Validation acc: 0.7405
Time taken: 39.30s
Start of epoch 20
Training loss (for one batch) at step 0: 268.7421, Accuracy: 0.8500
Training loss (for one batch) at step 10: 279.8432, Accuracy: 0.8191
Training loss (for one batch) at step 20: 260.8879, Accuracy: 0.8110
Training loss (for one batch) at step 30: 266.0927, Accuracy: 0.8139
Training loss (for one batch) at step 40: 279.5359, Accuracy: 0.8146
Training loss (for one batch) at step 50: 264.3742, Accuracy: 0.8169
Training loss (for one batch) at step 60: 264.5847, Accuracy: 0.8172
Training loss (for one batch) at step 70: 271.3310, Accuracy: 0.8135
Training loss (for one batch) at step 80: 279.0677, Accuracy: 0.8077
Training loss (for one batch) at step 90: 289.4846, Accuracy: 0.8086
Training loss (for one batch) at step 100: 283.0888, Accuracy: 0.8064
Training loss (for one batch) at step 110: 255.7741, Accuracy: 0.8073
Training loss (for one batch) at step 120: 251.3377, Accuracy: 0.8063
Training loss (for one batch) at step 130: 269.0388, Accuracy: 0.8066
Training loss (for one batch) at step 140: 273.2766, Accuracy: 0.8072
---- Training ----
Training loss: 249.3883
Training acc over epoch: 0.8062
---- Validation ----
Validation loss: 68.2650
Validation acc: 0.7423
Time taken: 69.16s
Start of epoch 21
Training loss (for one batch) at step 0: 262.3608, Accuracy: 0.9000
Training loss (for one batch) at step 10: 267.2453, Accuracy: 0.8318
Training loss (for one batch) at step 20: 275.9139, Accuracy: 0.8276
Training loss (for one batch) at step 30: 267.5163, Accuracy: 0.8223
Training loss (for one batch) at step 40: 252.0900, Accuracy: 0.8239
Training loss (for one batch) at step 50: 276.4173, Accuracy: 0.8235
Training loss (for one batch) at step 60: 279.0793, Accuracy: 0.8225
Training loss (for one batch) at step 70: 271.0175, Accuracy: 0.8211
Training loss (for one batch) at step 80: 289.7311, Accuracy: 0.8173
Training loss (for one batch) at step 90: 267.0374, Accuracy: 0.8148
Training loss (for one batch) at step 100: 271.7969, Accuracy: 0.8125
Training loss (for one batch) at step 110: 260.9313, Accuracy: 0.8110
Training loss (for one batch) at step 120: 261.1982, Accuracy: 0.8116
Training loss (for one batch) at step 130: 265.5856, Accuracy: 0.8099
Training loss (for one batch) at step 140: 269.0215, Accuracy: 0.8089
---- Training ----
Training loss: 245.0160
Training acc over epoch: 0.8093
---- Validation ----
Validation loss: 64.2289
Validation acc: 0.7421
Time taken: 40.34s
Start of epoch 22
Training loss (for one batch) at step 0: 273.0718, Accuracy: 0.7400
Training loss (for one batch) at step 10: 244.1673, Accuracy: 0.8073
Training loss (for one batch) at step 20: 259.3272, Accuracy: 0.8124
Training loss (for one batch) at step 30: 253.3538, Accuracy: 0.8113
Training loss (for one batch) at step 40: 267.5478, Accuracy: 0.8129
Training loss (for one batch) at step 50: 243.8898, Accuracy: 0.8198
Training loss (for one batch) at step 60: 262.6335, Accuracy: 0.8238
Training loss (for one batch) at step 70: 258.3515, Accuracy: 0.8215
Training loss (for one batch) at step 80: 265.3261, Accuracy: 0.8190
Training loss (for one batch) at step 90: 257.6531, Accuracy: 0.8192
Training loss (for one batch) at step 100: 263.3476, Accuracy: 0.8191
Training loss (for one batch) at step 110: 246.4431, Accuracy: 0.8176
Training loss (for one batch) at step 120: 274.0035, Accuracy: 0.8176
Training loss (for one batch) at step 130: 265.0739, Accuracy: 0.8163
Training loss (for one batch) at step 140: 274.4476, Accuracy: 0.8178
---- Training ----
Training loss: 220.1254
Training acc over epoch: 0.8167
---- Validation ----
Validation loss: 66.7004
Validation acc: 0.7286
Time taken: 67.41s
Start of epoch 23
Training loss (for one batch) at step 0: 257.4083, Accuracy: 0.7900
Training loss (for one batch) at step 10: 275.4130, Accuracy: 0.8191
Training loss (for one batch) at step 20: 271.4077, Accuracy: 0.8148
Training loss (for one batch) at step 30: 261.8566, Accuracy: 0.8265
Training loss (for one batch) at step 40: 241.5290, Accuracy: 0.8178
Training loss (for one batch) at step 50: 263.5068, Accuracy: 0.8208
Training loss (for one batch) at step 60: 241.3083, Accuracy: 0.8213
Training loss (for one batch) at step 70: 255.6885, Accuracy: 0.8177
Training loss (for one batch) at step 80: 262.2799, Accuracy: 0.8163
Training loss (for one batch) at step 90: 265.9784, Accuracy: 0.8155
Training loss (for one batch) at step 100: 263.2055, Accuracy: 0.8149
Training loss (for one batch) at step 110: 262.9856, Accuracy: 0.8157
Training loss (for one batch) at step 120: 254.1056, Accuracy: 0.8175
Training loss (for one batch) at step 130: 274.9526, Accuracy: 0.8164
Training loss (for one batch) at step 140: 266.5182, Accuracy: 0.8160
---- Training ----
Training loss: 233.3312
Training acc over epoch: 0.8145
---- Validation ----
Validation loss: 63.3822
Validation acc: 0.7340
Time taken: 39.84s
Start of epoch 24
Training loss (for one batch) at step 0: 254.2055, Accuracy: 0.7800
Training loss (for one batch) at step 10: 247.4862, Accuracy: 0.8264
Training loss (for one batch) at step 20: 255.9895, Accuracy: 0.8219
Training loss (for one batch) at step 30: 260.7684, Accuracy: 0.8171
Training loss (for one batch) at step 40: 267.2823, Accuracy: 0.8185
Training loss (for one batch) at step 50: 242.2910, Accuracy: 0.8218
Training loss (for one batch) at step 60: 249.2315, Accuracy: 0.8189
Training loss (for one batch) at step 70: 258.3653, Accuracy: 0.8190
Training loss (for one batch) at step 80: 249.1047, Accuracy: 0.8142
Training loss (for one batch) at step 90: 249.5768, Accuracy: 0.8133
Training loss (for one batch) at step 100: 254.9669, Accuracy: 0.8129
Training loss (for one batch) at step 110: 261.9439, Accuracy: 0.8123
Training loss (for one batch) at step 120: 251.5510, Accuracy: 0.8139
Training loss (for one batch) at step 130: 259.7151, Accuracy: 0.8135
Training loss (for one batch) at step 140: 244.7859, Accuracy: 0.8135
---- Training ----
Training loss: 244.8899
Training acc over epoch: 0.8131
---- Validation ----
Validation loss: 67.6783
Validation acc: 0.7402
Time taken: 66.52s
Start of epoch 25
Training loss (for one batch) at step 0: 271.7541, Accuracy: 0.7000
Training loss (for one batch) at step 10: 250.2131, Accuracy: 0.8100
Training loss (for one batch) at step 20: 269.1364, Accuracy: 0.8152
Training loss (for one batch) at step 30: 250.2654, Accuracy: 0.8177
Training loss (for one batch) at step 40: 255.4646, Accuracy: 0.8215
Training loss (for one batch) at step 50: 239.5324, Accuracy: 0.8278
Training loss (for one batch) at step 60: 238.1224, Accuracy: 0.8277
Training loss (for one batch) at step 70: 272.1417, Accuracy: 0.8256
Training loss (for one batch) at step 80: 285.5933, Accuracy: 0.8230
Training loss (for one batch) at step 90: 264.5500, Accuracy: 0.8211
Training loss (for one batch) at step 100: 251.5467, Accuracy: 0.8206
Training loss (for one batch) at step 110: 273.7039, Accuracy: 0.8205
Training loss (for one batch) at step 120: 267.1697, Accuracy: 0.8201
Training loss (for one batch) at step 130: 252.5623, Accuracy: 0.8205
Training loss (for one batch) at step 140: 243.4520, Accuracy: 0.8207
---- Training ----
Training loss: 226.5033
Training acc over epoch: 0.8202
---- Validation ----
Validation loss: 77.5648
Validation acc: 0.7082
Time taken: 38.66s
Start of epoch 26
Training loss (for one batch) at step 0: 257.2159, Accuracy: 0.8100
Training loss (for one batch) at step 10: 252.5316, Accuracy: 0.8245
Training loss (for one batch) at step 20: 230.7280, Accuracy: 0.8233
Training loss (for one batch) at step 30: 248.6595, Accuracy: 0.8252
Training loss (for one batch) at step 40: 238.3183, Accuracy: 0.8212
Training loss (for one batch) at step 50: 242.4891, Accuracy: 0.8247
Training loss (for one batch) at step 60: 245.5096, Accuracy: 0.8254
Training loss (for one batch) at step 70: 254.9247, Accuracy: 0.8245
Training loss (for one batch) at step 80: 237.3269, Accuracy: 0.8217
Training loss (for one batch) at step 90: 257.7325, Accuracy: 0.8202
Training loss (for one batch) at step 100: 253.9294, Accuracy: 0.8179
Training loss (for one batch) at step 110: 230.6475, Accuracy: 0.8197
Training loss (for one batch) at step 120: 252.7300, Accuracy: 0.8196
Training loss (for one batch) at step 130: 251.4202, Accuracy: 0.8191
Training loss (for one batch) at step 140: 238.5717, Accuracy: 0.8190
---- Training ----
Training loss: 210.5622
Training acc over epoch: 0.8195
---- Validation ----
Validation loss: 68.6347
Validation acc: 0.7286
Time taken: 66.17s
Start of epoch 27
Training loss (for one batch) at step 0: 251.4485, Accuracy: 0.8500
Training loss (for one batch) at step 10: 262.4686, Accuracy: 0.8309
Training loss (for one batch) at step 20: 242.3026, Accuracy: 0.8224
Training loss (for one batch) at step 30: 255.3805, Accuracy: 0.8229
Training loss (for one batch) at step 40: 237.1309, Accuracy: 0.8278
Training loss (for one batch) at step 50: 238.7290, Accuracy: 0.8308
Training loss (for one batch) at step 60: 257.7961, Accuracy: 0.8303
Training loss (for one batch) at step 70: 245.1685, Accuracy: 0.8301
Training loss (for one batch) at step 80: 232.4911, Accuracy: 0.8275
Training loss (for one batch) at step 90: 246.3766, Accuracy: 0.8257
Training loss (for one batch) at step 100: 232.2862, Accuracy: 0.8247
Training loss (for one batch) at step 110: 267.6193, Accuracy: 0.8241
Training loss (for one batch) at step 120: 249.8596, Accuracy: 0.8239
Training loss (for one batch) at step 130: 244.6400, Accuracy: 0.8234
Training loss (for one batch) at step 140: 246.4631, Accuracy: 0.8247
---- Training ----
Training loss: 233.5717
Training acc over epoch: 0.8246
---- Validation ----
Validation loss: 77.0562
Validation acc: 0.7278
Time taken: 37.80s
Start of epoch 28
Training loss (for one batch) at step 0: 241.8055, Accuracy: 0.8200
Training loss (for one batch) at step 10: 236.7795, Accuracy: 0.8391
Training loss (for one batch) at step 20: 250.2951, Accuracy: 0.8281
Training loss (for one batch) at step 30: 246.6147, Accuracy: 0.8352
Training loss (for one batch) at step 40: 229.5624, Accuracy: 0.8327
Training loss (for one batch) at step 50: 242.7495, Accuracy: 0.8345
Training loss (for one batch) at step 60: 228.4476, Accuracy: 0.8367
Training loss (for one batch) at step 70: 250.2468, Accuracy: 0.8342
Training loss (for one batch) at step 80: 222.4829, Accuracy: 0.8322
Training loss (for one batch) at step 90: 253.5771, Accuracy: 0.8300
Training loss (for one batch) at step 100: 237.4852, Accuracy: 0.8299
Training loss (for one batch) at step 110: 251.9977, Accuracy: 0.8305
Training loss (for one batch) at step 120: 247.5957, Accuracy: 0.8291
Training loss (for one batch) at step 130: 254.8130, Accuracy: 0.8282
Training loss (for one batch) at step 140: 255.3269, Accuracy: 0.8285
---- Training ----
Training loss: 225.3828
Training acc over epoch: 0.8291
---- Validation ----
Validation loss: 77.9171
Validation acc: 0.7225
Time taken: 66.16s
Start of epoch 29
Training loss (for one batch) at step 0: 248.6280, Accuracy: 0.8200
Training loss (for one batch) at step 10: 237.9588, Accuracy: 0.8409
Training loss (for one batch) at step 20: 240.9394, Accuracy: 0.8300
Training loss (for one batch) at step 30: 233.1646, Accuracy: 0.8306
Training loss (for one batch) at step 40: 243.2694, Accuracy: 0.8293
Training loss (for one batch) at step 50: 234.6088, Accuracy: 0.8304
Training loss (for one batch) at step 60: 249.5234, Accuracy: 0.8290
Training loss (for one batch) at step 70: 270.0999, Accuracy: 0.8259
Training loss (for one batch) at step 80: 237.4460, Accuracy: 0.8265
Training loss (for one batch) at step 90: 242.5996, Accuracy: 0.8255
Training loss (for one batch) at step 100: 232.5282, Accuracy: 0.8251
Training loss (for one batch) at step 110: 227.0333, Accuracy: 0.8250
Training loss (for one batch) at step 120: 238.8944, Accuracy: 0.8267
Training loss (for one batch) at step 130: 238.5324, Accuracy: 0.8266
Training loss (for one batch) at step 140: 244.7831, Accuracy: 0.8257
---- Training ----
Training loss: 210.1142
Training acc over epoch: 0.8262
---- Validation ----
Validation loss: 69.6281
Validation acc: 0.7340
Time taken: 38.25s
Start of epoch 30
Training loss (for one batch) at step 0: 233.6135, Accuracy: 0.8300
Training loss (for one batch) at step 10: 233.8873, Accuracy: 0.8173
Training loss (for one batch) at step 20: 240.2188, Accuracy: 0.8238
Training loss (for one batch) at step 30: 240.7806, Accuracy: 0.8265
Training loss (for one batch) at step 40: 223.9757, Accuracy: 0.8283
Training loss (for one batch) at step 50: 230.0146, Accuracy: 0.8327
Training loss (for one batch) at step 60: 219.8729, Accuracy: 0.8344
Training loss (for one batch) at step 70: 224.4239, Accuracy: 0.8339
Training loss (for one batch) at step 80: 244.8968, Accuracy: 0.8340
Training loss (for one batch) at step 90: 259.6616, Accuracy: 0.8308
Training loss (for one batch) at step 100: 246.4970, Accuracy: 0.8285
Training loss (for one batch) at step 110: 247.7748, Accuracy: 0.8295
Training loss (for one batch) at step 120: 240.6470, Accuracy: 0.8318
Training loss (for one batch) at step 130: 262.2714, Accuracy: 0.8318
Training loss (for one batch) at step 140: 246.1143, Accuracy: 0.8319
---- Training ----
Training loss: 207.9803
Training acc over epoch: 0.8336
---- Validation ----
Validation loss: 67.8286
Validation acc: 0.7276
Time taken: 66.41s
Start of epoch 31
Training loss (for one batch) at step 0: 241.2563, Accuracy: 0.8500
Training loss (for one batch) at step 10: 240.9076, Accuracy: 0.8373
Training loss (for one batch) at step 20: 254.9875, Accuracy: 0.8195
Training loss (for one batch) at step 30: 246.0640, Accuracy: 0.8255
Training loss (for one batch) at step 40: 252.9467, Accuracy: 0.8283
Training loss (for one batch) at step 50: 242.9528, Accuracy: 0.8343
Training loss (for one batch) at step 60: 239.5927, Accuracy: 0.8351
Training loss (for one batch) at step 70: 235.7014, Accuracy: 0.8338
Training loss (for one batch) at step 80: 256.6302, Accuracy: 0.8314
Training loss (for one batch) at step 90: 256.1144, Accuracy: 0.8301
Training loss (for one batch) at step 100: 229.5471, Accuracy: 0.8302
Training loss (for one batch) at step 110: 230.6715, Accuracy: 0.8305
Training loss (for one batch) at step 120: 237.2350, Accuracy: 0.8318
Training loss (for one batch) at step 130: 257.2697, Accuracy: 0.8324
Training loss (for one batch) at step 140: 230.1773, Accuracy: 0.8326
---- Training ----
Training loss: 192.6903
Training acc over epoch: 0.8319
---- Validation ----
Validation loss: 63.3008
Validation acc: 0.7225
Time taken: 39.03s
Start of epoch 32
Training loss (for one batch) at step 0: 232.2381, Accuracy: 0.8700
Training loss (for one batch) at step 10: 239.3753, Accuracy: 0.8382
Training loss (for one batch) at step 20: 238.6921, Accuracy: 0.8419
Training loss (for one batch) at step 30: 241.6314, Accuracy: 0.8390
Training loss (for one batch) at step 40: 242.8974, Accuracy: 0.8390
Training loss (for one batch) at step 50: 242.6385, Accuracy: 0.8412
Training loss (for one batch) at step 60: 245.8264, Accuracy: 0.8411
Training loss (for one batch) at step 70: 221.6444, Accuracy: 0.8399
Training loss (for one batch) at step 80: 241.2882, Accuracy: 0.8375
Training loss (for one batch) at step 90: 232.6875, Accuracy: 0.8352
Training loss (for one batch) at step 100: 238.3613, Accuracy: 0.8343
Training loss (for one batch) at step 110: 248.7412, Accuracy: 0.8366
Training loss (for one batch) at step 120: 234.3875, Accuracy: 0.8347
Training loss (for one batch) at step 130: 261.5870, Accuracy: 0.8340
Training loss (for one batch) at step 140: 238.9021, Accuracy: 0.8324
---- Training ----
Training loss: 235.2799
Training acc over epoch: 0.8325
---- Validation ----
Validation loss: 75.8381
Validation acc: 0.7485
Time taken: 73.74s
Start of epoch 33
Training loss (for one batch) at step 0: 248.4104, Accuracy: 0.7900
Training loss (for one batch) at step 10: 253.4173, Accuracy: 0.8427
Training loss (for one batch) at step 20: 247.0567, Accuracy: 0.8338
Training loss (for one batch) at step 30: 238.4016, Accuracy: 0.8335
Training loss (for one batch) at step 40: 216.2316, Accuracy: 0.8341
Training loss (for one batch) at step 50: 224.9815, Accuracy: 0.8396
Training loss (for one batch) at step 60: 243.2229, Accuracy: 0.8402
Training loss (for one batch) at step 70: 227.6909, Accuracy: 0.8377
Training loss (for one batch) at step 80: 237.3974, Accuracy: 0.8373
Training loss (for one batch) at step 90: 244.4578, Accuracy: 0.8374
Training loss (for one batch) at step 100: 234.5686, Accuracy: 0.8346
Training loss (for one batch) at step 110: 227.9779, Accuracy: 0.8366
Training loss (for one batch) at step 120: 244.4337, Accuracy: 0.8367
Training loss (for one batch) at step 130: 224.9476, Accuracy: 0.8362
Training loss (for one batch) at step 140: 255.3190, Accuracy: 0.8370
---- Training ----
Training loss: 211.1508
Training acc over epoch: 0.8365
---- Validation ----
Validation loss: 77.4379
Validation acc: 0.7190
Time taken: 38.40s
Start of epoch 34
Training loss (for one batch) at step 0: 233.1902, Accuracy: 0.8400
Training loss (for one batch) at step 10: 246.9863, Accuracy: 0.8400
Training loss (for one batch) at step 20: 251.7082, Accuracy: 0.8333
Training loss (for one batch) at step 30: 228.3527, Accuracy: 0.8313
Training loss (for one batch) at step 40: 221.5476, Accuracy: 0.8310
Training loss (for one batch) at step 50: 226.1902, Accuracy: 0.8351
Training loss (for one batch) at step 60: 225.6232, Accuracy: 0.8354
Training loss (for one batch) at step 70: 212.3603, Accuracy: 0.8345
Training loss (for one batch) at step 80: 232.3126, Accuracy: 0.8311
Training loss (for one batch) at step 90: 222.8756, Accuracy: 0.8319
Training loss (for one batch) at step 100: 234.7067, Accuracy: 0.8310
Training loss (for one batch) at step 110: 228.3782, Accuracy: 0.8316
Training loss (for one batch) at step 120: 223.0344, Accuracy: 0.8330
Training loss (for one batch) at step 130: 236.8021, Accuracy: 0.8321
Training loss (for one batch) at step 140: 270.8794, Accuracy: 0.8331
---- Training ----
Training loss: 213.0676
Training acc over epoch: 0.8322
---- Validation ----
Validation loss: 56.4675
Validation acc: 0.7238
Time taken: 66.45s
Start of epoch 35
Training loss (for one batch) at step 0: 247.4802, Accuracy: 0.7500
Training loss (for one batch) at step 10: 218.9325, Accuracy: 0.8273
Training loss (for one batch) at step 20: 218.0782, Accuracy: 0.8295
Training loss (for one batch) at step 30: 210.6562, Accuracy: 0.8342
Training loss (for one batch) at step 40: 220.0275, Accuracy: 0.8434
Training loss (for one batch) at step 50: 223.3390, Accuracy: 0.8449
Training loss (for one batch) at step 60: 234.4108, Accuracy: 0.8462
Training loss (for one batch) at step 70: 211.9288, Accuracy: 0.8446
Training loss (for one batch) at step 80: 239.6106, Accuracy: 0.8401
Training loss (for one batch) at step 90: 238.7744, Accuracy: 0.8377
Training loss (for one batch) at step 100: 226.8386, Accuracy: 0.8374
Training loss (for one batch) at step 110: 222.4343, Accuracy: 0.8390
Training loss (for one batch) at step 120: 235.5725, Accuracy: 0.8394
Training loss (for one batch) at step 130: 228.8778, Accuracy: 0.8392
Training loss (for one batch) at step 140: 225.0631, Accuracy: 0.8394
---- Training ----
Training loss: 214.2887
Training acc over epoch: 0.8395
---- Validation ----
Validation loss: 71.4977
Validation acc: 0.7286
Time taken: 38.68s
Start of epoch 36
Training loss (for one batch) at step 0: 224.3214, Accuracy: 0.8800
Training loss (for one batch) at step 10: 222.0542, Accuracy: 0.8427
Training loss (for one batch) at step 20: 215.8426, Accuracy: 0.8400
Training loss (for one batch) at step 30: 237.9241, Accuracy: 0.8413
Training loss (for one batch) at step 40: 202.2120, Accuracy: 0.8429
Training loss (for one batch) at step 50: 263.9170, Accuracy: 0.8410
Training loss (for one batch) at step 60: 220.6350, Accuracy: 0.8402
Training loss (for one batch) at step 70: 214.9335, Accuracy: 0.8408
Training loss (for one batch) at step 80: 228.2008, Accuracy: 0.8401
Training loss (for one batch) at step 90: 220.2791, Accuracy: 0.8385
Training loss (for one batch) at step 100: 232.5062, Accuracy: 0.8364
