[Classification] e-mail 분류하기

Author : tmlab / Date : 2017. 2. 17. 14:58 / Category : Text Mining/Python

from __future__ import division
import tensorflow as tf
import numpy as np
import tarfile
import os
import matplotlib.pyplot as plt
import time

DATA import

data.tar 다운받아서 쥬피터에서 업로드하심 됩니다

In [2]:

###################
### IMPORT DATA ###
###################

def csv_to_numpy_array(filePath, delimiter):
    return np.genfromtxt(filePath, delimiter=delimiter, dtype=None)

def import_data():
    if "data" not in os.listdir(os.getcwd()):
        # Untar directory of data if we haven't already
        tarObject = tarfile.open("data.tar.gz")
        tarObject.extractall()
        tarObject.close()
        print("Extracted tar to current directory")
    else:
        # we've already extracted the files
        pass

    print("loading training data")
    trainX = csv_to_numpy_array("data/trainX.csv", delimiter="\t")
    trainY = csv_to_numpy_array("data/trainY.csv", delimiter="\t")
    print("loading test data")
    testX = csv_to_numpy_array("data/testX.csv", delimiter="\t")
    testY = csv_to_numpy_array("data/testY.csv", delimiter="\t")
    return trainX,trainY,testX,testY

trainX,trainY,testX,testY = import_data()

Extracted tar to current directory
loading training data
loading test data

DATA는 메일의 내용을 단어단위로 구분하여 그 빈도를 벡터화한 메트릭스
trainX 는 메트릭스
trainY 는 햄, 스팸여부에 대한 결과값(1,0 은 스팸메일 / 0,1은 햄메일)

In [4]:

trainX

Out[4]:

array([[ 0.        ,  0.        ,  0.        , ...,  0.        ,
         0.        ,  0.        ],
       [ 0.        ,  0.        ,  0.        , ...,  0.        ,
         0.        ,  0.        ],
       [ 0.        ,  0.        ,  0.        , ...,  0.        ,
         0.        ,  0.        ],
       ..., 
       [ 0.        ,  0.        ,  0.        , ...,  0.        ,
         0.        ,  0.        ],
       [ 0.        ,  0.        ,  0.        , ...,  0.        ,
         0.        ,  0.        ],
       [ 0.01425178,  0.        ,  0.        , ...,  0.        ,
         0.        ,  0.        ]])

In [5]:

trainY

Out[5]:

array([[ 1.,  0.],
       [ 1.,  0.],
       [ 0.,  1.],
       ..., 
       [ 1.,  0.],
       [ 0.,  1.],
       [ 1.,  0.]])

학습 파라미터 세팅

In [6]:

#########################
### GLOBAL PARAMETERS ###
#########################

# DATA SET PARAMETERS
# Get our dimensions for our different variables and placeholders:
# numFeatures = the number of words extracted from each email
numFeatures = trainX.shape[1]
# numLabels = number of classes we are predicting (here just 2: Ham or Spam)
numLabels = trainY.shape[1]

# TRAINING SESSION PARAMETERS
# number of times we iterate through training data
# tensorboard shows that accuracy plateaus at ~25k epochs
numEpochs = 27000
# a smarter learning rate for gradientOptimizer
learningRate = tf.train.exponential_decay(learning_rate=0.0008,
                                          global_step= 1,
                                          decay_steps=trainX.shape[0],
                                          decay_rate= 0.95,
                                          staircase=True)

In [8]:

trainX.shape[1]  #컬럼수 : 피쳐로 사용한 단어의 개수

Out[8]:

In [9]:

trainY.shape[1] #스팸, 햄 구분 : 0,1 or 1.0

Out[9]:

TensorFlow Placeholders

Placeholders 데이터를 삽입하는 빈 쉘 (빈 텐서)로 생각할 수 있음
따라서 Placeholder를 정의 할 때 데이터의 모양과 일치하는 모양을 제공해야함
TensorFlow는 Placeholder에 "feed_dict"를 통해 데이터를 공급함
이것은 '배치'데이터를 트레이닝시킬때 쉽게 조정할수 있고, 다양한 샘플들을 학습시킬때 전체 알고리즘이 아닌 일부단계만 수정하여 분석이 가능하게 함

In [10]:

####################
### PLACEHOLDERS ###
####################

# X = X-matrix / feature-matrix / data-matrix... It's a tensor to hold our email
# data. 'None' here means that we can hold any number of emails
X = tf.placeholder(tf.float32, [None, numFeatures])
# yGold = Y-matrix / label-matrix / labels... This will be our correct answers
# matrix. Every row has either [1,0] for SPAM or [0,1] for HAM. 'None' here 
# means that we can hold any number of emails
yGold = tf.placeholder(tf.float32, [None, numLabels])

TensorFlow variables

변수는 로지스틱 회귀 분석의 가중치 및 편차를 담고있으며 학습을 통해 지속적으로 업데이트됨
변수는 로지스틱 회귀모델에서 정의한 객체로 학습후에 저장되어 재사용할 수 있다.
TensorFlow의 모든 데이터와 마찬가지로 변수는 텐서로 표시됩니다. 그러나placeholder와 달리 TensorFlow 변수는 초기화해야합니다.
변수초기화를 위해 tf.random_normal를 사용하여 정규 분포에서 샘플을 추출하여 임의의 값으로 텐서 형태의 "모양"을 채웁니다.

In [11]:

#################
### VARIABLES ###
#################

# Values are randomly sampled from a Gaussian with a standard deviation of:
#     sqrt(6 / (numInputNodes + numOutputNodes + 1))

weights = tf.Variable(tf.random_normal([numFeatures,numLabels],
                                       mean=0,
                                       stddev=(np.sqrt(6/numFeatures+
                                                         numLabels+1)),
                                       name="weights"))

bias = tf.Variable(tf.random_normal([1,numLabels],
                                    mean=0,
                                    stddev=(np.sqrt(6/numFeatures+numLabels+1)),
                                    name="bias"))

오퍼레이션 : 가중치곱 / 바이어스합/ 시그모이드

In [16]:

######################
### PREDICTION OPS ###
######################

# INITIALIZE our weights and biases
init_OP = tf.initialize_all_variables()

# PREDICTION ALGORITHM i.e. FEEDFORWARD ALGORITHM
apply_weights_OP = tf.matmul(X, weights, name="apply_weights")
add_bias_OP = tf.add(apply_weights_OP, bias, name="add_bias") 
activation_OP = tf.nn.sigmoid(add_bias_OP, name="activation")

In [17]:

#####################
### EVALUATION OP ###
#####################

# COST FUNCTION i.e. MEAN SQUARED ERROR
cost_OP = tf.nn.l2_loss(activation_OP-yGold, name="squared_error_cost")

In [18]:

#######################
### OPTIMIZATION OP ###
#######################

# OPTIMIZATION ALGORITHM i.e. GRADIENT DESCENT
training_OP = tf.train.GradientDescentOptimizer(learningRate).minimize(cost_OP)

In [19]:

###########################
### GRAPH LIVE UPDATING ###
###########################

epoch_values=[]
accuracy_values=[]
cost_values=[]
# Turn on interactive plotting
plt.ion()
# Create the main, super plot
fig = plt.figure()
# Create two subplots on their own axes and give titles
ax1 = plt.subplot("211")
ax1.set_title("TRAINING ACCURACY", fontsize=18)
ax2 = plt.subplot("212")
ax2.set_title("TRAINING COST", fontsize=18)
plt.tight_layout()

In [20]:

#####################
### RUN THE GRAPH ###
#####################

# Create a tensorflow session
sess = tf.Session()

# Initialize all tensorflow variables
sess.run(init_OP)

## Ops for vizualization
# argmax(activation_OP, 1) gives the label our model thought was most likely
# argmax(yGold, 1) is the correct label
correct_predictions_OP = tf.equal(tf.argmax(activation_OP,1),tf.argmax(yGold,1))
# False is 0 and True is 1, what was our average?
accuracy_OP = tf.reduce_mean(tf.cast(correct_predictions_OP, "float"))
# Summary op for regression output
activation_summary_OP = tf.histogram_summary("output", activation_OP)
# Summary op for accuracy
accuracy_summary_OP = tf.scalar_summary("accuracy", accuracy_OP)
# Summary op for cost
cost_summary_OP = tf.scalar_summary("cost", cost_OP)
# Summary ops to check how variables (W, b) are updating after each iteration
weightSummary = tf.histogram_summary("weights", weights.eval(session=sess))
biasSummary = tf.histogram_summary("biases", bias.eval(session=sess))
# Merge all summaries
all_summary_OPS = tf.merge_all_summaries()
# Summary writer
writer = tf.train.SummaryWriter("summary_logs", sess.graph_def)

# Initialize reporting variables
cost = 0
diff = 1

# Training epochs
for i in range(numEpochs):
    if i > 1 and diff < .0001:
        print("change in cost %g; convergence."%diff)
        break
    else:
        # Run training step
        step = sess.run(training_OP, feed_dict={X: trainX, yGold: trainY})
        # Report occasional stats
        if i % 10 == 0:
            # Add epoch to epoch_values
            epoch_values.append(i)
            # Generate accuracy stats on test data
            summary_results, train_accuracy, newCost = sess.run(
                [all_summary_OPS, accuracy_OP, cost_OP], 
                feed_dict={X: trainX, yGold: trainY}
            )
            # Add accuracy to live graphing variable
            accuracy_values.append(train_accuracy)
            # Add cost to live graphing variable
            cost_values.append(newCost)
            # Write summary stats to writer
            writer.add_summary(summary_results, i)
            # Re-assign values for variables
            diff = abs(newCost - cost)
            cost = newCost

            #generate print statements
            print("step %d, training accuracy %g"%(i, train_accuracy))
            print("step %d, cost %g"%(i, newCost))
            print("step %d, change in cost %g"%(i, diff))

            # Plot progress to our two subplots
            accuracyLine, = ax1.plot(epoch_values, accuracy_values)
            costLine, = ax2.plot(epoch_values, cost_values)
            fig.canvas.draw()
            time.sleep(1)


# How well do we perform on held-out test data?
print("final accuracy on test set: %s" %str(sess.run(accuracy_OP, 
                                                     feed_dict={X: testX, 
                                                                yGold: testY})))

WARNING:tensorflow:Passing a `GraphDef` to the SummaryWriter is deprecated. Pass a `Graph` object instead, such as `sess.graph`.