Training loss (for one batch) at step 110: 236.0719, Accuracy: 0.8369
Training loss (for one batch) at step 120: 226.4164, Accuracy: 0.8387
Training loss (for one batch) at step 130: 240.0429, Accuracy: 0.8377
Training loss (for one batch) at step 140: 234.6554, Accuracy: 0.8390
---- Training ----
Training loss: 197.9826
Training acc over epoch: 0.8385
---- Validation ----
Validation loss: 63.7251
Validation acc: 0.7276
Time taken: 71.50s
Start of epoch 37
Training loss (for one batch) at step 0: 238.2913, Accuracy: 0.8100
Training loss (for one batch) at step 10: 224.3347, Accuracy: 0.8555
Training loss (for one batch) at step 20: 203.8954, Accuracy: 0.8476
Training loss (for one batch) at step 30: 222.8010, Accuracy: 0.8445
Training loss (for one batch) at step 40: 223.6740, Accuracy: 0.8451
Training loss (for one batch) at step 50: 243.3969, Accuracy: 0.8473
Training loss (for one batch) at step 60: 243.6061, Accuracy: 0.8464
Training loss (for one batch) at step 70: 215.6344, Accuracy: 0.8459
Training loss (for one batch) at step 80: 225.6990, Accuracy: 0.8425
Training loss (for one batch) at step 90: 245.5683, Accuracy: 0.8423
Training loss (for one batch) at step 100: 225.5854, Accuracy: 0.8419
Training loss (for one batch) at step 110: 217.6298, Accuracy: 0.8434
Training loss (for one batch) at step 120: 227.2699, Accuracy: 0.8435
Training loss (for one batch) at step 130: 235.5796, Accuracy: 0.8440
Training loss (for one batch) at step 140: 209.1396, Accuracy: 0.8429
---- Training ----
Training loss: 203.7095
Training acc over epoch: 0.8431
---- Validation ----
Validation loss: 61.8124
Validation acc: 0.7329
Time taken: 39.98s
Start of epoch 38
Training loss (for one batch) at step 0: 215.1196, Accuracy: 0.8000
Training loss (for one batch) at step 10: 198.2802, Accuracy: 0.8336
Training loss (for one batch) at step 20: 225.3797, Accuracy: 0.8252
Training loss (for one batch) at step 30: 217.7483, Accuracy: 0.8342
Training loss (for one batch) at step 40: 239.9542, Accuracy: 0.8383
Training loss (for one batch) at step 50: 210.3000, Accuracy: 0.8427
Training loss (for one batch) at step 60: 233.6455, Accuracy: 0.8431
Training loss (for one batch) at step 70: 236.0225, Accuracy: 0.8414
Training loss (for one batch) at step 80: 212.6491, Accuracy: 0.8396
Training loss (for one batch) at step 90: 229.1385, Accuracy: 0.8368
Training loss (for one batch) at step 100: 225.1716, Accuracy: 0.8358
Training loss (for one batch) at step 110: 220.5824, Accuracy: 0.8377
Training loss (for one batch) at step 120: 235.9718, Accuracy: 0.8381
Training loss (for one batch) at step 130: 222.7420, Accuracy: 0.8369
Training loss (for one batch) at step 140: 239.5012, Accuracy: 0.8384
---- Training ----
Training loss: 220.2077
Training acc over epoch: 0.8377
---- Validation ----
Validation loss: 79.4969
Validation acc: 0.7329
Time taken: 70.09s
Start of epoch 39
Training loss (for one batch) at step 0: 216.5461, Accuracy: 0.8600
Training loss (for one batch) at step 10: 217.5700, Accuracy: 0.8518
Training loss (for one batch) at step 20: 236.8207, Accuracy: 0.8514
Training loss (for one batch) at step 30: 213.5869, Accuracy: 0.8474
Training loss (for one batch) at step 40: 221.4973, Accuracy: 0.8515
Training loss (for one batch) at step 50: 224.2252, Accuracy: 0.8547
Training loss (for one batch) at step 60: 225.5640, Accuracy: 0.8548
Training loss (for one batch) at step 70: 224.0919, Accuracy: 0.8521
Training loss (for one batch) at step 80: 234.3618, Accuracy: 0.8501
Training loss (for one batch) at step 90: 223.9925, Accuracy: 0.8493
Training loss (for one batch) at step 100: 220.8312, Accuracy: 0.8503
Training loss (for one batch) at step 110: 212.5044, Accuracy: 0.8485
Training loss (for one batch) at step 120: 234.4184, Accuracy: 0.8492
Training loss (for one batch) at step 130: 205.7584, Accuracy: 0.8490
Training loss (for one batch) at step 140: 249.1141, Accuracy: 0.8462
---- Training ----
Training loss: 203.4588
Training acc over epoch: 0.8459
---- Validation ----
Validation loss: 68.3646
Validation acc: 0.7273
Time taken: 40.03s
Start of epoch 40
Training loss (for one batch) at step 0: 207.0736, Accuracy: 0.8300
Training loss (for one batch) at step 10: 214.0502, Accuracy: 0.8591
Training loss (for one batch) at step 20: 210.8700, Accuracy: 0.8614
Training loss (for one batch) at step 30: 235.9331, Accuracy: 0.8535
Training loss (for one batch) at step 40: 207.5699, Accuracy: 0.8517
Training loss (for one batch) at step 50: 238.5863, Accuracy: 0.8508
Training loss (for one batch) at step 60: 243.9941, Accuracy: 0.8503
Training loss (for one batch) at step 70: 211.1517, Accuracy: 0.8458
Training loss (for one batch) at step 80: 240.7458, Accuracy: 0.8423
Training loss (for one batch) at step 90: 222.4530, Accuracy: 0.8431
Training loss (for one batch) at step 100: 237.1585, Accuracy: 0.8437
Training loss (for one batch) at step 110: 216.8531, Accuracy: 0.8438
Training loss (for one batch) at step 120: 210.6289, Accuracy: 0.8437
Training loss (for one batch) at step 130: 213.6298, Accuracy: 0.8440
Training loss (for one batch) at step 140: 238.6382, Accuracy: 0.8440
---- Training ----
Training loss: 200.5159
Training acc over epoch: 0.8430
---- Validation ----
Validation loss: 84.7053
Validation acc: 0.7319
Time taken: 69.85s
Start of epoch 41
Training loss (for one batch) at step 0: 207.8313, Accuracy: 0.8100
Training loss (for one batch) at step 10: 236.1079, Accuracy: 0.8527
Training loss (for one batch) at step 20: 204.2724, Accuracy: 0.8581
Training loss (for one batch) at step 30: 212.2942, Accuracy: 0.8558
Training loss (for one batch) at step 40: 211.1664, Accuracy: 0.8541
Training loss (for one batch) at step 50: 216.6557, Accuracy: 0.8531
Training loss (for one batch) at step 60: 225.1491, Accuracy: 0.8525
Training loss (for one batch) at step 70: 225.2661, Accuracy: 0.8542
Training loss (for one batch) at step 80: 220.9398, Accuracy: 0.8517
Training loss (for one batch) at step 90: 238.8253, Accuracy: 0.8504
Training loss (for one batch) at step 100: 244.8823, Accuracy: 0.8493
Training loss (for one batch) at step 110: 207.8640, Accuracy: 0.8487
Training loss (for one batch) at step 120: 211.6888, Accuracy: 0.8492
Training loss (for one batch) at step 130: 224.4673, Accuracy: 0.8482
Training loss (for one batch) at step 140: 228.4979, Accuracy: 0.8483
---- Training ----
Training loss: 179.6018
Training acc over epoch: 0.8483
---- Validation ----
Validation loss: 74.0037
Validation acc: 0.7155
Time taken: 40.68s
Start of epoch 42
Training loss (for one batch) at step 0: 221.7879, Accuracy: 0.8600
Training loss (for one batch) at step 10: 221.4762, Accuracy: 0.8373
Training loss (for one batch) at step 20: 217.2875, Accuracy: 0.8457
Training loss (for one batch) at step 30: 245.0501, Accuracy: 0.8487
Training loss (for one batch) at step 40: 213.3032, Accuracy: 0.8522
Training loss (for one batch) at step 50: 202.1278, Accuracy: 0.8549
Training loss (for one batch) at step 60: 218.2595, Accuracy: 0.8526
Training loss (for one batch) at step 70: 220.5493, Accuracy: 0.8487
Training loss (for one batch) at step 80: 231.2080, Accuracy: 0.8480
Training loss (for one batch) at step 90: 225.4845, Accuracy: 0.8464
Training loss (for one batch) at step 100: 197.8558, Accuracy: 0.8458
Training loss (for one batch) at step 110: 236.8451, Accuracy: 0.8467
Training loss (for one batch) at step 120: 220.1480, Accuracy: 0.8467
Training loss (for one batch) at step 130: 211.5138, Accuracy: 0.8469
Training loss (for one batch) at step 140: 231.1666, Accuracy: 0.8464
---- Training ----
Training loss: 201.4454
Training acc over epoch: 0.8470
---- Validation ----
Validation loss: 81.4464
Validation acc: 0.7437
Time taken: 71.35s
Start of epoch 43
Training loss (for one batch) at step 0: 194.5206, Accuracy: 0.8700
Training loss (for one batch) at step 10: 224.5829, Accuracy: 0.8491
Training loss (for one batch) at step 20: 215.9561, Accuracy: 0.8443
Training loss (for one batch) at step 30: 198.9930, Accuracy: 0.8390
Training loss (for one batch) at step 40: 197.3947, Accuracy: 0.8422
Training loss (for one batch) at step 50: 229.8982, Accuracy: 0.8445
Training loss (for one batch) at step 60: 186.8010, Accuracy: 0.8462
Training loss (for one batch) at step 70: 199.8136, Accuracy: 0.8432
Training loss (for one batch) at step 80: 212.4547, Accuracy: 0.8441
Training loss (for one batch) at step 90: 216.2263, Accuracy: 0.8429
Training loss (for one batch) at step 100: 205.6213, Accuracy: 0.8432
Training loss (for one batch) at step 110: 224.1265, Accuracy: 0.8440
Training loss (for one batch) at step 120: 214.5001, Accuracy: 0.8452
Training loss (for one batch) at step 130: 220.8449, Accuracy: 0.8446
Training loss (for one batch) at step 140: 223.0911, Accuracy: 0.8441
---- Training ----
Training loss: 184.6183
Training acc over epoch: 0.8448
---- Validation ----
Validation loss: 77.2103
Validation acc: 0.7356
Time taken: 41.75s
Start of epoch 44
Training loss (for one batch) at step 0: 214.0957, Accuracy: 0.7800
Training loss (for one batch) at step 10: 240.9085, Accuracy: 0.8436
Training loss (for one batch) at step 20: 212.6217, Accuracy: 0.8476
Training loss (for one batch) at step 30: 236.5129, Accuracy: 0.8477
Training loss (for one batch) at step 40: 199.8845, Accuracy: 0.8520
Training loss (for one batch) at step 50: 212.4296, Accuracy: 0.8518
Training loss (for one batch) at step 60: 226.4954, Accuracy: 0.8538
Training loss (for one batch) at step 70: 219.4652, Accuracy: 0.8524
Training loss (for one batch) at step 80: 251.5585, Accuracy: 0.8510
Training loss (for one batch) at step 90: 228.5986, Accuracy: 0.8498
Training loss (for one batch) at step 100: 237.5534, Accuracy: 0.8472
Training loss (for one batch) at step 110: 202.6739, Accuracy: 0.8486
Training loss (for one batch) at step 120: 222.9687, Accuracy: 0.8499
Training loss (for one batch) at step 130: 188.5314, Accuracy: 0.8494
Training loss (for one batch) at step 140: 221.6092, Accuracy: 0.8491
---- Training ----
Training loss: 195.2994
Training acc over epoch: 0.8499
---- Validation ----
Validation loss: 90.1100
Validation acc: 0.7351
Time taken: 67.99s
Start of epoch 45
Training loss (for one batch) at step 0: 219.2766, Accuracy: 0.7500
Training loss (for one batch) at step 10: 212.6325, Accuracy: 0.8500
Training loss (for one batch) at step 20: 204.4203, Accuracy: 0.8514
Training loss (for one batch) at step 30: 213.6050, Accuracy: 0.8487
Training loss (for one batch) at step 40: 209.1707, Accuracy: 0.8532
Training loss (for one batch) at step 50: 206.0803, Accuracy: 0.8561
Training loss (for one batch) at step 60: 213.8869, Accuracy: 0.8561
Training loss (for one batch) at step 70: 188.3012, Accuracy: 0.8569
Training loss (for one batch) at step 80: 217.0327, Accuracy: 0.8538
Training loss (for one batch) at step 90: 228.1745, Accuracy: 0.8531
Training loss (for one batch) at step 100: 207.0130, Accuracy: 0.8513
Training loss (for one batch) at step 110: 233.9702, Accuracy: 0.8523
Training loss (for one batch) at step 120: 217.2225, Accuracy: 0.8525
Training loss (for one batch) at step 130: 226.7958, Accuracy: 0.8525
Training loss (for one batch) at step 140: 205.3326, Accuracy: 0.8514
---- Training ----
Training loss: 193.7800
Training acc over epoch: 0.8518
---- Validation ----
Validation loss: 84.7421
Validation acc: 0.7372
Time taken: 39.09s
Start of epoch 46
Training loss (for one batch) at step 0: 204.7727, Accuracy: 0.8600
Training loss (for one batch) at step 10: 194.8137, Accuracy: 0.8436
Training loss (for one batch) at step 20: 196.2810, Accuracy: 0.8462
Training loss (for one batch) at step 30: 201.8774, Accuracy: 0.8465
Training loss (for one batch) at step 40: 216.2102, Accuracy: 0.8488
Training loss (for one batch) at step 50: 209.0035, Accuracy: 0.8563
Training loss (for one batch) at step 60: 217.0860, Accuracy: 0.8539
Training loss (for one batch) at step 70: 240.4558, Accuracy: 0.8537
Training loss (for one batch) at step 80: 224.0364, Accuracy: 0.8491
Training loss (for one batch) at step 90: 235.5498, Accuracy: 0.8498
Training loss (for one batch) at step 100: 204.9197, Accuracy: 0.8494
Training loss (for one batch) at step 110: 206.5943, Accuracy: 0.8513
Training loss (for one batch) at step 120: 235.6870, Accuracy: 0.8507
Training loss (for one batch) at step 130: 219.3824, Accuracy: 0.8505
Training loss (for one batch) at step 140: 206.5710, Accuracy: 0.8505
---- Training ----
Training loss: 186.8335
Training acc over epoch: 0.8504
---- Validation ----
Validation loss: 68.8416
Validation acc: 0.7327
Time taken: 69.10s
Start of epoch 47
Training loss (for one batch) at step 0: 197.0962, Accuracy: 0.8900
Training loss (for one batch) at step 10: 214.1809, Accuracy: 0.8418
Training loss (for one batch) at step 20: 215.9879, Accuracy: 0.8486
Training loss (for one batch) at step 30: 193.4881, Accuracy: 0.8465
Training loss (for one batch) at step 40: 194.3926, Accuracy: 0.8488
Training loss (for one batch) at step 50: 211.4341, Accuracy: 0.8545
Training loss (for one batch) at step 60: 206.2711, Accuracy: 0.8564
Training loss (for one batch) at step 70: 226.2395, Accuracy: 0.8559
Training loss (for one batch) at step 80: 221.5688, Accuracy: 0.8532
Training loss (for one batch) at step 90: 202.2805, Accuracy: 0.8534
Training loss (for one batch) at step 100: 219.2630, Accuracy: 0.8532
Training loss (for one batch) at step 110: 200.3749, Accuracy: 0.8528
Training loss (for one batch) at step 120: 200.5491, Accuracy: 0.8535
Training loss (for one batch) at step 130: 208.0553, Accuracy: 0.8535
Training loss (for one batch) at step 140: 204.2602, Accuracy: 0.8529
---- Training ----
Training loss: 169.6255
Training acc over epoch: 0.8528
---- Validation ----
Validation loss: 84.4762
Validation acc: 0.7337
Time taken: 40.12s
Start of epoch 48
Training loss (for one batch) at step 0: 199.8466, Accuracy: 0.8300
Training loss (for one batch) at step 10: 205.3529, Accuracy: 0.8555
Training loss (for one batch) at step 20: 214.1979, Accuracy: 0.8481
Training loss (for one batch) at step 30: 206.1032, Accuracy: 0.8513
Training loss (for one batch) at step 40: 201.5154, Accuracy: 0.8551
Training loss (for one batch) at step 50: 200.6290, Accuracy: 0.8571
Training loss (for one batch) at step 60: 206.0937, Accuracy: 0.8569
Training loss (for one batch) at step 70: 198.7331, Accuracy: 0.8563
Training loss (for one batch) at step 80: 203.3989, Accuracy: 0.8517
Training loss (for one batch) at step 90: 223.7517, Accuracy: 0.8507
Training loss (for one batch) at step 100: 214.3686, Accuracy: 0.8515
Training loss (for one batch) at step 110: 202.7460, Accuracy: 0.8518
Training loss (for one batch) at step 120: 197.4067, Accuracy: 0.8515
Training loss (for one batch) at step 130: 218.1138, Accuracy: 0.8525
Training loss (for one batch) at step 140: 197.6913, Accuracy: 0.8524
---- Training ----
Training loss: 183.0899
Training acc over epoch: 0.8518
---- Validation ----
Validation loss: 80.1335
Validation acc: 0.7394
Time taken: 67.78s
Start of epoch 49
Training loss (for one batch) at step 0: 202.8529, Accuracy: 0.8600
Training loss (for one batch) at step 10: 243.1760, Accuracy: 0.8455
Training loss (for one batch) at step 20: 209.5426, Accuracy: 0.8557
Training loss (for one batch) at step 30: 198.4198, Accuracy: 0.8516
Training loss (for one batch) at step 40: 225.9464, Accuracy: 0.8559
Training loss (for one batch) at step 50: 184.5410, Accuracy: 0.8582
Training loss (for one batch) at step 60: 198.5051, Accuracy: 0.8551
Training loss (for one batch) at step 70: 199.0852, Accuracy: 0.8530
Training loss (for one batch) at step 80: 204.0543, Accuracy: 0.8525
Training loss (for one batch) at step 90: 217.9881, Accuracy: 0.8527
Training loss (for one batch) at step 100: 215.2215, Accuracy: 0.8511
Training loss (for one batch) at step 110: 191.7576, Accuracy: 0.8526
Training loss (for one batch) at step 120: 224.1674, Accuracy: 0.8517
Training loss (for one batch) at step 130: 208.4098, Accuracy: 0.8518
Training loss (for one batch) at step 140: 199.1145, Accuracy: 0.8521
---- Training ----
Training loss: 181.0873
Training acc over epoch: 0.8523
---- Validation ----
Validation loss: 126.0566
Validation acc: 0.7294
Time taken: 39.32s
===== Q: 0.0001
Validation acc: 0.7405
Validation AUC: 0.7375
Validation Balanced_ACC: 0.4771
Validation MI: 0.1371
Validation Normalized MI: 0.2055
Validation Adjusted MI: 0.2055
Validation aUc_Sklearn: 0.8306
Start of epoch 0
Training loss (for one batch) at step 0: 527.5986, Accuracy: 0.4500
Training loss (for one batch) at step 10: 484.1193, Accuracy: 0.5245
Training loss (for one batch) at step 20: 423.7532, Accuracy: 0.5390
Training loss (for one batch) at step 30: 473.1334, Accuracy: 0.5374
Training loss (for one batch) at step 40: 439.4325, Accuracy: 0.5417
Training loss (for one batch) at step 50: 413.2783, Accuracy: 0.5516
Training loss (for one batch) at step 60: 461.4584, Accuracy: 0.5533
Training loss (for one batch) at step 70: 454.9706, Accuracy: 0.5532
Training loss (for one batch) at step 80: 437.4577, Accuracy: 0.5530
Training loss (for one batch) at step 90: 438.3748, Accuracy: 0.5577
Training loss (for one batch) at step 100: 424.0384, Accuracy: 0.5617
Training loss (for one batch) at step 110: 457.6732, Accuracy: 0.5630
Training loss (for one batch) at step 120: 480.1002, Accuracy: 0.5647
Training loss (for one batch) at step 130: 398.5197, Accuracy: 0.5636
Training loss (for one batch) at step 140: 432.7971, Accuracy: 0.5642
---- Training ----
Training loss: 375.6224
Training acc over epoch: 0.5666
---- Validation ----
Validation loss: 91.1115
Validation acc: 0.5134
Time taken: 73.22s
Start of epoch 1
Training loss (for one batch) at step 0: 413.8096, Accuracy: 0.5700
Training loss (for one batch) at step 10: 422.1463, Accuracy: 0.6336
Training loss (for one batch) at step 20: 382.1375, Accuracy: 0.6271
Training loss (for one batch) at step 30: 451.0680, Accuracy: 0.6145
Training loss (for one batch) at step 40: 395.3322, Accuracy: 0.6122
Training loss (for one batch) at step 50: 374.7920, Accuracy: 0.6143
Training loss (for one batch) at step 60: 407.3326, Accuracy: 0.6125
Training loss (for one batch) at step 70: 404.7450, Accuracy: 0.6128
Training loss (for one batch) at step 80: 422.5473, Accuracy: 0.6096
Training loss (for one batch) at step 90: 401.6553, Accuracy: 0.6096
Training loss (for one batch) at step 100: 403.7165, Accuracy: 0.6102
Training loss (for one batch) at step 110: 383.1733, Accuracy: 0.6107
Training loss (for one batch) at step 120: 365.2353, Accuracy: 0.6144
Training loss (for one batch) at step 130: 405.7834, Accuracy: 0.6135
Training loss (for one batch) at step 140: 386.5142, Accuracy: 0.6152
---- Training ----
Training loss: 329.8106
Training acc over epoch: 0.6157
---- Validation ----
Validation loss: 86.8742
Validation acc: 0.5419
Time taken: 54.71s
Start of epoch 2
Training loss (for one batch) at step 0: 352.3314, Accuracy: 0.7200
Training loss (for one batch) at step 10: 399.2194, Accuracy: 0.6418
Training loss (for one batch) at step 20: 391.8743, Accuracy: 0.6443
Training loss (for one batch) at step 30: 349.3879, Accuracy: 0.6435
Training loss (for one batch) at step 40: 372.9807, Accuracy: 0.6415
Training loss (for one batch) at step 50: 383.5710, Accuracy: 0.6355
Training loss (for one batch) at step 60: 359.5850, Accuracy: 0.6367
Training loss (for one batch) at step 70: 368.7337, Accuracy: 0.6369
Training loss (for one batch) at step 80: 366.2454, Accuracy: 0.6401
Training loss (for one batch) at step 90: 374.5175, Accuracy: 0.6407
Training loss (for one batch) at step 100: 368.7094, Accuracy: 0.6396
Training loss (for one batch) at step 110: 346.9746, Accuracy: 0.6403
Training loss (for one batch) at step 120: 355.5563, Accuracy: 0.6407
Training loss (for one batch) at step 130: 353.6960, Accuracy: 0.6397
Training loss (for one batch) at step 140: 367.8237, Accuracy: 0.6401
---- Training ----
Training loss: 312.4849
Training acc over epoch: 0.6398
---- Validation ----
Validation loss: 72.9293
Validation acc: 0.6883
Time taken: 69.75s
Start of epoch 3
Training loss (for one batch) at step 0: 364.5711, Accuracy: 0.6500
Training loss (for one batch) at step 10: 325.1856, Accuracy: 0.6600
Training loss (for one batch) at step 20: 369.5432, Accuracy: 0.6438
Training loss (for one batch) at step 30: 364.5819, Accuracy: 0.6439
Training loss (for one batch) at step 40: 362.9421, Accuracy: 0.6434
Training loss (for one batch) at step 50: 373.7325, Accuracy: 0.6459
Training loss (for one batch) at step 60: 373.2194, Accuracy: 0.6466
Training loss (for one batch) at step 70: 373.3829, Accuracy: 0.6438
Training loss (for one batch) at step 80: 354.2503, Accuracy: 0.6463
Training loss (for one batch) at step 90: 350.3513, Accuracy: 0.6446
Training loss (for one batch) at step 100: 348.6281, Accuracy: 0.6455
Training loss (for one batch) at step 110: 349.3692, Accuracy: 0.6452
Training loss (for one batch) at step 120: 350.9062, Accuracy: 0.6464
Training loss (for one batch) at step 130: 360.9332, Accuracy: 0.6481
Training loss (for one batch) at step 140: 360.2079, Accuracy: 0.6495
---- Training ----
Training loss: 291.6394
Training acc over epoch: 0.6500
---- Validation ----
Validation loss: 71.8143
Validation acc: 0.7045
Time taken: 42.81s
Start of epoch 4
Training loss (for one batch) at step 0: 337.0139, Accuracy: 0.7500
Training loss (for one batch) at step 10: 342.7183, Accuracy: 0.6955
Training loss (for one batch) at step 20: 344.1118, Accuracy: 0.6733
Training loss (for one batch) at step 30: 393.5463, Accuracy: 0.6787
Training loss (for one batch) at step 40: 347.4206, Accuracy: 0.6700
Training loss (for one batch) at step 50: 345.5345, Accuracy: 0.6678
Training loss (for one batch) at step 60: 366.5325, Accuracy: 0.6689
Training loss (for one batch) at step 70: 357.7226, Accuracy: 0.6679
Training loss (for one batch) at step 80: 345.2705, Accuracy: 0.6693
Training loss (for one batch) at step 90: 338.4989, Accuracy: 0.6704
Training loss (for one batch) at step 100: 351.0471, Accuracy: 0.6714
Training loss (for one batch) at step 110: 350.8172, Accuracy: 0.6715
Training loss (for one batch) at step 120: 353.2338, Accuracy: 0.6725
Training loss (for one batch) at step 130: 355.8357, Accuracy: 0.6730
Training loss (for one batch) at step 140: 347.5988, Accuracy: 0.6713
---- Training ----
Training loss: 295.2686
Training acc over epoch: 0.6722
---- Validation ----
Validation loss: 68.5381
Validation acc: 0.7058
Time taken: 69.90s
Start of epoch 5
Training loss (for one batch) at step 0: 343.3817, Accuracy: 0.7000
Training loss (for one batch) at step 10: 348.3071, Accuracy: 0.6764
Training loss (for one batch) at step 20: 320.7846, Accuracy: 0.6929
Training loss (for one batch) at step 30: 307.7726, Accuracy: 0.6958
Training loss (for one batch) at step 40: 335.4081, Accuracy: 0.6915
Training loss (for one batch) at step 50: 325.5349, Accuracy: 0.6931
Training loss (for one batch) at step 60: 341.6328, Accuracy: 0.6867
Training loss (for one batch) at step 70: 339.9951, Accuracy: 0.6866
Training loss (for one batch) at step 80: 367.3934, Accuracy: 0.6879
Training loss (for one batch) at step 90: 318.5256, Accuracy: 0.6852
Training loss (for one batch) at step 100: 336.7431, Accuracy: 0.6858
Training loss (for one batch) at step 110: 340.6738, Accuracy: 0.6884
Training loss (for one batch) at step 120: 321.7924, Accuracy: 0.6887
Training loss (for one batch) at step 130: 323.8076, Accuracy: 0.6900
Training loss (for one batch) at step 140: 345.3731, Accuracy: 0.6877
---- Training ----
Training loss: 295.0337
Training acc over epoch: 0.6879
---- Validation ----
Validation loss: 71.8220
Validation acc: 0.6948
Time taken: 41.04s
Start of epoch 6
Training loss (for one batch) at step 0: 340.8214, Accuracy: 0.5500
Training loss (for one batch) at step 10: 305.7826, Accuracy: 0.7045
Training loss (for one batch) at step 20: 327.1538, Accuracy: 0.7038
Training loss (for one batch) at step 30: 323.5538, Accuracy: 0.6968
Training loss (for one batch) at step 40: 313.0959, Accuracy: 0.6929
Training loss (for one batch) at step 50: 308.3360, Accuracy: 0.6959
Training loss (for one batch) at step 60: 305.5701, Accuracy: 0.7020
Training loss (for one batch) at step 70: 324.3964, Accuracy: 0.6997
Training loss (for one batch) at step 80: 325.6848, Accuracy: 0.7001
Training loss (for one batch) at step 90: 354.5569, Accuracy: 0.6992
Training loss (for one batch) at step 100: 309.4525, Accuracy: 0.7020
Training loss (for one batch) at step 110: 329.3031, Accuracy: 0.7034
Training loss (for one batch) at step 120: 329.6373, Accuracy: 0.7027
Training loss (for one batch) at step 130: 335.3600, Accuracy: 0.7027
Training loss (for one batch) at step 140: 344.4106, Accuracy: 0.7011
---- Training ----
Training loss: 266.7270
Training acc over epoch: 0.7029
---- Validation ----
Validation loss: 66.3468
Validation acc: 0.7243
Time taken: 70.81s
Start of epoch 7
Training loss (for one batch) at step 0: 320.9370, Accuracy: 0.7400
Training loss (for one batch) at step 10: 311.2009, Accuracy: 0.7327
Training loss (for one batch) at step 20: 315.4370, Accuracy: 0.7290
Training loss (for one batch) at step 30: 311.0593, Accuracy: 0.7200
Training loss (for one batch) at step 40: 328.2438, Accuracy: 0.7163
Training loss (for one batch) at step 50: 327.8079, Accuracy: 0.7229
Training loss (for one batch) at step 60: 322.7364, Accuracy: 0.7231
Training loss (for one batch) at step 70: 317.3809, Accuracy: 0.7249
Training loss (for one batch) at step 80: 347.7578, Accuracy: 0.7228
Training loss (for one batch) at step 90: 318.2994, Accuracy: 0.7212
Training loss (for one batch) at step 100: 309.1743, Accuracy: 0.7188
Training loss (for one batch) at step 110: 309.0551, Accuracy: 0.7209
Training loss (for one batch) at step 120: 315.9899, Accuracy: 0.7217
Training loss (for one batch) at step 130: 299.7163, Accuracy: 0.7217
Training loss (for one batch) at step 140: 323.9458, Accuracy: 0.7212
---- Training ----
Training loss: 299.3060
Training acc over epoch: 0.7205
---- Validation ----
Validation loss: 68.6832
Validation acc: 0.6985
Time taken: 41.78s
Start of epoch 8
Training loss (for one batch) at step 0: 308.3130, Accuracy: 0.7200
Training loss (for one batch) at step 10: 309.3829, Accuracy: 0.7245
Training loss (for one batch) at step 20: 309.8175, Accuracy: 0.7300
Training loss (for one batch) at step 30: 332.4000, Accuracy: 0.7300
Training loss (for one batch) at step 40: 316.0762, Accuracy: 0.7349
Training loss (for one batch) at step 50: 312.5413, Accuracy: 0.7327
Training loss (for one batch) at step 60: 316.2780, Accuracy: 0.7321
Training loss (for one batch) at step 70: 330.6068, Accuracy: 0.7330
Training loss (for one batch) at step 80: 303.6725, Accuracy: 0.7315
Training loss (for one batch) at step 90: 308.1479, Accuracy: 0.7303
Training loss (for one batch) at step 100: 316.3442, Accuracy: 0.7291
Training loss (for one batch) at step 110: 290.3474, Accuracy: 0.7309
Training loss (for one batch) at step 120: 311.6064, Accuracy: 0.7305
Training loss (for one batch) at step 130: 340.3927, Accuracy: 0.7287
Training loss (for one batch) at step 140: 314.1768, Accuracy: 0.7272
---- Training ----
Training loss: 286.2704
Training acc over epoch: 0.7274
---- Validation ----
Validation loss: 68.6967
Validation acc: 0.7184
Time taken: 70.11s
Start of epoch 9
Training loss (for one batch) at step 0: 311.5945, Accuracy: 0.6900
Training loss (for one batch) at step 10: 307.1524, Accuracy: 0.7364
Training loss (for one batch) at step 20: 304.2565, Accuracy: 0.7424
Training loss (for one batch) at step 30: 327.4934, Accuracy: 0.7303
Training loss (for one batch) at step 40: 299.0128, Accuracy: 0.7383
Training loss (for one batch) at step 50: 296.2714, Accuracy: 0.7416
Training loss (for one batch) at step 60: 307.7940, Accuracy: 0.7449
Training loss (for one batch) at step 70: 324.5955, Accuracy: 0.7432
Training loss (for one batch) at step 80: 300.6064, Accuracy: 0.7422
Training loss (for one batch) at step 90: 311.1146, Accuracy: 0.7404
Training loss (for one batch) at step 100: 323.5187, Accuracy: 0.7400
Training loss (for one batch) at step 110: 309.1363, Accuracy: 0.7444
Training loss (for one batch) at step 120: 324.5754, Accuracy: 0.7433
Training loss (for one batch) at step 130: 294.1886, Accuracy: 0.7443
Training loss (for one batch) at step 140: 317.5658, Accuracy: 0.7418
---- Training ----
Training loss: 284.1860
Training acc over epoch: 0.7416
---- Validation ----
Validation loss: 68.1358
Validation acc: 0.7219
Time taken: 40.62s
Start of epoch 10
Training loss (for one batch) at step 0: 310.2429, Accuracy: 0.7100
Training loss (for one batch) at step 10: 300.2978, Accuracy: 0.7345
Training loss (for one batch) at step 20: 308.7468, Accuracy: 0.7495
Training loss (for one batch) at step 30: 294.1781, Accuracy: 0.7513
Training loss (for one batch) at step 40: 302.7307, Accuracy: 0.7539
Training loss (for one batch) at step 50: 293.5435, Accuracy: 0.7551
Training loss (for one batch) at step 60: 312.3758, Accuracy: 0.7531
Training loss (for one batch) at step 70: 322.9708, Accuracy: 0.7515
Training loss (for one batch) at step 80: 304.9420, Accuracy: 0.7499
Training loss (for one batch) at step 90: 293.7973, Accuracy: 0.7468
Training loss (for one batch) at step 100: 296.7136, Accuracy: 0.7474
Training loss (for one batch) at step 110: 301.3406, Accuracy: 0.7470
Training loss (for one batch) at step 120: 297.8398, Accuracy: 0.7482
Training loss (for one batch) at step 130: 302.2362, Accuracy: 0.7482
Training loss (for one batch) at step 140: 292.7247, Accuracy: 0.7496
---- Training ----
Training loss: 268.6658
Training acc over epoch: 0.7500
---- Validation ----
Validation loss: 68.7989
Validation acc: 0.7136
Time taken: 70.19s
Start of epoch 11
Training loss (for one batch) at step 0: 287.5923, Accuracy: 0.7400
Training loss (for one batch) at step 10: 288.5273, Accuracy: 0.7527
Training loss (for one batch) at step 20: 305.5519, Accuracy: 0.7524
Training loss (for one batch) at step 30: 303.8906, Accuracy: 0.7542
Training loss (for one batch) at step 40: 296.3648, Accuracy: 0.7505
Training loss (for one batch) at step 50: 299.0033, Accuracy: 0.7525
Training loss (for one batch) at step 60: 281.5224, Accuracy: 0.7552
Training loss (for one batch) at step 70: 287.0531, Accuracy: 0.7561
Training loss (for one batch) at step 80: 296.0748, Accuracy: 0.7562
Training loss (for one batch) at step 90: 295.1410, Accuracy: 0.7520
Training loss (for one batch) at step 100: 302.4777, Accuracy: 0.7530
Training loss (for one batch) at step 110: 306.0706, Accuracy: 0.7561
Training loss (for one batch) at step 120: 302.4145, Accuracy: 0.7572
Training loss (for one batch) at step 130: 305.3720, Accuracy: 0.7569
Training loss (for one batch) at step 140: 305.5745, Accuracy: 0.7562
---- Training ----
Training loss: 266.9718
Training acc over epoch: 0.7559
---- Validation ----
Validation loss: 78.2004
Validation acc: 0.7246
Time taken: 39.85s
Start of epoch 12
Training loss (for one batch) at step 0: 293.1306, Accuracy: 0.7800
Training loss (for one batch) at step 10: 302.3249, Accuracy: 0.7855
Training loss (for one batch) at step 20: 300.7356, Accuracy: 0.7776
Training loss (for one batch) at step 30: 292.7833, Accuracy: 0.7671
Training loss (for one batch) at step 40: 279.1132, Accuracy: 0.7654
Training loss (for one batch) at step 50: 292.5120, Accuracy: 0.7665
Training loss (for one batch) at step 60: 267.4562, Accuracy: 0.7682
Training loss (for one batch) at step 70: 302.8366, Accuracy: 0.7642
Training loss (for one batch) at step 80: 307.7257, Accuracy: 0.7623
Training loss (for one batch) at step 90: 289.1540, Accuracy: 0.7612
Training loss (for one batch) at step 100: 289.1783, Accuracy: 0.7598
Training loss (for one batch) at step 110: 276.9674, Accuracy: 0.7635
Training loss (for one batch) at step 120: 288.2111, Accuracy: 0.7645
Training loss (for one batch) at step 130: 281.3755, Accuracy: 0.7636
Training loss (for one batch) at step 140: 299.9367, Accuracy: 0.7632
---- Training ----
Training loss: 255.3710
Training acc over epoch: 0.7648
---- Validation ----
Validation loss: 71.0651
Validation acc: 0.7144
Time taken: 68.70s
Start of epoch 13
Training loss (for one batch) at step 0: 296.2023, Accuracy: 0.7100
Training loss (for one batch) at step 10: 286.3375, Accuracy: 0.7727
Training loss (for one batch) at step 20: 287.4993, Accuracy: 0.7757
Training loss (for one batch) at step 30: 304.1086, Accuracy: 0.7726
Training loss (for one batch) at step 40: 292.0949, Accuracy: 0.7722
Training loss (for one batch) at step 50: 312.6662, Accuracy: 0.7747
Training loss (for one batch) at step 60: 332.0870, Accuracy: 0.7785
Training loss (for one batch) at step 70: 310.1657, Accuracy: 0.7751
Training loss (for one batch) at step 80: 290.9586, Accuracy: 0.7741
Training loss (for one batch) at step 90: 289.1524, Accuracy: 0.7743
Training loss (for one batch) at step 100: 296.8352, Accuracy: 0.7736
Training loss (for one batch) at step 110: 288.9477, Accuracy: 0.7740
Training loss (for one batch) at step 120: 272.5976, Accuracy: 0.7734
Training loss (for one batch) at step 130: 286.6409, Accuracy: 0.7753
Training loss (for one batch) at step 140: 289.9462, Accuracy: 0.7751
---- Training ----
Training loss: 267.0271
Training acc over epoch: 0.7744
---- Validation ----
Validation loss: 66.8267
Validation acc: 0.7200
Time taken: 40.01s
Start of epoch 14
Training loss (for one batch) at step 0: 274.4041, Accuracy: 0.7800
Training loss (for one batch) at step 10: 281.1799, Accuracy: 0.7891
Training loss (for one batch) at step 20: 284.5640, Accuracy: 0.7800
Training loss (for one batch) at step 30: 294.5078, Accuracy: 0.7755
Training loss (for one batch) at step 40: 271.8104, Accuracy: 0.7810
Training loss (for one batch) at step 50: 271.3116, Accuracy: 0.7814
Training loss (for one batch) at step 60: 280.5667, Accuracy: 0.7848
Training loss (for one batch) at step 70: 285.3317, Accuracy: 0.7831
Training loss (for one batch) at step 80: 288.6895, Accuracy: 0.7806
Training loss (for one batch) at step 90: 278.3506, Accuracy: 0.7800
Training loss (for one batch) at step 100: 281.6333, Accuracy: 0.7789
Training loss (for one batch) at step 110: 273.0492, Accuracy: 0.7787
Training loss (for one batch) at step 120: 281.7909, Accuracy: 0.7792
Training loss (for one batch) at step 130: 285.3177, Accuracy: 0.7799
Training loss (for one batch) at step 140: 278.3694, Accuracy: 0.7800
---- Training ----
Training loss: 251.5452
Training acc over epoch: 0.7786
---- Validation ----
Validation loss: 63.7223
Validation acc: 0.7117
Time taken: 67.39s
Start of epoch 15
Training loss (for one batch) at step 0: 282.8299, Accuracy: 0.7400
Training loss (for one batch) at step 10: 282.6632, Accuracy: 0.7955
Training loss (for one batch) at step 20: 270.8885, Accuracy: 0.7952
Training loss (for one batch) at step 30: 270.3952, Accuracy: 0.7881
Training loss (for one batch) at step 40: 278.5165, Accuracy: 0.7824
Training loss (for one batch) at step 50: 281.3590, Accuracy: 0.7816
Training loss (for one batch) at step 60: 276.5955, Accuracy: 0.7834
Training loss (for one batch) at step 70: 268.8085, Accuracy: 0.7855
Training loss (for one batch) at step 80: 298.5565, Accuracy: 0.7827
Training loss (for one batch) at step 90: 283.5504, Accuracy: 0.7825
Training loss (for one batch) at step 100: 275.0671, Accuracy: 0.7821
Training loss (for one batch) at step 110: 307.2551, Accuracy: 0.7803
Training loss (for one batch) at step 120: 278.6587, Accuracy: 0.7806
Training loss (for one batch) at step 130: 264.2073, Accuracy: 0.7812
Training loss (for one batch) at step 140: 284.8057, Accuracy: 0.7811
---- Training ----
Training loss: 263.4333
Training acc over epoch: 0.7808
---- Validation ----
Validation loss: 69.0696
Validation acc: 0.7243
Time taken: 40.04s
Start of epoch 16
Training loss (for one batch) at step 0: 276.6978, Accuracy: 0.7700
Training loss (for one batch) at step 10: 274.0174, Accuracy: 0.8027
Training loss (for one batch) at step 20: 291.4106, Accuracy: 0.7995
Training loss (for one batch) at step 30: 291.3621, Accuracy: 0.7942
Training loss (for one batch) at step 40: 268.0268, Accuracy: 0.7978
Training loss (for one batch) at step 50: 294.2104, Accuracy: 0.7976
Training loss (for one batch) at step 60: 266.9556, Accuracy: 0.7962
Training loss (for one batch) at step 70: 261.6482, Accuracy: 0.7961
Training loss (for one batch) at step 80: 274.3265, Accuracy: 0.7941
Training loss (for one batch) at step 90: 296.7287, Accuracy: 0.7929
Training loss (for one batch) at step 100: 264.4135, Accuracy: 0.7897
Training loss (for one batch) at step 110: 286.2061, Accuracy: 0.7897
Training loss (for one batch) at step 120: 268.6156, Accuracy: 0.7893
Training loss (for one batch) at step 130: 272.5456, Accuracy: 0.7899
Training loss (for one batch) at step 140: 279.2499, Accuracy: 0.7891
---- Training ----
Training loss: 248.6221
Training acc over epoch: 0.7884
---- Validation ----
Validation loss: 65.3632
Validation acc: 0.7351
Time taken: 67.92s
Start of epoch 17
Training loss (for one batch) at step 0: 293.9531, Accuracy: 0.7800
Training loss (for one batch) at step 10: 260.4293, Accuracy: 0.8136
Training loss (for one batch) at step 20: 302.4777, Accuracy: 0.8033
Training loss (for one batch) at step 30: 278.2523, Accuracy: 0.8016
Training loss (for one batch) at step 40: 265.1810, Accuracy: 0.7983
Training loss (for one batch) at step 50: 276.9482, Accuracy: 0.8018
Training loss (for one batch) at step 60: 285.3032, Accuracy: 0.7998
Training loss (for one batch) at step 70: 290.5832, Accuracy: 0.8010
Training loss (for one batch) at step 80: 273.2991, Accuracy: 0.7999
Training loss (for one batch) at step 90: 281.3829, Accuracy: 0.7990
Training loss (for one batch) at step 100: 284.4566, Accuracy: 0.7980
Training loss (for one batch) at step 110: 265.6779, Accuracy: 0.7982
Training loss (for one batch) at step 120: 289.3460, Accuracy: 0.7983
Training loss (for one batch) at step 130: 295.4084, Accuracy: 0.7972
Training loss (for one batch) at step 140: 284.2663, Accuracy: 0.7961
---- Training ----
Training loss: 268.2704
Training acc over epoch: 0.7957
---- Validation ----
Validation loss: 65.8764
Validation acc: 0.7050
Time taken: 39.67s
Start of epoch 18
Training loss (for one batch) at step 0: 284.8246, Accuracy: 0.7600
Training loss (for one batch) at step 10: 271.8633, Accuracy: 0.8191
Training loss (for one batch) at step 20: 275.5013, Accuracy: 0.8043
Training loss (for one batch) at step 30: 260.5297, Accuracy: 0.7977
Training loss (for one batch) at step 40: 260.0592, Accuracy: 0.7988
Training loss (for one batch) at step 50: 258.6255, Accuracy: 0.8051
Training loss (for one batch) at step 60: 257.3095, Accuracy: 0.8044
Training loss (for one batch) at step 70: 272.2793, Accuracy: 0.8035
Training loss (for one batch) at step 80: 269.2575, Accuracy: 0.7999
Training loss (for one batch) at step 90: 271.0782, Accuracy: 0.7979
Training loss (for one batch) at step 100: 280.7691, Accuracy: 0.7980
Training loss (for one batch) at step 110: 279.1541, Accuracy: 0.7986
Training loss (for one batch) at step 120: 278.8455, Accuracy: 0.7979
Training loss (for one batch) at step 130: 274.0843, Accuracy: 0.7979
Training loss (for one batch) at step 140: 268.4203, Accuracy: 0.7978
---- Training ----
Training loss: 248.5131
Training acc over epoch: 0.7965
---- Validation ----
Validation loss: 55.4152
Validation acc: 0.7203
Time taken: 68.64s
Start of epoch 19
Training loss (for one batch) at step 0: 257.9449, Accuracy: 0.8100
Training loss (for one batch) at step 10: 260.1976, Accuracy: 0.8064
Training loss (for one batch) at step 20: 261.0378, Accuracy: 0.8076
Training loss (for one batch) at step 30: 289.5254, Accuracy: 0.8071
Training loss (for one batch) at step 40: 254.1850, Accuracy: 0.8044
Training loss (for one batch) at step 50: 281.1477, Accuracy: 0.8057
Training loss (for one batch) at step 60: 249.8896, Accuracy: 0.8044
Training loss (for one batch) at step 70: 270.9515, Accuracy: 0.8027
Training loss (for one batch) at step 80: 261.9622, Accuracy: 0.8038
Training loss (for one batch) at step 90: 263.8286, Accuracy: 0.8045
Training loss (for one batch) at step 100: 256.3716, Accuracy: 0.8048
Training loss (for one batch) at step 110: 266.3513, Accuracy: 0.8030
Training loss (for one batch) at step 120: 273.1258, Accuracy: 0.8012