step 0, training accuracy 0.533054
step 0, cost 335.832
step 0, change in cost 335.832
step 10, training accuracy 0.533054
step 10, cost 335.086
step 10, change in cost 0.745667
step 20, training accuracy 0.533054
step 20, cost 334.345
step 20, change in cost 0.740997
step 30, training accuracy 0.533054
step 30, cost 333.543
step 30, change in cost 0.802002
step 40, training accuracy 0.533054
step 40, cost 332.64
step 40, change in cost 0.90332
step 50, training accuracy 0.533054
step 50, cost 331.6
step 50, change in cost 1.04007
step 60, training accuracy 0.533054
step 60, cost 330.383
step 60, change in cost 1.21613
step 70, training accuracy 0.533054
step 70, cost 328.942
step 70, change in cost 1.44186
step 80, training accuracy 0.533054
step 80, cost 327.207
step 80, change in cost 1.73477
step 90, training accuracy 0.533054
step 90, cost 325.085
step 90, change in cost 2.12146
step 100, training accuracy 0.533054
step 100, cost 322.445
step 100, change in cost 2.64017
step 110, training accuracy 0.533054
step 110, cost 319.095
step 110, change in cost 3.34979
step 120, training accuracy 0.533054
step 120, cost 314.76
step 120, change in cost 4.33563
step 130, training accuracy 0.533054
step 130, cost 309.043
step 130, change in cost 5.71634
step 140, training accuracy 0.533054
step 140, cost 301.42
step 140, change in cost 7.62329
step 150, training accuracy 0.533054
step 150, cost 291.328
step 150, change in cost 10.0921
step 160, training accuracy 0.533054
step 160, cost 278.606
step 160, change in cost 12.722
step 170, training accuracy 0.533054
step 170, cost 264.424
step 170, change in cost 14.1816
step 180, training accuracy 0.533054
step 180, cost 251.689
step 180, change in cost 12.7355
step 190, training accuracy 0.533054
step 190, cost 242.986
step 190, change in cost 8.70335
step 200, training accuracy 0.533054
step 200, cost 238.334
step 200, change in cost 4.65115
step 210, training accuracy 0.544596
step 210, cost 236.213
step 210, change in cost 2.12172
step 220, training accuracy 0.551941
step 220, cost 235.305
step 220, change in cost 0.90773
step 230, training accuracy 0.558237
step 230, cost 234.907
step 230, change in cost 0.397812
step 240, training accuracy 0.570829
step 240, cost 234.711
step 240, change in cost 0.196609
step 250, training accuracy 0.56978
step 250, cost 234.591
step 250, change in cost 0.119858
step 260, training accuracy 0.573977
step 260, cost 234.5
step 260, change in cost 0.0906677
step 270, training accuracy 0.575026
step 270, cost 234.42
step 270, change in cost 0.0798492
step 280, training accuracy 0.578174
step 280, cost 234.345
step 280, change in cost 0.0756226
step 290, training accuracy 0.579224
step 290, cost 234.27
step 290, change in cost 0.0743256
step 300, training accuracy 0.580273
step 300, cost 234.197
step 300, change in cost 0.0735474
step 310, training accuracy 0.582371
step 310, cost 234.123
step 310, change in cost 0.0733948
step 320, training accuracy 0.58447
step 320, cost 234.05
step 320, change in cost 0.0731812
step 330, training accuracy 0.58447
step 330, cost 233.977
step 330, change in cost 0.0732727
step 340, training accuracy 0.58447
step 340, cost 233.904
step 340, change in cost 0.0730743
step 350, training accuracy 0.58447
step 350, cost 233.831
step 350, change in cost 0.0730743
step 360, training accuracy 0.585519
step 360, cost 233.758
step 360, change in cost 0.0729828
step 370, training accuracy 0.586569
step 370, cost 233.685
step 370, change in cost 0.072998
step 380, training accuracy 0.586569
step 380, cost 233.612
step 380, change in cost 0.0729828
step 390, training accuracy 0.588667
step 390, cost 233.539
step 390, change in cost 0.0729218
step 400, training accuracy 0.588667
step 400, cost 233.466
step 400, change in cost 0.0728302
step 410, training accuracy 0.588667
step 410, cost 233.393
step 410, change in cost 0.0728455
step 420, training accuracy 0.588667
step 420, cost 233.32
step 420, change in cost 0.0727692
step 430, training accuracy 0.588667
step 430, cost 233.248
step 430, change in cost 0.0727539
step 440, training accuracy 0.588667
step 440, cost 233.175
step 440, change in cost 0.0727234
step 450, training accuracy 0.589717
step 450, cost 233.102
step 450, change in cost 0.0726624
step 460, training accuracy 0.589717
step 460, cost 233.03
step 460, change in cost 0.0726013
step 470, training accuracy 0.590766
step 470, cost 232.957
step 470, change in cost 0.0726013
step 480, training accuracy 0.591815
step 480, cost 232.884
step 480, change in cost 0.072525
step 490, training accuracy 0.591815
step 490, cost 232.812
step 490, change in cost 0.0724792
step 500, training accuracy 0.591815
step 500, cost 232.74
step 500, change in cost 0.0724945
step 510, training accuracy 0.591815
step 510, cost 232.667
step 510, change in cost 0.0724487
step 520, training accuracy 0.592865
step 520, cost 232.595
step 520, change in cost 0.0723419
step 530, training accuracy 0.592865
step 530, cost 232.522
step 530, change in cost 0.0723877
step 540, training accuracy 0.592865
step 540, cost 232.45
step 540, change in cost 0.0722961
step 550, training accuracy 0.592865
step 550, cost 232.378
step 550, change in cost 0.0722351
step 560, training accuracy 0.593914
step 560, cost 232.306
step 560, change in cost 0.0722656
step 570, training accuracy 0.593914
step 570, cost 232.233
step 570, change in cost 0.0722504
step 580, training accuracy 0.594963
step 580, cost 232.161
step 580, change in cost 0.0720978
step 590, training accuracy 0.594963
step 590, cost 232.089
step 590, change in cost 0.0721436
step 600, training accuracy 0.596013
step 600, cost 232.017
step 600, change in cost 0.0720673
step 610, training accuracy 0.596013
step 610, cost 231.945
step 610, change in cost 0.0720215
step 620, training accuracy 0.596013
step 620, cost 231.873
step 620, change in cost 0.0719757
step 630, training accuracy 0.596013
step 630, cost 231.801
step 630, change in cost 0.0719147
step 640, training accuracy 0.596013
step 640, cost 231.729
step 640, change in cost 0.0718994
step 650, training accuracy 0.597062
step 650, cost 231.657
step 650, change in cost 0.0718842
step 660, training accuracy 0.597062
step 660, cost 231.585
step 660, change in cost 0.0718994
step 670, training accuracy 0.597062
step 670, cost 231.514
step 670, change in cost 0.0717316
step 680, training accuracy 0.598111
step 680, cost 231.442
step 680, change in cost 0.0717468
step 690, training accuracy 0.60021
step 690, cost 231.37
step 690, change in cost 0.0717316
step 700, training accuracy 0.60021
step 700, cost 231.298
step 700, change in cost 0.0717468
step 710, training accuracy 0.60021
step 710, cost 231.227
step 710, change in cost 0.071579
step 720, training accuracy 0.60021
step 720, cost 231.155
step 720, change in cost 0.071579
step 730, training accuracy 0.60021
step 730, cost 231.084
step 730, change in cost 0.0715637
step 740, training accuracy 0.60021
step 740, cost 231.012
step 740, change in cost 0.0715485
step 750, training accuracy 0.60021
step 750, cost 230.941
step 750, change in cost 0.0714264
step 760, training accuracy 0.60021
step 760, cost 230.869
step 760, change in cost 0.0714874
step 770, training accuracy 0.60021
step 770, cost 230.798
step 770, change in cost 0.0713959
step 780, training accuracy 0.60021
step 780, cost 230.727
step 780, change in cost 0.0712738
step 790, training accuracy 0.602309
step 790, cost 230.655
step 790, change in cost 0.0714111
step 800, training accuracy 0.602309
step 800, cost 230.584
step 800, change in cost 0.0712433
step 810, training accuracy 0.603358
step 810, cost 230.513
step 810, change in cost 0.071228
step 820, training accuracy 0.603358
step 820, cost 230.441
step 820, change in cost 0.0711823
step 830, training accuracy 0.603358
step 830, cost 230.37
step 830, change in cost 0.0711365
step 840, training accuracy 0.604407
step 840, cost 230.299
step 840, change in cost 0.0712585
step 850, training accuracy 0.605456
step 850, cost 230.228
step 850, change in cost 0.0709534
step 860, training accuracy 0.606506
step 860, cost 230.157
step 860, change in cost 0.0711212
step 870, training accuracy 0.606506
step 870, cost 230.086
step 870, change in cost 0.0709991
step 880, training accuracy 0.607555
step 880, cost 230.015
step 880, change in cost 0.0709839
step 890, training accuracy 0.607555
step 890, cost 229.944
step 890, change in cost 0.0708008
step 900, training accuracy 0.607555
step 900, cost 229.873
step 900, change in cost 0.0709534
step 910, training accuracy 0.607555
step 910, cost 229.802
step 910, change in cost 0.0708466
step 920, training accuracy 0.608604
step 920, cost 229.732
step 920, change in cost 0.0707397
step 930, training accuracy 0.608604
step 930, cost 229.661
step 930, change in cost 0.0708008
step 940, training accuracy 0.608604
step 940, cost 229.59
step 940, change in cost 0.070694
step 950, training accuracy 0.608604
step 950, cost 229.52
step 950, change in cost 0.070694
step 960, training accuracy 0.608604
step 960, cost 229.449
step 960, change in cost 0.0707092
step 970, training accuracy 0.608604
step 970, cost 229.378
step 970, change in cost 0.0706177
step 980, training accuracy 0.610703
step 980, cost 229.308
step 980, change in cost 0.0704803
step 990, training accuracy 0.610703
step 990, cost 229.237
step 990, change in cost 0.0705566
step 1000, training accuracy 0.611752
step 1000, cost 229.167
step 1000, change in cost 0.0704956
step 1010, training accuracy 0.612802
step 1010, cost 229.096
step 1010, change in cost 0.0705109
step 1020, training accuracy 0.61595
step 1020, cost 229.026
step 1020, change in cost 0.0704498
step 1030, training accuracy 0.616999
step 1030, cost 228.955
step 1030, change in cost 0.0703125
step 1040, training accuracy 0.618048
step 1040, cost 228.885
step 1040, change in cost 0.0703278
step 1050, training accuracy 0.618048
step 1050, cost 228.815
step 1050, change in cost 0.070282
step 1060, training accuracy 0.619098
step 1060, cost 228.744
step 1060, change in cost 0.070282
step 1070, training accuracy 0.619098
step 1070, cost 228.674
step 1070, change in cost 0.0701599
step 1080, training accuracy 0.619098
step 1080, cost 228.604
step 1080, change in cost 0.0702362
step 1090, training accuracy 0.619098
step 1090, cost 228.534
step 1090, change in cost 0.0700531
step 1100, training accuracy 0.619098
step 1100, cost 228.464
step 1100, change in cost 0.0700836
step 1110, training accuracy 0.619098
step 1110, cost 228.394
step 1110, change in cost 0.0700836
step 1120, training accuracy 0.619098
step 1120, cost 228.324
step 1120, change in cost 0.0701141
step 1130, training accuracy 0.619098
step 1130, cost 228.254
step 1130, change in cost 0.0699158
step 1140, training accuracy 0.620147
step 1140, cost 228.184
step 1140, change in cost 0.0699005
step 1150, training accuracy 0.620147
step 1150, cost 228.114
step 1150, change in cost 0.0699005
step 1160, training accuracy 0.620147
step 1160, cost 228.044
step 1160, change in cost 0.0699158
step 1170, training accuracy 0.621196
step 1170, cost 227.974
step 1170, change in cost 0.0697632
step 1180, training accuracy 0.622246
step 1180, cost 227.904
step 1180, change in cost 0.0698853
step 1190, training accuracy 0.623295
step 1190, cost 227.835
step 1190, change in cost 0.0697174
step 1200, training accuracy 0.624344
step 1200, cost 227.765
step 1200, change in cost 0.0697021
step 1210, training accuracy 0.624344
step 1210, cost 227.695
step 1210, change in cost 0.0695953
step 1220, training accuracy 0.624344
step 1220, cost 227.626
step 1220, change in cost 0.0696869
step 1230, training accuracy 0.624344
step 1230, cost 227.556
step 1230, change in cost 0.0695496
step 1240, training accuracy 0.624344
step 1240, cost 227.487
step 1240, change in cost 0.0695953
step 1250, training accuracy 0.626443
step 1250, cost 227.417
step 1250, change in cost 0.0695496
step 1260, training accuracy 0.626443
step 1260, cost 227.348
step 1260, change in cost 0.0694733
step 1270, training accuracy 0.627492
step 1270, cost 227.278
step 1270, change in cost 0.0694122
step 1280, training accuracy 0.628541
step 1280, cost 227.209
step 1280, change in cost 0.0693817
step 1290, training accuracy 0.628541
step 1290, cost 227.139
step 1290, change in cost 0.0693207
step 1300, training accuracy 0.628541
step 1300, cost 227.07
step 1300, change in cost 0.0693512
step 1310, training accuracy 0.629591
step 1310, cost 227.001
step 1310, change in cost 0.0693054
step 1320, training accuracy 0.63064
step 1320, cost 226.932
step 1320, change in cost 0.0692139
step 1330, training accuracy 0.63064
step 1330, cost 226.862
step 1330, change in cost 0.0691986
step 1340, training accuracy 0.63064
step 1340, cost 226.793
step 1340, change in cost 0.0691528
step 1350, training accuracy 0.631689
step 1350, cost 226.724
step 1350, change in cost 0.0691071
step 1360, training accuracy 0.631689
step 1360, cost 226.655
step 1360, change in cost 0.0691223
step 1370, training accuracy 0.632739
step 1370, cost 226.586
step 1370, change in cost 0.069046
step 1380, training accuracy 0.631689
step 1380, cost 226.517
step 1380, change in cost 0.069046
step 1390, training accuracy 0.632739
step 1390, cost 226.448
step 1390, change in cost 0.0689087
step 1400, training accuracy 0.633788
step 1400, cost 226.379
step 1400, change in cost 0.068924
step 1410, training accuracy 0.634837
step 1410, cost 226.31
step 1410, change in cost 0.0689697
step 1420, training accuracy 0.635887
step 1420, cost 226.241
step 1420, change in cost 0.0688171
step 1430, training accuracy 0.634837
step 1430, cost 226.173
step 1430, change in cost 0.0688019
step 1440, training accuracy 0.634837
step 1440, cost 226.104
step 1440, change in cost 0.0688171
step 1450, training accuracy 0.634837
step 1450, cost 226.035
step 1450, change in cost 0.0687256
step 1460, training accuracy 0.634837
step 1460, cost 225.966
step 1460, change in cost 0.068634
step 1470, training accuracy 0.635887
step 1470, cost 225.898
step 1470, change in cost 0.0687408
step 1480, training accuracy 0.635887
step 1480, cost 225.829
step 1480, change in cost 0.068634
step 1490, training accuracy 0.635887
step 1490, cost 225.761
step 1490, change in cost 0.0684509
step 1500, training accuracy 0.635887
step 1500, cost 225.692
step 1500, change in cost 0.0686188
step 1510, training accuracy 0.636936
step 1510, cost 225.623
step 1510, change in cost 0.0684814
step 1520, training accuracy 0.636936
step 1520, cost 225.555
step 1520, change in cost 0.0685425
step 1530, training accuracy 0.639035
step 1530, cost 225.486
step 1530, change in cost 0.0683899
step 1540, training accuracy 0.639035
step 1540, cost 225.418
step 1540, change in cost 0.0684052
step 1550, training accuracy 0.640084
step 1550, cost 225.35
step 1550, change in cost 0.0683441
step 1560, training accuracy 0.640084
step 1560, cost 225.281
step 1560, change in cost 0.0682831
step 1570, training accuracy 0.642183
step 1570, cost 225.213
step 1570, change in cost 0.0682373
step 1580, training accuracy 0.642183
step 1580, cost 225.145
step 1580, change in cost 0.0682983
step 1590, training accuracy 0.643232
step 1590, cost 225.077
step 1590, change in cost 0.068222
step 1600, training accuracy 0.643232
step 1600, cost 225.009
step 1600, change in cost 0.068161
step 1610, training accuracy 0.644281
step 1610, cost 224.94
step 1610, change in cost 0.0680542
step 1620, training accuracy 0.645331
step 1620, cost 224.872
step 1620, change in cost 0.0681458
step 1630, training accuracy 0.64638
step 1630, cost 224.804
step 1630, change in cost 0.0680542
step 1640, training accuracy 0.647429
step 1640, cost 224.736
step 1640, change in cost 0.0679932
step 1650, training accuracy 0.648479
step 1650, cost 224.668
step 1650, change in cost 0.0680084
step 1660, training accuracy 0.650577
step 1660, cost 224.6
step 1660, change in cost 0.0679321
step 1670, training accuracy 0.650577
step 1670, cost 224.532
step 1670, change in cost 0.0679016
step 1680, training accuracy 0.650577
step 1680, cost 224.465
step 1680, change in cost 0.0677795
step 1690, training accuracy 0.651626
step 1690, cost 224.397
step 1690, change in cost 0.0678101
step 1700, training accuracy 0.651626
step 1700, cost 224.329
step 1700, change in cost 0.0678406
step 1710, training accuracy 0.652676
step 1710, cost 224.261
step 1710, change in cost 0.0677032
step 1720, training accuracy 0.652676
step 1720, cost 224.194
step 1720, change in cost 0.067749
step 1730, training accuracy 0.653725
step 1730, cost 224.126
step 1730, change in cost 0.0676117
step 1740, training accuracy 0.653725
step 1740, cost 224.058
step 1740, change in cost 0.0676117
step 1750, training accuracy 0.654774
step 1750, cost 223.991
step 1750, change in cost 0.0676422
step 1760, training accuracy 0.654774
step 1760, cost 223.923
step 1760, change in cost 0.0675659
step 1770, training accuracy 0.655824
step 1770, cost 223.856
step 1770, change in cost 0.0675049
step 1780, training accuracy 0.656873
step 1780, cost 223.788
step 1780, change in cost 0.0673828
step 1790, training accuracy 0.656873
step 1790, cost 223.721
step 1790, change in cost 0.0674438
step 1800, training accuracy 0.656873
step 1800, cost 223.653
step 1800, change in cost 0.0674133
step 1810, training accuracy 0.657922
step 1810, cost 223.586
step 1810, change in cost 0.0673676
step 1820, training accuracy 0.657922
step 1820, cost 223.519
step 1820, change in cost 0.0673523
step 1830, training accuracy 0.657922
step 1830, cost 223.451
step 1830, change in cost 0.0672607
step 1840, training accuracy 0.657922
step 1840, cost 223.384
step 1840, change in cost 0.0672607
step 1850, training accuracy 0.660021
step 1850, cost 223.317
step 1850, change in cost 0.067215
step 1860, training accuracy 0.66107
step 1860, cost 223.25
step 1860, change in cost 0.0671844
step 1870, training accuracy 0.66212
step 1870, cost 223.183
step 1870, change in cost 0.0671234
step 1880, training accuracy 0.663169
step 1880, cost 223.116
step 1880, change in cost 0.0670471
step 1890, training accuracy 0.663169
step 1890, cost 223.049
step 1890, change in cost 0.0670013
step 1900, training accuracy 0.664218
step 1900, cost 222.981
step 1900, change in cost 0.0670929
step 1910, training accuracy 0.664218
step 1910, cost 222.914
step 1910, change in cost 0.0670013
step 1920, training accuracy 0.666317
step 1920, cost 222.848
step 1920, change in cost 0.0669708
step 1930, training accuracy 0.668416
step 1930, cost 222.781
step 1930, change in cost 0.0667725
step 1940, training accuracy 0.668416
step 1940, cost 222.714
step 1940, change in cost 0.0669098
step 1950, training accuracy 0.669465
step 1950, cost 222.647
step 1950, change in cost 0.0668793
step 1960, training accuracy 0.670514
step 1960, cost 222.58
step 1960, change in cost 0.0666962
step 1970, training accuracy 0.670514
step 1970, cost 222.513
step 1970, change in cost 0.066803
step 1980, training accuracy 0.670514
step 1980, cost 222.447
step 1980, change in cost 0.0667267
step 1990, training accuracy 0.670514
step 1990, cost 222.38
step 1990, change in cost 0.0666046
step 2000, training accuracy 0.671564
step 2000, cost 222.313
step 2000, change in cost 0.0667419
step 2010, training accuracy 0.673662
step 2010, cost 222.247
step 2010, change in cost 0.0665588
step 2020, training accuracy 0.673662
step 2020, cost 222.18
step 2020, change in cost 0.0665894
step 2030, training accuracy 0.673662
step 2030, cost 222.114
step 2030, change in cost 0.0665131
step 2040, training accuracy 0.673662
step 2040, cost 222.047
step 2040, change in cost 0.0664978
step 2050, training accuracy 0.673662
step 2050, cost 221.981
step 2050, change in cost 0.0664673
step 2060, training accuracy 0.675761
step 2060, cost 221.914
step 2060, change in cost 0.0664368
step 2070, training accuracy 0.675761
step 2070, cost 221.848
step 2070, change in cost 0.0663452
step 2080, training accuracy 0.675761
step 2080, cost 221.782
step 2080, change in cost 0.0663757
step 2090, training accuracy 0.675761
step 2090, cost 221.715
step 2090, change in cost 0.0662689
step 2100, training accuracy 0.67681
step 2100, cost 221.649
step 2100, change in cost 0.0662842
step 2110, training accuracy 0.677859
step 2110, cost 221.583
step 2110, change in cost 0.0662079
step 2120, training accuracy 0.677859
step 2120, cost 221.517
step 2120, change in cost 0.0661926
step 2130, training accuracy 0.677859
step 2130, cost 221.451