Training loss (for one batch) at step 130: 286.5981, Accuracy: 0.8012
Training loss (for one batch) at step 140: 261.0540, Accuracy: 0.8005
---- Training ----
Training loss: 230.6163
Training acc over epoch: 0.8000
---- Validation ----
Validation loss: 64.9526
Validation acc: 0.7440
Time taken: 41.48s
Start of epoch 20
Training loss (for one batch) at step 0: 267.9152, Accuracy: 0.8300
Training loss (for one batch) at step 10: 254.0600, Accuracy: 0.8300
Training loss (for one batch) at step 20: 261.5470, Accuracy: 0.8143
Training loss (for one batch) at step 30: 276.0139, Accuracy: 0.8135
Training loss (for one batch) at step 40: 262.6774, Accuracy: 0.8156
Training loss (for one batch) at step 50: 247.3437, Accuracy: 0.8171
Training loss (for one batch) at step 60: 259.3389, Accuracy: 0.8166
Training loss (for one batch) at step 70: 264.9327, Accuracy: 0.8176
Training loss (for one batch) at step 80: 273.5218, Accuracy: 0.8130
Training loss (for one batch) at step 90: 271.5909, Accuracy: 0.8109
Training loss (for one batch) at step 100: 266.0771, Accuracy: 0.8087
Training loss (for one batch) at step 110: 274.3528, Accuracy: 0.8113
Training loss (for one batch) at step 120: 260.2671, Accuracy: 0.8098
Training loss (for one batch) at step 130: 251.9205, Accuracy: 0.8089
Training loss (for one batch) at step 140: 271.0856, Accuracy: 0.8074
---- Training ----
Training loss: 230.5191
Training acc over epoch: 0.8085
---- Validation ----
Validation loss: 74.2728
Validation acc: 0.7397
Time taken: 69.62s
Start of epoch 21
Training loss (for one batch) at step 0: 253.1247, Accuracy: 0.7400
Training loss (for one batch) at step 10: 277.1951, Accuracy: 0.7973
Training loss (for one batch) at step 20: 267.5648, Accuracy: 0.8014
Training loss (for one batch) at step 30: 262.8856, Accuracy: 0.8032
Training loss (for one batch) at step 40: 267.0617, Accuracy: 0.8029
Training loss (for one batch) at step 50: 285.0754, Accuracy: 0.8041
Training loss (for one batch) at step 60: 262.1369, Accuracy: 0.8075
Training loss (for one batch) at step 70: 257.0067, Accuracy: 0.8044
Training loss (for one batch) at step 80: 287.7650, Accuracy: 0.8042
Training loss (for one batch) at step 90: 287.9579, Accuracy: 0.8033
Training loss (for one batch) at step 100: 266.3832, Accuracy: 0.8025
Training loss (for one batch) at step 110: 241.1022, Accuracy: 0.8051
Training loss (for one batch) at step 120: 272.0801, Accuracy: 0.8068
Training loss (for one batch) at step 130: 280.6678, Accuracy: 0.8050
Training loss (for one batch) at step 140: 274.8775, Accuracy: 0.8043
---- Training ----
Training loss: 227.7609
Training acc over epoch: 0.8047
---- Validation ----
Validation loss: 78.8511
Validation acc: 0.7270
Time taken: 40.34s
Start of epoch 22
Training loss (for one batch) at step 0: 272.7597, Accuracy: 0.7300
Training loss (for one batch) at step 10: 249.4550, Accuracy: 0.8273
Training loss (for one batch) at step 20: 248.4739, Accuracy: 0.8148
Training loss (for one batch) at step 30: 259.5974, Accuracy: 0.8148
Training loss (for one batch) at step 40: 240.2643, Accuracy: 0.8154
Training loss (for one batch) at step 50: 276.4696, Accuracy: 0.8178
Training loss (for one batch) at step 60: 273.6829, Accuracy: 0.8159
Training loss (for one batch) at step 70: 248.0798, Accuracy: 0.8159
Training loss (for one batch) at step 80: 243.1398, Accuracy: 0.8135
Training loss (for one batch) at step 90: 271.6026, Accuracy: 0.8103
Training loss (for one batch) at step 100: 261.9376, Accuracy: 0.8099
Training loss (for one batch) at step 110: 250.8972, Accuracy: 0.8095
Training loss (for one batch) at step 120: 244.6315, Accuracy: 0.8087
Training loss (for one batch) at step 130: 259.9177, Accuracy: 0.8085
Training loss (for one batch) at step 140: 263.2397, Accuracy: 0.8103
---- Training ----
Training loss: 223.7578
Training acc over epoch: 0.8090
---- Validation ----
Validation loss: 75.1391
Validation acc: 0.7023
Time taken: 68.94s
Start of epoch 23
Training loss (for one batch) at step 0: 257.3048, Accuracy: 0.8300
Training loss (for one batch) at step 10: 262.6836, Accuracy: 0.8009
Training loss (for one batch) at step 20: 241.6714, Accuracy: 0.8124
Training loss (for one batch) at step 30: 252.3996, Accuracy: 0.8194
Training loss (for one batch) at step 40: 258.4375, Accuracy: 0.8117
Training loss (for one batch) at step 50: 253.2057, Accuracy: 0.8208
Training loss (for one batch) at step 60: 244.6396, Accuracy: 0.8211
Training loss (for one batch) at step 70: 259.3844, Accuracy: 0.8200
Training loss (for one batch) at step 80: 263.0248, Accuracy: 0.8174
Training loss (for one batch) at step 90: 244.5002, Accuracy: 0.8164
Training loss (for one batch) at step 100: 242.4540, Accuracy: 0.8135
Training loss (for one batch) at step 110: 256.4812, Accuracy: 0.8144
Training loss (for one batch) at step 120: 247.5332, Accuracy: 0.8141
Training loss (for one batch) at step 130: 252.2279, Accuracy: 0.8145
Training loss (for one batch) at step 140: 250.0751, Accuracy: 0.8133
---- Training ----
Training loss: 253.3919
Training acc over epoch: 0.8123
---- Validation ----
Validation loss: 71.9518
Validation acc: 0.7303
Time taken: 39.49s
Start of epoch 24
Training loss (for one batch) at step 0: 245.2087, Accuracy: 0.8000
Training loss (for one batch) at step 10: 267.4346, Accuracy: 0.8318
Training loss (for one batch) at step 20: 251.3935, Accuracy: 0.8210
Training loss (for one batch) at step 30: 252.2466, Accuracy: 0.8106
Training loss (for one batch) at step 40: 252.9010, Accuracy: 0.8085
Training loss (for one batch) at step 50: 232.7100, Accuracy: 0.8114
Training loss (for one batch) at step 60: 268.0819, Accuracy: 0.8143
Training loss (for one batch) at step 70: 282.3296, Accuracy: 0.8128
Training loss (for one batch) at step 80: 252.0375, Accuracy: 0.8138
Training loss (for one batch) at step 90: 266.4183, Accuracy: 0.8127
Training loss (for one batch) at step 100: 244.3384, Accuracy: 0.8115
Training loss (for one batch) at step 110: 246.0055, Accuracy: 0.8122
Training loss (for one batch) at step 120: 268.5213, Accuracy: 0.8127
Training loss (for one batch) at step 130: 259.4031, Accuracy: 0.8114
Training loss (for one batch) at step 140: 257.0443, Accuracy: 0.8101
---- Training ----
Training loss: 226.5488
Training acc over epoch: 0.8102
---- Validation ----
Validation loss: 71.7860
Validation acc: 0.7324
Time taken: 67.82s
Start of epoch 25
Training loss (for one batch) at step 0: 262.8203, Accuracy: 0.8500
Training loss (for one batch) at step 10: 257.3925, Accuracy: 0.8218
Training loss (for one batch) at step 20: 257.8772, Accuracy: 0.8157
Training loss (for one batch) at step 30: 259.9486, Accuracy: 0.8145
Training loss (for one batch) at step 40: 244.5385, Accuracy: 0.8161
Training loss (for one batch) at step 50: 237.1455, Accuracy: 0.8239
Training loss (for one batch) at step 60: 248.1620, Accuracy: 0.8262
Training loss (for one batch) at step 70: 258.5505, Accuracy: 0.8214
Training loss (for one batch) at step 80: 250.6287, Accuracy: 0.8190
Training loss (for one batch) at step 90: 266.9042, Accuracy: 0.8195
Training loss (for one batch) at step 100: 227.0710, Accuracy: 0.8206
Training loss (for one batch) at step 110: 249.6210, Accuracy: 0.8199
Training loss (for one batch) at step 120: 249.4041, Accuracy: 0.8193
Training loss (for one batch) at step 130: 233.6432, Accuracy: 0.8187
Training loss (for one batch) at step 140: 240.1343, Accuracy: 0.8178
---- Training ----
Training loss: 238.7665
Training acc over epoch: 0.8173
---- Validation ----
Validation loss: 80.8421
Validation acc: 0.7243
Time taken: 39.57s
Start of epoch 26
Training loss (for one batch) at step 0: 263.9147, Accuracy: 0.7800
Training loss (for one batch) at step 10: 248.9690, Accuracy: 0.8227
Training loss (for one batch) at step 20: 236.7310, Accuracy: 0.8057
Training loss (for one batch) at step 30: 267.7122, Accuracy: 0.8135
Training loss (for one batch) at step 40: 242.1663, Accuracy: 0.8149
Training loss (for one batch) at step 50: 257.6321, Accuracy: 0.8188
Training loss (for one batch) at step 60: 262.6769, Accuracy: 0.8205
Training loss (for one batch) at step 70: 249.0263, Accuracy: 0.8177
Training loss (for one batch) at step 80: 245.6560, Accuracy: 0.8147
Training loss (for one batch) at step 90: 249.2642, Accuracy: 0.8142
Training loss (for one batch) at step 100: 250.1915, Accuracy: 0.8135
Training loss (for one batch) at step 110: 247.5913, Accuracy: 0.8159
Training loss (for one batch) at step 120: 243.0607, Accuracy: 0.8152
Training loss (for one batch) at step 130: 253.2670, Accuracy: 0.8141
Training loss (for one batch) at step 140: 248.5217, Accuracy: 0.8152
---- Training ----
Training loss: 230.1614
Training acc over epoch: 0.8143
---- Validation ----
Validation loss: 83.7253
Validation acc: 0.7337
Time taken: 67.79s
Start of epoch 27
Training loss (for one batch) at step 0: 242.9598, Accuracy: 0.8800
Training loss (for one batch) at step 10: 247.8664, Accuracy: 0.8182
Training loss (for one batch) at step 20: 266.8396, Accuracy: 0.8205
Training loss (for one batch) at step 30: 268.3075, Accuracy: 0.8152
Training loss (for one batch) at step 40: 238.9866, Accuracy: 0.8149
Training loss (for one batch) at step 50: 249.5738, Accuracy: 0.8175
Training loss (for one batch) at step 60: 250.1463, Accuracy: 0.8211
Training loss (for one batch) at step 70: 260.3709, Accuracy: 0.8185
Training loss (for one batch) at step 80: 241.2599, Accuracy: 0.8225
Training loss (for one batch) at step 90: 246.5519, Accuracy: 0.8192
Training loss (for one batch) at step 100: 223.7112, Accuracy: 0.8182
Training loss (for one batch) at step 110: 244.1425, Accuracy: 0.8186
Training loss (for one batch) at step 120: 250.3738, Accuracy: 0.8179
Training loss (for one batch) at step 130: 251.2357, Accuracy: 0.8176
Training loss (for one batch) at step 140: 230.5758, Accuracy: 0.8165
---- Training ----
Training loss: 214.3449
Training acc over epoch: 0.8163
---- Validation ----
Validation loss: 59.5852
Validation acc: 0.7241
Time taken: 39.51s
Start of epoch 28
Training loss (for one batch) at step 0: 237.0660, Accuracy: 0.8000
Training loss (for one batch) at step 10: 245.4707, Accuracy: 0.8327
Training loss (for one batch) at step 20: 242.7267, Accuracy: 0.8381
Training loss (for one batch) at step 30: 245.9972, Accuracy: 0.8281
Training loss (for one batch) at step 40: 245.9184, Accuracy: 0.8263
Training loss (for one batch) at step 50: 220.9484, Accuracy: 0.8255
Training loss (for one batch) at step 60: 248.2107, Accuracy: 0.8254
Training loss (for one batch) at step 70: 251.8611, Accuracy: 0.8246
Training loss (for one batch) at step 80: 239.7196, Accuracy: 0.8216
Training loss (for one batch) at step 90: 269.4475, Accuracy: 0.8216
Training loss (for one batch) at step 100: 256.1593, Accuracy: 0.8198
Training loss (for one batch) at step 110: 245.2618, Accuracy: 0.8204
Training loss (for one batch) at step 120: 248.5625, Accuracy: 0.8217
Training loss (for one batch) at step 130: 253.1875, Accuracy: 0.8224
Training loss (for one batch) at step 140: 240.0161, Accuracy: 0.8216
---- Training ----
Training loss: 225.6399
Training acc over epoch: 0.8217
---- Validation ----
Validation loss: 70.6175
Validation acc: 0.7225
Time taken: 68.55s
Start of epoch 29
Training loss (for one batch) at step 0: 240.8630, Accuracy: 0.8400
Training loss (for one batch) at step 10: 235.8821, Accuracy: 0.8427
Training loss (for one batch) at step 20: 231.8177, Accuracy: 0.8305
Training loss (for one batch) at step 30: 223.3994, Accuracy: 0.8290
Training loss (for one batch) at step 40: 241.3397, Accuracy: 0.8261
Training loss (for one batch) at step 50: 223.1221, Accuracy: 0.8288
Training loss (for one batch) at step 60: 265.8589, Accuracy: 0.8293
Training loss (for one batch) at step 70: 225.0162, Accuracy: 0.8311
Training loss (for one batch) at step 80: 240.6938, Accuracy: 0.8296
Training loss (for one batch) at step 90: 239.2405, Accuracy: 0.8287
Training loss (for one batch) at step 100: 239.6542, Accuracy: 0.8273
Training loss (for one batch) at step 110: 238.0170, Accuracy: 0.8279
Training loss (for one batch) at step 120: 248.8421, Accuracy: 0.8271
Training loss (for one batch) at step 130: 246.5535, Accuracy: 0.8254
Training loss (for one batch) at step 140: 250.7069, Accuracy: 0.8265
---- Training ----
Training loss: 219.0284
Training acc over epoch: 0.8263
---- Validation ----
Validation loss: 77.5206
Validation acc: 0.7286
Time taken: 41.22s
Start of epoch 30
Training loss (for one batch) at step 0: 226.5276, Accuracy: 0.8500
Training loss (for one batch) at step 10: 229.6468, Accuracy: 0.8355
Training loss (for one batch) at step 20: 238.3939, Accuracy: 0.8286
Training loss (for one batch) at step 30: 257.4417, Accuracy: 0.8235
Training loss (for one batch) at step 40: 228.6163, Accuracy: 0.8207
Training loss (for one batch) at step 50: 249.1708, Accuracy: 0.8251
Training loss (for one batch) at step 60: 250.3281, Accuracy: 0.8257
Training loss (for one batch) at step 70: 254.4079, Accuracy: 0.8270
Training loss (for one batch) at step 80: 227.8084, Accuracy: 0.8253
Training loss (for one batch) at step 90: 234.0721, Accuracy: 0.8254
Training loss (for one batch) at step 100: 238.3733, Accuracy: 0.8225
Training loss (for one batch) at step 110: 242.7036, Accuracy: 0.8243
Training loss (for one batch) at step 120: 238.0730, Accuracy: 0.8234
Training loss (for one batch) at step 130: 239.5786, Accuracy: 0.8220
Training loss (for one batch) at step 140: 224.7513, Accuracy: 0.8224
---- Training ----
Training loss: 215.7556
Training acc over epoch: 0.8233
---- Validation ----
Validation loss: 71.5662
Validation acc: 0.7300
Time taken: 70.37s
Start of epoch 31
Training loss (for one batch) at step 0: 235.6551, Accuracy: 0.8200
Training loss (for one batch) at step 10: 238.6401, Accuracy: 0.8509
Training loss (for one batch) at step 20: 216.5951, Accuracy: 0.8400
Training loss (for one batch) at step 30: 229.7914, Accuracy: 0.8371
Training loss (for one batch) at step 40: 225.5388, Accuracy: 0.8363
Training loss (for one batch) at step 50: 238.1542, Accuracy: 0.8382
Training loss (for one batch) at step 60: 243.2029, Accuracy: 0.8362
Training loss (for one batch) at step 70: 255.1176, Accuracy: 0.8334
Training loss (for one batch) at step 80: 244.8720, Accuracy: 0.8331
Training loss (for one batch) at step 90: 247.1379, Accuracy: 0.8302
Training loss (for one batch) at step 100: 213.0667, Accuracy: 0.8274
Training loss (for one batch) at step 110: 231.0137, Accuracy: 0.8275
Training loss (for one batch) at step 120: 244.9926, Accuracy: 0.8288
Training loss (for one batch) at step 130: 248.2804, Accuracy: 0.8283
Training loss (for one batch) at step 140: 245.8166, Accuracy: 0.8280
---- Training ----
Training loss: 229.2257
Training acc over epoch: 0.8274
---- Validation ----
Validation loss: 76.1638
Validation acc: 0.7104
Time taken: 39.44s
Start of epoch 32
Training loss (for one batch) at step 0: 236.0183, Accuracy: 0.8500
Training loss (for one batch) at step 10: 253.8174, Accuracy: 0.8264
Training loss (for one batch) at step 20: 229.4721, Accuracy: 0.8338
Training loss (for one batch) at step 30: 239.5814, Accuracy: 0.8306
Training loss (for one batch) at step 40: 212.8347, Accuracy: 0.8288
Training loss (for one batch) at step 50: 245.4743, Accuracy: 0.8341
Training loss (for one batch) at step 60: 238.6210, Accuracy: 0.8321
Training loss (for one batch) at step 70: 244.3634, Accuracy: 0.8310
Training loss (for one batch) at step 80: 238.9213, Accuracy: 0.8305
Training loss (for one batch) at step 90: 218.5621, Accuracy: 0.8297
Training loss (for one batch) at step 100: 221.9011, Accuracy: 0.8298
Training loss (for one batch) at step 110: 255.0074, Accuracy: 0.8310
Training loss (for one batch) at step 120: 246.5009, Accuracy: 0.8303
Training loss (for one batch) at step 130: 258.9625, Accuracy: 0.8285
Training loss (for one batch) at step 140: 219.8892, Accuracy: 0.8279
---- Training ----
Training loss: 217.1170
Training acc over epoch: 0.8283
---- Validation ----
Validation loss: 66.0604
Validation acc: 0.7254
Time taken: 68.67s
Start of epoch 33
Training loss (for one batch) at step 0: 239.2077, Accuracy: 0.8400
Training loss (for one batch) at step 10: 225.8753, Accuracy: 0.8264
Training loss (for one batch) at step 20: 232.7441, Accuracy: 0.8181
Training loss (for one batch) at step 30: 218.0241, Accuracy: 0.8242
Training loss (for one batch) at step 40: 201.7208, Accuracy: 0.8261
Training loss (for one batch) at step 50: 223.0511, Accuracy: 0.8324
Training loss (for one batch) at step 60: 223.9432, Accuracy: 0.8300
Training loss (for one batch) at step 70: 235.0882, Accuracy: 0.8293
Training loss (for one batch) at step 80: 223.1269, Accuracy: 0.8281
Training loss (for one batch) at step 90: 227.8282, Accuracy: 0.8263
Training loss (for one batch) at step 100: 229.0715, Accuracy: 0.8251
Training loss (for one batch) at step 110: 226.9343, Accuracy: 0.8261
Training loss (for one batch) at step 120: 247.7040, Accuracy: 0.8256
Training loss (for one batch) at step 130: 239.1024, Accuracy: 0.8251
Training loss (for one batch) at step 140: 217.4137, Accuracy: 0.8263
---- Training ----
Training loss: 225.1393
Training acc over epoch: 0.8262
---- Validation ----
Validation loss: 80.4782
Validation acc: 0.7211
Time taken: 39.64s
Start of epoch 34
Training loss (for one batch) at step 0: 222.8628, Accuracy: 0.8700
Training loss (for one batch) at step 10: 227.8701, Accuracy: 0.8355
Training loss (for one batch) at step 20: 253.4336, Accuracy: 0.8376
Training loss (for one batch) at step 30: 229.9568, Accuracy: 0.8468
Training loss (for one batch) at step 40: 233.7424, Accuracy: 0.8415
Training loss (for one batch) at step 50: 222.1446, Accuracy: 0.8382
Training loss (for one batch) at step 60: 233.4369, Accuracy: 0.8416
Training loss (for one batch) at step 70: 237.5229, Accuracy: 0.8400
Training loss (for one batch) at step 80: 232.7879, Accuracy: 0.8364
Training loss (for one batch) at step 90: 223.6380, Accuracy: 0.8348
Training loss (for one batch) at step 100: 241.2095, Accuracy: 0.8346
Training loss (for one batch) at step 110: 239.6607, Accuracy: 0.8337
Training loss (for one batch) at step 120: 253.2920, Accuracy: 0.8329
Training loss (for one batch) at step 130: 229.1096, Accuracy: 0.8324
Training loss (for one batch) at step 140: 244.7189, Accuracy: 0.8318
---- Training ----
Training loss: 222.2115
Training acc over epoch: 0.8325
---- Validation ----
Validation loss: 60.3246
Validation acc: 0.7254
Time taken: 67.82s
Start of epoch 35
Training loss (for one batch) at step 0: 229.1195, Accuracy: 0.8300
Training loss (for one batch) at step 10: 245.0390, Accuracy: 0.8373
Training loss (for one batch) at step 20: 225.3183, Accuracy: 0.8367
Training loss (for one batch) at step 30: 233.7484, Accuracy: 0.8361
Training loss (for one batch) at step 40: 224.6992, Accuracy: 0.8407
Training loss (for one batch) at step 50: 221.2496, Accuracy: 0.8443
Training loss (for one batch) at step 60: 232.3968, Accuracy: 0.8433
Training loss (for one batch) at step 70: 224.8880, Accuracy: 0.8437
Training loss (for one batch) at step 80: 239.7034, Accuracy: 0.8394
Training loss (for one batch) at step 90: 212.1902, Accuracy: 0.8373
Training loss (for one batch) at step 100: 221.4440, Accuracy: 0.8360
Training loss (for one batch) at step 110: 229.4701, Accuracy: 0.8356
Training loss (for one batch) at step 120: 213.8596, Accuracy: 0.8351
Training loss (for one batch) at step 130: 242.9181, Accuracy: 0.8354
Training loss (for one batch) at step 140: 207.2268, Accuracy: 0.8352
---- Training ----
Training loss: 225.0655
Training acc over epoch: 0.8346
---- Validation ----
Validation loss: 91.1490
Validation acc: 0.7260
Time taken: 41.42s
Start of epoch 36
Training loss (for one batch) at step 0: 235.7008, Accuracy: 0.8500
Training loss (for one batch) at step 10: 230.7741, Accuracy: 0.8345
Training loss (for one batch) at step 20: 223.9419, Accuracy: 0.8414
Training loss (for one batch) at step 30: 224.5347, Accuracy: 0.8406
Training loss (for one batch) at step 40: 230.9295, Accuracy: 0.8422
Training loss (for one batch) at step 50: 226.3443, Accuracy: 0.8398
Training loss (for one batch) at step 60: 229.7600, Accuracy: 0.8385
Training loss (for one batch) at step 70: 246.8531, Accuracy: 0.8344
Training loss (for one batch) at step 80: 224.3175, Accuracy: 0.8354
Training loss (for one batch) at step 90: 233.3235, Accuracy: 0.8324
Training loss (for one batch) at step 100: 212.0814, Accuracy: 0.8318
Training loss (for one batch) at step 110: 228.0109, Accuracy: 0.8311
Training loss (for one batch) at step 120: 238.6880, Accuracy: 0.8310
Training loss (for one batch) at step 130: 222.6603, Accuracy: 0.8317
Training loss (for one batch) at step 140: 225.4574, Accuracy: 0.8306
---- Training ----
Training loss: 197.5582
Training acc over epoch: 0.8309
---- Validation ----
Validation loss: 83.5725
Validation acc: 0.7082
Time taken: 69.35s
Start of epoch 37
Training loss (for one batch) at step 0: 221.4831, Accuracy: 0.8500
Training loss (for one batch) at step 10: 233.2910, Accuracy: 0.8336
Training loss (for one batch) at step 20: 217.4336, Accuracy: 0.8405
Training loss (for one batch) at step 30: 231.5486, Accuracy: 0.8294
Training loss (for one batch) at step 40: 227.9975, Accuracy: 0.8307
Training loss (for one batch) at step 50: 217.0598, Accuracy: 0.8343
Training loss (for one batch) at step 60: 224.3783, Accuracy: 0.8369
Training loss (for one batch) at step 70: 234.3911, Accuracy: 0.8352
Training loss (for one batch) at step 80: 228.3164, Accuracy: 0.8343
Training loss (for one batch) at step 90: 248.6913, Accuracy: 0.8331
Training loss (for one batch) at step 100: 211.7279, Accuracy: 0.8343
Training loss (for one batch) at step 110: 234.5641, Accuracy: 0.8332
Training loss (for one batch) at step 120: 205.3940, Accuracy: 0.8337
Training loss (for one batch) at step 130: 231.4433, Accuracy: 0.8318
Training loss (for one batch) at step 140: 216.6727, Accuracy: 0.8325
---- Training ----
Training loss: 226.7897
Training acc over epoch: 0.8328
---- Validation ----
Validation loss: 61.6835
Validation acc: 0.7300
Time taken: 39.96s
Start of epoch 38
Training loss (for one batch) at step 0: 232.8424, Accuracy: 0.7800
Training loss (for one batch) at step 10: 207.6014, Accuracy: 0.8473
Training loss (for one batch) at step 20: 240.2044, Accuracy: 0.8414
Training loss (for one batch) at step 30: 225.2753, Accuracy: 0.8384
Training loss (for one batch) at step 40: 208.2381, Accuracy: 0.8351
Training loss (for one batch) at step 50: 225.3697, Accuracy: 0.8347
Training loss (for one batch) at step 60: 224.4839, Accuracy: 0.8366
Training loss (for one batch) at step 70: 214.2183, Accuracy: 0.8385
Training loss (for one batch) at step 80: 235.1060, Accuracy: 0.8391
Training loss (for one batch) at step 90: 212.5414, Accuracy: 0.8359
Training loss (for one batch) at step 100: 212.0553, Accuracy: 0.8359
Training loss (for one batch) at step 110: 209.2366, Accuracy: 0.8375
Training loss (for one batch) at step 120: 214.8327, Accuracy: 0.8356
Training loss (for one batch) at step 130: 222.3927, Accuracy: 0.8356
Training loss (for one batch) at step 140: 232.0658, Accuracy: 0.8349
---- Training ----
Training loss: 212.5442
Training acc over epoch: 0.8352
---- Validation ----
Validation loss: 83.5780
Validation acc: 0.7311
Time taken: 66.64s
Start of epoch 39
Training loss (for one batch) at step 0: 242.4544, Accuracy: 0.7600
Training loss (for one batch) at step 10: 212.1332, Accuracy: 0.8427
Training loss (for one batch) at step 20: 221.3975, Accuracy: 0.8429
Training loss (for one batch) at step 30: 225.3596, Accuracy: 0.8390
Training loss (for one batch) at step 40: 215.4514, Accuracy: 0.8373
Training loss (for one batch) at step 50: 218.8609, Accuracy: 0.8380
Training loss (for one batch) at step 60: 196.1662, Accuracy: 0.8370
Training loss (for one batch) at step 70: 243.7510, Accuracy: 0.8365
Training loss (for one batch) at step 80: 216.1343, Accuracy: 0.8365
Training loss (for one batch) at step 90: 246.3656, Accuracy: 0.8371
Training loss (for one batch) at step 100: 218.1694, Accuracy: 0.8349
Training loss (for one batch) at step 110: 207.9781, Accuracy: 0.8377
Training loss (for one batch) at step 120: 232.6457, Accuracy: 0.8364
Training loss (for one batch) at step 130: 230.9736, Accuracy: 0.8355
Training loss (for one batch) at step 140: 244.2299, Accuracy: 0.8357
---- Training ----
Training loss: 195.0882
Training acc over epoch: 0.8361
---- Validation ----
Validation loss: 68.8239
Validation acc: 0.7190
Time taken: 38.96s
Start of epoch 40
Training loss (for one batch) at step 0: 251.9922, Accuracy: 0.7200
Training loss (for one batch) at step 10: 228.0255, Accuracy: 0.8255
Training loss (for one batch) at step 20: 221.6367, Accuracy: 0.8329
Training loss (for one batch) at step 30: 246.8746, Accuracy: 0.8384
Training loss (for one batch) at step 40: 211.6280, Accuracy: 0.8337
Training loss (for one batch) at step 50: 218.8036, Accuracy: 0.8367
Training loss (for one batch) at step 60: 231.2731, Accuracy: 0.8380
Training loss (for one batch) at step 70: 230.3606, Accuracy: 0.8359
Training loss (for one batch) at step 80: 219.1460, Accuracy: 0.8354
Training loss (for one batch) at step 90: 237.6003, Accuracy: 0.8336
Training loss (for one batch) at step 100: 204.3558, Accuracy: 0.8338
Training loss (for one batch) at step 110: 235.5454, Accuracy: 0.8347
Training loss (for one batch) at step 120: 233.1591, Accuracy: 0.8349
Training loss (for one batch) at step 130: 213.6061, Accuracy: 0.8353
Training loss (for one batch) at step 140: 214.1243, Accuracy: 0.8348
---- Training ----
Training loss: 199.8629
Training acc over epoch: 0.8352
---- Validation ----
Validation loss: 80.6648
Validation acc: 0.6977
Time taken: 66.30s
Start of epoch 41
Training loss (for one batch) at step 0: 239.5077, Accuracy: 0.8300
Training loss (for one batch) at step 10: 208.6214, Accuracy: 0.8527
Training loss (for one batch) at step 20: 222.0068, Accuracy: 0.8467
Training loss (for one batch) at step 30: 211.4985, Accuracy: 0.8348
Training loss (for one batch) at step 40: 208.8649, Accuracy: 0.8383
Training loss (for one batch) at step 50: 216.1010, Accuracy: 0.8400
Training loss (for one batch) at step 60: 222.7691, Accuracy: 0.8421
Training loss (for one batch) at step 70: 229.5697, Accuracy: 0.8428
Training loss (for one batch) at step 80: 224.2507, Accuracy: 0.8399
Training loss (for one batch) at step 90: 222.4402, Accuracy: 0.8389
Training loss (for one batch) at step 100: 219.7448, Accuracy: 0.8378
Training loss (for one batch) at step 110: 210.3742, Accuracy: 0.8389
Training loss (for one batch) at step 120: 213.9673, Accuracy: 0.8391
Training loss (for one batch) at step 130: 222.9016, Accuracy: 0.8378
Training loss (for one batch) at step 140: 231.4823, Accuracy: 0.8370
---- Training ----
Training loss: 194.4597
Training acc over epoch: 0.8366
---- Validation ----
Validation loss: 83.3201
Validation acc: 0.7195
Time taken: 38.77s
Start of epoch 42
Training loss (for one batch) at step 0: 227.1496, Accuracy: 0.7500
Training loss (for one batch) at step 10: 235.9235, Accuracy: 0.8145
Training loss (for one batch) at step 20: 211.7898, Accuracy: 0.8400
Training loss (for one batch) at step 30: 230.5993, Accuracy: 0.8452
Training loss (for one batch) at step 40: 187.0107, Accuracy: 0.8441
Training loss (for one batch) at step 50: 227.6437, Accuracy: 0.8457
Training loss (for one batch) at step 60: 228.1958, Accuracy: 0.8444
Training loss (for one batch) at step 70: 196.7518, Accuracy: 0.8441
Training loss (for one batch) at step 80: 211.7545, Accuracy: 0.8430
Training loss (for one batch) at step 90: 228.6042, Accuracy: 0.8400
Training loss (for one batch) at step 100: 203.0460, Accuracy: 0.8399
Training loss (for one batch) at step 110: 225.6506, Accuracy: 0.8409
Training loss (for one batch) at step 120: 208.2786, Accuracy: 0.8404
Training loss (for one batch) at step 130: 219.2914, Accuracy: 0.8384
Training loss (for one batch) at step 140: 213.9928, Accuracy: 0.8394
---- Training ----
Training loss: 216.3493
Training acc over epoch: 0.8398
---- Validation ----
Validation loss: 70.2902
Validation acc: 0.7211
Time taken: 65.73s
Start of epoch 43
Training loss (for one batch) at step 0: 204.2107, Accuracy: 0.8600
Training loss (for one batch) at step 10: 217.1594, Accuracy: 0.8327
Training loss (for one batch) at step 20: 217.0646, Accuracy: 0.8362
Training loss (for one batch) at step 30: 200.9380, Accuracy: 0.8355
Training loss (for one batch) at step 40: 215.5766, Accuracy: 0.8337
Training loss (for one batch) at step 50: 222.0875, Accuracy: 0.8343
Training loss (for one batch) at step 60: 219.3408, Accuracy: 0.8356
Training loss (for one batch) at step 70: 225.2516, Accuracy: 0.8362
Training loss (for one batch) at step 80: 215.6136, Accuracy: 0.8374
Training loss (for one batch) at step 90: 224.7710, Accuracy: 0.8367
Training loss (for one batch) at step 100: 211.5945, Accuracy: 0.8362
Training loss (for one batch) at step 110: 202.5294, Accuracy: 0.8380
Training loss (for one batch) at step 120: 213.5864, Accuracy: 0.8386
Training loss (for one batch) at step 130: 207.4355, Accuracy: 0.8374
Training loss (for one batch) at step 140: 236.6324, Accuracy: 0.8376
---- Training ----
Training loss: 207.1648
Training acc over epoch: 0.8389
---- Validation ----
Validation loss: 89.5880
Validation acc: 0.7233
Time taken: 38.74s
Start of epoch 44
Training loss (for one batch) at step 0: 227.7313, Accuracy: 0.8400
Training loss (for one batch) at step 10: 224.0572, Accuracy: 0.8427
Training loss (for one batch) at step 20: 238.5672, Accuracy: 0.8414
Training loss (for one batch) at step 30: 211.0170, Accuracy: 0.8397
Training loss (for one batch) at step 40: 197.7054, Accuracy: 0.8429
Training loss (for one batch) at step 50: 203.5685, Accuracy: 0.8441
Training loss (for one batch) at step 60: 214.0670, Accuracy: 0.8430
Training loss (for one batch) at step 70: 238.4037, Accuracy: 0.8408
Training loss (for one batch) at step 80: 203.5839, Accuracy: 0.8398
Training loss (for one batch) at step 90: 203.3805, Accuracy: 0.8392
Training loss (for one batch) at step 100: 209.8988, Accuracy: 0.8400
Training loss (for one batch) at step 110: 223.6774, Accuracy: 0.8405
Training loss (for one batch) at step 120: 196.0912, Accuracy: 0.8415
Training loss (for one batch) at step 130: 219.5452, Accuracy: 0.8411
Training loss (for one batch) at step 140: 230.2804, Accuracy: 0.8409
---- Training ----
Training loss: 186.0357
Training acc over epoch: 0.8404
---- Validation ----
Validation loss: 80.6919
Validation acc: 0.7227
Time taken: 64.74s
Start of epoch 45
Training loss (for one batch) at step 0: 238.4560, Accuracy: 0.8300
Training loss (for one batch) at step 10: 194.8253, Accuracy: 0.8318
Training loss (for one batch) at step 20: 219.9620, Accuracy: 0.8381
Training loss (for one batch) at step 30: 215.7497, Accuracy: 0.8368
Training loss (for one batch) at step 40: 212.7747, Accuracy: 0.8412
Training loss (for one batch) at step 50: 198.8821, Accuracy: 0.8445
Training loss (for one batch) at step 60: 236.3775, Accuracy: 0.8433
Training loss (for one batch) at step 70: 221.9170, Accuracy: 0.8428
Training loss (for one batch) at step 80: 213.0508, Accuracy: 0.8419
Training loss (for one batch) at step 90: 197.8893, Accuracy: 0.8401
Training loss (for one batch) at step 100: 236.7733, Accuracy: 0.8394
Training loss (for one batch) at step 110: 222.2791, Accuracy: 0.8410
Training loss (for one batch) at step 120: 218.5991, Accuracy: 0.8412
Training loss (for one batch) at step 130: 224.9725, Accuracy: 0.8408
Training loss (for one batch) at step 140: 215.9731, Accuracy: 0.8409
---- Training ----
Training loss: 191.4384
Training acc over epoch: 0.8406
---- Validation ----
Validation loss: 97.7488
Validation acc: 0.7093
Time taken: 39.65s
Start of epoch 46
Training loss (for one batch) at step 0: 208.7464, Accuracy: 0.8200
Training loss (for one batch) at step 10: 210.0813, Accuracy: 0.8527
Training loss (for one batch) at step 20: 221.9965, Accuracy: 0.8571
Training loss (for one batch) at step 30: 230.0991, Accuracy: 0.8552
Training loss (for one batch) at step 40: 204.7511, Accuracy: 0.8515
Training loss (for one batch) at step 50: 218.2825, Accuracy: 0.8502
Training loss (for one batch) at step 60: 200.8929, Accuracy: 0.8516
Training loss (for one batch) at step 70: 206.2544, Accuracy: 0.8499
Training loss (for one batch) at step 80: 225.0881, Accuracy: 0.8469
Training loss (for one batch) at step 90: 202.8847, Accuracy: 0.8449
Training loss (for one batch) at step 100: 211.7589, Accuracy: 0.8445
Training loss (for one batch) at step 110: 221.0608, Accuracy: 0.8442
Training loss (for one batch) at step 120: 200.0377, Accuracy: 0.8441
Training loss (for one batch) at step 130: 210.6541, Accuracy: 0.8440
Training loss (for one batch) at step 140: 210.0822, Accuracy: 0.8430
---- Training ----
Training loss: 194.9838
Training acc over epoch: 0.8432
---- Validation ----
Validation loss: 80.0137
Validation acc: 0.7327
Time taken: 69.23s
Start of epoch 47
Training loss (for one batch) at step 0: 197.1196, Accuracy: 0.8700
Training loss (for one batch) at step 10: 218.8996, Accuracy: 0.8527
Training loss (for one batch) at step 20: 198.7946, Accuracy: 0.8533
Training loss (for one batch) at step 30: 225.0060, Accuracy: 0.8516
Training loss (for one batch) at step 40: 194.3371, Accuracy: 0.8495
Training loss (for one batch) at step 50: 188.4386, Accuracy: 0.8522
Training loss (for one batch) at step 60: 206.9781, Accuracy: 0.8511
Training loss (for one batch) at step 70: 199.8394, Accuracy: 0.8500
Training loss (for one batch) at step 80: 217.7742, Accuracy: 0.8436
Training loss (for one batch) at step 90: 215.6571, Accuracy: 0.8431
Training loss (for one batch) at step 100: 210.4976, Accuracy: 0.8419
Training loss (for one batch) at step 110: 212.9773, Accuracy: 0.8405
Training loss (for one batch) at step 120: 220.7786, Accuracy: 0.8416
Training loss (for one batch) at step 130: 225.8599, Accuracy: 0.8400
Training loss (for one batch) at step 140: 213.2460, Accuracy: 0.8414
---- Training ----
Training loss: 186.6153
Training acc over epoch: 0.8409
---- Validation ----
Validation loss: 85.3479
Validation acc: 0.7276
Time taken: 39.92s
Start of epoch 48
Training loss (for one batch) at step 0: 203.1366, Accuracy: 0.9200
Training loss (for one batch) at step 10: 185.1559, Accuracy: 0.8600
Training loss (for one batch) at step 20: 218.7847, Accuracy: 0.8576
Training loss (for one batch) at step 30: 191.1471, Accuracy: 0.8539
Training loss (for one batch) at step 40: 215.8734, Accuracy: 0.8488
Training loss (for one batch) at step 50: 203.3598, Accuracy: 0.8496
Training loss (for one batch) at step 60: 206.9249, Accuracy: 0.8485
Training loss (for one batch) at step 70: 207.5836, Accuracy: 0.8492
Training loss (for one batch) at step 80: 206.9852, Accuracy: 0.8470
Training loss (for one batch) at step 90: 227.6345, Accuracy: 0.8455
Training loss (for one batch) at step 100: 225.8455, Accuracy: 0.8462
Training loss (for one batch) at step 110: 193.3656, Accuracy: 0.8459
Training loss (for one batch) at step 120: 212.8346, Accuracy: 0.8462
Training loss (for one batch) at step 130: 202.9728, Accuracy: 0.8460
Training loss (for one batch) at step 140: 237.0291, Accuracy: 0.8449
---- Training ----
Training loss: 190.6838
Training acc over epoch: 0.8448
---- Validation ----
Validation loss: 86.4745
Validation acc: 0.7128
Time taken: 67.08s
Start of epoch 49
Training loss (for one batch) at step 0: 212.2415, Accuracy: 0.8400
Training loss (for one batch) at step 10: 215.1803, Accuracy: 0.8555
Training loss (for one batch) at step 20: 222.5047, Accuracy: 0.8562
Training loss (for one batch) at step 30: 205.4164, Accuracy: 0.8471
Training loss (for one batch) at step 40: 195.6361, Accuracy: 0.8502
Training loss (for one batch) at step 50: 184.7876, Accuracy: 0.8504
Training loss (for one batch) at step 60: 211.4153, Accuracy: 0.8502
Training loss (for one batch) at step 70: 203.3895, Accuracy: 0.8490
Training loss (for one batch) at step 80: 208.5665, Accuracy: 0.8458
Training loss (for one batch) at step 90: 209.2171, Accuracy: 0.8455
Training loss (for one batch) at step 100: 169.9461, Accuracy: 0.8443
Training loss (for one batch) at step 110: 215.6253, Accuracy: 0.8439
Training loss (for one batch) at step 120: 209.8707, Accuracy: 0.8425
Training loss (for one batch) at step 130: 213.9637, Accuracy: 0.8440
Training loss (for one batch) at step 140: 206.6464, Accuracy: 0.8435
---- Training ----
Training loss: 184.4137
Training acc over epoch: 0.8425
---- Validation ----
Validation loss: 88.4317
Validation acc: 0.7343
Time taken: 38.73s
===== Q: 0.0001
Validation acc: 0.7451
Validation AUC: 0.7429
Validation Balanced_ACC: 0.4780
Validation MI: 0.1371
Validation Normalized MI: 0.2052
Validation Adjusted MI: 0.2052
Validation aUc_Sklearn: 0.8301
Start of epoch 0
Training loss (for one batch) at step 0: 508.5974, Accuracy: 0.4800
Training loss (for one batch) at step 10: 458.0078, Accuracy: 0.5473
Training loss (for one batch) at step 20: 448.5780, Accuracy: 0.5438
Training loss (for one batch) at step 30: 462.5913, Accuracy: 0.5481
Training loss (for one batch) at step 40: 442.1822, Accuracy: 0.5461
Training loss (for one batch) at step 50: 417.6051, Accuracy: 0.5561
Training loss (for one batch) at step 60: 399.7358, Accuracy: 0.5559
Training loss (for one batch) at step 70: 449.2058, Accuracy: 0.5603
Training loss (for one batch) at step 80: 450.9482, Accuracy: 0.5606
Training loss (for one batch) at step 90: 445.9540, Accuracy: 0.5627
Training loss (for one batch) at step 100: 433.6956, Accuracy: 0.5602
Training loss (for one batch) at step 110: 419.5416, Accuracy: 0.5618
Training loss (for one batch) at step 120: 427.3777, Accuracy: 0.5631
Training loss (for one batch) at step 130: 440.2532, Accuracy: 0.5645
Training loss (for one batch) at step 140: 413.0059, Accuracy: 0.5667
---- Training ----
Training loss: 346.2086
Training acc over epoch: 0.5678
---- Validation ----
Validation loss: 73.1438
Validation acc: 0.5134
Time taken: 68.20s
Start of epoch 1
Training loss (for one batch) at step 0: 404.1805, Accuracy: 0.5800
Training loss (for one batch) at step 10: 382.2370, Accuracy: 0.5891
Training loss (for one batch) at step 20: 377.0359, Accuracy: 0.5967
Training loss (for one batch) at step 30: 399.3176, Accuracy: 0.5913
Training loss (for one batch) at step 40: 370.0696, Accuracy: 0.5966
Training loss (for one batch) at step 50: 387.0901, Accuracy: 0.5861
Training loss (for one batch) at step 60: 380.5152, Accuracy: 0.5875
Training loss (for one batch) at step 70: 378.6922, Accuracy: 0.5872
Training loss (for one batch) at step 80: 393.3930, Accuracy: 0.5901
Training loss (for one batch) at step 90: 377.2006, Accuracy: 0.5922
Training loss (for one batch) at step 100: 396.8155, Accuracy: 0.5935
Training loss (for one batch) at step 110: 346.6873, Accuracy: 0.5940
Training loss (for one batch) at step 120: 394.1331, Accuracy: 0.5931
Training loss (for one batch) at step 130: 351.9615, Accuracy: 0.5918
Training loss (for one batch) at step 140: 377.4893, Accuracy: 0.5926
---- Training ----
Training loss: 320.4898
Training acc over epoch: 0.5939
---- Validation ----
Validation loss: 84.0435
Validation acc: 0.5247
Time taken: 51.47s
Start of epoch 2
Training loss (for one batch) at step 0: 399.1810, Accuracy: 0.5500
Training loss (for one batch) at step 10: 367.5387, Accuracy: 0.5918
Training loss (for one batch) at step 20: 351.4949, Accuracy: 0.6071
Training loss (for one batch) at step 30: 357.1950, Accuracy: 0.6071
Training loss (for one batch) at step 40: 360.3572, Accuracy: 0.6054
Training loss (for one batch) at step 50: 378.9421, Accuracy: 0.6018
Training loss (for one batch) at step 60: 360.5316, Accuracy: 0.6064
Training loss (for one batch) at step 70: 357.5767, Accuracy: 0.6035
Training loss (for one batch) at step 80: 360.2577, Accuracy: 0.6091
Training loss (for one batch) at step 90: 367.6597, Accuracy: 0.6087
Training loss (for one batch) at step 100: 338.8252, Accuracy: 0.6089
Training loss (for one batch) at step 110: 339.9271, Accuracy: 0.6092
Training loss (for one batch) at step 120: 346.8394, Accuracy: 0.6112
Training loss (for one batch) at step 130: 378.6564, Accuracy: 0.6139
Training loss (for one batch) at step 140: 381.2691, Accuracy: 0.6144