step 2130, change in cost 0.0661316
step 2140, training accuracy 0.677859
step 2140, cost 221.384
step 2140, change in cost 0.0661316
step 2150, training accuracy 0.677859
step 2150, cost 221.318
step 2150, change in cost 0.0660248
step 2160, training accuracy 0.677859
step 2160, cost 221.252
step 2160, change in cost 0.0660706
step 2170, training accuracy 0.678909
step 2170, cost 221.186
step 2170, change in cost 0.06604
step 2180, training accuracy 0.679958
step 2180, cost 221.12
step 2180, change in cost 0.065918
step 2190, training accuracy 0.681007
step 2190, cost 221.054
step 2190, change in cost 0.0660095
step 2200, training accuracy 0.682057
step 2200, cost 220.988
step 2200, change in cost 0.0658569
step 2210, training accuracy 0.682057
step 2210, cost 220.923
step 2210, change in cost 0.0658264
step 2220, training accuracy 0.683106
step 2220, cost 220.857
step 2220, change in cost 0.0657806
step 2230, training accuracy 0.685205
step 2230, cost 220.791
step 2230, change in cost 0.0658417
step 2240, training accuracy 0.685205
step 2240, cost 220.725
step 2240, change in cost 0.0657806
step 2250, training accuracy 0.685205
step 2250, cost 220.66
step 2250, change in cost 0.0656586
step 2260, training accuracy 0.686254
step 2260, cost 220.594
step 2260, change in cost 0.0656128
step 2270, training accuracy 0.686254
step 2270, cost 220.528
step 2270, change in cost 0.0656281
step 2280, training accuracy 0.686254
step 2280, cost 220.463
step 2280, change in cost 0.0656128
step 2290, training accuracy 0.687303
step 2290, cost 220.397
step 2290, change in cost 0.0655975
step 2300, training accuracy 0.687303
step 2300, cost 220.332
step 2300, change in cost 0.0655518
step 2310, training accuracy 0.687303
step 2310, cost 220.266
step 2310, change in cost 0.0654449
step 2320, training accuracy 0.687303
step 2320, cost 220.201
step 2320, change in cost 0.0654144
step 2330, training accuracy 0.688353
step 2330, cost 220.135
step 2330, change in cost 0.0653992
step 2340, training accuracy 0.689402
step 2340, cost 220.07
step 2340, change in cost 0.0654297
step 2350, training accuracy 0.689402
step 2350, cost 220.005
step 2350, change in cost 0.0652924
step 2360, training accuracy 0.689402
step 2360, cost 219.939
step 2360, change in cost 0.0653076
step 2370, training accuracy 0.691501
step 2370, cost 219.874
step 2370, change in cost 0.0652313
step 2380, training accuracy 0.69255
step 2380, cost 219.809
step 2380, change in cost 0.0652466
step 2390, training accuracy 0.691501
step 2390, cost 219.744
step 2390, change in cost 0.065155
step 2400, training accuracy 0.690451
step 2400, cost 219.678
step 2400, change in cost 0.0651855
step 2410, training accuracy 0.691501
step 2410, cost 219.613
step 2410, change in cost 0.065033
step 2420, training accuracy 0.69255
step 2420, cost 219.548
step 2420, change in cost 0.065094
step 2430, training accuracy 0.69255
step 2430, cost 219.483
step 2430, change in cost 0.065094
step 2440, training accuracy 0.693599
step 2440, cost 219.418
step 2440, change in cost 0.0650177
step 2450, training accuracy 0.694649
step 2450, cost 219.353
step 2450, change in cost 0.0648956
step 2460, training accuracy 0.694649
step 2460, cost 219.288
step 2460, change in cost 0.065033
step 2470, training accuracy 0.694649
step 2470, cost 219.223
step 2470, change in cost 0.0648804
step 2480, training accuracy 0.694649
step 2480, cost 219.159
step 2480, change in cost 0.0647888
step 2490, training accuracy 0.694649
step 2490, cost 219.094
step 2490, change in cost 0.0648193
step 2500, training accuracy 0.695698
step 2500, cost 219.029
step 2500, change in cost 0.0647888
step 2510, training accuracy 0.697796
step 2510, cost 218.964
step 2510, change in cost 0.064682
step 2520, training accuracy 0.697796
step 2520, cost 218.9
step 2520, change in cost 0.0647583
step 2530, training accuracy 0.699895
step 2530, cost 218.835
step 2530, change in cost 0.0646515
step 2540, training accuracy 0.700944
step 2540, cost 218.77
step 2540, change in cost 0.0646515
step 2550, training accuracy 0.700944
step 2550, cost 218.706
step 2550, change in cost 0.0645905
step 2560, training accuracy 0.700944
step 2560, cost 218.641
step 2560, change in cost 0.0645905
step 2570, training accuracy 0.700944
step 2570, cost 218.577
step 2570, change in cost 0.0644836
step 2580, training accuracy 0.703043
step 2580, cost 218.512
step 2580, change in cost 0.0644684
step 2590, training accuracy 0.701994
step 2590, cost 218.448
step 2590, change in cost 0.0645142
step 2600, training accuracy 0.701994
step 2600, cost 218.383
step 2600, change in cost 0.0643921
step 2610, training accuracy 0.703043
step 2610, cost 218.319
step 2610, change in cost 0.0643616
step 2620, training accuracy 0.703043
step 2620, cost 218.255
step 2620, change in cost 0.0643768
step 2630, training accuracy 0.703043
step 2630, cost 218.19
step 2630, change in cost 0.0642853
step 2640, training accuracy 0.704092
step 2640, cost 218.126
step 2640, change in cost 0.0642548
step 2650, training accuracy 0.705142
step 2650, cost 218.062
step 2650, change in cost 0.0642242
step 2660, training accuracy 0.706191
step 2660, cost 217.998
step 2660, change in cost 0.0641785
step 2670, training accuracy 0.70724
step 2670, cost 217.933
step 2670, change in cost 0.0641022
step 2680, training accuracy 0.70724
step 2680, cost 217.869
step 2680, change in cost 0.0641632
step 2690, training accuracy 0.70724
step 2690, cost 217.805
step 2690, change in cost 0.0641327
step 2700, training accuracy 0.70829
step 2700, cost 217.741
step 2700, change in cost 0.0640411
step 2710, training accuracy 0.70829
step 2710, cost 217.677
step 2710, change in cost 0.0639343
step 2720, training accuracy 0.70829
step 2720, cost 217.613
step 2720, change in cost 0.0640564
step 2730, training accuracy 0.70829
step 2730, cost 217.549
step 2730, change in cost 0.0639343
step 2740, training accuracy 0.709339
step 2740, cost 217.485
step 2740, change in cost 0.0638733
step 2750, training accuracy 0.709339
step 2750, cost 217.421
step 2750, change in cost 0.0638428
step 2760, training accuracy 0.712487
step 2760, cost 217.358
step 2760, change in cost 0.0638123
step 2770, training accuracy 0.712487
step 2770, cost 217.294
step 2770, change in cost 0.0638123
step 2780, training accuracy 0.712487
step 2780, cost 217.23
step 2780, change in cost 0.063736
step 2790, training accuracy 0.712487
step 2790, cost 217.166
step 2790, change in cost 0.0637512
step 2800, training accuracy 0.713536
step 2800, cost 217.103
step 2800, change in cost 0.0636292
step 2810, training accuracy 0.714586
step 2810, cost 217.039
step 2810, change in cost 0.0637207
step 2820, training accuracy 0.714586
step 2820, cost 216.975
step 2820, change in cost 0.0636139
step 2830, training accuracy 0.716684
step 2830, cost 216.912
step 2830, change in cost 0.0635681
step 2840, training accuracy 0.716684
step 2840, cost 216.848
step 2840, change in cost 0.0635834
step 2850, training accuracy 0.717734
step 2850, cost 216.785
step 2850, change in cost 0.0634918
step 2860, training accuracy 0.717734
step 2860, cost 216.721
step 2860, change in cost 0.0633545
step 2870, training accuracy 0.718783
step 2870, cost 216.658
step 2870, change in cost 0.0635529
step 2880, training accuracy 0.719832
step 2880, cost 216.594
step 2880, change in cost 0.0633545
step 2890, training accuracy 0.719832
step 2890, cost 216.531
step 2890, change in cost 0.0633545
step 2900, training accuracy 0.720881
step 2900, cost 216.468
step 2900, change in cost 0.0633392
step 2910, training accuracy 0.719832
step 2910, cost 216.405
step 2910, change in cost 0.0632782
step 2920, training accuracy 0.719832
step 2920, cost 216.341
step 2920, change in cost 0.0632629
step 2930, training accuracy 0.719832
step 2930, cost 216.278
step 2930, change in cost 0.0632019
step 2940, training accuracy 0.719832
step 2940, cost 216.215
step 2940, change in cost 0.0631561
step 2950, training accuracy 0.72298
step 2950, cost 216.152
step 2950, change in cost 0.0632019
step 2960, training accuracy 0.72298
step 2960, cost 216.089
step 2960, change in cost 0.0630493
step 2970, training accuracy 0.72298
step 2970, cost 216.026
step 2970, change in cost 0.0631409
step 2980, training accuracy 0.724029
step 2980, cost 215.963
step 2980, change in cost 0.0629578
step 2990, training accuracy 0.724029
step 2990, cost 215.9
step 2990, change in cost 0.0630035
step 3000, training accuracy 0.724029
step 3000, cost 215.837
step 3000, change in cost 0.0629425
step 3010, training accuracy 0.725079
step 3010, cost 215.774
step 3010, change in cost 0.0630951
step 3020, training accuracy 0.725079
step 3020, cost 215.711
step 3020, change in cost 0.0627747
step 3030, training accuracy 0.726128
step 3030, cost 215.648
step 3030, change in cost 0.0628662
step 3040, training accuracy 0.726128
step 3040, cost 215.585
step 3040, change in cost 0.062851
step 3050, training accuracy 0.727177
step 3050, cost 215.522
step 3050, change in cost 0.0627747
step 3060, training accuracy 0.727177
step 3060, cost 215.46
step 3060, change in cost 0.0627136
step 3070, training accuracy 0.727177
step 3070, cost 215.397
step 3070, change in cost 0.0627594
step 3080, training accuracy 0.728227
step 3080, cost 215.334
step 3080, change in cost 0.0627136
step 3090, training accuracy 0.730325
step 3090, cost 215.271
step 3090, change in cost 0.0626068
step 3100, training accuracy 0.731375
step 3100, cost 215.209
step 3100, change in cost 0.0626068
step 3110, training accuracy 0.731375
step 3110, cost 215.146
step 3110, change in cost 0.0625305
step 3120, training accuracy 0.731375
step 3120, cost 215.084
step 3120, change in cost 0.0625305
step 3130, training accuracy 0.731375
step 3130, cost 215.021
step 3130, change in cost 0.0624847
step 3140, training accuracy 0.733473
step 3140, cost 214.959
step 3140, change in cost 0.0625763
step 3150, training accuracy 0.733473
step 3150, cost 214.896
step 3150, change in cost 0.0623322
step 3160, training accuracy 0.733473
step 3160, cost 214.834
step 3160, change in cost 0.0623932
step 3170, training accuracy 0.734523
step 3170, cost 214.772
step 3170, change in cost 0.0623474
step 3180, training accuracy 0.734523
step 3180, cost 214.709
step 3180, change in cost 0.0623474
step 3190, training accuracy 0.734523
step 3190, cost 214.647
step 3190, change in cost 0.0623016
step 3200, training accuracy 0.735572
step 3200, cost 214.585
step 3200, change in cost 0.0622559
step 3210, training accuracy 0.737671
step 3210, cost 214.523
step 3210, change in cost 0.0621948
step 3220, training accuracy 0.73872
step 3220, cost 214.46
step 3220, change in cost 0.0622253
step 3230, training accuracy 0.739769
step 3230, cost 214.398
step 3230, change in cost 0.0621033
step 3240, training accuracy 0.740818
step 3240, cost 214.336
step 3240, change in cost 0.0621185
step 3250, training accuracy 0.740818
step 3250, cost 214.274
step 3250, change in cost 0.0620117
step 3260, training accuracy 0.742917
step 3260, cost 214.212
step 3260, change in cost 0.062149
step 3270, training accuracy 0.742917
step 3270, cost 214.15
step 3270, change in cost 0.0619965
step 3280, training accuracy 0.743966
step 3280, cost 214.088
step 3280, change in cost 0.0619812
step 3290, training accuracy 0.745016
step 3290, cost 214.026
step 3290, change in cost 0.0619202
step 3300, training accuracy 0.745016
step 3300, cost 213.964
step 3300, change in cost 0.0618744
step 3310, training accuracy 0.745016
step 3310, cost 213.902
step 3310, change in cost 0.0618591
step 3320, training accuracy 0.746065
step 3320, cost 213.84
step 3320, change in cost 0.0618286
step 3330, training accuracy 0.747114
step 3330, cost 213.779
step 3330, change in cost 0.0618134
step 3340, training accuracy 0.747114
step 3340, cost 213.717
step 3340, change in cost 0.0617065
step 3350, training accuracy 0.747114
step 3350, cost 213.655
step 3350, change in cost 0.0617676
step 3360, training accuracy 0.747114
step 3360, cost 213.593
step 3360, change in cost 0.0617218
step 3370, training accuracy 0.748164
step 3370, cost 213.532
step 3370, change in cost 0.061676
step 3380, training accuracy 0.748164
step 3380, cost 213.47
step 3380, change in cost 0.061615
step 3390, training accuracy 0.749213
step 3390, cost 213.409
step 3390, change in cost 0.0614777
step 3400, training accuracy 0.749213
step 3400, cost 213.347
step 3400, change in cost 0.0615845
step 3410, training accuracy 0.749213
step 3410, cost 213.286
step 3410, change in cost 0.0615387
step 3420, training accuracy 0.749213
step 3420, cost 213.224
step 3420, change in cost 0.0614624
step 3430, training accuracy 0.750262
step 3430, cost 213.163
step 3430, change in cost 0.0614319
step 3440, training accuracy 0.750262
step 3440, cost 213.101
step 3440, change in cost 0.0614624
step 3450, training accuracy 0.750262
step 3450, cost 213.04
step 3450, change in cost 0.0613403
step 3460, training accuracy 0.750262
step 3460, cost 212.979
step 3460, change in cost 0.0613098
step 3470, training accuracy 0.751312
step 3470, cost 212.917
step 3470, change in cost 0.0613556
step 3480, training accuracy 0.751312
step 3480, cost 212.856
step 3480, change in cost 0.0612488
step 3490, training accuracy 0.75341
step 3490, cost 212.795
step 3490, change in cost 0.0612793
step 3500, training accuracy 0.75341
step 3500, cost 212.734
step 3500, change in cost 0.0611877
step 3510, training accuracy 0.75446
step 3510, cost 212.672
step 3510, change in cost 0.0611267
step 3520, training accuracy 0.75446
step 3520, cost 212.611
step 3520, change in cost 0.061142
step 3530, training accuracy 0.75446
step 3530, cost 212.55
step 3530, change in cost 0.0611725
step 3540, training accuracy 0.75446
step 3540, cost 212.489
step 3540, change in cost 0.0610046
step 3550, training accuracy 0.75446
step 3550, cost 212.428
step 3550, change in cost 0.0610199
step 3560, training accuracy 0.75446
step 3560, cost 212.367
step 3560, change in cost 0.0610199
step 3570, training accuracy 0.75446
step 3570, cost 212.306
step 3570, change in cost 0.0608978
step 3580, training accuracy 0.75446
step 3580, cost 212.245
step 3580, change in cost 0.0609283
step 3590, training accuracy 0.75446
step 3590, cost 212.184
step 3590, change in cost 0.0608215
step 3600, training accuracy 0.755509
step 3600, cost 212.123
step 3600, change in cost 0.0608978
step 3610, training accuracy 0.755509
step 3610, cost 212.063
step 3610, change in cost 0.0608826
step 3620, training accuracy 0.755509
step 3620, cost 212.002
step 3620, change in cost 0.0607147
step 3630, training accuracy 0.755509
step 3630, cost 211.941
step 3630, change in cost 0.0607605
step 3640, training accuracy 0.755509
step 3640, cost 211.88
step 3640, change in cost 0.0607605
step 3650, training accuracy 0.755509
step 3650, cost 211.82
step 3650, change in cost 0.0606232
step 3660, training accuracy 0.755509
step 3660, cost 211.759
step 3660, change in cost 0.0606689
step 3670, training accuracy 0.755509
step 3670, cost 211.698
step 3670, change in cost 0.0605774
step 3680, training accuracy 0.756558
step 3680, cost 211.638
step 3680, change in cost 0.0606232
step 3690, training accuracy 0.757608
step 3690, cost 211.577
step 3690, change in cost 0.0604858
step 3700, training accuracy 0.757608
step 3700, cost 211.517
step 3700, change in cost 0.0605316
step 3710, training accuracy 0.757608
step 3710, cost 211.456
step 3710, change in cost 0.0604858
step 3720, training accuracy 0.758657
step 3720, cost 211.396
step 3720, change in cost 0.0604706
step 3730, training accuracy 0.758657
step 3730, cost 211.336
step 3730, change in cost 0.0603333
step 3740, training accuracy 0.758657
step 3740, cost 211.275
step 3740, change in cost 0.060379
step 3750, training accuracy 0.759706
step 3750, cost 211.215
step 3750, change in cost 0.0603638
step 3760, training accuracy 0.759706
step 3760, cost 211.155
step 3760, change in cost 0.0602417
step 3770, training accuracy 0.759706
step 3770, cost 211.094
step 3770, change in cost 0.0602875
step 3780, training accuracy 0.759706
step 3780, cost 211.034
step 3780, change in cost 0.0602112
step 3790, training accuracy 0.759706
step 3790, cost 210.974
step 3790, change in cost 0.0602417
step 3800, training accuracy 0.759706
step 3800, cost 210.914
step 3800, change in cost 0.0602112
step 3810, training accuracy 0.759706
step 3810, cost 210.854
step 3810, change in cost 0.0600739
step 3820, training accuracy 0.759706
step 3820, cost 210.793
step 3820, change in cost 0.0600739
step 3830, training accuracy 0.761805
step 3830, cost 210.733
step 3830, change in cost 0.0600891
step 3840, training accuracy 0.761805
step 3840, cost 210.673
step 3840, change in cost 0.0600281
step 3850, training accuracy 0.761805
step 3850, cost 210.613
step 3850, change in cost 0.059967
step 3860, training accuracy 0.762854
step 3860, cost 210.553
step 3860, change in cost 0.0599823
step 3870, training accuracy 0.763903
step 3870, cost 210.493
step 3870, change in cost 0.0599518
step 3880, training accuracy 0.763903
step 3880, cost 210.434
step 3880, change in cost 0.0598602
step 3890, training accuracy 0.763903
step 3890, cost 210.374
step 3890, change in cost 0.0598755
step 3900, training accuracy 0.763903
step 3900, cost 210.314
step 3900, change in cost 0.0597839
step 3910, training accuracy 0.764953
step 3910, cost 210.254
step 3910, change in cost 0.0597992
step 3920, training accuracy 0.766002
step 3920, cost 210.194
step 3920, change in cost 0.0597992
step 3930, training accuracy 0.767051
step 3930, cost 210.135
step 3930, change in cost 0.0596619
step 3940, training accuracy 0.767051
step 3940, cost 210.075
step 3940, change in cost 0.0596924
step 3950, training accuracy 0.767051
step 3950, cost 210.015
step 3950, change in cost 0.0597076
step 3960, training accuracy 0.767051
step 3960, cost 209.956
step 3960, change in cost 0.0595703
step 3970, training accuracy 0.767051
step 3970, cost 209.896
step 3970, change in cost 0.0595551
step 3980, training accuracy 0.768101
step 3980, cost 209.836
step 3980, change in cost 0.0596619
step 3990, training accuracy 0.768101
step 3990, cost 209.777
step 3990, change in cost 0.0594788
step 4000, training accuracy 0.768101
step 4000, cost 209.718
step 4000, change in cost 0.059433
step 4010, training accuracy 0.76915
step 4010, cost 209.658
step 4010, change in cost 0.0594788
step 4020, training accuracy 0.76915
step 4020, cost 209.599
step 4020, change in cost 0.0594025
step 4030, training accuracy 0.770199
step 4030, cost 209.539
step 4030, change in cost 0.0594177
step 4040, training accuracy 0.770199
step 4040, cost 209.48
step 4040, change in cost 0.0593719
step 4050, training accuracy 0.770199
step 4050, cost 209.421
step 4050, change in cost 0.0593109
step 4060, training accuracy 0.770199
step 4060, cost 209.361
step 4060, change in cost 0.0592651
step 4070, training accuracy 0.771249
step 4070, cost 209.302
step 4070, change in cost 0.0592346
step 4080, training accuracy 0.771249
step 4080, cost 209.243
step 4080, change in cost 0.0592194
step 4090, training accuracy 0.772298
step 4090, cost 209.184
step 4090, change in cost 0.0590973
step 4100, training accuracy 0.772298
step 4100, cost 209.125
step 4100, change in cost 0.0591888
step 4110, training accuracy 0.772298
step 4110, cost 209.065
step 4110, change in cost 0.0592194
step 4120, training accuracy 0.772298
step 4120, cost 209.006
step 4120, change in cost 0.059082
step 4130, training accuracy 0.772298
step 4130, cost 208.947
step 4130, change in cost 0.0589752
step 4140, training accuracy 0.772298
step 4140, cost 208.888
step 4140, change in cost 0.0590057
step 4150, training accuracy 0.775446
step 4150, cost 208.829
step 4150, change in cost 0.05896
step 4160, training accuracy 0.775446
step 4160, cost 208.77
step 4160, change in cost 0.0588989
step 4170, training accuracy 0.775446
step 4170, cost 208.711
step 4170, change in cost 0.0590057
step 4180, training accuracy 0.777545
step 4180, cost 208.653
step 4180, change in cost 0.0588531
step 4190, training accuracy 0.777545
step 4190, cost 208.594
step 4190, change in cost 0.0588226
step 4200, training accuracy 0.776495
step 4200, cost 208.535
step 4200, change in cost 0.0588379
step 4210, training accuracy 0.777545
step 4210, cost 208.476
step 4210, change in cost 0.0587311
step 4220, training accuracy 0.777545
step 4220, cost 208.417
step 4220, change in cost 0.0588226
step 4230, training accuracy 0.777545
step 4230, cost 208.359