---- Training ----
Training loss: 308.0762
Training acc over epoch: 0.6147
---- Validation ----
Validation loss: 72.1242
Validation acc: 0.6609
Time taken: 65.82s
Start of epoch 3
Training loss (for one batch) at step 0: 360.2675, Accuracy: 0.5900
Training loss (for one batch) at step 10: 360.1122, Accuracy: 0.6245
Training loss (for one batch) at step 20: 355.1179, Accuracy: 0.6367
Training loss (for one batch) at step 30: 339.0730, Accuracy: 0.6265
Training loss (for one batch) at step 40: 347.0356, Accuracy: 0.6244
Training loss (for one batch) at step 50: 345.8120, Accuracy: 0.6263
Training loss (for one batch) at step 60: 352.2596, Accuracy: 0.6254
Training loss (for one batch) at step 70: 364.5292, Accuracy: 0.6215
Training loss (for one batch) at step 80: 348.7098, Accuracy: 0.6248
Training loss (for one batch) at step 90: 359.7964, Accuracy: 0.6270
Training loss (for one batch) at step 100: 346.9525, Accuracy: 0.6274
Training loss (for one batch) at step 110: 335.0567, Accuracy: 0.6300
Training loss (for one batch) at step 120: 345.9951, Accuracy: 0.6304
Training loss (for one batch) at step 130: 349.5189, Accuracy: 0.6303
Training loss (for one batch) at step 140: 336.6990, Accuracy: 0.6325
---- Training ----
Training loss: 306.4244
Training acc over epoch: 0.6320
---- Validation ----
Validation loss: 68.0483
Validation acc: 0.6738
Time taken: 37.45s
Start of epoch 4
Training loss (for one batch) at step 0: 332.3388, Accuracy: 0.7200
Training loss (for one batch) at step 10: 341.0671, Accuracy: 0.6691
Training loss (for one batch) at step 20: 353.7315, Accuracy: 0.6486
Training loss (for one batch) at step 30: 348.4402, Accuracy: 0.6526
Training loss (for one batch) at step 40: 325.0941, Accuracy: 0.6527
Training loss (for one batch) at step 50: 356.5078, Accuracy: 0.6545
Training loss (for one batch) at step 60: 325.4660, Accuracy: 0.6539
Training loss (for one batch) at step 70: 335.8635, Accuracy: 0.6528
Training loss (for one batch) at step 80: 345.9576, Accuracy: 0.6512
Training loss (for one batch) at step 90: 363.9437, Accuracy: 0.6523
Training loss (for one batch) at step 100: 330.3947, Accuracy: 0.6522
Training loss (for one batch) at step 110: 338.1832, Accuracy: 0.6521
Training loss (for one batch) at step 120: 339.5884, Accuracy: 0.6521
Training loss (for one batch) at step 130: 332.8400, Accuracy: 0.6550
Training loss (for one batch) at step 140: 336.9701, Accuracy: 0.6530
---- Training ----
Training loss: 284.4721
Training acc over epoch: 0.6543
---- Validation ----
Validation loss: 73.5025
Validation acc: 0.6754
Time taken: 65.55s
Start of epoch 5
Training loss (for one batch) at step 0: 333.8199, Accuracy: 0.7100
Training loss (for one batch) at step 10: 346.2613, Accuracy: 0.6855
Training loss (for one batch) at step 20: 334.9705, Accuracy: 0.6852
Training loss (for one batch) at step 30: 347.4201, Accuracy: 0.6774
Training loss (for one batch) at step 40: 330.2390, Accuracy: 0.6724
Training loss (for one batch) at step 50: 328.3431, Accuracy: 0.6733
Training loss (for one batch) at step 60: 338.9945, Accuracy: 0.6746
Training loss (for one batch) at step 70: 322.7267, Accuracy: 0.6735
Training loss (for one batch) at step 80: 350.3091, Accuracy: 0.6752
Training loss (for one batch) at step 90: 330.6779, Accuracy: 0.6752
Training loss (for one batch) at step 100: 338.9464, Accuracy: 0.6739
Training loss (for one batch) at step 110: 320.3706, Accuracy: 0.6739
Training loss (for one batch) at step 120: 328.9919, Accuracy: 0.6736
Training loss (for one batch) at step 130: 363.3353, Accuracy: 0.6741
Training loss (for one batch) at step 140: 327.6748, Accuracy: 0.6749
---- Training ----
Training loss: 305.5929
Training acc over epoch: 0.6756
---- Validation ----
Validation loss: 67.2395
Validation acc: 0.6703
Time taken: 37.26s
Start of epoch 6
Training loss (for one batch) at step 0: 343.1735, Accuracy: 0.6600
Training loss (for one batch) at step 10: 325.7560, Accuracy: 0.6691
Training loss (for one batch) at step 20: 335.8406, Accuracy: 0.6729
Training loss (for one batch) at step 30: 333.3301, Accuracy: 0.6800
Training loss (for one batch) at step 40: 343.4000, Accuracy: 0.6788
Training loss (for one batch) at step 50: 315.4317, Accuracy: 0.6818
Training loss (for one batch) at step 60: 324.9974, Accuracy: 0.6857
Training loss (for one batch) at step 70: 327.0681, Accuracy: 0.6870
Training loss (for one batch) at step 80: 324.9272, Accuracy: 0.6863
Training loss (for one batch) at step 90: 323.9642, Accuracy: 0.6856
Training loss (for one batch) at step 100: 320.3303, Accuracy: 0.6866
Training loss (for one batch) at step 110: 314.5504, Accuracy: 0.6882
Training loss (for one batch) at step 120: 325.6480, Accuracy: 0.6890
Training loss (for one batch) at step 130: 318.8661, Accuracy: 0.6894
Training loss (for one batch) at step 140: 342.5833, Accuracy: 0.6894
---- Training ----
Training loss: 299.2988
Training acc over epoch: 0.6893
---- Validation ----
Validation loss: 76.6859
Validation acc: 0.6875
Time taken: 63.98s
Start of epoch 7
Training loss (for one batch) at step 0: 320.1844, Accuracy: 0.6900
Training loss (for one batch) at step 10: 314.7481, Accuracy: 0.6955
Training loss (for one batch) at step 20: 325.0437, Accuracy: 0.6952
Training loss (for one batch) at step 30: 324.3101, Accuracy: 0.6974
Training loss (for one batch) at step 40: 326.5236, Accuracy: 0.7000
Training loss (for one batch) at step 50: 323.1355, Accuracy: 0.6990
Training loss (for one batch) at step 60: 317.6761, Accuracy: 0.7016
Training loss (for one batch) at step 70: 304.0023, Accuracy: 0.6990
Training loss (for one batch) at step 80: 330.1042, Accuracy: 0.6995
Training loss (for one batch) at step 90: 308.8157, Accuracy: 0.6991
Training loss (for one batch) at step 100: 334.6648, Accuracy: 0.6991
Training loss (for one batch) at step 110: 301.3532, Accuracy: 0.7010
Training loss (for one batch) at step 120: 311.9257, Accuracy: 0.7020
Training loss (for one batch) at step 130: 313.4055, Accuracy: 0.7019
Training loss (for one batch) at step 140: 319.5402, Accuracy: 0.7030
---- Training ----
Training loss: 281.0635
Training acc over epoch: 0.7038
---- Validation ----
Validation loss: 70.9094
Validation acc: 0.7123
Time taken: 37.31s
Start of epoch 8
Training loss (for one batch) at step 0: 309.1336, Accuracy: 0.7100
Training loss (for one batch) at step 10: 319.8992, Accuracy: 0.7127
Training loss (for one batch) at step 20: 315.8604, Accuracy: 0.7157
Training loss (for one batch) at step 30: 322.7823, Accuracy: 0.7155
Training loss (for one batch) at step 40: 314.8830, Accuracy: 0.7110
Training loss (for one batch) at step 50: 316.9138, Accuracy: 0.7159
Training loss (for one batch) at step 60: 315.5333, Accuracy: 0.7187
Training loss (for one batch) at step 70: 312.9787, Accuracy: 0.7200
Training loss (for one batch) at step 80: 309.6868, Accuracy: 0.7184
Training loss (for one batch) at step 90: 329.1476, Accuracy: 0.7178
Training loss (for one batch) at step 100: 315.8623, Accuracy: 0.7177
Training loss (for one batch) at step 110: 322.8610, Accuracy: 0.7168
Training loss (for one batch) at step 120: 331.7591, Accuracy: 0.7166
Training loss (for one batch) at step 130: 312.1237, Accuracy: 0.7166
Training loss (for one batch) at step 140: 334.0428, Accuracy: 0.7172
---- Training ----
Training loss: 275.5959
Training acc over epoch: 0.7163
---- Validation ----
Validation loss: 74.7744
Validation acc: 0.7074
Time taken: 63.31s
Start of epoch 9
Training loss (for one batch) at step 0: 314.8957, Accuracy: 0.7100
Training loss (for one batch) at step 10: 295.9340, Accuracy: 0.7345
Training loss (for one batch) at step 20: 312.0419, Accuracy: 0.7310
Training loss (for one batch) at step 30: 312.0032, Accuracy: 0.7248
Training loss (for one batch) at step 40: 306.9714, Accuracy: 0.7315
Training loss (for one batch) at step 50: 317.9958, Accuracy: 0.7306
Training loss (for one batch) at step 60: 317.4953, Accuracy: 0.7315
Training loss (for one batch) at step 70: 305.0045, Accuracy: 0.7325
Training loss (for one batch) at step 80: 314.2765, Accuracy: 0.7307
Training loss (for one batch) at step 90: 307.3655, Accuracy: 0.7315
Training loss (for one batch) at step 100: 307.3610, Accuracy: 0.7306
Training loss (for one batch) at step 110: 313.3464, Accuracy: 0.7321
Training loss (for one batch) at step 120: 308.6949, Accuracy: 0.7330
Training loss (for one batch) at step 130: 303.7930, Accuracy: 0.7324
Training loss (for one batch) at step 140: 296.3772, Accuracy: 0.7326
---- Training ----
Training loss: 278.9771
Training acc over epoch: 0.7319
---- Validation ----
Validation loss: 63.7425
Validation acc: 0.7131
Time taken: 37.63s
Start of epoch 10
Training loss (for one batch) at step 0: 306.8549, Accuracy: 0.7800
Training loss (for one batch) at step 10: 303.9464, Accuracy: 0.7636
Training loss (for one batch) at step 20: 312.3066, Accuracy: 0.7567
Training loss (for one batch) at step 30: 312.0124, Accuracy: 0.7506
Training loss (for one batch) at step 40: 305.7898, Accuracy: 0.7541
Training loss (for one batch) at step 50: 307.4559, Accuracy: 0.7531
Training loss (for one batch) at step 60: 300.6809, Accuracy: 0.7508
Training loss (for one batch) at step 70: 294.0911, Accuracy: 0.7469
Training loss (for one batch) at step 80: 319.8887, Accuracy: 0.7448
Training loss (for one batch) at step 90: 321.5705, Accuracy: 0.7438
Training loss (for one batch) at step 100: 333.7896, Accuracy: 0.7413
Training loss (for one batch) at step 110: 326.5966, Accuracy: 0.7395
Training loss (for one batch) at step 120: 315.4378, Accuracy: 0.7394
Training loss (for one batch) at step 130: 300.6674, Accuracy: 0.7397
Training loss (for one batch) at step 140: 306.2756, Accuracy: 0.7399
---- Training ----
Training loss: 275.2092
Training acc over epoch: 0.7410
---- Validation ----
Validation loss: 68.4777
Validation acc: 0.7171
Time taken: 63.21s
Start of epoch 11
Training loss (for one batch) at step 0: 284.5240, Accuracy: 0.8200
Training loss (for one batch) at step 10: 305.3553, Accuracy: 0.7564
Training loss (for one batch) at step 20: 295.9024, Accuracy: 0.7519
Training loss (for one batch) at step 30: 291.3435, Accuracy: 0.7535
Training loss (for one batch) at step 40: 299.9389, Accuracy: 0.7563
Training loss (for one batch) at step 50: 311.8002, Accuracy: 0.7555
Training loss (for one batch) at step 60: 316.4650, Accuracy: 0.7539
Training loss (for one batch) at step 70: 299.7022, Accuracy: 0.7535
Training loss (for one batch) at step 80: 312.1980, Accuracy: 0.7536
Training loss (for one batch) at step 90: 321.8228, Accuracy: 0.7535
Training loss (for one batch) at step 100: 287.9266, Accuracy: 0.7531
Training loss (for one batch) at step 110: 302.3047, Accuracy: 0.7522
Training loss (for one batch) at step 120: 281.2987, Accuracy: 0.7540
Training loss (for one batch) at step 130: 288.7980, Accuracy: 0.7527
Training loss (for one batch) at step 140: 308.8012, Accuracy: 0.7511
---- Training ----
Training loss: 275.6327
Training acc over epoch: 0.7509
---- Validation ----
Validation loss: 71.5102
Validation acc: 0.7200
Time taken: 37.20s
Start of epoch 12
Training loss (for one batch) at step 0: 297.7991, Accuracy: 0.8000
Training loss (for one batch) at step 10: 289.0128, Accuracy: 0.7636
Training loss (for one batch) at step 20: 288.5938, Accuracy: 0.7633
Training loss (for one batch) at step 30: 305.1466, Accuracy: 0.7590
Training loss (for one batch) at step 40: 298.4105, Accuracy: 0.7600
Training loss (for one batch) at step 50: 304.1230, Accuracy: 0.7596
Training loss (for one batch) at step 60: 284.1893, Accuracy: 0.7620
Training loss (for one batch) at step 70: 287.9382, Accuracy: 0.7592
Training loss (for one batch) at step 80: 317.1934, Accuracy: 0.7574
Training loss (for one batch) at step 90: 298.5917, Accuracy: 0.7569
Training loss (for one batch) at step 100: 311.7328, Accuracy: 0.7562
Training loss (for one batch) at step 110: 283.5586, Accuracy: 0.7565
Training loss (for one batch) at step 120: 305.9326, Accuracy: 0.7548
Training loss (for one batch) at step 130: 321.1220, Accuracy: 0.7557
Training loss (for one batch) at step 140: 313.6051, Accuracy: 0.7563
---- Training ----
Training loss: 253.2909
Training acc over epoch: 0.7566
---- Validation ----
Validation loss: 69.5159
Validation acc: 0.7268
Time taken: 63.72s
Start of epoch 13
Training loss (for one batch) at step 0: 316.8684, Accuracy: 0.7500
Training loss (for one batch) at step 10: 292.1217, Accuracy: 0.7709
Training loss (for one batch) at step 20: 307.6197, Accuracy: 0.7752
Training loss (for one batch) at step 30: 305.2005, Accuracy: 0.7681
Training loss (for one batch) at step 40: 300.3006, Accuracy: 0.7705
Training loss (for one batch) at step 50: 290.2699, Accuracy: 0.7741
Training loss (for one batch) at step 60: 286.2824, Accuracy: 0.7734
Training loss (for one batch) at step 70: 301.1277, Accuracy: 0.7720
Training loss (for one batch) at step 80: 312.2248, Accuracy: 0.7710
Training loss (for one batch) at step 90: 315.1287, Accuracy: 0.7704
Training loss (for one batch) at step 100: 290.1288, Accuracy: 0.7692
Training loss (for one batch) at step 110: 292.6099, Accuracy: 0.7689
Training loss (for one batch) at step 120: 301.6969, Accuracy: 0.7707
Training loss (for one batch) at step 130: 290.4244, Accuracy: 0.7697
Training loss (for one batch) at step 140: 281.0597, Accuracy: 0.7681
---- Training ----
Training loss: 260.8023
Training acc over epoch: 0.7683
---- Validation ----
Validation loss: 79.8110
Validation acc: 0.7286
Time taken: 37.25s
Start of epoch 14
Training loss (for one batch) at step 0: 287.8112, Accuracy: 0.8000
Training loss (for one batch) at step 10: 288.2426, Accuracy: 0.8091
Training loss (for one batch) at step 20: 292.5754, Accuracy: 0.7829
Training loss (for one batch) at step 30: 312.2665, Accuracy: 0.7810
Training loss (for one batch) at step 40: 308.0047, Accuracy: 0.7763
Training loss (for one batch) at step 50: 287.4537, Accuracy: 0.7731
Training loss (for one batch) at step 60: 281.3705, Accuracy: 0.7764
Training loss (for one batch) at step 70: 299.7886, Accuracy: 0.7751
Training loss (for one batch) at step 80: 308.2660, Accuracy: 0.7733
Training loss (for one batch) at step 90: 286.6970, Accuracy: 0.7710
Training loss (for one batch) at step 100: 284.2503, Accuracy: 0.7702
Training loss (for one batch) at step 110: 290.3051, Accuracy: 0.7722
Training loss (for one batch) at step 120: 303.9724, Accuracy: 0.7717
Training loss (for one batch) at step 130: 276.2774, Accuracy: 0.7712
Training loss (for one batch) at step 140: 291.7871, Accuracy: 0.7713
---- Training ----
Training loss: 261.9626
Training acc over epoch: 0.7700
---- Validation ----
Validation loss: 71.8812
Validation acc: 0.7289
Time taken: 65.26s
Start of epoch 15
Training loss (for one batch) at step 0: 287.7042, Accuracy: 0.7500
Training loss (for one batch) at step 10: 283.0758, Accuracy: 0.7809
Training loss (for one batch) at step 20: 303.7711, Accuracy: 0.7781
Training loss (for one batch) at step 30: 296.5357, Accuracy: 0.7800
Training loss (for one batch) at step 40: 276.4526, Accuracy: 0.7793
Training loss (for one batch) at step 50: 285.5413, Accuracy: 0.7757
Training loss (for one batch) at step 60: 293.6508, Accuracy: 0.7780
Training loss (for one batch) at step 70: 277.4861, Accuracy: 0.7813
Training loss (for one batch) at step 80: 295.0909, Accuracy: 0.7786
Training loss (for one batch) at step 90: 277.5001, Accuracy: 0.7771
Training loss (for one batch) at step 100: 279.5539, Accuracy: 0.7763
Training loss (for one batch) at step 110: 290.8910, Accuracy: 0.7779
Training loss (for one batch) at step 120: 297.6659, Accuracy: 0.7778
Training loss (for one batch) at step 130: 287.6613, Accuracy: 0.7775
Training loss (for one batch) at step 140: 307.6964, Accuracy: 0.7779
---- Training ----
Training loss: 254.6772
Training acc over epoch: 0.7772
---- Validation ----
Validation loss: 69.2148
Validation acc: 0.7203
Time taken: 38.95s
Start of epoch 16
Training loss (for one batch) at step 0: 294.9160, Accuracy: 0.7100
Training loss (for one batch) at step 10: 272.7084, Accuracy: 0.7809
Training loss (for one batch) at step 20: 275.8794, Accuracy: 0.7914
Training loss (for one batch) at step 30: 296.7771, Accuracy: 0.7919
Training loss (for one batch) at step 40: 280.2924, Accuracy: 0.7915
Training loss (for one batch) at step 50: 306.4625, Accuracy: 0.7876
Training loss (for one batch) at step 60: 311.4976, Accuracy: 0.7898
Training loss (for one batch) at step 70: 269.2623, Accuracy: 0.7904
Training loss (for one batch) at step 80: 296.5634, Accuracy: 0.7878
Training loss (for one batch) at step 90: 306.0323, Accuracy: 0.7892
Training loss (for one batch) at step 100: 305.9054, Accuracy: 0.7853
Training loss (for one batch) at step 110: 292.7782, Accuracy: 0.7850
Training loss (for one batch) at step 120: 281.2682, Accuracy: 0.7859
Training loss (for one batch) at step 130: 291.4432, Accuracy: 0.7859
Training loss (for one batch) at step 140: 279.6985, Accuracy: 0.7847
---- Training ----
Training loss: 259.8427
Training acc over epoch: 0.7846
---- Validation ----
Validation loss: 67.0372
Validation acc: 0.7348
Time taken: 72.62s
Start of epoch 17
Training loss (for one batch) at step 0: 281.1983, Accuracy: 0.8200
Training loss (for one batch) at step 10: 272.8742, Accuracy: 0.8091
Training loss (for one batch) at step 20: 268.1824, Accuracy: 0.8038
Training loss (for one batch) at step 30: 278.9404, Accuracy: 0.7974
Training loss (for one batch) at step 40: 275.6928, Accuracy: 0.7934
Training loss (for one batch) at step 50: 279.3399, Accuracy: 0.7965
Training loss (for one batch) at step 60: 280.3119, Accuracy: 0.7967
Training loss (for one batch) at step 70: 269.3581, Accuracy: 0.7946
Training loss (for one batch) at step 80: 288.1784, Accuracy: 0.7915
Training loss (for one batch) at step 90: 269.9775, Accuracy: 0.7908
Training loss (for one batch) at step 100: 281.1439, Accuracy: 0.7895
Training loss (for one batch) at step 110: 269.7388, Accuracy: 0.7923
Training loss (for one batch) at step 120: 291.3633, Accuracy: 0.7908
Training loss (for one batch) at step 130: 271.1154, Accuracy: 0.7906
Training loss (for one batch) at step 140: 268.1231, Accuracy: 0.7883
---- Training ----
Training loss: 253.1129
Training acc over epoch: 0.7889
---- Validation ----
Validation loss: 66.7540
Validation acc: 0.7273
Time taken: 37.52s
Start of epoch 18
Training loss (for one batch) at step 0: 274.9939, Accuracy: 0.7900
Training loss (for one batch) at step 10: 266.5839, Accuracy: 0.8145
Training loss (for one batch) at step 20: 264.9141, Accuracy: 0.8010
Training loss (for one batch) at step 30: 275.9818, Accuracy: 0.8006
Training loss (for one batch) at step 40: 265.4191, Accuracy: 0.7973
Training loss (for one batch) at step 50: 277.9290, Accuracy: 0.7976
Training loss (for one batch) at step 60: 288.5861, Accuracy: 0.7954
Training loss (for one batch) at step 70: 290.2316, Accuracy: 0.7930
Training loss (for one batch) at step 80: 277.1662, Accuracy: 0.7893
Training loss (for one batch) at step 90: 274.7190, Accuracy: 0.7907
Training loss (for one batch) at step 100: 283.0499, Accuracy: 0.7920
Training loss (for one batch) at step 110: 268.4845, Accuracy: 0.7919
Training loss (for one batch) at step 120: 286.5542, Accuracy: 0.7925
Training loss (for one batch) at step 130: 268.5126, Accuracy: 0.7935
Training loss (for one batch) at step 140: 284.2941, Accuracy: 0.7923
---- Training ----
Training loss: 263.3982
Training acc over epoch: 0.7914
---- Validation ----
Validation loss: 73.8680
Validation acc: 0.7090
Time taken: 97.48s
Start of epoch 19
Training loss (for one batch) at step 0: 272.6887, Accuracy: 0.7800
Training loss (for one batch) at step 10: 283.7895, Accuracy: 0.7845
Training loss (for one batch) at step 20: 283.6638, Accuracy: 0.8014
Training loss (for one batch) at step 30: 286.4651, Accuracy: 0.8010
Training loss (for one batch) at step 40: 268.7259, Accuracy: 0.8012
Training loss (for one batch) at step 50: 288.4514, Accuracy: 0.8016
Training loss (for one batch) at step 60: 281.6421, Accuracy: 0.8010
Training loss (for one batch) at step 70: 275.3656, Accuracy: 0.7986
Training loss (for one batch) at step 80: 260.4134, Accuracy: 0.7993
Training loss (for one batch) at step 90: 283.1617, Accuracy: 0.7976
Training loss (for one batch) at step 100: 267.3992, Accuracy: 0.7985
Training loss (for one batch) at step 110: 265.6535, Accuracy: 0.7988
Training loss (for one batch) at step 120: 267.9982, Accuracy: 0.7996
Training loss (for one batch) at step 130: 274.6024, Accuracy: 0.7982
Training loss (for one batch) at step 140: 274.4560, Accuracy: 0.7979
---- Training ----
Training loss: 250.8210
Training acc over epoch: 0.7986
---- Validation ----
Validation loss: 77.5652
Validation acc: 0.7294
Time taken: 38.54s
Start of epoch 20
Training loss (for one batch) at step 0: 285.4881, Accuracy: 0.8200
Training loss (for one batch) at step 10: 273.3094, Accuracy: 0.8100
Training loss (for one batch) at step 20: 265.6511, Accuracy: 0.8133
Training loss (for one batch) at step 30: 274.4017, Accuracy: 0.8065
Training loss (for one batch) at step 40: 267.3605, Accuracy: 0.8073
Training loss (for one batch) at step 50: 250.6604, Accuracy: 0.8086
Training loss (for one batch) at step 60: 264.2548, Accuracy: 0.8098
Training loss (for one batch) at step 70: 292.1863, Accuracy: 0.8117
Training loss (for one batch) at step 80: 285.2086, Accuracy: 0.8069
Training loss (for one batch) at step 90: 287.3350, Accuracy: 0.8069
Training loss (for one batch) at step 100: 279.5231, Accuracy: 0.8064
Training loss (for one batch) at step 110: 265.5845, Accuracy: 0.8086
Training loss (for one batch) at step 120: 283.3965, Accuracy: 0.8076
Training loss (for one batch) at step 130: 261.0379, Accuracy: 0.8065
Training loss (for one batch) at step 140: 278.5376, Accuracy: 0.8063
---- Training ----
Training loss: 237.2046
Training acc over epoch: 0.8049
---- Validation ----
Validation loss: 60.6623
Validation acc: 0.7292
Time taken: 65.08s
Start of epoch 21
Training loss (for one batch) at step 0: 266.2986, Accuracy: 0.8300
Training loss (for one batch) at step 10: 262.8398, Accuracy: 0.7964
Training loss (for one batch) at step 20: 263.0704, Accuracy: 0.8019
Training loss (for one batch) at step 30: 281.2963, Accuracy: 0.8026
Training loss (for one batch) at step 40: 265.4263, Accuracy: 0.8061
Training loss (for one batch) at step 50: 257.4875, Accuracy: 0.8110
Training loss (for one batch) at step 60: 259.3760, Accuracy: 0.8110
Training loss (for one batch) at step 70: 276.1196, Accuracy: 0.8083
Training loss (for one batch) at step 80: 260.8427, Accuracy: 0.8074
Training loss (for one batch) at step 90: 267.3523, Accuracy: 0.8073
Training loss (for one batch) at step 100: 274.2722, Accuracy: 0.8073
Training loss (for one batch) at step 110: 260.6438, Accuracy: 0.8075
Training loss (for one batch) at step 120: 278.0892, Accuracy: 0.8073
Training loss (for one batch) at step 130: 272.5726, Accuracy: 0.8069
Training loss (for one batch) at step 140: 252.9618, Accuracy: 0.8077
---- Training ----
Training loss: 233.6352
Training acc over epoch: 0.8068
---- Validation ----
Validation loss: 60.8059
Validation acc: 0.7303
Time taken: 38.91s
Start of epoch 22
Training loss (for one batch) at step 0: 268.2427, Accuracy: 0.7800
Training loss (for one batch) at step 10: 253.6900, Accuracy: 0.8082
Training loss (for one batch) at step 20: 273.1879, Accuracy: 0.8086
Training loss (for one batch) at step 30: 267.2325, Accuracy: 0.8042
Training loss (for one batch) at step 40: 272.8430, Accuracy: 0.8032
Training loss (for one batch) at step 50: 253.9477, Accuracy: 0.8082
Training loss (for one batch) at step 60: 273.7468, Accuracy: 0.8098
Training loss (for one batch) at step 70: 264.7573, Accuracy: 0.8114
Training loss (for one batch) at step 80: 258.3544, Accuracy: 0.8130
Training loss (for one batch) at step 90: 261.5009, Accuracy: 0.8126
Training loss (for one batch) at step 100: 260.4083, Accuracy: 0.8108
Training loss (for one batch) at step 110: 279.3535, Accuracy: 0.8109
Training loss (for one batch) at step 120: 288.3007, Accuracy: 0.8097
Training loss (for one batch) at step 130: 259.0465, Accuracy: 0.8102
Training loss (for one batch) at step 140: 269.2532, Accuracy: 0.8095
---- Training ----
Training loss: 230.9660
Training acc over epoch: 0.8091
---- Validation ----
Validation loss: 59.1527
Validation acc: 0.7270
Time taken: 65.09s
Start of epoch 23
Training loss (for one batch) at step 0: 248.9862, Accuracy: 0.7900
Training loss (for one batch) at step 10: 255.4626, Accuracy: 0.8127
Training loss (for one batch) at step 20: 258.1023, Accuracy: 0.8138
Training loss (for one batch) at step 30: 245.6452, Accuracy: 0.8219
Training loss (for one batch) at step 40: 245.3950, Accuracy: 0.8227
Training loss (for one batch) at step 50: 259.8107, Accuracy: 0.8243
Training loss (for one batch) at step 60: 251.9704, Accuracy: 0.8218
Training loss (for one batch) at step 70: 251.1315, Accuracy: 0.8187
Training loss (for one batch) at step 80: 265.2372, Accuracy: 0.8169
Training loss (for one batch) at step 90: 269.7393, Accuracy: 0.8166
Training loss (for one batch) at step 100: 270.7150, Accuracy: 0.8157
Training loss (for one batch) at step 110: 245.3944, Accuracy: 0.8162
Training loss (for one batch) at step 120: 258.1553, Accuracy: 0.8147
Training loss (for one batch) at step 130: 249.0710, Accuracy: 0.8131
Training loss (for one batch) at step 140: 287.2775, Accuracy: 0.8131
---- Training ----
Training loss: 230.6401
Training acc over epoch: 0.8119
---- Validation ----
Validation loss: 71.3853
Validation acc: 0.7348
Time taken: 37.27s
Start of epoch 24
Training loss (for one batch) at step 0: 280.7761, Accuracy: 0.7400
Training loss (for one batch) at step 10: 247.2088, Accuracy: 0.8118
Training loss (for one batch) at step 20: 258.5042, Accuracy: 0.8190
Training loss (for one batch) at step 30: 255.4095, Accuracy: 0.8213
Training loss (for one batch) at step 40: 254.1044, Accuracy: 0.8198
Training loss (for one batch) at step 50: 256.7287, Accuracy: 0.8220
Training loss (for one batch) at step 60: 239.2006, Accuracy: 0.8248
Training loss (for one batch) at step 70: 255.1565, Accuracy: 0.8214
Training loss (for one batch) at step 80: 268.6976, Accuracy: 0.8196
Training loss (for one batch) at step 90: 267.7452, Accuracy: 0.8207
Training loss (for one batch) at step 100: 242.0513, Accuracy: 0.8187
Training loss (for one batch) at step 110: 258.2068, Accuracy: 0.8193
Training loss (for one batch) at step 120: 259.0406, Accuracy: 0.8165
Training loss (for one batch) at step 130: 259.2622, Accuracy: 0.8170
Training loss (for one batch) at step 140: 253.7758, Accuracy: 0.8154
---- Training ----
Training loss: 216.2381
Training acc over epoch: 0.8168
---- Validation ----
Validation loss: 69.9500
Validation acc: 0.7337
Time taken: 64.98s
Start of epoch 25
Training loss (for one batch) at step 0: 263.5667, Accuracy: 0.8300
Training loss (for one batch) at step 10: 242.4959, Accuracy: 0.8236
Training loss (for one batch) at step 20: 266.2979, Accuracy: 0.8324
Training loss (for one batch) at step 30: 275.1202, Accuracy: 0.8281
Training loss (for one batch) at step 40: 286.8224, Accuracy: 0.8229
Training loss (for one batch) at step 50: 264.0530, Accuracy: 0.8214
Training loss (for one batch) at step 60: 229.2666, Accuracy: 0.8233
Training loss (for one batch) at step 70: 251.2237, Accuracy: 0.8228
Training loss (for one batch) at step 80: 283.0322, Accuracy: 0.8209
Training loss (for one batch) at step 90: 236.1496, Accuracy: 0.8215
Training loss (for one batch) at step 100: 265.8540, Accuracy: 0.8175
Training loss (for one batch) at step 110: 261.0361, Accuracy: 0.8185
Training loss (for one batch) at step 120: 248.9195, Accuracy: 0.8175
Training loss (for one batch) at step 130: 262.9921, Accuracy: 0.8182
Training loss (for one batch) at step 140: 252.1389, Accuracy: 0.8158
---- Training ----
Training loss: 225.3495
Training acc over epoch: 0.8159
---- Validation ----
Validation loss: 68.0859
Validation acc: 0.7386
Time taken: 37.04s
Start of epoch 26
Training loss (for one batch) at step 0: 244.2747, Accuracy: 0.8400
Training loss (for one batch) at step 10: 269.4955, Accuracy: 0.8227
Training loss (for one batch) at step 20: 251.7158, Accuracy: 0.8233
Training loss (for one batch) at step 30: 253.7226, Accuracy: 0.8177
Training loss (for one batch) at step 40: 256.1872, Accuracy: 0.8190
Training loss (for one batch) at step 50: 239.9199, Accuracy: 0.8208
Training loss (for one batch) at step 60: 243.3136, Accuracy: 0.8205
Training loss (for one batch) at step 70: 250.4837, Accuracy: 0.8203
Training loss (for one batch) at step 80: 246.3107, Accuracy: 0.8204
Training loss (for one batch) at step 90: 273.5338, Accuracy: 0.8181
Training loss (for one batch) at step 100: 265.1309, Accuracy: 0.8182
Training loss (for one batch) at step 110: 258.0261, Accuracy: 0.8179
Training loss (for one batch) at step 120: 256.5783, Accuracy: 0.8200
Training loss (for one batch) at step 130: 269.6611, Accuracy: 0.8206
Training loss (for one batch) at step 140: 250.4053, Accuracy: 0.8213
---- Training ----
Training loss: 230.8903
Training acc over epoch: 0.8210
---- Validation ----
Validation loss: 69.6143
Validation acc: 0.7378
Time taken: 64.62s
Start of epoch 27
Training loss (for one batch) at step 0: 268.1575, Accuracy: 0.7800
Training loss (for one batch) at step 10: 243.6683, Accuracy: 0.8418
Training loss (for one batch) at step 20: 259.9017, Accuracy: 0.8310
Training loss (for one batch) at step 30: 262.0081, Accuracy: 0.8258
Training loss (for one batch) at step 40: 240.2090, Accuracy: 0.8249
Training loss (for one batch) at step 50: 250.9597, Accuracy: 0.8259
Training loss (for one batch) at step 60: 239.2737, Accuracy: 0.8264
Training loss (for one batch) at step 70: 255.2901, Accuracy: 0.8269
Training loss (for one batch) at step 80: 250.7855, Accuracy: 0.8256
Training loss (for one batch) at step 90: 260.3322, Accuracy: 0.8242
Training loss (for one batch) at step 100: 251.3374, Accuracy: 0.8237
Training loss (for one batch) at step 110: 239.4255, Accuracy: 0.8251
Training loss (for one batch) at step 120: 254.1772, Accuracy: 0.8233
Training loss (for one batch) at step 130: 250.1361, Accuracy: 0.8244
Training loss (for one batch) at step 140: 271.4566, Accuracy: 0.8230
---- Training ----
Training loss: 234.3596
Training acc over epoch: 0.8226
---- Validation ----
Validation loss: 68.2509
Validation acc: 0.7397
Time taken: 38.04s
Start of epoch 28
Training loss (for one batch) at step 0: 256.8065, Accuracy: 0.7800
Training loss (for one batch) at step 10: 259.0746, Accuracy: 0.8273
Training loss (for one batch) at step 20: 254.2482, Accuracy: 0.8390
Training loss (for one batch) at step 30: 237.0715, Accuracy: 0.8339
Training loss (for one batch) at step 40: 232.1272, Accuracy: 0.8327
Training loss (for one batch) at step 50: 264.7004, Accuracy: 0.8369
Training loss (for one batch) at step 60: 256.0864, Accuracy: 0.8370
Training loss (for one batch) at step 70: 255.6832, Accuracy: 0.8356
Training loss (for one batch) at step 80: 264.9968, Accuracy: 0.8319
Training loss (for one batch) at step 90: 239.4745, Accuracy: 0.8301
Training loss (for one batch) at step 100: 257.3871, Accuracy: 0.8294
Training loss (for one batch) at step 110: 235.6699, Accuracy: 0.8315
Training loss (for one batch) at step 120: 263.0173, Accuracy: 0.8307
Training loss (for one batch) at step 130: 253.0811, Accuracy: 0.8285
Training loss (for one batch) at step 140: 238.1869, Accuracy: 0.8269
---- Training ----
Training loss: 220.7025
Training acc over epoch: 0.8266
---- Validation ----
Validation loss: 60.2272
Validation acc: 0.7284
Time taken: 63.69s
Start of epoch 29
Training loss (for one batch) at step 0: 239.6166, Accuracy: 0.8300
Training loss (for one batch) at step 10: 248.6262, Accuracy: 0.8227
Training loss (for one batch) at step 20: 273.5850, Accuracy: 0.8305
Training loss (for one batch) at step 30: 240.8706, Accuracy: 0.8371
Training loss (for one batch) at step 40: 236.1343, Accuracy: 0.8393
Training loss (for one batch) at step 50: 246.2553, Accuracy: 0.8371
Training loss (for one batch) at step 60: 251.3374, Accuracy: 0.8367
Training loss (for one batch) at step 70: 245.7694, Accuracy: 0.8345
Training loss (for one batch) at step 80: 291.0194, Accuracy: 0.8326
Training loss (for one batch) at step 90: 236.9663, Accuracy: 0.8305
Training loss (for one batch) at step 100: 240.1508, Accuracy: 0.8291
Training loss (for one batch) at step 110: 230.7787, Accuracy: 0.8299
Training loss (for one batch) at step 120: 220.4176, Accuracy: 0.8301
Training loss (for one batch) at step 130: 246.2157, Accuracy: 0.8292
Training loss (for one batch) at step 140: 235.0207, Accuracy: 0.8296
---- Training ----
Training loss: 237.4948
Training acc over epoch: 0.8289
---- Validation ----
Validation loss: 63.4127
Validation acc: 0.7359
Time taken: 37.58s
Start of epoch 30
Training loss (for one batch) at step 0: 249.9998, Accuracy: 0.8500
Training loss (for one batch) at step 10: 256.5432, Accuracy: 0.8364
Training loss (for one batch) at step 20: 225.3118, Accuracy: 0.8290
Training loss (for one batch) at step 30: 247.8926, Accuracy: 0.8319
Training loss (for one batch) at step 40: 257.8928, Accuracy: 0.8285
Training loss (for one batch) at step 50: 235.7447, Accuracy: 0.8322
Training loss (for one batch) at step 60: 235.9714, Accuracy: 0.8285
Training loss (for one batch) at step 70: 246.7176, Accuracy: 0.8285
Training loss (for one batch) at step 80: 263.3189, Accuracy: 0.8274
Training loss (for one batch) at step 90: 266.1384, Accuracy: 0.8288
Training loss (for one batch) at step 100: 237.1652, Accuracy: 0.8278
Training loss (for one batch) at step 110: 252.4740, Accuracy: 0.8286
Training loss (for one batch) at step 120: 256.4757, Accuracy: 0.8282
Training loss (for one batch) at step 130: 244.4297, Accuracy: 0.8285
Training loss (for one batch) at step 140: 261.2341, Accuracy: 0.8283
---- Training ----
Training loss: 220.5580
Training acc over epoch: 0.8291
---- Validation ----
Validation loss: 75.0448
Validation acc: 0.7316
Time taken: 66.29s
Start of epoch 31
Training loss (for one batch) at step 0: 251.4385, Accuracy: 0.8300
Training loss (for one batch) at step 10: 251.8913, Accuracy: 0.8518
Training loss (for one batch) at step 20: 253.3459, Accuracy: 0.8538
Training loss (for one batch) at step 30: 240.6779, Accuracy: 0.8510
Training loss (for one batch) at step 40: 241.1933, Accuracy: 0.8456
Training loss (for one batch) at step 50: 227.3037, Accuracy: 0.8424
Training loss (for one batch) at step 60: 253.6561, Accuracy: 0.8425
Training loss (for one batch) at step 70: 263.4375, Accuracy: 0.8406
Training loss (for one batch) at step 80: 233.6592, Accuracy: 0.8394
Training loss (for one batch) at step 90: 219.3743, Accuracy: 0.8376
Training loss (for one batch) at step 100: 242.1733, Accuracy: 0.8357
Training loss (for one batch) at step 110: 240.5328, Accuracy: 0.8348
Training loss (for one batch) at step 120: 249.4072, Accuracy: 0.8351
Training loss (for one batch) at step 130: 247.5589, Accuracy: 0.8340
Training loss (for one batch) at step 140: 246.7546, Accuracy: 0.8338
---- Training ----
Training loss: 206.1685
Training acc over epoch: 0.8342
---- Validation ----
Validation loss: 67.4582
Validation acc: 0.7391
Time taken: 37.34s
Start of epoch 32
Training loss (for one batch) at step 0: 244.0874, Accuracy: 0.8700
Training loss (for one batch) at step 10: 249.9423, Accuracy: 0.8436
Training loss (for one batch) at step 20: 224.8503, Accuracy: 0.8448
Training loss (for one batch) at step 30: 271.1761, Accuracy: 0.8381
Training loss (for one batch) at step 40: 226.0677, Accuracy: 0.8388
Training loss (for one batch) at step 50: 238.1078, Accuracy: 0.8378
Training loss (for one batch) at step 60: 218.3685, Accuracy: 0.8415
Training loss (for one batch) at step 70: 243.2289, Accuracy: 0.8393
Training loss (for one batch) at step 80: 272.7744, Accuracy: 0.8396
Training loss (for one batch) at step 90: 246.1263, Accuracy: 0.8374
Training loss (for one batch) at step 100: 274.7321, Accuracy: 0.8365
Training loss (for one batch) at step 110: 238.6112, Accuracy: 0.8357
Training loss (for one batch) at step 120: 227.7910, Accuracy: 0.8351
Training loss (for one batch) at step 130: 248.2352, Accuracy: 0.8344
Training loss (for one batch) at step 140: 242.2143, Accuracy: 0.8343
---- Training ----
Training loss: 213.9095
Training acc over epoch: 0.8340
---- Validation ----
Validation loss: 70.6265
Validation acc: 0.7316
Time taken: 63.81s
Start of epoch 33
Training loss (for one batch) at step 0: 252.0213, Accuracy: 0.8100
Training loss (for one batch) at step 10: 268.9356, Accuracy: 0.8273
Training loss (for one batch) at step 20: 237.1700, Accuracy: 0.8319
Training loss (for one batch) at step 30: 245.6171, Accuracy: 0.8323
Training loss (for one batch) at step 40: 247.8270, Accuracy: 0.8349
Training loss (for one batch) at step 50: 233.1107, Accuracy: 0.8392
Training loss (for one batch) at step 60: 247.4797, Accuracy: 0.8413
Training loss (for one batch) at step 70: 234.9050, Accuracy: 0.8411
Training loss (for one batch) at step 80: 239.3679, Accuracy: 0.8388
Training loss (for one batch) at step 90: 262.2308, Accuracy: 0.8351
Training loss (for one batch) at step 100: 235.6351, Accuracy: 0.8360
Training loss (for one batch) at step 110: 248.0833, Accuracy: 0.8376
Training loss (for one batch) at step 120: 240.7896, Accuracy: 0.8377
Training loss (for one batch) at step 130: 249.3737, Accuracy: 0.8367
Training loss (for one batch) at step 140: 242.9373, Accuracy: 0.8365
---- Training ----
Training loss: 200.1012
Training acc over epoch: 0.8367
---- Validation ----
Validation loss: 65.7357
Validation acc: 0.7340
Time taken: 37.16s
Start of epoch 34
Training loss (for one batch) at step 0: 239.9920, Accuracy: 0.8500
Training loss (for one batch) at step 10: 231.5297, Accuracy: 0.8373
Training loss (for one batch) at step 20: 236.1277, Accuracy: 0.8348
Training loss (for one batch) at step 30: 247.3307, Accuracy: 0.8319
Training loss (for one batch) at step 40: 220.9794, Accuracy: 0.8368
Training loss (for one batch) at step 50: 236.2619, Accuracy: 0.8371
Training loss (for one batch) at step 60: 229.9386, Accuracy: 0.8393
Training loss (for one batch) at step 70: 254.0831, Accuracy: 0.8390
Training loss (for one batch) at step 80: 249.4035, Accuracy: 0.8402
Training loss (for one batch) at step 90: 239.7289, Accuracy: 0.8390
Training loss (for one batch) at step 100: 219.4154, Accuracy: 0.8385
Training loss (for one batch) at step 110: 230.9085, Accuracy: 0.8400
Training loss (for one batch) at step 120: 221.5601, Accuracy: 0.8402
Training loss (for one batch) at step 130: 217.7972, Accuracy: 0.8397
Training loss (for one batch) at step 140: 229.0081, Accuracy: 0.8384
---- Training ----
Training loss: 209.9108
Training acc over epoch: 0.8387
---- Validation ----
Validation loss: 76.8264
Validation acc: 0.7405
Time taken: 64.45s
Start of epoch 35
Training loss (for one batch) at step 0: 242.5919, Accuracy: 0.8300
Training loss (for one batch) at step 10: 237.5313, Accuracy: 0.8355
Training loss (for one batch) at step 20: 228.6697, Accuracy: 0.8495
Training loss (for one batch) at step 30: 254.7920, Accuracy: 0.8500
Training loss (for one batch) at step 40: 222.7771, Accuracy: 0.8495
Training loss (for one batch) at step 50: 227.3997, Accuracy: 0.8494
Training loss (for one batch) at step 60: 234.4799, Accuracy: 0.8508
Training loss (for one batch) at step 70: 250.8777, Accuracy: 0.8451
Training loss (for one batch) at step 80: 229.2868, Accuracy: 0.8428
Training loss (for one batch) at step 90: 234.9481, Accuracy: 0.8437
Training loss (for one batch) at step 100: 232.2818, Accuracy: 0.8422
Training loss (for one batch) at step 110: 217.0236, Accuracy: 0.8419
Training loss (for one batch) at step 120: 241.6974, Accuracy: 0.8408
Training loss (for one batch) at step 130: 225.0129, Accuracy: 0.8398
Training loss (for one batch) at step 140: 252.2674, Accuracy: 0.8409
---- Training ----
Training loss: 206.1332
Training acc over epoch: 0.8405
---- Validation ----
Validation loss: 81.2987
Validation acc: 0.7262
Time taken: 37.71s
Start of epoch 36
Training loss (for one batch) at step 0: 235.5174, Accuracy: 0.8700
Training loss (for one batch) at step 10: 230.6322, Accuracy: 0.8555
Training loss (for one batch) at step 20: 266.4501, Accuracy: 0.8467