step 4230, change in cost 0.05867
step 4240, training accuracy 0.778594
step 4240, cost 208.3
step 4240, change in cost 0.0586853
step 4250, training accuracy 0.778594
step 4250, cost 208.241
step 4250, change in cost 0.0585785
step 4260, training accuracy 0.778594
step 4260, cost 208.183
step 4260, change in cost 0.0586548
step 4270, training accuracy 0.778594
step 4270, cost 208.124
step 4270, change in cost 0.0586548
step 4280, training accuracy 0.779643
step 4280, cost 208.066
step 4280, change in cost 0.0585022
step 4290, training accuracy 0.779643
step 4290, cost 208.007
step 4290, change in cost 0.0584564
step 4300, training accuracy 0.779643
step 4300, cost 207.949
step 4300, change in cost 0.0585022
step 4310, training accuracy 0.779643
step 4310, cost 207.89
step 4310, change in cost 0.0584412
step 4320, training accuracy 0.780693
step 4320, cost 207.832
step 4320, change in cost 0.0584564
step 4330, training accuracy 0.781742
step 4330, cost 207.773
step 4330, change in cost 0.0583649
step 4340, training accuracy 0.781742
step 4340, cost 207.715
step 4340, change in cost 0.0583191
step 4350, training accuracy 0.781742
step 4350, cost 207.657
step 4350, change in cost 0.0583649
step 4360, training accuracy 0.781742
step 4360, cost 207.599
step 4360, change in cost 0.0582123
step 4370, training accuracy 0.781742
step 4370, cost 207.54
step 4370, change in cost 0.0582733
step 4380, training accuracy 0.781742
step 4380, cost 207.482
step 4380, change in cost 0.0582275
step 4390, training accuracy 0.781742
step 4390, cost 207.424
step 4390, change in cost 0.0582123
step 4400, training accuracy 0.781742
step 4400, cost 207.366
step 4400, change in cost 0.058136
step 4410, training accuracy 0.781742
step 4410, cost 207.308
step 4410, change in cost 0.0581055
step 4420, training accuracy 0.781742
step 4420, cost 207.249
step 4420, change in cost 0.0580902
step 4430, training accuracy 0.782791
step 4430, cost 207.191
step 4430, change in cost 0.0580444
step 4440, training accuracy 0.782791
step 4440, cost 207.133
step 4440, change in cost 0.0580597
step 4450, training accuracy 0.78384
step 4450, cost 207.075
step 4450, change in cost 0.0578766
step 4460, training accuracy 0.78489
step 4460, cost 207.017
step 4460, change in cost 0.0579987
step 4470, training accuracy 0.78489
step 4470, cost 206.96
step 4470, change in cost 0.0578613
step 4480, training accuracy 0.785939
step 4480, cost 206.902
step 4480, change in cost 0.0579834
step 4490, training accuracy 0.785939
step 4490, cost 206.844
step 4490, change in cost 0.057785
step 4500, training accuracy 0.785939
step 4500, cost 206.786
step 4500, change in cost 0.0578156
step 4510, training accuracy 0.786988
step 4510, cost 206.728
step 4510, change in cost 0.0578003
step 4520, training accuracy 0.788038
step 4520, cost 206.67
step 4520, change in cost 0.057785
step 4530, training accuracy 0.788038
step 4530, cost 206.613
step 4530, change in cost 0.0577545
step 4540, training accuracy 0.788038
step 4540, cost 206.555
step 4540, change in cost 0.0576477
step 4550, training accuracy 0.789087
step 4550, cost 206.497
step 4550, change in cost 0.0576477
step 4560, training accuracy 0.791186
step 4560, cost 206.44
step 4560, change in cost 0.0576324
step 4570, training accuracy 0.791186
step 4570, cost 206.382
step 4570, change in cost 0.0575104
step 4580, training accuracy 0.791186
step 4580, cost 206.325
step 4580, change in cost 0.057663
step 4590, training accuracy 0.791186
step 4590, cost 206.267
step 4590, change in cost 0.0574799
step 4600, training accuracy 0.792235
step 4600, cost 206.21
step 4600, change in cost 0.0574951
step 4610, training accuracy 0.792235
step 4610, cost 206.152
step 4610, change in cost 0.0574493
step 4620, training accuracy 0.793284
step 4620, cost 206.095
step 4620, change in cost 0.0574646
step 4630, training accuracy 0.793284
step 4630, cost 206.037
step 4630, change in cost 0.0573578
step 4640, training accuracy 0.793284
step 4640, cost 205.98
step 4640, change in cost 0.0573578
step 4650, training accuracy 0.793284
step 4650, cost 205.923
step 4650, change in cost 0.057373
step 4660, training accuracy 0.793284
step 4660, cost 205.865
step 4660, change in cost 0.0572357
step 4670, training accuracy 0.794334
step 4670, cost 205.808
step 4670, change in cost 0.0572815
step 4680, training accuracy 0.794334
step 4680, cost 205.751
step 4680, change in cost 0.0572815
step 4690, training accuracy 0.797482
step 4690, cost 205.694
step 4690, change in cost 0.0571594
step 4700, training accuracy 0.797482
step 4700, cost 205.636
step 4700, change in cost 0.0571899
step 4710, training accuracy 0.797482
step 4710, cost 205.579
step 4710, change in cost 0.0571594
step 4720, training accuracy 0.797482
step 4720, cost 205.522
step 4720, change in cost 0.0570831
step 4730, training accuracy 0.797482
step 4730, cost 205.465
step 4730, change in cost 0.0570221
step 4740, training accuracy 0.797482
step 4740, cost 205.408
step 4740, change in cost 0.0571136
step 4750, training accuracy 0.797482
step 4750, cost 205.351
step 4750, change in cost 0.0569763
step 4760, training accuracy 0.797482
step 4760, cost 205.294
step 4760, change in cost 0.0569916
step 4770, training accuracy 0.797482
step 4770, cost 205.237
step 4770, change in cost 0.0568848
step 4780, training accuracy 0.797482
step 4780, cost 205.18
step 4780, change in cost 0.0570068
step 4790, training accuracy 0.797482
step 4790, cost 205.123
step 4790, change in cost 0.056839
step 4800, training accuracy 0.798531
step 4800, cost 205.067
step 4800, change in cost 0.0568848
step 4810, training accuracy 0.798531
step 4810, cost 205.01
step 4810, change in cost 0.0567169
step 4820, training accuracy 0.798531
step 4820, cost 204.953
step 4820, change in cost 0.0568237
step 4830, training accuracy 0.798531
step 4830, cost 204.896
step 4830, change in cost 0.0566864
step 4840, training accuracy 0.79958
step 4840, cost 204.84
step 4840, change in cost 0.056778
step 4850, training accuracy 0.80063
step 4850, cost 204.783
step 4850, change in cost 0.0566864
step 4860, training accuracy 0.80063
step 4860, cost 204.726
step 4860, change in cost 0.0566406
step 4870, training accuracy 0.80063
step 4870, cost 204.67
step 4870, change in cost 0.0566101
step 4880, training accuracy 0.79958
step 4880, cost 204.613
step 4880, change in cost 0.0566254
step 4890, training accuracy 0.80063
step 4890, cost 204.556
step 4890, change in cost 0.056488
step 4900, training accuracy 0.80063
step 4900, cost 204.5
step 4900, change in cost 0.0565796
step 4910, training accuracy 0.802728
step 4910, cost 204.443
step 4910, change in cost 0.0564728
step 4920, training accuracy 0.802728
step 4920, cost 204.387
step 4920, change in cost 0.0564575
step 4930, training accuracy 0.802728
step 4930, cost 204.331
step 4930, change in cost 0.0564117
step 4940, training accuracy 0.803778
step 4940, cost 204.274
step 4940, change in cost 0.0564117
step 4950, training accuracy 0.803778
step 4950, cost 204.218
step 4950, change in cost 0.0564575
step 4960, training accuracy 0.803778
step 4960, cost 204.161
step 4960, change in cost 0.0562592
step 4970, training accuracy 0.804827
step 4970, cost 204.105
step 4970, change in cost 0.0562897
step 4980, training accuracy 0.804827
step 4980, cost 204.049
step 4980, change in cost 0.0562286
step 4990, training accuracy 0.804827
step 4990, cost 203.993
step 4990, change in cost 0.0562744
step 5000, training accuracy 0.803778
step 5000, cost 203.936
step 5000, change in cost 0.0561981
step 5010, training accuracy 0.803778
step 5010, cost 203.88
step 5010, change in cost 0.0561676
step 5020, training accuracy 0.803778
step 5020, cost 203.824
step 5020, change in cost 0.0561676
step 5030, training accuracy 0.803778
step 5030, cost 203.768
step 5030, change in cost 0.0560608
step 5040, training accuracy 0.803778
step 5040, cost 203.712
step 5040, change in cost 0.0561066
step 5050, training accuracy 0.805876
step 5050, cost 203.656
step 5050, change in cost 0.0560455
step 5060, training accuracy 0.805876
step 5060, cost 203.6
step 5060, change in cost 0.056076
step 5070, training accuracy 0.805876
step 5070, cost 203.544
step 5070, change in cost 0.0559692
step 5080, training accuracy 0.805876
step 5080, cost 203.488
step 5080, change in cost 0.055954
step 5090, training accuracy 0.805876
step 5090, cost 203.432
step 5090, change in cost 0.0559235
step 5100, training accuracy 0.806925
step 5100, cost 203.376
step 5100, change in cost 0.0559387
step 5110, training accuracy 0.807975
step 5110, cost 203.32
step 5110, change in cost 0.0557861
step 5120, training accuracy 0.809024
step 5120, cost 203.264
step 5120, change in cost 0.0558624
step 5130, training accuracy 0.809024
step 5130, cost 203.209
step 5130, change in cost 0.0558167
step 5140, training accuracy 0.810073
step 5140, cost 203.153
step 5140, change in cost 0.0557251
step 5150, training accuracy 0.810073
step 5150, cost 203.097
step 5150, change in cost 0.0557251
step 5160, training accuracy 0.810073
step 5160, cost 203.041
step 5160, change in cost 0.0557251
step 5170, training accuracy 0.810073
step 5170, cost 202.986
step 5170, change in cost 0.0557098
step 5180, training accuracy 0.811123
step 5180, cost 202.93
step 5180, change in cost 0.0555725
step 5190, training accuracy 0.811123
step 5190, cost 202.874
step 5190, change in cost 0.0556641
step 5200, training accuracy 0.811123
step 5200, cost 202.819
step 5200, change in cost 0.0555725
step 5210, training accuracy 0.811123
step 5210, cost 202.763
step 5210, change in cost 0.0554657
step 5220, training accuracy 0.811123
step 5220, cost 202.708
step 5220, change in cost 0.0555573
step 5230, training accuracy 0.811123
step 5230, cost 202.652
step 5230, change in cost 0.0555115
step 5240, training accuracy 0.811123
step 5240, cost 202.597
step 5240, change in cost 0.0554199
step 5250, training accuracy 0.811123
step 5250, cost 202.542
step 5250, change in cost 0.0553589
step 5260, training accuracy 0.811123
step 5260, cost 202.486
step 5260, change in cost 0.0554657
step 5270, training accuracy 0.811123
step 5270, cost 202.431
step 5270, change in cost 0.0553589
step 5280, training accuracy 0.810073
step 5280, cost 202.375
step 5280, change in cost 0.0552826
step 5290, training accuracy 0.811123
step 5290, cost 202.32
step 5290, change in cost 0.0552979
step 5300, training accuracy 0.811123
step 5300, cost 202.265
step 5300, change in cost 0.0552979
step 5310, training accuracy 0.812172
step 5310, cost 202.21
step 5310, change in cost 0.055191
step 5320, training accuracy 0.813221
step 5320, cost 202.154
step 5320, change in cost 0.0552063
step 5330, training accuracy 0.813221
step 5330, cost 202.099
step 5330, change in cost 0.0551605
step 5340, training accuracy 0.813221
step 5340, cost 202.044
step 5340, change in cost 0.055191
step 5350, training accuracy 0.813221
step 5350, cost 201.989
step 5350, change in cost 0.0550842
step 5360, training accuracy 0.814271
step 5360, cost 201.934
step 5360, change in cost 0.055069
step 5370, training accuracy 0.814271
step 5370, cost 201.879
step 5370, change in cost 0.0550385
step 5380, training accuracy 0.814271
step 5380, cost 201.824
step 5380, change in cost 0.0550232
step 5390, training accuracy 0.814271
step 5390, cost 201.769
step 5390, change in cost 0.0549622
step 5400, training accuracy 0.814271
step 5400, cost 201.714
step 5400, change in cost 0.0550385
step 5410, training accuracy 0.814271
step 5410, cost 201.659
step 5410, change in cost 0.0547943
step 5420, training accuracy 0.814271
step 5420, cost 201.604
step 5420, change in cost 0.0549011
step 5430, training accuracy 0.814271
step 5430, cost 201.549
step 5430, change in cost 0.0549622
step 5440, training accuracy 0.814271
step 5440, cost 201.494
step 5440, change in cost 0.0547485
step 5450, training accuracy 0.81532
step 5450, cost 201.44
step 5450, change in cost 0.0548248
step 5460, training accuracy 0.81532
step 5460, cost 201.385
step 5460, change in cost 0.0547638
step 5470, training accuracy 0.81532
step 5470, cost 201.33
step 5470, change in cost 0.0547485
step 5480, training accuracy 0.81532
step 5480, cost 201.275
step 5480, change in cost 0.054718
step 5490, training accuracy 0.816369
step 5490, cost 201.221
step 5490, change in cost 0.054657
step 5500, training accuracy 0.816369
step 5500, cost 201.166
step 5500, change in cost 0.0547638
step 5510, training accuracy 0.817419
step 5510, cost 201.111
step 5510, change in cost 0.0545502
step 5520, training accuracy 0.817419
step 5520, cost 201.057
step 5520, change in cost 0.0545502
step 5530, training accuracy 0.817419
step 5530, cost 201.002
step 5530, change in cost 0.0545044
step 5540, training accuracy 0.817419
step 5540, cost 200.948
step 5540, change in cost 0.0545197
step 5550, training accuracy 0.818468
step 5550, cost 200.893
step 5550, change in cost 0.0545502
step 5560, training accuracy 0.819517
step 5560, cost 200.839
step 5560, change in cost 0.0544586
step 5570, training accuracy 0.819517
step 5570, cost 200.784
step 5570, change in cost 0.0543823
step 5580, training accuracy 0.819517
step 5580, cost 200.73
step 5580, change in cost 0.0545044
step 5590, training accuracy 0.819517
step 5590, cost 200.676
step 5590, change in cost 0.054306
step 5600, training accuracy 0.819517
step 5600, cost 200.621
step 5600, change in cost 0.0542908
step 5610, training accuracy 0.819517
step 5610, cost 200.567
step 5610, change in cost 0.0543365
step 5620, training accuracy 0.819517
step 5620, cost 200.513
step 5620, change in cost 0.054306
step 5630, training accuracy 0.819517
step 5630, cost 200.458
step 5630, change in cost 0.0542297
step 5640, training accuracy 0.819517
step 5640, cost 200.404
step 5640, change in cost 0.0541687
step 5650, training accuracy 0.819517
step 5650, cost 200.35
step 5650, change in cost 0.0542755
step 5660, training accuracy 0.818468
step 5660, cost 200.296
step 5660, change in cost 0.0541077
step 5670, training accuracy 0.819517
step 5670, cost 200.242
step 5670, change in cost 0.0541687
step 5680, training accuracy 0.820567
step 5680, cost 200.188
step 5680, change in cost 0.0540619
step 5690, training accuracy 0.820567
step 5690, cost 200.134
step 5690, change in cost 0.0540771
step 5700, training accuracy 0.822665
step 5700, cost 200.08
step 5700, change in cost 0.0541077
step 5710, training accuracy 0.822665
step 5710, cost 200.026
step 5710, change in cost 0.0539551
step 5720, training accuracy 0.822665
step 5720, cost 199.972
step 5720, change in cost 0.0539093
step 5730, training accuracy 0.823715
step 5730, cost 199.918
step 5730, change in cost 0.0539551
step 5740, training accuracy 0.823715
step 5740, cost 199.864
step 5740, change in cost 0.053894
step 5750, training accuracy 0.823715
step 5750, cost 199.81
step 5750, change in cost 0.0539093
step 5760, training accuracy 0.823715
step 5760, cost 199.756
step 5760, change in cost 0.0538483
step 5770, training accuracy 0.823715
step 5770, cost 199.702
step 5770, change in cost 0.053894
step 5780, training accuracy 0.823715
step 5780, cost 199.648
step 5780, change in cost 0.0537109
step 5790, training accuracy 0.824764
step 5790, cost 199.595
step 5790, change in cost 0.053772
step 5800, training accuracy 0.824764
step 5800, cost 199.541
step 5800, change in cost 0.0537415
step 5810, training accuracy 0.824764
step 5810, cost 199.487
step 5810, change in cost 0.0536499
step 5820, training accuracy 0.824764
step 5820, cost 199.434
step 5820, change in cost 0.0537567
step 5830, training accuracy 0.824764
step 5830, cost 199.38
step 5830, change in cost 0.0535736
step 5840, training accuracy 0.824764
step 5840, cost 199.326
step 5840, change in cost 0.0536957
step 5850, training accuracy 0.824764
step 5850, cost 199.273
step 5850, change in cost 0.0535889
step 5860, training accuracy 0.824764
step 5860, cost 199.219
step 5860, change in cost 0.0534973
step 5870, training accuracy 0.824764
step 5870, cost 199.166
step 5870, change in cost 0.0535736
step 5880, training accuracy 0.824764
step 5880, cost 199.112
step 5880, change in cost 0.053421
step 5890, training accuracy 0.828961
step 5890, cost 199.059
step 5890, change in cost 0.0535278
step 5900, training accuracy 0.828961
step 5900, cost 199.005
step 5900, change in cost 0.0533905
step 5910, training accuracy 0.83001
step 5910, cost 198.952
step 5910, change in cost 0.0534363
step 5920, training accuracy 0.83001
step 5920, cost 198.898
step 5920, change in cost 0.0533905
step 5930, training accuracy 0.83106
step 5930, cost 198.845
step 5930, change in cost 0.0533142
step 5940, training accuracy 0.83106
step 5940, cost 198.792
step 5940, change in cost 0.0533142
step 5950, training accuracy 0.83106
step 5950, cost 198.739
step 5950, change in cost 0.0532837
step 5960, training accuracy 0.83106
step 5960, cost 198.685
step 5960, change in cost 0.0532379
step 5970, training accuracy 0.83106
step 5970, cost 198.632
step 5970, change in cost 0.0532379
step 5980, training accuracy 0.832109
step 5980, cost 198.579
step 5980, change in cost 0.0531616
step 5990, training accuracy 0.832109
step 5990, cost 198.526
step 5990, change in cost 0.0531921
step 6000, training accuracy 0.832109
step 6000, cost 198.473
step 6000, change in cost 0.0531311
step 6010, training accuracy 0.832109
step 6010, cost 198.419
step 6010, change in cost 0.0531311
step 6020, training accuracy 0.832109
step 6020, cost 198.366
step 6020, change in cost 0.0531158
step 6030, training accuracy 0.832109
step 6030, cost 198.313
step 6030, change in cost 0.0529938
step 6040, training accuracy 0.832109
step 6040, cost 198.26
step 6040, change in cost 0.0530243
step 6050, training accuracy 0.832109
step 6050, cost 198.207
step 6050, change in cost 0.0529633
step 6060, training accuracy 0.832109
step 6060, cost 198.154
step 6060, change in cost 0.053009
step 6070, training accuracy 0.832109
step 6070, cost 198.101
step 6070, change in cost 0.0529022
step 6080, training accuracy 0.832109
step 6080, cost 198.049
step 6080, change in cost 0.0528412
step 6090, training accuracy 0.832109
step 6090, cost 197.996
step 6090, change in cost 0.0529022
step 6100, training accuracy 0.832109
step 6100, cost 197.943
step 6100, change in cost 0.0528717
step 6110, training accuracy 0.834208
step 6110, cost 197.89
step 6110, change in cost 0.0527802
step 6120, training accuracy 0.834208
step 6120, cost 197.837
step 6120, change in cost 0.0527496
step 6130, training accuracy 0.834208
step 6130, cost 197.784
step 6130, change in cost 0.0528107
step 6140, training accuracy 0.834208
step 6140, cost 197.732
step 6140, change in cost 0.0526733
step 6150, training accuracy 0.834208
step 6150, cost 197.679
step 6150, change in cost 0.0526428
step 6160, training accuracy 0.834208
step 6160, cost 197.626
step 6160, change in cost 0.0527191
step 6170, training accuracy 0.834208
step 6170, cost 197.574
step 6170, change in cost 0.0526733
step 6180, training accuracy 0.834208
step 6180, cost 197.521
step 6180, change in cost 0.0525665
step 6190, training accuracy 0.835257
step 6190, cost 197.469
step 6190, change in cost 0.052536
step 6200, training accuracy 0.836306
step 6200, cost 197.416
step 6200, change in cost 0.0526123
step 6210, training accuracy 0.837356
step 6210, cost 197.364
step 6210, change in cost 0.0525055
step 6220, training accuracy 0.837356
step 6220, cost 197.311
step 6220, change in cost 0.0524597
step 6230, training accuracy 0.837356
step 6230, cost 197.259
step 6230, change in cost 0.0524597
step 6240, training accuracy 0.837356
step 6240, cost 197.206
step 6240, change in cost 0.052475
step 6250, training accuracy 0.838405
step 6250, cost 197.154
step 6250, change in cost 0.0523529
step 6260, training accuracy 0.838405
step 6260, cost 197.101
step 6260, change in cost 0.0524292
step 6270, training accuracy 0.839454
step 6270, cost 197.049
step 6270, change in cost 0.0522461
step 6280, training accuracy 0.839454
step 6280, cost 196.997
step 6280, change in cost 0.0523529
step 6290, training accuracy 0.839454
step 6290, cost 196.944
step 6290, change in cost 0.0522919
step 6300, training accuracy 0.840504
step 6300, cost 196.892
step 6300, change in cost 0.0522308
step 6310, training accuracy 0.840504
step 6310, cost 196.84
step 6310, change in cost 0.0522308
step 6320, training accuracy 0.840504
step 6320, cost 196.788
step 6320, change in cost 0.0522461
step 6330, training accuracy 0.840504
step 6330, cost 196.736
step 6330, change in cost 0.0520935
step 6340, training accuracy 0.840504
step 6340, cost 196.684
step 6340, change in cost 0.0521698
step 6350, training accuracy 0.840504
step 6350, cost 196.631
step 6350, change in cost 0.0521393
step 6360, training accuracy 0.840504
step 6360, cost 196.579
step 6360, change in cost 0.0520172
step 6370, training accuracy 0.840504
step 6370, cost 196.527
step 6370, change in cost 0.0521088
step 6380, training accuracy 0.840504
step 6380, cost 196.475
step 6380, change in cost 0.0519867
step 6390, training accuracy 0.840504
step 6390, cost 196.423
step 6390, change in cost 0.0519714
step 6400, training accuracy 0.840504
step 6400, cost 196.371
step 6400, change in cost 0.0519714
step 6410, training accuracy 0.840504
step 6410, cost 196.319
step 6410, change in cost 0.0519562
step 6420, training accuracy 0.840504
step 6420, cost 196.267
step 6420, change in cost 0.0518646
step 6430, training accuracy 0.840504
step 6430, cost 196.216
step 6430, change in cost 0.0519257
step 6440, training accuracy 0.840504
step 6440, cost 196.164
step 6440, change in cost 0.0518799
step 6450, training accuracy 0.840504
step 6450, cost 196.112
step 6450, change in cost 0.0517731
step 6460, training accuracy 0.840504
step 6460, cost 196.06
step 6460, change in cost 0.0518188
step 6470, training accuracy 0.840504
step 6470, cost 196.008
step 6470, change in cost 0.0517273
step 6480, training accuracy 0.840504
step 6480, cost 195.957
step 6480, change in cost 0.0516968
step 6490, training accuracy 0.840504
step 6490, cost 195.905
step 6490, change in cost 0.0518036
step 6500, training accuracy 0.840504
step 6500, cost 195.853
step 6500, change in cost 0.05159
step 6510, training accuracy 0.840504
step 6510, cost 195.802
step 6510, change in cost 0.0516357
step 6520, training accuracy 0.840504
step 6520, cost 195.75
step 6520, change in cost 0.0516815
step 6530, training accuracy 0.840504
step 6530, cost 195.698
step 6530, change in cost 0.0515594
step 6540, training accuracy 0.840504
step 6540, cost 195.647
step 6540, change in cost 0.0514984
step 6550, training accuracy 0.840504
step 6550, cost 195.595
step 6550, change in cost 0.05159
step 6560, training accuracy 0.841553
step 6560, cost 195.544
step 6560, change in cost 0.0514221
step 6570, training accuracy 0.841553
step 6570, cost 195.492
step 6570, change in cost 0.0515747
step 6580, training accuracy 0.841553
step 6580, cost 195.441
step 6580, change in cost 0.0514221
step 6590, training accuracy 0.842602
step 6590, cost 195.389
step 6590, change in cost 0.0514221
step 6600, training accuracy 0.842602
step 6600, cost 195.338
step 6600, change in cost 0.0513916
step 6610, training accuracy 0.842602
step 6610, cost 195.287
step 6610, change in cost 0.0513153
step 6620, training accuracy 0.842602
step 6620, cost 195.235
step 6620, change in cost 0.0513763
step 6630, training accuracy 0.842602
step 6630, cost 195.184
step 6630, change in cost 0.0513
step 6640, training accuracy 0.842602
step 6640, cost 195.133
step 6640, change in cost 0.0513153
step 6650, training accuracy 0.842602
step 6650, cost 195.082
step 6650, change in cost 0.051239
step 6660, training accuracy 0.843652
step 6660, cost 195.03
step 6660, change in cost 0.051178
step 6670, training accuracy 0.844701
step 6670, cost 194.979
step 6670, change in cost 0.0511932
step 6680, training accuracy 0.844701
step 6680, cost 194.928
step 6680, change in cost 0.0511475
step 6690, training accuracy 0.844701
step 6690, cost 194.877
step 6690, change in cost 0.0511322
step 6700, training accuracy 0.844701
step 6700, cost 194.826
step 6700, change in cost 0.0511169
step 6710, training accuracy 0.84575
step 6710, cost 194.775
step 6710, change in cost 0.0510406
step 6720, training accuracy 0.84575
step 6720, cost 194.724
step 6720, change in cost 0.0511475
step 6730, training accuracy 0.8468
step 6730, cost 194.673
step 6730, change in cost 0.0509491
step 6740, training accuracy 0.847849
step 6740, cost 194.622
step 6740, change in cost 0.0509796
step 6750, training accuracy 0.847849
step 6750, cost 194.571
step 6750, change in cost 0.0510559
step 6760, training accuracy 0.847849
step 6760, cost 194.52
step 6760, change in cost 0.0508881
step 6770, training accuracy 0.847849
step 6770, cost 194.469
step 6770, change in cost 0.0508575
step 6780, training accuracy 0.8468
step 6780, cost 194.418
step 6780, change in cost 0.0509338
step 6790, training accuracy 0.847849
step 6790, cost 194.367
step 6790, change in cost 0.0507812
step 6800, training accuracy 0.847849
step 6800, cost 194.316
step 6800, change in cost 0.0508575
step 6810, training accuracy 0.847849
step 6810, cost 194.265
step 6810, change in cost 0.0508423
step 6820, training accuracy 0.847849
step 6820, cost 194.215
step 6820, change in cost 0.0506897
step 6830, training accuracy 0.847849
step 6830, cost 194.164
step 6830, change in cost 0.050827
step 6840, training accuracy 0.847849
step 6840, cost 194.113
step 6840, change in cost 0.0506287
step 6850, training accuracy 0.847849
step 6850, cost 194.063
step 6850, change in cost 0.050705
step 6860, training accuracy 0.848898
step 6860, cost 194.012
step 6860, change in cost 0.0507202
step 6870, training accuracy 0.848898
step 6870, cost 193.961
step 6870, change in cost 0.0505371
step 6880, training accuracy 0.848898
step 6880, cost 193.911
step 6880, change in cost 0.0506134
step 6890, training accuracy 0.848898
step 6890, cost 193.86
step 6890, change in cost 0.0505981
step 6900, training accuracy 0.848898
step 6900, cost 193.81
step 6900, change in cost 0.0505371
step 6910, training accuracy 0.850997
step 6910, cost 193.759
step 6910, change in cost 0.0505219
step 6920, training accuracy 0.850997
step 6920, cost 193.709
step 6920, change in cost 0.0505371
step 6930, training accuracy 0.850997
step 6930, cost 193.658
step 6930, change in cost 0.0504761
step 6940, training accuracy 0.850997
step 6940, cost 193.608
step 6940, change in cost 0.0503693
step 6950, training accuracy 0.850997
step 6950, cost 193.557
step 6950, change in cost 0.050415
step 6960, training accuracy 0.850997
step 6960, cost 193.507
step 6960, change in cost 0.0503387
step 6970, training accuracy 0.850997
step 6970, cost 193.457
step 6970, change in cost 0.0503387
step 6980, training accuracy 0.850997
step 6980, cost 193.406
step 6980, change in cost 0.0503387
step 6990, training accuracy 0.850997
step 6990, cost 193.356
step 6990, change in cost 0.0502625
step 7000, training accuracy 0.850997
step 7000, cost 193.306
step 7000, change in cost 0.050354
step 7010, training accuracy 0.850997
step 7010, cost 193.255
step 7010, change in cost 0.0501862
step 7020, training accuracy 0.850997
step 7020, cost 193.205
step 7020, change in cost 0.0501556
step 7030, training accuracy 0.850997
step 7030, cost 193.155
step 7030, change in cost 0.0502014
step 7040, training accuracy 0.850997
step 7040, cost 193.105
step 7040, change in cost 0.0502167
step 7050, training accuracy 0.850997
step 7050, cost 193.055
step 7050, change in cost 0.0500793
step 7060, training accuracy 0.850997
step 7060, cost 193.005
step 7060, change in cost 0.0500946
step 7070, training accuracy 0.853095
step 7070, cost 192.955
step 7070, change in cost 0.0501099
step 7080, training accuracy 0.853095
step 7080, cost 192.905
step 7080, change in cost 0.0499878
step 7090, training accuracy 0.853095
step 7090, cost 192.854
step 7090, change in cost 0.0500946
step 7100, training accuracy 0.853095
step 7100, cost 192.805
step 7100, change in cost 0.049942
step 7110, training accuracy 0.853095
step 7110, cost 192.755
step 7110, change in cost 0.049942
step 7120, training accuracy 0.853095
step 7120, cost 192.705
step 7120, change in cost 0.0499268
step 7130, training accuracy 0.853095
step 7130, cost 192.655
step 7130, change in cost 0.0499573
step 7140, training accuracy 0.853095
step 7140, cost 192.605
step 7140, change in cost 0.0498657
step 7150, training accuracy 0.853095
step 7150, cost 192.555
step 7150, change in cost 0.0498505
step 7160, training accuracy 0.853095
step 7160, cost 192.505
step 7160, change in cost 0.0498352
step 7170, training accuracy 0.853095
step 7170, cost 192.455
step 7170, change in cost 0.0498657
step 7180, training accuracy 0.853095
step 7180, cost 192.406
step 7180, change in cost 0.0496826
step 7190, training accuracy 0.853095
step 7190, cost 192.356
step 7190, change in cost 0.0497437
step 7200, training accuracy 0.853095
step 7200, cost 192.306
step 7200, change in cost 0.0496979
step 7210, training accuracy 0.853095
step 7210, cost 192.256
step 7210, change in cost 0.0497131
step 7220, training accuracy 0.854145
step 7220, cost 192.207
step 7220, change in cost 0.0495758
step 7230, training accuracy 0.854145
step 7230, cost 192.157
step 7230, change in cost 0.0497742
step 7240, training accuracy 0.854145
step 7240, cost 192.108
step 7240, change in cost 0.04953
step 7250, training accuracy 0.854145
step 7250, cost 192.058
step 7250, change in cost 0.0496063
step 7260, training accuracy 0.856243
step 7260, cost 192.008
step 7260, change in cost 0.0494995
step 7270, training accuracy 0.856243
step 7270, cost 191.959
step 7270, change in cost 0.0495148
step 7280, training accuracy 0.856243
step 7280, cost 191.909
step 7280, change in cost 0.0496063
step 7290, training accuracy 0.856243
step 7290, cost 191.86
step 7290, change in cost 0.0493774
step 7300, training accuracy 0.856243
step 7300, cost 191.811
step 7300, change in cost 0.0494385
step 7310, training accuracy 0.856243
step 7310, cost 191.761
step 7310, change in cost 0.0494385
step 7320, training accuracy 0.857293
step 7320, cost 191.712
step 7320, change in cost 0.0493774
step 7330, training accuracy 0.857293
step 7330, cost 191.662
step 7330, change in cost 0.0493317
step 7340, training accuracy 0.857293
step 7340, cost 191.613
step 7340, change in cost 0.0493927
step 7350, training accuracy 0.857293
step 7350, cost 191.564
step 7350, change in cost 0.0493774
step 7360, training accuracy 0.857293
step 7360, cost 191.514
step 7360, change in cost 0.0491791
step 7370, training accuracy 0.857293
step 7370, cost 191.465
step 7370, change in cost 0.0492401
step 7380, training accuracy 0.857293
step 7380, cost 191.416
step 7380, change in cost 0.0492249
step 7390, training accuracy 0.858342
step 7390, cost 191.367
step 7390, change in cost 0.0492554
step 7400, training accuracy 0.858342
step 7400, cost 191.318
step 7400, change in cost 0.0491638
step 7410, training accuracy 0.858342
step 7410, cost 191.268
step 7410, change in cost 0.0491028
step 7420, training accuracy 0.859391
step 7420, cost 191.219
step 7420, change in cost 0.0491638
step 7430, training accuracy 0.859391
step 7430, cost 191.17
step 7430, change in cost 0.0490417
step 7440, training accuracy 0.859391
step 7440, cost 191.121
step 7440, change in cost 0.0491028
step 7450, training accuracy 0.859391
step 7450, cost 191.072
step 7450, change in cost 0.0490112
step 7460, training accuracy 0.859391
step 7460, cost 191.023
step 7460, change in cost 0.048996
step 7470, training accuracy 0.86149
step 7470, cost 190.974
step 7470, change in cost 0.0489502
step 7480, training accuracy 0.860441
step 7480, cost 190.925
step 7480, change in cost 0.048996
step 7490, training accuracy 0.860441
step 7490, cost 190.876
step 7490, change in cost 0.0489349
step 7500, training accuracy 0.860441
step 7500, cost 190.827
step 7500, change in cost 0.0488892
step 7510, training accuracy 0.86149
step 7510, cost 190.779
step 7510, change in cost 0.0488739
step 7520, training accuracy 0.86149
step 7520, cost 190.73
step 7520, change in cost 0.0488586
step 7530, training accuracy 0.86149
step 7530, cost 190.681
step 7530, change in cost 0.0487671
step 7540, training accuracy 0.860441
step 7540, cost 190.632
step 7540, change in cost 0.0488281
step 7550, training accuracy 0.860441
step 7550, cost 190.583
step 7550, change in cost 0.0488129
step 7560, training accuracy 0.860441
step 7560, cost 190.535
step 7560, change in cost 0.0487213
step 7570, training accuracy 0.860441
step 7570, cost 190.486
step 7570, change in cost 0.0487061
step 7580, training accuracy 0.860441
step 7580, cost 190.437
step 7580, change in cost 0.0487213
step 7590, training accuracy 0.860441
step 7590, cost 190.388
step 7590, change in cost 0.0486908
step 7600, training accuracy 0.860441
step 7600, cost 190.34
step 7600, change in cost 0.0485687
step 7610, training accuracy 0.860441
step 7610, cost 190.291
step 7610, change in cost 0.048645
step 7620, training accuracy 0.860441
step 7620, cost 190.243
step 7620, change in cost 0.0485687
step 7630, training accuracy 0.860441
step 7630, cost 190.194
step 7630, change in cost 0.0484924
step 7640, training accuracy 0.860441
step 7640, cost 190.146
step 7640, change in cost 0.0486145
step 7650, training accuracy 0.860441
step 7650, cost 190.097
step 7650, change in cost 0.0485077
step 7660, training accuracy 0.860441
step 7660, cost 190.049
step 7660, change in cost 0.0483856
step 7670, training accuracy 0.860441
step 7670, cost 190
step 7670, change in cost 0.0484619
step 7680, training accuracy 0.860441
step 7680, cost 189.952
step 7680, change in cost 0.0484772
step 7690, training accuracy 0.860441
step 7690, cost 189.903
step 7690, change in cost 0.0483856
step 7700, training accuracy 0.860441
step 7700, cost 189.855
step 7700, change in cost 0.0483246
step 7710, training accuracy 0.860441
step 7710, cost 189.807
step 7710, change in cost 0.0484009
step 7720, training accuracy 0.86149
step 7720, cost 189.758
step 7720, change in cost 0.0483246
step 7730, training accuracy 0.86149
step 7730, cost 189.71
step 7730, change in cost 0.0482635
step 7740, training accuracy 0.862539
step 7740, cost 189.662
step 7740, change in cost 0.0483398
step 7750, training accuracy 0.862539
step 7750, cost 189.613
step 7750, change in cost 0.0481873
step 7760, training accuracy 0.862539
step 7760, cost 189.565
step 7760, change in cost 0.0482483
step 7770, training accuracy 0.862539
step 7770, cost 189.517
step 7770, change in cost 0.0481262
step 7780, training accuracy 0.862539
step 7780, cost 189.469
step 7780, change in cost 0.0481415
step 7790, training accuracy 0.862539
step 7790, cost 189.421
step 7790, change in cost 0.0482178
step 7800, training accuracy 0.862539
step 7800, cost 189.373
step 7800, change in cost 0.0480957
step 7810, training accuracy 0.862539
step 7810, cost 189.325
step 7810, change in cost 0.0480804
step 7820, training accuracy 0.862539
step 7820, cost 189.276
step 7820, change in cost 0.0480499
step 7830, training accuracy 0.862539
step 7830, cost 189.229
step 7830, change in cost 0.0479736
step 7840, training accuracy 0.862539
step 7840, cost 189.18
step 7840, change in cost 0.0480652
step 7850, training accuracy 0.862539
step 7850, cost 189.132
step 7850, change in cost 0.0479889
step 7860, training accuracy 0.863589
step 7860, cost 189.085
step 7860, change in cost 0.0479279
step 7870, training accuracy 0.864638
step 7870, cost 189.037
step 7870, change in cost 0.0479431
step 7880, training accuracy 0.864638
step 7880, cost 188.989
step 7880, change in cost 0.0478516
step 7890, training accuracy 0.864638
step 7890, cost 188.941
step 7890, change in cost 0.0479584
step 7900, training accuracy 0.864638
step 7900, cost 188.893
step 7900, change in cost 0.0477295
step 7910, training accuracy 0.864638
step 7910, cost 188.845
step 7910, change in cost 0.0479126
step 7920, training accuracy 0.864638
step 7920, cost 188.797
step 7920, change in cost 0.0477753
step 7930, training accuracy 0.864638
step 7930, cost 188.75
step 7930, change in cost 0.0477753
step 7940, training accuracy 0.865687
step 7940, cost 188.702
step 7940, change in cost 0.0476685
step 7950, training accuracy 0.865687
step 7950, cost 188.654
step 7950, change in cost 0.0477753
step 7960, training accuracy 0.865687
step 7960, cost 188.606
step 7960, change in cost 0.0477142
step 7970, training accuracy 0.865687
step 7970, cost 188.559
step 7970, change in cost 0.0475922
step 7980, training accuracy 0.865687
step 7980, cost 188.511
step 7980, change in cost 0.047699
step 7990, training accuracy 0.865687
step 7990, cost 188.464
step 7990, change in cost 0.0475769
step 8000, training accuracy 0.865687
step 8000, cost 188.416
step 8000, change in cost 0.0476379
step 8010, training accuracy 0.865687
step 8010, cost 188.368
step 8010, change in cost 0.0475159
step 8020, training accuracy 0.865687
step 8020, cost 188.321
step 8020, change in cost 0.0475464
step 8030, training accuracy 0.865687
step 8030, cost 188.273
step 8030, change in cost 0.0475159
step 8040, training accuracy 0.865687
step 8040, cost 188.226
step 8040, change in cost 0.0474854
step 8050, training accuracy 0.865687
step 8050, cost 188.178
step 8050, change in cost 0.0474701
step 8060, training accuracy 0.866737
step 8060, cost 188.131
step 8060, change in cost 0.0474396
step 8070, training accuracy 0.866737
step 8070, cost 188.084
step 8070, change in cost 0.0474243
step 8080, training accuracy 0.867786
step 8080, cost 188.036
step 8080, change in cost 0.047348
step 8090, training accuracy 0.867786
step 8090, cost 187.989
step 8090, change in cost 0.0473328
step 8100, training accuracy 0.867786
step 8100, cost 187.942
step 8100, change in cost 0.0473175
step 8110, training accuracy 0.867786
step 8110, cost 187.894
step 8110, change in cost 0.0473175
step 8120, training accuracy 0.867786
step 8120, cost 187.847
step 8120, change in cost 0.0473328
step 8130, training accuracy 0.867786
step 8130, cost 187.8
step 8130, change in cost 0.047226
step 8140, training accuracy 0.867786
step 8140, cost 187.752
step 8140, change in cost 0.0471954
step 8150, training accuracy 0.867786
step 8150, cost 187.705
step 8150, change in cost 0.0472412
step 8160, training accuracy 0.867786
step 8160, cost 187.658
step 8160, change in cost 0.0472412
step 8170, training accuracy 0.867786
step 8170, cost 187.611
step 8170, change in cost 0.0470886
step 8180, training accuracy 0.867786
step 8180, cost 187.564
step 8180, change in cost 0.0471344
step 8190, training accuracy 0.867786
step 8190, cost 187.517
step 8190, change in cost 0.0471497
step 8200, training accuracy 0.867786
step 8200, cost 187.47
step 8200, change in cost 0.0470734
step 8210, training accuracy 0.867786
step 8210, cost 187.423
step 8210, change in cost 0.0469818
step 8220, training accuracy 0.867786
step 8220, cost 187.376
step 8220, change in cost 0.0470276
step 8230, training accuracy 0.867786
step 8230, cost 187.329
step 8230, change in cost 0.0469818
step 8240, training accuracy 0.867786
step 8240, cost 187.282
step 8240, change in cost 0.0469971
step 8250, training accuracy 0.867786
step 8250, cost 187.235
step 8250, change in cost 0.0469513
step 8260, training accuracy 0.867786
step 8260, cost 187.188
step 8260, change in cost 0.0469513
step 8270, training accuracy 0.867786
step 8270, cost 187.141
step 8270, change in cost 0.046814
step 8280, training accuracy 0.867786
step 8280, cost 187.094
step 8280, change in cost 0.046936
step 8290, training accuracy 0.867786
step 8290, cost 187.047
step 8290, change in cost 0.0468445
step 8300, training accuracy 0.868835
step 8300, cost 187
step 8300, change in cost 0.046814
step 8310, training accuracy 0.868835
step 8310, cost 186.953
step 8310, change in cost 0.0467834
step 8320, training accuracy 0.868835
step 8320, cost 186.907
step 8320, change in cost 0.0468597
step 8330, training accuracy 0.868835
step 8330, cost 186.86
step 8330, change in cost 0.0467072
step 8340, training accuracy 0.868835
step 8340, cost 186.813
step 8340, change in cost 0.0466919
step 8350, training accuracy 0.868835
step 8350, cost 186.766
step 8350, change in cost 0.0467072
step 8360, training accuracy 0.868835
step 8360, cost 186.72
step 8360, change in cost 0.0467072
step 8370, training accuracy 0.869885
step 8370, cost 186.673
step 8370, change in cost 0.0466003
step 8380, training accuracy 0.869885
step 8380, cost 186.627
step 8380, change in cost 0.0465851
step 8390, training accuracy 0.869885
step 8390, cost 186.58
step 8390, change in cost 0.0465698
step 8400, training accuracy 0.869885
step 8400, cost 186.533
step 8400, change in cost 0.0465851
step 8410, training accuracy 0.870934
step 8410, cost 186.487
step 8410, change in cost 0.0465546
step 8420, training accuracy 0.870934
step 8420, cost 186.44
step 8420, change in cost 0.0465851
step 8430, training accuracy 0.870934
step 8430, cost 186.394
step 8430, change in cost 0.0464172
step 8440, training accuracy 0.870934
step 8440, cost 186.347
step 8440, change in cost 0.046463
step 8450, training accuracy 0.870934
step 8450, cost 186.301
step 8450, change in cost 0.046524
step 8460, training accuracy 0.870934
step 8460, cost 186.254
step 8460, change in cost 0.0464478
step 8470, training accuracy 0.870934
step 8470, cost 186.208
step 8470, change in cost 0.0463409
step 8480, training accuracy 0.870934
step 8480, cost 186.162
step 8480, change in cost 0.0463257
step 8490, training accuracy 0.870934
step 8490, cost 186.115
step 8490, change in cost 0.0463867
step 8500, training accuracy 0.870934
step 8500, cost 186.069
step 8500, change in cost 0.0463562
step 8510, training accuracy 0.870934
step 8510, cost 186.023
step 8510, change in cost 0.0462341
step 8520, training accuracy 0.870934
step 8520, cost 185.976
step 8520, change in cost 0.0463104
step 8530, training accuracy 0.870934
step 8530, cost 185.93
step 8530, change in cost 0.0461884
step 8540, training accuracy 0.870934
step 8540, cost 185.884
step 8540, change in cost 0.0462646
step 8550, training accuracy 0.870934
step 8550, cost 185.838
step 8550, change in cost 0.0462189
step 8560, training accuracy 0.870934
step 8560, cost 185.792
step 8560, change in cost 0.0462036
step 8570, training accuracy 0.871983
step 8570, cost 185.745
step 8570, change in cost 0.0461426
step 8580, training accuracy 0.873033
step 8580, cost 185.699
step 8580, change in cost 0.0461273
step 8590, training accuracy 0.873033
step 8590, cost 185.653
step 8590, change in cost 0.0460815
step 8600, training accuracy 0.873033
step 8600, cost 185.607
step 8600, change in cost 0.046051
step 8610, training accuracy 0.873033
step 8610, cost 185.561
step 8610, change in cost 0.0460815
step 8620, training accuracy 0.873033
step 8620, cost 185.515
step 8620, change in cost 0.046051
step 8630, training accuracy 0.873033
step 8630, cost 185.469
step 8630, change in cost 0.0458832
step 8640, training accuracy 0.873033
step 8640, cost 185.423
step 8640, change in cost 0.04599
step 8650, training accuracy 0.873033
step 8650, cost 185.377
step 8650, change in cost 0.0459747
step 8660, training accuracy 0.873033
step 8660, cost 185.331
step 8660, change in cost 0.0459442
step 8670, training accuracy 0.873033
step 8670, cost 185.285
step 8670, change in cost 0.0458832
step 8680, training accuracy 0.873033
step 8680, cost 185.24
step 8680, change in cost 0.0458374
step 8690, training accuracy 0.873033
step 8690, cost 185.194
step 8690, change in cost 0.0457611
step 8700, training accuracy 0.873033
step 8700, cost 185.148
step 8700, change in cost 0.0458679
step 8710, training accuracy 0.873033
step 8710, cost 185.102
step 8710, change in cost 0.0458679
step 8720, training accuracy 0.873033
step 8720, cost 185.056
step 8720, change in cost 0.0457458
step 8730, training accuracy 0.875131
step 8730, cost 185.011
step 8730, change in cost 0.0457458
step 8740, training accuracy 0.87618
step 8740, cost 184.965
step 8740, change in cost 0.0457916
step 8750, training accuracy 0.87618
step 8750, cost 184.919
step 8750, change in cost 0.0456543
step 8760, training accuracy 0.87618
step 8760, cost 184.873
step 8760, change in cost 0.0457916
step 8770, training accuracy 0.87618
step 8770, cost 184.828
step 8770, change in cost 0.0455933
step 8780, training accuracy 0.87618
step 8780, cost 184.782
step 8780, change in cost 0.0456543
step 8790, training accuracy 0.87618
step 8790, cost 184.737
step 8790, change in cost 0.045517
step 8800, training accuracy 0.87618
step 8800, cost 184.691
step 8800, change in cost 0.045578
step 8810, training accuracy 0.87618
step 8810, cost 184.645
step 8810, change in cost 0.0455322
step 8820, training accuracy 0.87618
step 8820, cost 184.6
step 8820, change in cost 0.0455933
step 8830, training accuracy 0.87618
step 8830, cost 184.554
step 8830, change in cost 0.0454407
step 8840, training accuracy 0.87618
step 8840, cost 184.509
step 8840, change in cost 0.0455017
step 8850, training accuracy 0.87618
step 8850, cost 184.463
step 8850, change in cost 0.045517
step 8860, training accuracy 0.87618
step 8860, cost 184.418
step 8860, change in cost 0.0454407
step 8870, training accuracy 0.87618
step 8870, cost 184.373
step 8870, change in cost 0.0453491
step 8880, training accuracy 0.87723
step 8880, cost 184.327
step 8880, change in cost 0.0454407
step 8890, training accuracy 0.87723
step 8890, cost 184.282
step 8890, change in cost 0.0453186
step 8900, training accuracy 0.87723
step 8900, cost 184.236
step 8900, change in cost 0.0453796
step 8910, training accuracy 0.87723
step 8910, cost 184.191
step 8910, change in cost 0.0453033
step 8920, training accuracy 0.87723
step 8920, cost 184.146
step 8920, change in cost 0.0452881
step 8930, training accuracy 0.87723
step 8930, cost 184.101
step 8930, change in cost 0.0452423
step 8940, training accuracy 0.87723
step 8940, cost 184.055
step 8940, change in cost 0.0451813
step 8950, training accuracy 0.87723
step 8950, cost 184.01
step 8950, change in cost 0.0452576
step 8960, training accuracy 0.87723
step 8960, cost 183.965
step 8960, change in cost 0.0451813
step 8970, training accuracy 0.87723
step 8970, cost 183.92
step 8970, change in cost 0.0452118
step 8980, training accuracy 0.87723
step 8980, cost 183.875
step 8980, change in cost 0.0451355
step 8990, training accuracy 0.87723
step 8990, cost 183.83
step 8990, change in cost 0.0451355
step 9000, training accuracy 0.878279
step 9000, cost 183.784
step 9000, change in cost 0.0450897
step 9010, training accuracy 0.878279
step 9010, cost 183.739
step 9010, change in cost 0.0449829
step 9020, training accuracy 0.878279
step 9020, cost 183.694
step 9020, change in cost 0.0452118
step 9030, training accuracy 0.878279
step 9030, cost 183.649
step 9030, change in cost 0.0449524
step 9040, training accuracy 0.879328
step 9040, cost 183.604
step 9040, change in cost 0.0450287
step 9050, training accuracy 0.879328
step 9050, cost 183.559
step 9050, change in cost 0.0449524
step 9060, training accuracy 0.879328
step 9060, cost 183.514
step 9060, change in cost 0.0449677
step 9070, training accuracy 0.879328
step 9070, cost 183.469
step 9070, change in cost 0.0449524
step 9080, training accuracy 0.879328
step 9080, cost 183.425
step 9080, change in cost 0.0448608
step 9090, training accuracy 0.879328
step 9090, cost 183.38
step 9090, change in cost 0.0448761
step 9100, training accuracy 0.879328
step 9100, cost 183.335
step 9100, change in cost 0.0449066
step 9110, training accuracy 0.879328
step 9110, cost 183.29
step 9110, change in cost 0.0447693
step 9120, training accuracy 0.879328
step 9120, cost 183.245
step 9120, change in cost 0.0449219
step 9130, training accuracy 0.879328
step 9130, cost 183.2
step 9130, change in cost 0.044693
step 9140, training accuracy 0.879328
step 9140, cost 183.156
step 9140, change in cost 0.0447235
step 9150, training accuracy 0.880378
step 9150, cost 183.111
step 9150, change in cost 0.0447693
step 9160, training accuracy 0.880378
step 9160, cost 183.066
step 9160, change in cost 0.0447083
step 9170, training accuracy 0.880378
step 9170, cost 183.021
step 9170, change in cost 0.0447388
step 9180, training accuracy 0.880378
step 9180, cost 182.977
step 9180, change in cost 0.0446167
step 9190, training accuracy 0.880378
step 9190, cost 182.932
step 9190, change in cost 0.0446625
step 9200, training accuracy 0.880378
step 9200, cost 182.888
step 9200, change in cost 0.0446167
step 9210, training accuracy 0.880378
step 9210, cost 182.843
step 9210, change in cost 0.0446167
step 9220, training accuracy 0.880378
step 9220, cost 182.798
step 9220, change in cost 0.0445404
step 9230, training accuracy 0.880378
step 9230, cost 182.754
step 9230, change in cost 0.0445557
step 9240, training accuracy 0.880378
step 9240, cost 182.709
step 9240, change in cost 0.0445709
step 9250, training accuracy 0.880378
step 9250, cost 182.665
step 9250, change in cost 0.0444336
step 9260, training accuracy 0.880378
step 9260, cost 182.62
step 9260, change in cost 0.0445404
step 9270, training accuracy 0.880378
step 9270, cost 182.576
step 9270, change in cost 0.0444336
step 9280, training accuracy 0.880378
step 9280, cost 182.531
step 9280, change in cost 0.0444489
step 9290, training accuracy 0.880378
step 9290, cost 182.487
step 9290, change in cost 0.0444031
step 9300, training accuracy 0.880378
step 9300, cost 182.443
step 9300, change in cost 0.044342
step 9310, training accuracy 0.881427
step 9310, cost 182.398
step 9310, change in cost 0.0443878
step 9320, training accuracy 0.880378
step 9320, cost 182.354
step 9320, change in cost 0.0443878
step 9330, training accuracy 0.880378
step 9330, cost 182.31
step 9330, change in cost 0.0442505
step 9340, training accuracy 0.880378
step 9340, cost 182.265
step 9340, change in cost 0.0443268
step 9350, training accuracy 0.880378
step 9350, cost 182.221
step 9350, change in cost 0.0442657
step 9360, training accuracy 0.880378
step 9360, cost 182.177
step 9360, change in cost 0.0442505
step 9370, training accuracy 0.880378
step 9370, cost 182.133
step 9370, change in cost 0.0442352
step 9380, training accuracy 0.882476
step 9380, cost 182.088
step 9380, change in cost 0.0441895
step 9390, training accuracy 0.882476
step 9390, cost 182.044
step 9390, change in cost 0.0441284
step 9400, training accuracy 0.882476
step 9400, cost 182
step 9400, change in cost 0.0441284
step 9410, training accuracy 0.882476
step 9410, cost 181.956
step 9410, change in cost 0.0441895
step 9420, training accuracy 0.882476
step 9420, cost 181.912
step 9420, change in cost 0.0441132
step 9430, training accuracy 0.882476
step 9430, cost 181.868
step 9430, change in cost 0.0441132
step 9440, training accuracy 0.882476
step 9440, cost 181.824
step 9440, change in cost 0.0439453
step 9450, training accuracy 0.882476
step 9450, cost 181.78
step 9450, change in cost 0.0441284
step 9460, training accuracy 0.882476
step 9460, cost 181.736
step 9460, change in cost 0.0439758
step 9470, training accuracy 0.882476
step 9470, cost 181.692
step 9470, change in cost 0.0439453
step 9480, training accuracy 0.882476
step 9480, cost 181.648
step 9480, change in cost 0.0439911
step 9490, training accuracy 0.882476
step 9490, cost 181.604
step 9490, change in cost 0.0440063
step 9500, training accuracy 0.882476
step 9500, cost 181.56
step 9500, change in cost 0.0438232
step 9510, training accuracy 0.882476
step 9510, cost 181.516
step 9510, change in cost 0.043869
step 9520, training accuracy 0.883526
step 9520, cost 181.472
step 9520, change in cost 0.0439758
step 9530, training accuracy 0.883526
step 9530, cost 181.428
step 9530, change in cost 0.0437927
step 9540, training accuracy 0.884575
step 9540, cost 181.384
step 9540, change in cost 0.0438385
step 9550, training accuracy 0.884575
step 9550, cost 181.341
step 9550, change in cost 0.0437622
step 9560, training accuracy 0.884575
step 9560, cost 181.297
step 9560, change in cost 0.0437622
step 9570, training accuracy 0.885624
step 9570, cost 181.253
step 9570, change in cost 0.0438232
step 9580, training accuracy 0.886674
step 9580, cost 181.209
step 9580, change in cost 0.0437622
step 9590, training accuracy 0.886674
step 9590, cost 181.166
step 9590, change in cost 0.0437012
step 9600, training accuracy 0.886674
step 9600, cost 181.122
step 9600, change in cost 0.0436249
step 9610, training accuracy 0.886674
step 9610, cost 181.078
step 9610, change in cost 0.0437012
step 9620, training accuracy 0.886674
step 9620, cost 181.035
step 9620, change in cost 0.0435791
step 9630, training accuracy 0.887723
step 9630, cost 180.991
step 9630, change in cost 0.0436859
step 9640, training accuracy 0.887723
step 9640, cost 180.947
step 9640, change in cost 0.0435791
step 9650, training accuracy 0.887723
step 9650, cost 180.904
step 9650, change in cost 0.0435028
step 9660, training accuracy 0.887723
step 9660, cost 180.86
step 9660, change in cost 0.0435791
step 9670, training accuracy 0.887723
step 9670, cost 180.817
step 9670, change in cost 0.0435486
step 9680, training accuracy 0.887723
step 9680, cost 180.773
step 9680, change in cost 0.0435333
step 9690, training accuracy 0.887723
step 9690, cost 180.73
step 9690, change in cost 0.0434265
step 9700, training accuracy 0.887723
step 9700, cost 180.686
step 9700, change in cost 0.0434418
step 9710, training accuracy 0.887723
step 9710, cost 180.643
step 9710, change in cost 0.0434875
step 9720, training accuracy 0.887723
step 9720, cost 180.599
step 9720, change in cost 0.0434113
step 9730, training accuracy 0.887723
step 9730, cost 180.556
step 9730, change in cost 0.043335
step 9740, training accuracy 0.887723
step 9740, cost 180.513
step 9740, change in cost 0.0433807
step 9750, training accuracy 0.887723
step 9750, cost 180.469
step 9750, change in cost 0.0434265
step 9760, training accuracy 0.887723
step 9760, cost 180.426
step 9760, change in cost 0.0432587
step 9770, training accuracy 0.887723
step 9770, cost 180.383
step 9770, change in cost 0.0433197
step 9780, training accuracy 0.887723
step 9780, cost 180.339
step 9780, change in cost 0.0433044
step 9790, training accuracy 0.887723
step 9790, cost 180.296
step 9790, change in cost 0.0431976
step 9800, training accuracy 0.887723
step 9800, cost 180.253
step 9800, change in cost 0.0432587
step 9810, training accuracy 0.888772
step 9810, cost 180.21
step 9810, change in cost 0.0431976
step 9820, training accuracy 0.888772
step 9820, cost 180.167
step 9820, change in cost 0.0431519
step 9830, training accuracy 0.888772
step 9830, cost 180.124
step 9830, change in cost 0.0431213
step 9840, training accuracy 0.888772
step 9840, cost 180.08
step 9840, change in cost 0.0431671
step 9850, training accuracy 0.888772
step 9850, cost 180.037
step 9850, change in cost 0.0431366
step 9860, training accuracy 0.888772
step 9860, cost 179.994
step 9860, change in cost 0.0431671
step 9870, training accuracy 0.888772
step 9870, cost 179.951
step 9870, change in cost 0.0430298
step 9880, training accuracy 0.888772
step 9880, cost 179.908
step 9880, change in cost 0.0429688
step 9890, training accuracy 0.888772
step 9890, cost 179.865
step 9890, change in cost 0.0430908
step 9900, training accuracy 0.888772
step 9900, cost 179.822
step 9900, change in cost 0.0430145
step 9910, training accuracy 0.888772
step 9910, cost 179.779
step 9910, change in cost 0.0429688
step 9920, training accuracy 0.888772
step 9920, cost 179.736
step 9920, change in cost 0.0429077
step 9930, training accuracy 0.888772
step 9930, cost 179.693
step 9930, change in cost 0.0429382
step 9940, training accuracy 0.888772
step 9940, cost 179.65
step 9940, change in cost 0.0429382
step 9950, training accuracy 0.888772
step 9950, cost 179.607
step 9950, change in cost 0.042923
step 9960, training accuracy 0.888772
step 9960, cost 179.564
step 9960, change in cost 0.0428467
step 9970, training accuracy 0.888772
step 9970, cost 179.522
step 9970, change in cost 0.0428314
step 9980, training accuracy 0.888772
step 9980, cost 179.479
step 9980, change in cost 0.0428619
step 9990, training accuracy 0.888772
step 9990, cost 179.436
step 9990, change in cost 0.0428009
step 10000, training accuracy 0.888772
step 10000, cost 179.393
step 10000, change in cost 0.0427399
step 10010, training accuracy 0.888772
step 10010, cost 179.35
step 10010, change in cost 0.0427551
step 10020, training accuracy 0.888772
step 10020, cost 179.308
step 10020, change in cost 0.0427551
step 10030, training accuracy 0.888772
step 10030, cost 179.265
step 10030, change in cost 0.0427246
step 10040, training accuracy 0.888772
step 10040, cost 179.222
step 10040, change in cost 0.0426483
step 10050, training accuracy 0.888772
step 10050, cost 179.18
step 10050, change in cost 0.0426788
step 10060, training accuracy 0.888772
step 10060, cost 179.137
step 10060, change in cost 0.0426178
step 10070, training accuracy 0.888772
step 10070, cost 179.094
step 10070, change in cost 0.0426483
step 10080, training accuracy 0.888772
step 10080, cost 179.052
step 10080, change in cost 0.0426178
step 10090, training accuracy 0.888772
step 10090, cost 179.009
step 10090, change in cost 0.0426331
step 10100, training accuracy 0.888772
step 10100, cost 178.967
step 10100, change in cost 0.0424805
step 10110, training accuracy 0.888772
step 10110, cost 178.924
step 10110, change in cost 0.0425568
step 10120, training accuracy 0.888772
step 10120, cost 178.882
step 10120, change in cost 0.0424957
step 10130, training accuracy 0.888772
step 10130, cost 178.839
step 10130, change in cost 0.0424652
step 10140, training accuracy 0.888772
step 10140, cost 178.797
step 10140, change in cost 0.042511
step 10150, training accuracy 0.888772
step 10150, cost 178.754
step 10150, change in cost 0.0424194
step 10160, training accuracy 0.888772
step 10160, cost 178.712
step 10160, change in cost 0.0424194
step 10170, training accuracy 0.888772
step 10170, cost 178.669
step 10170, change in cost 0.04245
step 10180, training accuracy 0.887723
step 10180, cost 178.627
step 10180, change in cost 0.0423279
step 10190, training accuracy 0.887723
step 10190, cost 178.585
step 10190, change in cost 0.0423737
step 10200, training accuracy 0.887723
step 10200, cost 178.542
step 10200, change in cost 0.0423584
step 10210, training accuracy 0.887723
step 10210, cost 178.5
step 10210, change in cost 0.0423126
step 10220, training accuracy 0.887723
step 10220, cost 178.458
step 10220, change in cost 0.0423279
step 10230, training accuracy 0.887723
step 10230, cost 178.415
step 10230, change in cost 0.0422211
step 10240, training accuracy 0.887723
step 10240, cost 178.373
step 10240, change in cost 0.0422974
step 10250, training accuracy 0.887723
step 10250, cost 178.331
step 10250, change in cost 0.0422363
step 10260, training accuracy 0.887723
step 10260, cost 178.289
step 10260, change in cost 0.0421906
step 10270, training accuracy 0.887723
step 10270, cost 178.247
step 10270, change in cost 0.0421295
step 10280, training accuracy 0.887723
step 10280, cost 178.204
step 10280, change in cost 0.0422058
step 10290, training accuracy 0.887723
step 10290, cost 178.162
step 10290, change in cost 0.0420532
step 10300, training accuracy 0.887723
step 10300, cost 178.12
step 10300, change in cost 0.0421906
step 10310, training accuracy 0.887723
step 10310, cost 178.078
step 10310, change in cost 0.0420685
step 10320, training accuracy 0.887723
step 10320, cost 178.036
step 10320, change in cost 0.0421143
step 10330, training accuracy 0.887723
step 10330, cost 177.994
step 10330, change in cost 0.042038
step 10340, training accuracy 0.887723
step 10340, cost 177.952
step 10340, change in cost 0.0419922
step 10350, training accuracy 0.887723
step 10350, cost 177.91
step 10350, change in cost 0.0420227
step 10360, training accuracy 0.887723
step 10360, cost 177.868
step 10360, change in cost 0.0420074
step 10370, training accuracy 0.887723
step 10370, cost 177.826
step 10370, change in cost 0.0419617
step 10380, training accuracy 0.887723
step 10380, cost 177.784
step 10380, change in cost 0.0419159
step 10390, training accuracy 0.887723
step 10390, cost 177.742
step 10390, change in cost 0.0419464
step 10400, training accuracy 0.887723
step 10400, cost 177.7
step 10400, change in cost 0.0419159
step 10410, training accuracy 0.887723
step 10410, cost 177.658
step 10410, change in cost 0.0418091
step 10420, training accuracy 0.887723
step 10420, cost 177.616
step 10420, change in cost 0.0418549
step 10430, training accuracy 0.887723
step 10430, cost 177.575
step 10430, change in cost 0.0419159
step 10440, training accuracy 0.887723
step 10440, cost 177.533
step 10440, change in cost 0.0417175
step 10450, training accuracy 0.887723
step 10450, cost 177.491
step 10450, change in cost 0.0418701
step 10460, training accuracy 0.887723
step 10460, cost 177.449
step 10460, change in cost 0.0417786
step 10470, training accuracy 0.888772
step 10470, cost 177.407
step 10470, change in cost 0.041687
step 10480, training accuracy 0.888772
step 10480, cost 177.366
step 10480, change in cost 0.0417023
step 10490, training accuracy 0.888772
step 10490, cost 177.324
step 10490, change in cost 0.0417023
step 10500, training accuracy 0.888772
step 10500, cost 177.282
step 10500, change in cost 0.0416565
step 10510, training accuracy 0.888772
step 10510, cost 177.241
step 10510, change in cost 0.0417175
step 10520, training accuracy 0.888772
step 10520, cost 177.199
step 10520, change in cost 0.0416107
step 10530, training accuracy 0.888772
step 10530, cost 177.157
step 10530, change in cost 0.0416412
step 10540, training accuracy 0.888772
step 10540, cost 177.116
step 10540, change in cost 0.0416412
step 10550, training accuracy 0.888772
step 10550, cost 177.074
step 10550, change in cost 0.0415497
step 10560, training accuracy 0.889822
step 10560, cost 177.033
step 10560, change in cost 0.0415344
step 10570, training accuracy 0.889822
step 10570, cost 176.991
step 10570, change in cost 0.0415039
step 10580, training accuracy 0.889822
step 10580, cost 176.95
step 10580, change in cost 0.0415344
step 10590, training accuracy 0.889822
step 10590, cost 176.908
step 10590, change in cost 0.0415039
step 10600, training accuracy 0.889822
step 10600, cost 176.867
step 10600, change in cost 0.0414124
step 10610, training accuracy 0.889822
step 10610, cost 176.825
step 10610, change in cost 0.0414886
step 10620, training accuracy 0.889822
step 10620, cost 176.784
step 10620, change in cost 0.0414276
step 10630, training accuracy 0.889822
step 10630, cost 176.742
step 10630, change in cost 0.0413971
step 10640, training accuracy 0.889822
step 10640, cost 176.701
step 10640, change in cost 0.0413971
step 10650, training accuracy 0.889822
step 10650, cost 176.66
step 10650, change in cost 0.0413513
step 10660, training accuracy 0.889822
step 10660, cost 176.618
step 10660, change in cost 0.0413361
step 10670, training accuracy 0.889822
step 10670, cost 176.577
step 10670, change in cost 0.0413208
step 10680, training accuracy 0.889822
step 10680, cost 176.536
step 10680, change in cost 0.0412598
step 10690, training accuracy 0.889822
step 10690, cost 176.495
step 10690, change in cost 0.0412903
step 10700, training accuracy 0.889822
step 10700, cost 176.453
step 10700, change in cost 0.0412445
step 10710, training accuracy 0.889822
step 10710, cost 176.412
step 10710, change in cost 0.0412445
step 10720, training accuracy 0.889822
step 10720, cost 176.371
step 10720, change in cost 0.0411987
step 10730, training accuracy 0.889822
step 10730, cost 176.33
step 10730, change in cost 0.0412292
step 10740, training accuracy 0.889822
step 10740, cost 176.288
step 10740, change in cost 0.0411224
step 10750, training accuracy 0.889822
step 10750, cost 176.247
step 10750, change in cost 0.0411682
step 10760, training accuracy 0.889822
step 10760, cost 176.206
step 10760, change in cost 0.0411224
step 10770, training accuracy 0.889822
step 10770, cost 176.165
step 10770, change in cost 0.0411224
step 10780, training accuracy 0.889822
step 10780, cost 176.124
step 10780, change in cost 0.0410767
step 10790, training accuracy 0.890871
step 10790, cost 176.083
step 10790, change in cost 0.0410309
step 10800, training accuracy 0.890871
step 10800, cost 176.042
step 10800, change in cost 0.0410614
step 10810, training accuracy 0.890871
step 10810, cost 176.001
step 10810, change in cost 0.0410156
step 10820, training accuracy 0.890871
step 10820, cost 175.96
step 10820, change in cost 0.0410309
step 10830, training accuracy 0.890871
step 10830, cost 175.919
step 10830, change in cost 0.0409546
step 10840, training accuracy 0.890871
step 10840, cost 175.878
step 10840, change in cost 0.0409698
step 10850, training accuracy 0.890871
step 10850, cost 175.837
step 10850, change in cost 0.0409393
step 10860, training accuracy 0.890871
step 10860, cost 175.796
step 10860, change in cost 0.0408783
step 10870, training accuracy 0.890871
step 10870, cost 175.755
step 10870, change in cost 0.0408936
step 10880, training accuracy 0.890871
step 10880, cost 175.714
step 10880, change in cost 0.0409088
step 10890, training accuracy 0.890871
step 10890, cost 175.674
step 10890, change in cost 0.0407562
step 10900, training accuracy 0.890871
step 10900, cost 175.633
step 10900, change in cost 0.0409088
step 10910, training accuracy 0.89192
step 10910, cost 175.592
step 10910, change in cost 0.040802
step 10920, training accuracy 0.89192
step 10920, cost 175.551
step 10920, change in cost 0.0407715
step 10930, training accuracy 0.89192
step 10930, cost 175.51
step 10930, change in cost 0.0408173
step 10940, training accuracy 0.89192
step 10940, cost 175.47
step 10940, change in cost 0.0407104
step 10950, training accuracy 0.89192
step 10950, cost 175.429
step 10950, change in cost 0.0407867
step 10960, training accuracy 0.89192
step 10960, cost 175.388
step 10960, change in cost 0.0406647
step 10970, training accuracy 0.89192
step 10970, cost 175.347
step 10970, change in cost 0.0407562
step 10980, training accuracy 0.89192
step 10980, cost 175.307
step 10980, change in cost 0.0405731
step 10990, training accuracy 0.89192
step 10990, cost 175.266
step 10990, change in cost 0.0406952
step 11000, training accuracy 0.89297
step 11000, cost 175.225
step 11000, change in cost 0.0406036
step 11010, training accuracy 0.89297
step 11010, cost 175.185
step 11010, change in cost 0.0406342
step 11020, training accuracy 0.89297
step 11020, cost 175.144
step 11020, change in cost 0.0406189
step 11030, training accuracy 0.89297
step 11030, cost 175.104
step 11030, change in cost 0.0405121
step 11040, training accuracy 0.89297
step 11040, cost 175.063
step 11040, change in cost 0.0405121
step 11050, training accuracy 0.89297
step 11050, cost 175.023
step 11050, change in cost 0.0405731
step 11060, training accuracy 0.894019
step 11060, cost 174.982
step 11060, change in cost 0.0404816
step 11070, training accuracy 0.894019
step 11070, cost 174.942
step 11070, change in cost 0.0404053
step 11080, training accuracy 0.895068
step 11080, cost 174.901
step 11080, change in cost 0.0404663
step 11090, training accuracy 0.895068
step 11090, cost 174.861
step 11090, change in cost 0.0405273
step 11100, training accuracy 0.896118
step 11100, cost 174.82
step 11100, change in cost 0.0403442
step 11110, training accuracy 0.896118
step 11110, cost 174.78
step 11110, change in cost 0.04039
step 11120, training accuracy 0.896118
step 11120, cost 174.74
step 11120, change in cost 0.0404205
step 11130, training accuracy 0.896118
step 11130, cost 174.699
step 11130, change in cost 0.04039
step 11140, training accuracy 0.896118
step 11140, cost 174.659
step 11140, change in cost 0.0402679
step 11150, training accuracy 0.896118
step 11150, cost 174.619
step 11150, change in cost 0.0403442
step 11160, training accuracy 0.896118
step 11160, cost 174.578
step 11160, change in cost 0.0402679
step 11170, training accuracy 0.896118
step 11170, cost 174.538
step 11170, change in cost 0.0402832
step 11180, training accuracy 0.896118
step 11180, cost 174.498
step 11180, change in cost 0.0402374
step 11190, training accuracy 0.896118
step 11190, cost 174.458
step 11190, change in cost 0.0402679
step 11200, training accuracy 0.896118
step 11200, cost 174.417
step 11200, change in cost 0.0402222
step 11210, training accuracy 0.896118
step 11210, cost 174.377
step 11210, change in cost 0.0401611
step 11220, training accuracy 0.896118
step 11220, cost 174.337
step 11220, change in cost 0.0401611
step 11230, training accuracy 0.896118
step 11230, cost 174.297
step 11230, change in cost 0.0401154
step 11240, training accuracy 0.896118
step 11240, cost 174.257
step 11240, change in cost 0.0401611
step 11250, training accuracy 0.896118
step 11250, cost 174.217
step 11250, change in cost 0.0401154
step 11260, training accuracy 0.896118
step 11260, cost 174.176
step 11260, change in cost 0.0400696
step 11270, training accuracy 0.896118
step 11270, cost 174.136
step 11270, change in cost 0.0400696
step 11280, training accuracy 0.896118
step 11280, cost 174.096
step 11280, change in cost 0.0400391
step 11290, training accuracy 0.897167
step 11290, cost 174.056
step 11290, change in cost 0.0400543
step 11300, training accuracy 0.897167
step 11300, cost 174.016
step 11300, change in cost 0.0400238
step 11310, training accuracy 0.897167
step 11310, cost 173.976
step 11310, change in cost 0.039917
step 11320, training accuracy 0.897167
step 11320, cost 173.936
step 11320, change in cost 0.0400238
step 11330, training accuracy 0.897167
step 11330, cost 173.896
step 11330, change in cost 0.0398865
step 11340, training accuracy 0.897167
step 11340, cost 173.857
step 11340, change in cost 0.0399628
step 11350, training accuracy 0.897167
step 11350, cost 173.817
step 11350, change in cost 0.0398865
step 11360, training accuracy 0.897167
step 11360, cost 173.777
step 11360, change in cost 0.039917
step 11370, training accuracy 0.897167
step 11370, cost 173.737
step 11370, change in cost 0.039856
step 11380, training accuracy 0.898216
step 11380, cost 173.697
step 11380, change in cost 0.0398407
step 11390, training accuracy 0.898216
step 11390, cost 173.657
step 11390, change in cost 0.0398102
step 11400, training accuracy 0.898216
step 11400, cost 173.617
step 11400, change in cost 0.0398102
step 11410, training accuracy 0.898216
step 11410, cost 173.578
step 11410, change in cost 0.039856
step 11420, training accuracy 0.898216
step 11420, cost 173.538
step 11420, change in cost 0.0397186
step 11430, training accuracy 0.898216
step 11430, cost 173.498
step 11430, change in cost 0.0397339
step 11440, training accuracy 0.898216
step 11440, cost 173.458
step 11440, change in cost 0.0397644
step 11450, training accuracy 0.898216
step 11450, cost 173.419
step 11450, change in cost 0.0396118
step 11460, training accuracy 0.899265
step 11460, cost 173.379
step 11460, change in cost 0.0397186
step 11470, training accuracy 0.899265
step 11470, cost 173.339
step 11470, change in cost 0.0396423
step 11480, training accuracy 0.899265
step 11480, cost 173.3
step 11480, change in cost 0.0396118
step 11490, training accuracy 0.899265
step 11490, cost 173.26
step 11490, change in cost 0.0397186
step 11500, training accuracy 0.899265
step 11500, cost 173.221
step 11500, change in cost 0.0395203
step 11510, training accuracy 0.899265
step 11510, cost 173.181
step 11510, change in cost 0.0395813
step 11520, training accuracy 0.899265
step 11520, cost 173.141
step 11520, change in cost 0.0396576
step 11530, training accuracy 0.899265
step 11530, cost 173.102
step 11530, change in cost 0.0394592
step 11540, training accuracy 0.899265
step 11540, cost 173.062
step 11540, change in cost 0.0395508
step 11550, training accuracy 0.899265
step 11550, cost 173.023
step 11550, change in cost 0.0394745
step 11560, training accuracy 0.899265
step 11560, cost 172.983
step 11560, change in cost 0.0395203
step 11570, training accuracy 0.899265
step 11570, cost 172.944
step 11570, change in cost 0.0394135
step 11580, training accuracy 0.900315
step 11580, cost 172.904
step 11580, change in cost 0.0394745
step 11590, training accuracy 0.900315
step 11590, cost 172.865
step 11590, change in cost 0.0394287
step 11600, training accuracy 0.900315
step 11600, cost 172.826
step 11600, change in cost 0.0393982
step 11610, training accuracy 0.900315
step 11610, cost 172.786
step 11610, change in cost 0.0393829
step 11620, training accuracy 0.901364
step 11620, cost 172.747
step 11620, change in cost 0.0393066
step 11630, training accuracy 0.901364
step 11630, cost 172.707
step 11630, change in cost 0.0393829
step 11640, training accuracy 0.901364
step 11640, cost 172.668
step 11640, change in cost 0.0393066
step 11650, training accuracy 0.901364
step 11650, cost 172.629
step 11650, change in cost 0.0393524
step 11660, training accuracy 0.901364
step 11660, cost 172.589
step 11660, change in cost 0.0393372
step 11670, training accuracy 0.901364
step 11670, cost 172.55
step 11670, change in cost 0.0391541
step 11680, training accuracy 0.901364
step 11680, cost 172.511
step 11680, change in cost 0.0393219
step 11690, training accuracy 0.903463
step 11690, cost 172.472
step 11690, change in cost 0.0391541
step 11700, training accuracy 0.903463
step 11700, cost 172.433
step 11700, change in cost 0.0392151
step 11710, training accuracy 0.903463
step 11710, cost 172.393
step 11710, change in cost 0.0391846
step 11720, training accuracy 0.903463
step 11720, cost 172.354
step 11720, change in cost 0.0392303
step 11730, training accuracy 0.903463
step 11730, cost 172.315
step 11730, change in cost 0.0390778
step 11740, training accuracy 0.903463
step 11740, cost 172.276
step 11740, change in cost 0.0391846
step 11750, training accuracy 0.903463
step 11750, cost 172.237
step 11750, change in cost 0.039093
step 11760, training accuracy 0.903463
step 11760, cost 172.198
step 11760, change in cost 0.0390015
step 11770, training accuracy 0.903463
step 11770, cost 172.159
step 11770, change in cost 0.039093
step 11780, training accuracy 0.903463
step 11780, cost 172.12
step 11780, change in cost 0.039093
step 11790, training accuracy 0.903463
step 11790, cost 172.081
step 11790, change in cost 0.0390167
step 11800, training accuracy 0.903463
step 11800, cost 172.042
step 11800, change in cost 0.039032
step 11810, training accuracy 0.903463
step 11810, cost 172.003
step 11810, change in cost 0.0389862
step 11820, training accuracy 0.903463
step 11820, cost 171.964
step 11820, change in cost 0.0389557
step 11830, training accuracy 0.903463
step 11830, cost 171.925
step 11830, change in cost 0.0388947
step 11840, training accuracy 0.903463
step 11840, cost 171.886
step 11840, change in cost 0.0389862
step 11850, training accuracy 0.903463
step 11850, cost 171.847
step 11850, change in cost 0.0388489
step 11860, training accuracy 0.903463
step 11860, cost 171.808
step 11860, change in cost 0.0389099
step 11870, training accuracy 0.903463
step 11870, cost 171.769
step 11870, change in cost 0.0388489
step 11880, training accuracy 0.903463
step 11880, cost 171.73
step 11880, change in cost 0.0388641
step 11890, training accuracy 0.903463
step 11890, cost 171.691
step 11890, change in cost 0.0388794
step 11900, training accuracy 0.903463
step 11900, cost 171.653
step 11900, change in cost 0.0387878
step 11910, training accuracy 0.903463
step 11910, cost 171.614
step 11910, change in cost 0.0387878
step 11920, training accuracy 0.903463
step 11920, cost 171.575
step 11920, change in cost 0.0387878
step 11930, training accuracy 0.903463
step 11930, cost 171.536
step 11930, change in cost 0.0387268
step 11940, training accuracy 0.903463
step 11940, cost 171.498
step 11940, change in cost 0.0386963
step 11950, training accuracy 0.903463
step 11950, cost 171.459
step 11950, change in cost 0.0387573
step 11960, training accuracy 0.903463
step 11960, cost 171.42
step 11960, change in cost 0.0387268
step 11970, training accuracy 0.903463
step 11970, cost 171.381
step 11970, change in cost 0.0387115
step 11980, training accuracy 0.903463
step 11980, cost 171.343
step 11980, change in cost 0.0386658
step 11990, training accuracy 0.903463
step 11990, cost 171.304
step 11990, change in cost 0.0385132
step 12000, training accuracy 0.903463
step 12000, cost 171.266
step 12000, change in cost 0.0386963
step 12010, training accuracy 0.903463
step 12010, cost 171.227
step 12010, change in cost 0.0385742
step 12020, training accuracy 0.903463
step 12020, cost 171.189
step 12020, change in cost 0.0385132
step 12030, training accuracy 0.903463
step 12030, cost 171.15
step 12030, change in cost 0.0385742
step 12040, training accuracy 0.904512
step 12040, cost 171.111
step 12040, change in cost 0.0385742
step 12050, training accuracy 0.904512
step 12050, cost 171.073
step 12050, change in cost 0.0385132
step 12060, training accuracy 0.904512
step 12060, cost 171.034
step 12060, change in cost 0.0385284
step 12070, training accuracy 0.904512
step 12070, cost 170.996
step 12070, change in cost 0.0384674
step 12080, training accuracy 0.904512
step 12080, cost 170.957
step 12080, change in cost 0.0384369
step 12090, training accuracy 0.904512
step 12090, cost 170.919
step 12090, change in cost 0.0385132
step 12100, training accuracy 0.904512
step 12100, cost 170.881
step 12100, change in cost 0.0383606
step 12110, training accuracy 0.904512
step 12110, cost 170.842
step 12110, change in cost 0.0384521
step 12120, training accuracy 0.904512
step 12120, cost 170.804
step 12120, change in cost 0.0383148
step 12130, training accuracy 0.904512
step 12130, cost 170.765
step 12130, change in cost 0.0384674
step 12140, training accuracy 0.904512
step 12140, cost 170.727
step 12140, change in cost 0.038269
step 12150, training accuracy 0.904512
step 12150, cost 170.689
step 12150, change in cost 0.0384064
step 12160, training accuracy 0.903463
step 12160, cost 170.65
step 12160, change in cost 0.0382843
step 12170, training accuracy 0.903463
step 12170, cost 170.612
step 12170, change in cost 0.0382843
step 12180, training accuracy 0.903463
step 12180, cost 170.574
step 12180, change in cost 0.038269
step 12190, training accuracy 0.903463
step 12190, cost 170.536
step 12190, change in cost 0.0382385
step 12200, training accuracy 0.903463
step 12200, cost 170.497
step 12200, change in cost 0.038269
step 12210, training accuracy 0.903463
step 12210, cost 170.459
step 12210, change in cost 0.0381927
step 12220, training accuracy 0.903463
step 12220, cost 170.421
step 12220, change in cost 0.038208
step 12230, training accuracy 0.903463
step 12230, cost 170.383
step 12230, change in cost 0.0381775
step 12240, training accuracy 0.903463
step 12240, cost 170.345
step 12240, change in cost 0.0381317
step 12250, training accuracy 0.903463
step 12250, cost 170.306
step 12250, change in cost 0.0381317
step 12260, training accuracy 0.903463
step 12260, cost 170.268
step 12260, change in cost 0.038147
step 12270, training accuracy 0.903463
step 12270, cost 170.23
step 12270, change in cost 0.0381012
step 12280, training accuracy 0.903463
step 12280, cost 170.192
step 12280, change in cost 0.0380707
step 12290, training accuracy 0.903463
step 12290, cost 170.154
step 12290, change in cost 0.0380402
step 12300, training accuracy 0.903463
step 12300, cost 170.116
step 12300, change in cost 0.0381165
step 12310, training accuracy 0.903463
step 12310, cost 170.078
step 12310, change in cost 0.0380096
step 12320, training accuracy 0.903463
step 12320, cost 170.04
step 12320, change in cost 0.0379791
step 12330, training accuracy 0.903463
step 12330, cost 170.002
step 12330, change in cost 0.0379333
step 12340, training accuracy 0.903463
step 12340, cost 169.964
step 12340, change in cost 0.0380402
step 12350, training accuracy 0.904512
step 12350, cost 169.926
step 12350, change in cost 0.0379181
step 12360, training accuracy 0.904512
step 12360, cost 169.888
step 12360, change in cost 0.0379333
step 12370, training accuracy 0.904512
step 12370, cost 169.85
step 12370, change in cost 0.0379028
step 12380, training accuracy 0.904512
step 12380, cost 169.812
step 12380, change in cost 0.0378571
step 12390, training accuracy 0.904512
step 12390, cost 169.774
step 12390, change in cost 0.0379333
step 12400, training accuracy 0.904512
step 12400, cost 169.737
step 12400, change in cost 0.0378571
step 12410, training accuracy 0.904512
step 12410, cost 169.699
step 12410, change in cost 0.0378418
step 12420, training accuracy 0.904512
step 12420, cost 169.661
step 12420, change in cost 0.037796
step 12430, training accuracy 0.904512
step 12430, cost 169.623
step 12430, change in cost 0.0378265
step 12440, training accuracy 0.904512
step 12440, cost 169.585
step 12440, change in cost 0.0377045
step 12450, training accuracy 0.905561
step 12450, cost 169.548
step 12450, change in cost 0.0378113
step 12460, training accuracy 0.905561
step 12460, cost 169.51
step 12460, change in cost 0.0377502
step 12470, training accuracy 0.905561
step 12470, cost 169.472
step 12470, change in cost 0.0377197
step 12480, training accuracy 0.905561
step 12480, cost 169.434
step 12480, change in cost 0.0377197
step 12490, training accuracy 0.905561
step 12490, cost 169.397
step 12490, change in cost 0.037674
step 12500, training accuracy 0.905561
step 12500, cost 169.359
step 12500, change in cost 0.0376892
step 12510, training accuracy 0.905561
step 12510, cost 169.321
step 12510, change in cost 0.0376129
step 12520, training accuracy 0.905561
step 12520, cost 169.284
step 12520, change in cost 0.0376434
step 12530, training accuracy 0.90766
step 12530, cost 169.246
step 12530, change in cost 0.0376282
step 12540, training accuracy 0.90766
step 12540, cost 169.209
step 12540, change in cost 0.0376129
step 12550, training accuracy 0.90766
step 12550, cost 169.171
step 12550, change in cost 0.0375519
step 12560, training accuracy 0.90766
step 12560, cost 169.133
step 12560, change in cost 0.0375366
step 12570, training accuracy 0.90766
step 12570, cost 169.096
step 12570, change in cost 0.0375671
step 12580, training accuracy 0.90766
step 12580, cost 169.058
step 12580, change in cost 0.0374603
step 12590, training accuracy 0.90766
step 12590, cost 169.021
step 12590, change in cost 0.0375519
step 12600, training accuracy 0.90766
step 12600, cost 168.983
step 12600, change in cost 0.0374146
step 12610, training accuracy 0.90766
step 12610, cost 168.946
step 12610, change in cost 0.0375214
step 12620, training accuracy 0.90766
step 12620, cost 168.909
step 12620, change in cost 0.0374451
step 12630, training accuracy 0.90766
step 12630, cost 168.871
step 12630, change in cost 0.0374146
step 12640, training accuracy 0.90766
step 12640, cost 168.834
step 12640, change in cost 0.0373993
step 12650, training accuracy 0.90766
step 12650, cost 168.796
step 12650, change in cost 0.0374146
step 12660, training accuracy 0.908709
step 12660, cost 168.759
step 12660, change in cost 0.0373688
step 12670, training accuracy 0.908709
step 12670, cost 168.722
step 12670, change in cost 0.0372925
step 12680, training accuracy 0.908709
step 12680, cost 168.684
step 12680, change in cost 0.0374146
step 12690, training accuracy 0.908709
step 12690, cost 168.647
step 12690, change in cost 0.037262
step 12700, training accuracy 0.908709
step 12700, cost 168.61
step 12700, change in cost 0.0372925
step 12710, training accuracy 0.908709
step 12710, cost 168.572
step 12710, change in cost 0.0372925
step 12720, training accuracy 0.909759
step 12720, cost 168.535
step 12720, change in cost 0.0372925
step 12730, training accuracy 0.909759
step 12730, cost 168.498
step 12730, change in cost 0.0372467
step 12740, training accuracy 0.909759
step 12740, cost 168.461
step 12740, change in cost 0.0372467
step 12750, training accuracy 0.909759
step 12750, cost 168.423
step 12750, change in cost 0.0371704
step 12760, training accuracy 0.909759
step 12760, cost 168.386
step 12760, change in cost 0.0371704
step 12770, training accuracy 0.909759
step 12770, cost 168.349
step 12770, change in cost 0.037262
step 12780, training accuracy 0.909759
step 12780, cost 168.312
step 12780, change in cost 0.0370789
step 12790, training accuracy 0.909759
step 12790, cost 168.275
step 12790, change in cost 0.0371094
step 12800, training accuracy 0.909759
step 12800, cost 168.238
step 12800, change in cost 0.0371552
step 12810, training accuracy 0.909759
step 12810, cost 168.201
step 12810, change in cost 0.0370483
step 12820, training accuracy 0.909759
step 12820, cost 168.163
step 12820, change in cost 0.0371246
step 12830, training accuracy 0.909759
step 12830, cost 168.126
step 12830, change in cost 0.0370026
step 12840, training accuracy 0.909759
step 12840, cost 168.089
step 12840, change in cost 0.0371399
step 12850, training accuracy 0.909759
step 12850, cost 168.052
step 12850, change in cost 0.0370178
step 12860, training accuracy 0.909759
step 12860, cost 168.015
step 12860, change in cost 0.0369568
step 12870, training accuracy 0.909759
step 12870, cost 167.978
step 12870, change in cost 0.036972
step 12880, training accuracy 0.909759
step 12880, cost 167.941
step 12880, change in cost 0.0369873
step 12890, training accuracy 0.909759
step 12890, cost 167.904
step 12890, change in cost 0.036911
step 12900, training accuracy 0.909759
step 12900, cost 167.868
step 12900, change in cost 0.0369568
step 12910, training accuracy 0.909759
step 12910, cost 167.831
step 12910, change in cost 0.0369568
step 12920, training accuracy 0.909759
step 12920, cost 167.794
step 12920, change in cost 0.0368805
step 12930, training accuracy 0.909759
step 12930, cost 167.757
step 12930, change in cost 0.0368347
step 12940, training accuracy 0.909759
step 12940, cost 167.72
step 12940, change in cost 0.036911
step 12950, training accuracy 0.909759
step 12950, cost 167.683
step 12950, change in cost 0.0368652
step 12960, training accuracy 0.909759
step 12960, cost 167.646
step 12960, change in cost 0.0367584
step 12970, training accuracy 0.909759
step 12970, cost 167.609
step 12970, change in cost 0.0368195
step 12980, training accuracy 0.909759
step 12980, cost 167.573
step 12980, change in cost 0.0367737
step 12990, training accuracy 0.909759
step 12990, cost 167.536
step 12990, change in cost 0.0367584
step 13000, training accuracy 0.910808
step 13000, cost 167.499
step 13000, change in cost 0.0367889
step 13010, training accuracy 0.910808
step 13010, cost 167.462
step 13010, change in cost 0.0367279
step 13020, training accuracy 0.910808
step 13020, cost 167.426
step 13020, change in cost 0.0367279
step 13030, training accuracy 0.910808
step 13030, cost 167.389
step 13030, change in cost 0.0367279
step 13040, training accuracy 0.910808
step 13040, cost 167.352
step 13040, change in cost 0.0366516
step 13050, training accuracy 0.911857
step 13050, cost 167.316
step 13050, change in cost 0.0366821
step 13060, training accuracy 0.911857
step 13060, cost 167.279
step 13060, change in cost 0.0366364
step 13070, training accuracy 0.911857
step 13070, cost 167.242
step 13070, change in cost 0.0365601
step 13080, training accuracy 0.911857
step 13080, cost 167.206
step 13080, change in cost 0.0366669
step 13090, training accuracy 0.911857
step 13090, cost 167.169
step 13090, change in cost 0.0365601
step 13100, training accuracy 0.911857
step 13100, cost 167.133
step 13100, change in cost 0.0366058
step 13110, training accuracy 0.912907
step 13110, cost 167.096
step 13110, change in cost 0.0365448
step 13120, training accuracy 0.912907
step 13120, cost 167.06
step 13120, change in cost 0.036499
step 13130, training accuracy 0.912907
step 13130, cost 167.023
step 13130, change in cost 0.0365295
step 13140, training accuracy 0.912907
step 13140, cost 166.987
step 13140, change in cost 0.0365295
step 13150, training accuracy 0.912907
step 13150, cost 166.95
step 13150, change in cost 0.036438
step 13160, training accuracy 0.912907
step 13160, cost 166.914
step 13160, change in cost 0.036499
step 13170, training accuracy 0.912907
step 13170, cost 166.877
step 13170, change in cost 0.0363617
step 13180, training accuracy 0.912907
step 13180, cost 166.841
step 13180, change in cost 0.0364838
step 13190, training accuracy 0.912907
step 13190, cost 166.804
step 13190, change in cost 0.0364227
step 13200, training accuracy 0.912907
step 13200, cost 166.768
step 13200, change in cost 0.0364532
step 13210, training accuracy 0.912907
step 13210, cost 166.732
step 13210, change in cost 0.0363464
step 13220, training accuracy 0.912907
step 13220, cost 166.695
step 13220, change in cost 0.0363007
step 13230, training accuracy 0.912907
step 13230, cost 166.659
step 13230, change in cost 0.0363922
step 13240, training accuracy 0.912907
step 13240, cost 166.622
step 13240, change in cost 0.0363617
step 13250, training accuracy 0.912907
step 13250, cost 166.586
step 13250, change in cost 0.0362701
step 13260, training accuracy 0.912907
step 13260, cost 166.55
step 13260, change in cost 0.0362549
step 13270, training accuracy 0.912907
step 13270, cost 166.514
step 13270, change in cost 0.0363159
step 13280, training accuracy 0.912907
step 13280, cost 166.477
step 13280, change in cost 0.0361633
step 13290, training accuracy 0.912907
step 13290, cost 166.441
step 13290, change in cost 0.0362701
step 13300, training accuracy 0.912907
step 13300, cost 166.405
step 13300, change in cost 0.0361786
step 13310, training accuracy 0.912907
step 13310, cost 166.369
step 13310, change in cost 0.0362854
step 13320, training accuracy 0.912907
step 13320, cost 166.333
step 13320, change in cost 0.0361176
step 13330, training accuracy 0.912907
step 13330, cost 166.296
step 13330, change in cost 0.0362091
step 13340, training accuracy 0.912907
step 13340, cost 166.26
step 13340, change in cost 0.0361786
step 13350, training accuracy 0.912907
step 13350, cost 166.224
step 13350, change in cost 0.036087
step 13360, training accuracy 0.912907
step 13360, cost 166.188
step 13360, change in cost 0.0361328
step 13370, training accuracy 0.912907
step 13370, cost 166.152
step 13370, change in cost 0.0360565
step 13380, training accuracy 0.912907
step 13380, cost 166.116
step 13380, change in cost 0.036087
step 13390, training accuracy 0.912907
step 13390, cost 166.08
step 13390, change in cost 0.0360718
step 13400, training accuracy 0.912907
step 13400, cost 166.044
step 13400, change in cost 0.0360718
step 13410, training accuracy 0.912907
step 13410, cost 166.008
step 13410, change in cost 0.036026
step 13420, training accuracy 0.912907
step 13420, cost 165.972
step 13420, change in cost 0.0359802
step 13430, training accuracy 0.912907
step 13430, cost 165.936
step 13430, change in cost 0.0359344
step 13440, training accuracy 0.912907
step 13440, cost 165.9
step 13440, change in cost 0.0359802
step 13450, training accuracy 0.912907
step 13450, cost 165.864
step 13450, change in cost 0.0359497
step 13460, training accuracy 0.912907
step 13460, cost 165.828
step 13460, change in cost 0.035965
step 13470, training accuracy 0.912907
step 13470, cost 165.792
step 13470, change in cost 0.0360107
step 13480, training accuracy 0.912907
step 13480, cost 165.756
step 13480, change in cost 0.0357971
step 13490, training accuracy 0.912907
step 13490, cost 165.72
step 13490, change in cost 0.0359039
step 13500, training accuracy 0.912907
step 13500, cost 165.684
step 13500, change in cost 0.0358887
step 13510, training accuracy 0.912907
step 13510, cost 165.648
step 13510, change in cost 0.0358734
step 13520, training accuracy 0.912907
step 13520, cost 165.613
step 13520, change in cost 0.0358276
step 13530, training accuracy 0.912907
step 13530, cost 165.577
step 13530, change in cost 0.0358124
step 13540, training accuracy 0.912907
step 13540, cost 165.541
step 13540, change in cost 0.0357971
step 13550, training accuracy 0.912907
step 13550, cost 165.505
step 13550, change in cost 0.0357666
step 13560, training accuracy 0.912907
step 13560, cost 165.469
step 13560, change in cost 0.0357056
step 13570, training accuracy 0.912907
step 13570, cost 165.434
step 13570, change in cost 0.0357971
step 13580, training accuracy 0.911857
step 13580, cost 165.398
step 13580, change in cost 0.035675
step 13590, training accuracy 0.911857
step 13590, cost 165.362
step 13590, change in cost 0.0357971
step 13600, training accuracy 0.911857
step 13600, cost 165.327
step 13600, change in cost 0.035675
step 13610, training accuracy 0.911857
step 13610, cost 165.291
step 13610, change in cost 0.0357361
step 13620, training accuracy 0.911857
step 13620, cost 165.255
step 13620, change in cost 0.0355835
step 13630, training accuracy 0.911857
step 13630, cost 165.22
step 13630, change in cost 0.0356445
step 13640, training accuracy 0.911857
step 13640, cost 165.184
step 13640, change in cost 0.0356293
step 13650, training accuracy 0.911857
step 13650, cost 165.148
step 13650, change in cost 0.0356445
step 13660, training accuracy 0.911857
step 13660, cost 165.113
step 13660, change in cost 0.0355377
step 13670, training accuracy 0.911857
step 13670, cost 165.077
step 13670, change in cost 0.0356445
step 13680, training accuracy 0.911857
step 13680, cost 165.042
step 13680, change in cost 0.035553
step 13690, training accuracy 0.911857
step 13690, cost 165.006
step 13690, change in cost 0.0354767
step 13700, training accuracy 0.911857
step 13700, cost 164.971
step 13700, change in cost 0.035553
step 13710, training accuracy 0.911857
step 13710, cost 164.935
step 13710, change in cost 0.0354767
step 13720, training accuracy 0.911857
step 13720, cost 164.9
step 13720, change in cost 0.0354919
step 13730, training accuracy 0.911857
step 13730, cost 164.864
step 13730, change in cost 0.0354767
step 13740, training accuracy 0.911857
step 13740, cost 164.829
step 13740, change in cost 0.0354309
step 13750, training accuracy 0.911857
step 13750, cost 164.793
step 13750, change in cost 0.0354614
step 13760, training accuracy 0.911857
step 13760, cost 164.758
step 13760, change in cost 0.0354156
step 13770, training accuracy 0.911857
step 13770, cost 164.722
step 13770, change in cost 0.0354462
step 13780, training accuracy 0.911857
step 13780, cost 164.687
step 13780, change in cost 0.0353088
step 13790, training accuracy 0.911857
step 13790, cost 164.652
step 13790, change in cost 0.0354309
step 13800, training accuracy 0.911857
step 13800, cost 164.616
step 13800, change in cost 0.0353394
step 13810, training accuracy 0.911857
step 13810, cost 164.581
step 13810, change in cost 0.0353394
step 13820, training accuracy 0.911857
step 13820, cost 164.546
step 13820, change in cost 0.0353241
step 13830, training accuracy 0.911857
step 13830, cost 164.51
step 13830, change in cost 0.0352631
step 13840, training accuracy 0.911857
step 13840, cost 164.475
step 13840, change in cost 0.0353241
step 13850, training accuracy 0.912907
step 13850, cost 164.44
step 13850, change in cost 0.0352478
step 13860, training accuracy 0.912907
step 13860, cost 164.405
step 13860, change in cost 0.0352631
step 13870, training accuracy 0.912907
step 13870, cost 164.369
step 13870, change in cost 0.0352325
step 13880, training accuracy 0.913956
step 13880, cost 164.334
step 13880, change in cost 0.035202
step 13890, training accuracy 0.913956
step 13890, cost 164.299
step 13890, change in cost 0.0351868
step 13900, training accuracy 0.913956
step 13900, cost 164.264
step 13900, change in cost 0.0351868
step 13910, training accuracy 0.913956
step 13910, cost 164.229
step 13910, change in cost 0.035141
step 13920, training accuracy 0.913956
step 13920, cost 164.193
step 13920, change in cost 0.035141
step 13930, training accuracy 0.913956
step 13930, cost 164.158
step 13930, change in cost 0.0351868
step 13940, training accuracy 0.913956
step 13940, cost 164.123
step 13940, change in cost 0.03508
step 13950, training accuracy 0.913956
step 13950, cost 164.088
step 13950, change in cost 0.0350647
step 13960, training accuracy 0.913956
step 13960, cost 164.053
step 13960, change in cost 0.0350647
step 13970, training accuracy 0.913956
step 13970, cost 164.018
step 13970, change in cost 0.0351257
step 13980, training accuracy 0.913956
step 13980, cost 163.983
step 13980, change in cost 0.0349884
step 13990, training accuracy 0.913956
step 13990, cost 163.948
step 13990, change in cost 0.03508
step 14000, training accuracy 0.913956
step 14000, cost 163.913
step 14000, change in cost 0.0349731
step 14010, training accuracy 0.913956
step 14010, cost 163.878
step 14010, change in cost 0.0350342
step 14020, training accuracy 0.913956
step 14020, cost 163.843
step 14020, change in cost 0.0350342
step 14030, training accuracy 0.913956
step 14030, cost 163.808
step 14030, change in cost 0.0349579
step 14040, training accuracy 0.913956
step 14040, cost 163.773
step 14040, change in cost 0.0349274
step 14050, training accuracy 0.915005
step 14050, cost 163.738
step 14050, change in cost 0.0348969
step 14060, training accuracy 0.915005
step 14060, cost 163.703
step 14060, change in cost 0.0349884
step 14070, training accuracy 0.915005
step 14070, cost 163.668
step 14070, change in cost 0.0348511
step 14080, training accuracy 0.915005
step 14080, cost 163.633
step 14080, change in cost 0.0348816
step 14090, training accuracy 0.915005
step 14090, cost 163.598
step 14090, change in cost 0.0348358
step 14100, training accuracy 0.915005
step 14100, cost 163.564
step 14100, change in cost 0.0348358
step 14110, training accuracy 0.915005
step 14110, cost 163.529
step 14110, change in cost 0.0348511
step 14120, training accuracy 0.915005
step 14120, cost 163.494
step 14120, change in cost 0.0347443
step 14130, training accuracy 0.915005
step 14130, cost 163.459
step 14130, change in cost 0.0349121
step 14140, training accuracy 0.915005
step 14140, cost 163.424
step 14140, change in cost 0.034729
step 14150, training accuracy 0.915005
step 14150, cost 163.39
step 14150, change in cost 0.0347748
step 14160, training accuracy 0.915005
step 14160, cost 163.355
step 14160, change in cost 0.0347443
step 14170, training accuracy 0.915005
step 14170, cost 163.32
step 14170, change in cost 0.0347595
step 14180, training accuracy 0.915005
step 14180, cost 163.285
step 14180, change in cost 0.0346527
step 14190, training accuracy 0.915005
step 14190, cost 163.251
step 14190, change in cost 0.0347137
step 14200, training accuracy 0.915005
step 14200, cost 163.216
step 14200, change in cost 0.0347595
step 14210, training accuracy 0.915005
step 14210, cost 163.181
step 14210, change in cost 0.0346527
step 14220, training accuracy 0.915005
step 14220, cost 163.147
step 14220, change in cost 0.0346527
step 14230, training accuracy 0.916055
step 14230, cost 163.112
step 14230, change in cost 0.0345764
step 14240, training accuracy 0.916055
step 14240, cost 163.077
step 14240, change in cost 0.0345917
step 14250, training accuracy 0.916055
step 14250, cost 163.043
step 14250, change in cost 0.0345917
step 14260, training accuracy 0.916055
step 14260, cost 163.008
step 14260, change in cost 0.0346527
step 14270, training accuracy 0.916055
step 14270, cost 162.974
step 14270, change in cost 0.0345001
step 14280, training accuracy 0.916055
step 14280, cost 162.939
step 14280, change in cost 0.0344849
step 14290, training accuracy 0.916055
step 14290, cost 162.905
step 14290, change in cost 0.0346375
step 14300, training accuracy 0.916055
step 14300, cost 162.87
step 14300, change in cost 0.0345459
step 14310, training accuracy 0.916055
step 14310, cost 162.836
step 14310, change in cost 0.0344391
step 14320, training accuracy 0.916055
step 14320, cost 162.801
step 14320, change in cost 0.0345154
step 14330, training accuracy 0.916055
step 14330, cost 162.767
step 14330, change in cost 0.0344543
step 14340, training accuracy 0.916055
step 14340, cost 162.732
step 14340, change in cost 0.0344849
step 14350, training accuracy 0.916055
step 14350, cost 162.698
step 14350, change in cost 0.0343628
step 14360, training accuracy 0.916055
step 14360, cost 162.663
step 14360, change in cost 0.0344391
step 14370, training accuracy 0.916055
step 14370, cost 162.629
step 14370, change in cost 0.0344238
step 14380, training accuracy 0.916055
step 14380, cost 162.595
step 14380, change in cost 0.0343781
step 14390, training accuracy 0.916055
step 14390, cost 162.56
step 14390, change in cost 0.0343475
step 14400, training accuracy 0.916055
step 14400, cost 162.526
step 14400, change in cost 0.0343628
step 14410, training accuracy 0.916055
step 14410, cost 162.492
step 14410, change in cost 0.034317
step 14420, training accuracy 0.916055
step 14420, cost 162.457
step 14420, change in cost 0.0343781
step 14430, training accuracy 0.916055
step 14430, cost 162.423
step 14430, change in cost 0.0343018
step 14440, training accuracy 0.917104
step 14440, cost 162.389
step 14440, change in cost 0.0342407
step 14450, training accuracy 0.917104
step 14450, cost 162.354
step 14450, change in cost 0.0343018
step 14460, training accuracy 0.918153
step 14460, cost 162.32
step 14460, change in cost 0.034256
step 14470, training accuracy 0.918153
step 14470, cost 162.286
step 14470, change in cost 0.0342102
step 14480, training accuracy 0.918153
step 14480, cost 162.252
step 14480, change in cost 0.0342255
step 14490, training accuracy 0.918153
step 14490, cost 162.217
step 14490, change in cost 0.0341949
step 14500, training accuracy 0.918153
step 14500, cost 162.183
step 14500, change in cost 0.0342255
step 14510, training accuracy 0.918153
step 14510, cost 162.149
step 14510, change in cost 0.0341644
step 14520, training accuracy 0.918153
step 14520, cost 162.115
step 14520, change in cost 0.0341644
step 14530, training accuracy 0.918153
step 14530, cost 162.081
step 14530, change in cost 0.0341187
step 14540, training accuracy 0.918153
step 14540, cost 162.047
step 14540, change in cost 0.0341187
step 14550, training accuracy 0.918153
step 14550, cost 162.013
step 14550, change in cost 0.0341339
step 14560, training accuracy 0.918153
step 14560, cost 161.979
step 14560, change in cost 0.0340118
step 14570, training accuracy 0.918153
step 14570, cost 161.944
step 14570, change in cost 0.0341644
step 14580, training accuracy 0.918153
step 14580, cost 161.91
step 14580, change in cost