Training loss (for one batch) at step 30: 217.2251, Accuracy: 0.8442
Training loss (for one batch) at step 40: 214.7453, Accuracy: 0.8427
Training loss (for one batch) at step 50: 244.0030, Accuracy: 0.8431
Training loss (for one batch) at step 60: 239.7386, Accuracy: 0.8454
Training loss (for one batch) at step 70: 243.2821, Accuracy: 0.8428
Training loss (for one batch) at step 80: 256.6152, Accuracy: 0.8417
Training loss (for one batch) at step 90: 229.0629, Accuracy: 0.8409
Training loss (for one batch) at step 100: 233.9327, Accuracy: 0.8378
Training loss (for one batch) at step 110: 222.1417, Accuracy: 0.8383
Training loss (for one batch) at step 120: 223.3936, Accuracy: 0.8368
Training loss (for one batch) at step 130: 217.7450, Accuracy: 0.8370
Training loss (for one batch) at step 140: 253.1354, Accuracy: 0.8363
---- Training ----
Training loss: 218.2092
Training acc over epoch: 0.8366
---- Validation ----
Validation loss: 70.4013
Validation acc: 0.7249
Time taken: 64.66s
Start of epoch 37
Training loss (for one batch) at step 0: 238.1612, Accuracy: 0.8400
Training loss (for one batch) at step 10: 240.4977, Accuracy: 0.8491
Training loss (for one batch) at step 20: 226.4392, Accuracy: 0.8529
Training loss (for one batch) at step 30: 229.9782, Accuracy: 0.8439
Training loss (for one batch) at step 40: 219.1117, Accuracy: 0.8454
Training loss (for one batch) at step 50: 229.7378, Accuracy: 0.8457
Training loss (for one batch) at step 60: 248.6942, Accuracy: 0.8474
Training loss (for one batch) at step 70: 228.1802, Accuracy: 0.8487
Training loss (for one batch) at step 80: 238.2119, Accuracy: 0.8432
Training loss (for one batch) at step 90: 255.6680, Accuracy: 0.8413
Training loss (for one batch) at step 100: 218.1819, Accuracy: 0.8400
Training loss (for one batch) at step 110: 234.9074, Accuracy: 0.8397
Training loss (for one batch) at step 120: 234.5446, Accuracy: 0.8404
Training loss (for one batch) at step 130: 243.2114, Accuracy: 0.8398
Training loss (for one batch) at step 140: 237.8987, Accuracy: 0.8383
---- Training ----
Training loss: 191.9649
Training acc over epoch: 0.8379
---- Validation ----
Validation loss: 76.1656
Validation acc: 0.7316
Time taken: 38.09s
Start of epoch 38
Training loss (for one batch) at step 0: 234.5698, Accuracy: 0.8100
Training loss (for one batch) at step 10: 246.7703, Accuracy: 0.8418
Training loss (for one batch) at step 20: 226.3742, Accuracy: 0.8405
Training loss (for one batch) at step 30: 226.5220, Accuracy: 0.8410
Training loss (for one batch) at step 40: 229.8373, Accuracy: 0.8410
Training loss (for one batch) at step 50: 238.0048, Accuracy: 0.8412
Training loss (for one batch) at step 60: 213.0829, Accuracy: 0.8430
Training loss (for one batch) at step 70: 230.9470, Accuracy: 0.8452
Training loss (for one batch) at step 80: 223.7115, Accuracy: 0.8436
Training loss (for one batch) at step 90: 227.6953, Accuracy: 0.8446
Training loss (for one batch) at step 100: 237.4434, Accuracy: 0.8436
Training loss (for one batch) at step 110: 227.3601, Accuracy: 0.8431
Training loss (for one batch) at step 120: 206.1754, Accuracy: 0.8432
Training loss (for one batch) at step 130: 220.3103, Accuracy: 0.8424
Training loss (for one batch) at step 140: 230.6531, Accuracy: 0.8408
---- Training ----
Training loss: 199.6105
Training acc over epoch: 0.8413
---- Validation ----
Validation loss: 64.4639
Validation acc: 0.7284
Time taken: 65.31s
Start of epoch 39
Training loss (for one batch) at step 0: 237.1849, Accuracy: 0.8400
Training loss (for one batch) at step 10: 221.5655, Accuracy: 0.8573
Training loss (for one batch) at step 20: 229.8233, Accuracy: 0.8538
Training loss (for one batch) at step 30: 215.2856, Accuracy: 0.8542
Training loss (for one batch) at step 40: 233.7465, Accuracy: 0.8524
Training loss (for one batch) at step 50: 226.9208, Accuracy: 0.8559
Training loss (for one batch) at step 60: 231.8494, Accuracy: 0.8544
Training loss (for one batch) at step 70: 236.1653, Accuracy: 0.8545
Training loss (for one batch) at step 80: 230.4915, Accuracy: 0.8519
Training loss (for one batch) at step 90: 235.8813, Accuracy: 0.8485
Training loss (for one batch) at step 100: 232.1350, Accuracy: 0.8463
Training loss (for one batch) at step 110: 227.4958, Accuracy: 0.8460
Training loss (for one batch) at step 120: 219.5463, Accuracy: 0.8463
Training loss (for one batch) at step 130: 210.1389, Accuracy: 0.8458
Training loss (for one batch) at step 140: 225.8702, Accuracy: 0.8460
---- Training ----
Training loss: 184.7298
Training acc over epoch: 0.8458
---- Validation ----
Validation loss: 72.7286
Validation acc: 0.7246
Time taken: 38.61s
Start of epoch 40
Training loss (for one batch) at step 0: 216.3292, Accuracy: 0.8700
Training loss (for one batch) at step 10: 204.0099, Accuracy: 0.8564
Training loss (for one batch) at step 20: 223.7343, Accuracy: 0.8510
Training loss (for one batch) at step 30: 215.5699, Accuracy: 0.8448
Training loss (for one batch) at step 40: 223.6602, Accuracy: 0.8456
Training loss (for one batch) at step 50: 220.5359, Accuracy: 0.8498
Training loss (for one batch) at step 60: 227.9573, Accuracy: 0.8490
Training loss (for one batch) at step 70: 231.0555, Accuracy: 0.8500
Training loss (for one batch) at step 80: 243.2740, Accuracy: 0.8504
Training loss (for one batch) at step 90: 226.2473, Accuracy: 0.8512
Training loss (for one batch) at step 100: 239.3641, Accuracy: 0.8499
Training loss (for one batch) at step 110: 246.7267, Accuracy: 0.8505
Training loss (for one batch) at step 120: 206.5442, Accuracy: 0.8497
Training loss (for one batch) at step 130: 222.5840, Accuracy: 0.8496
Training loss (for one batch) at step 140: 251.3899, Accuracy: 0.8486
---- Training ----
Training loss: 201.5881
Training acc over epoch: 0.8481
---- Validation ----
Validation loss: 73.3230
Validation acc: 0.7238
Time taken: 65.30s
Start of epoch 41
Training loss (for one batch) at step 0: 227.7898, Accuracy: 0.8600
Training loss (for one batch) at step 10: 228.7859, Accuracy: 0.8545
Training loss (for one batch) at step 20: 219.9217, Accuracy: 0.8567
Training loss (for one batch) at step 30: 209.3973, Accuracy: 0.8577
Training loss (for one batch) at step 40: 208.9625, Accuracy: 0.8585
Training loss (for one batch) at step 50: 204.5882, Accuracy: 0.8592
Training loss (for one batch) at step 60: 229.8008, Accuracy: 0.8557
Training loss (for one batch) at step 70: 207.5428, Accuracy: 0.8577
Training loss (for one batch) at step 80: 230.2812, Accuracy: 0.8549
Training loss (for one batch) at step 90: 217.8126, Accuracy: 0.8543
Training loss (for one batch) at step 100: 223.2358, Accuracy: 0.8519
Training loss (for one batch) at step 110: 211.3388, Accuracy: 0.8500
Training loss (for one batch) at step 120: 226.7275, Accuracy: 0.8500
Training loss (for one batch) at step 130: 227.1644, Accuracy: 0.8492
Training loss (for one batch) at step 140: 224.5988, Accuracy: 0.8487
---- Training ----
Training loss: 195.1359
Training acc over epoch: 0.8483
---- Validation ----
Validation loss: 91.6706
Validation acc: 0.7423
Time taken: 38.84s
Start of epoch 42
Training loss (for one batch) at step 0: 217.0396, Accuracy: 0.8700
Training loss (for one batch) at step 10: 241.3870, Accuracy: 0.8455
Training loss (for one batch) at step 20: 233.0230, Accuracy: 0.8514
Training loss (for one batch) at step 30: 213.6895, Accuracy: 0.8484
Training loss (for one batch) at step 40: 220.4002, Accuracy: 0.8500
Training loss (for one batch) at step 50: 220.8671, Accuracy: 0.8506
Training loss (for one batch) at step 60: 241.0061, Accuracy: 0.8485
Training loss (for one batch) at step 70: 221.8659, Accuracy: 0.8486
Training loss (for one batch) at step 80: 215.1025, Accuracy: 0.8478
Training loss (for one batch) at step 90: 193.4868, Accuracy: 0.8451
Training loss (for one batch) at step 100: 242.7956, Accuracy: 0.8444
Training loss (for one batch) at step 110: 215.4156, Accuracy: 0.8434
Training loss (for one batch) at step 120: 212.0618, Accuracy: 0.8443
Training loss (for one batch) at step 130: 228.4740, Accuracy: 0.8446
Training loss (for one batch) at step 140: 196.6378, Accuracy: 0.8452
---- Training ----
Training loss: 202.2497
Training acc over epoch: 0.8446
---- Validation ----
Validation loss: 62.3941
Validation acc: 0.7370
Time taken: 66.41s
Start of epoch 43
Training loss (for one batch) at step 0: 224.0869, Accuracy: 0.8600
Training loss (for one batch) at step 10: 215.2218, Accuracy: 0.8409
Training loss (for one batch) at step 20: 211.4269, Accuracy: 0.8505
Training loss (for one batch) at step 30: 212.3453, Accuracy: 0.8442
Training loss (for one batch) at step 40: 215.4431, Accuracy: 0.8512
Training loss (for one batch) at step 50: 234.1839, Accuracy: 0.8488
Training loss (for one batch) at step 60: 224.4264, Accuracy: 0.8487
Training loss (for one batch) at step 70: 226.4505, Accuracy: 0.8496
Training loss (for one batch) at step 80: 226.6686, Accuracy: 0.8478
Training loss (for one batch) at step 90: 220.1757, Accuracy: 0.8486
Training loss (for one batch) at step 100: 210.6343, Accuracy: 0.8472
Training loss (for one batch) at step 110: 234.1867, Accuracy: 0.8475
Training loss (for one batch) at step 120: 218.1427, Accuracy: 0.8478
Training loss (for one batch) at step 130: 220.1093, Accuracy: 0.8482
Training loss (for one batch) at step 140: 227.0024, Accuracy: 0.8482
---- Training ----
Training loss: 179.8753
Training acc over epoch: 0.8475
---- Validation ----
Validation loss: 69.2404
Validation acc: 0.7171
Time taken: 39.09s
Start of epoch 44
Training loss (for one batch) at step 0: 216.6525, Accuracy: 0.9000
Training loss (for one batch) at step 10: 230.2513, Accuracy: 0.8609
Training loss (for one batch) at step 20: 248.1977, Accuracy: 0.8543
Training loss (for one batch) at step 30: 210.0165, Accuracy: 0.8490
Training loss (for one batch) at step 40: 202.3168, Accuracy: 0.8505
Training loss (for one batch) at step 50: 199.5857, Accuracy: 0.8549
Training loss (for one batch) at step 60: 208.3279, Accuracy: 0.8551
Training loss (for one batch) at step 70: 225.5777, Accuracy: 0.8555
Training loss (for one batch) at step 80: 205.6580, Accuracy: 0.8538
Training loss (for one batch) at step 90: 226.4557, Accuracy: 0.8516
Training loss (for one batch) at step 100: 235.0610, Accuracy: 0.8496
Training loss (for one batch) at step 110: 234.9911, Accuracy: 0.8496
Training loss (for one batch) at step 120: 213.8809, Accuracy: 0.8503
Training loss (for one batch) at step 130: 229.6682, Accuracy: 0.8499
Training loss (for one batch) at step 140: 229.0847, Accuracy: 0.8493
---- Training ----
Training loss: 213.2135
Training acc over epoch: 0.8491
---- Validation ----
Validation loss: 85.4744
Validation acc: 0.7249
Time taken: 63.66s
Start of epoch 45
Training loss (for one batch) at step 0: 222.7992, Accuracy: 0.8700
Training loss (for one batch) at step 10: 209.8335, Accuracy: 0.8600
Training loss (for one batch) at step 20: 219.3079, Accuracy: 0.8495
Training loss (for one batch) at step 30: 229.8299, Accuracy: 0.8552
Training loss (for one batch) at step 40: 224.5604, Accuracy: 0.8532
Training loss (for one batch) at step 50: 190.8823, Accuracy: 0.8563
Training loss (for one batch) at step 60: 208.4450, Accuracy: 0.8570
Training loss (for one batch) at step 70: 205.4574, Accuracy: 0.8527
Training loss (for one batch) at step 80: 238.4645, Accuracy: 0.8520
Training loss (for one batch) at step 90: 225.2913, Accuracy: 0.8479
Training loss (for one batch) at step 100: 213.0399, Accuracy: 0.8496
Training loss (for one batch) at step 110: 199.3031, Accuracy: 0.8495
Training loss (for one batch) at step 120: 243.4660, Accuracy: 0.8512
Training loss (for one batch) at step 130: 215.6532, Accuracy: 0.8498
Training loss (for one batch) at step 140: 232.8289, Accuracy: 0.8491
---- Training ----
Training loss: 185.4447
Training acc over epoch: 0.8505
---- Validation ----
Validation loss: 85.1821
Validation acc: 0.7450
Time taken: 36.82s
Start of epoch 46
Training loss (for one batch) at step 0: 216.4352, Accuracy: 0.8600
Training loss (for one batch) at step 10: 199.4248, Accuracy: 0.8536
Training loss (for one batch) at step 20: 212.3192, Accuracy: 0.8543
Training loss (for one batch) at step 30: 237.6564, Accuracy: 0.8494
Training loss (for one batch) at step 40: 226.7646, Accuracy: 0.8505
Training loss (for one batch) at step 50: 219.4993, Accuracy: 0.8561
Training loss (for one batch) at step 60: 217.4336, Accuracy: 0.8559
Training loss (for one batch) at step 70: 201.6453, Accuracy: 0.8554
Training loss (for one batch) at step 80: 211.7825, Accuracy: 0.8560
Training loss (for one batch) at step 90: 227.1505, Accuracy: 0.8538
Training loss (for one batch) at step 100: 234.6304, Accuracy: 0.8528
Training loss (for one batch) at step 110: 223.3621, Accuracy: 0.8532
Training loss (for one batch) at step 120: 205.4325, Accuracy: 0.8539
Training loss (for one batch) at step 130: 204.2140, Accuracy: 0.8516
Training loss (for one batch) at step 140: 236.2521, Accuracy: 0.8522
---- Training ----
Training loss: 186.7998
Training acc over epoch: 0.8522
---- Validation ----
Validation loss: 73.4796
Validation acc: 0.7391
Time taken: 64.83s
Start of epoch 47
Training loss (for one batch) at step 0: 203.1421, Accuracy: 0.8400
Training loss (for one batch) at step 10: 229.9133, Accuracy: 0.8491
Training loss (for one batch) at step 20: 234.2820, Accuracy: 0.8500
Training loss (for one batch) at step 30: 196.1465, Accuracy: 0.8513
Training loss (for one batch) at step 40: 228.6390, Accuracy: 0.8515
Training loss (for one batch) at step 50: 223.1656, Accuracy: 0.8533
Training loss (for one batch) at step 60: 217.5273, Accuracy: 0.8557
Training loss (for one batch) at step 70: 210.7529, Accuracy: 0.8548
Training loss (for one batch) at step 80: 191.5359, Accuracy: 0.8547
Training loss (for one batch) at step 90: 227.4669, Accuracy: 0.8518
Training loss (for one batch) at step 100: 211.4654, Accuracy: 0.8506
Training loss (for one batch) at step 110: 199.8591, Accuracy: 0.8514
Training loss (for one batch) at step 120: 225.0475, Accuracy: 0.8514
Training loss (for one batch) at step 130: 226.9212, Accuracy: 0.8512
Training loss (for one batch) at step 140: 213.8333, Accuracy: 0.8510
---- Training ----
Training loss: 179.8663
Training acc over epoch: 0.8505
---- Validation ----
Validation loss: 72.2260
Validation acc: 0.7254
Time taken: 36.92s
Start of epoch 48
Training loss (for one batch) at step 0: 215.3543, Accuracy: 0.8300
Training loss (for one batch) at step 10: 243.1009, Accuracy: 0.8555
Training loss (for one batch) at step 20: 212.6778, Accuracy: 0.8519
Training loss (for one batch) at step 30: 183.0243, Accuracy: 0.8497
Training loss (for one batch) at step 40: 206.0371, Accuracy: 0.8546
Training loss (for one batch) at step 50: 220.6156, Accuracy: 0.8578
Training loss (for one batch) at step 60: 202.1973, Accuracy: 0.8595
Training loss (for one batch) at step 70: 207.4982, Accuracy: 0.8572
Training loss (for one batch) at step 80: 211.1456, Accuracy: 0.8563
Training loss (for one batch) at step 90: 204.6884, Accuracy: 0.8556
Training loss (for one batch) at step 100: 220.5500, Accuracy: 0.8546
Training loss (for one batch) at step 110: 203.2720, Accuracy: 0.8551
Training loss (for one batch) at step 120: 216.8195, Accuracy: 0.8560
Training loss (for one batch) at step 130: 215.0410, Accuracy: 0.8547
Training loss (for one batch) at step 140: 238.3763, Accuracy: 0.8531
---- Training ----
Training loss: 187.8261
Training acc over epoch: 0.8536
---- Validation ----
Validation loss: 75.8526
Validation acc: 0.7246
Time taken: 63.56s
Start of epoch 49
Training loss (for one batch) at step 0: 227.0806, Accuracy: 0.8500
Training loss (for one batch) at step 10: 209.5961, Accuracy: 0.8591
Training loss (for one batch) at step 20: 214.4810, Accuracy: 0.8514
Training loss (for one batch) at step 30: 206.1188, Accuracy: 0.8539
Training loss (for one batch) at step 40: 200.9710, Accuracy: 0.8549
Training loss (for one batch) at step 50: 197.2371, Accuracy: 0.8567
Training loss (for one batch) at step 60: 215.7870, Accuracy: 0.8580
Training loss (for one batch) at step 70: 211.5255, Accuracy: 0.8563
Training loss (for one batch) at step 80: 238.5272, Accuracy: 0.8552
Training loss (for one batch) at step 90: 197.1934, Accuracy: 0.8522
Training loss (for one batch) at step 100: 220.6883, Accuracy: 0.8512
Training loss (for one batch) at step 110: 204.7271, Accuracy: 0.8545
Training loss (for one batch) at step 120: 210.4459, Accuracy: 0.8545
Training loss (for one batch) at step 130: 198.4456, Accuracy: 0.8540
Training loss (for one batch) at step 140: 208.7387, Accuracy: 0.8532
---- Training ----
Training loss: 205.0569
Training acc over epoch: 0.8534
---- Validation ----
Validation loss: 88.7453
Validation acc: 0.7311
Time taken: 36.91s
===== Q: 0.0001
Validation acc: 0.7384
Validation AUC: 0.7353
Validation Balanced_ACC: 0.4773
Validation MI: 0.1371
Validation Normalized MI: 0.2055
Validation Adjusted MI: 0.2055
Validation aUc_Sklearn: 0.8305
[19]:
# classification_report_r= []
# model = create_model()
# K=2
# R=5
# NUM_RUNS = 10
# N_EPOCHS = 30
# val_acc = np.zeros(NUM_RUNS)
# for i in range(NUM_RUNS):
# MA = MultipleAnnotators_Classification(K, R, 0.1)
# model = create_model()
# optimizer = tf.keras.optimizers.Adam(learning_rate=1e-3, clipnorm=1.0)
# model.compile(optimizer=optimizer, loss= MA.loss())
# history_model = model.fit(train_batches_MA, validation_data=val_batches_MA, epochs= N_EPOCHS, callbacks=callbacks, verbose=0)
# #model = MA.fit(model, Data_train_MA, N_EPOCHS)
# pred_2 = model.predict(X_test)
# lambda_R_ = pred_2[:, K:] #annotators reliability prediction N x R
# classification_report_r += [classification_report( pred_2[:,:K].argmax(axis=1),Y_true_test.ravel(),output_dict=True)]
# print(classification_report( pred_2[:,:K].argmax(axis=1),Y_true_test.ravel()))
# #val_acc[i] = MA.eval_model(test_batches_MA)
# #print("Validation acc: %.4f" % (float(val_acc[i]),))
# # Create the history figure
# plt.figure(figsize=(16,9))
# for i in history_model.history:
# plt.plot(history_model.history[i],label=i)
# plt.title('Model history')
# plt.legend()
# plt.grid()
# import pandas as pd
# df = pd.DataFrame(val_acc)
# #df.to_csimport pandas as pddf = pd.DataFrame(val_acc)#df.to_csv('/kaggle/working/CatDogs_MA_InceptionV3.csv',index=False) # save to notebook outputv('/kaggle/working/CatDogs_MA_InceptionV3.csv',index=False) # save to notebook output
[20]:
print('Average Accuracy: ', np.round( ACC.mean(),4)*100)
print('Average std: ',np.round(np.std( ACC),4)*100)
print('==============================================')
print('Average AUC: ', np.round( AUC.mean(),4)*100)
print('Average AUC std: ',np.round(np.std( AUC),4)*100)
print('==============================================')
print('Average AUC Sklearn: ', np.round( AUCSK.mean(),4)*100)
print('Average AUC SK std: ',np.round(np.std( AUCSK),4)*100)
print('==============================================')
print('Average Balanced Accuracy: ', np.round( BACC.mean(),4)*100)
print('Average std: ',np.round(np.std( BACC),4)*100)
print('==============================================')
print('Average MI: ', np.round( MI.mean(),4)*100)
print('Average std: ',np.round(np.std(MI),4)*100)
print('==============================================')
print('Average Normalized MI: ', np.round( NMI.mean(),4)*100)
print('Average std: ',np.round(np.std(NMI),4)*100)
print('==============================================')
print('Average Ajdusted MI: ', np.round( AMI.mean(),4)*100)
print('Average std: ',np.round(np.std(AMI),4)*100)
Average Accuracy: 74.15
Average std: 0.58
==============================================
Average AUC: 73.88
Average AUC std: 0.61
==============================================
Average AUC Sklearn: 82.99
Average AUC SK std: 0.3
==============================================
Average Balanced Accuracy: 48.24
Average std: 0.5499999999999999
==============================================
Average MI: 13.889999999999999
Average std: 0.22999999999999998
==============================================
Average Normalized MI: 20.79
Average std: 0.33999999999999997
==============================================
Average Ajdusted MI: 20.79
Average std: 0.33999999999999997
[ ]: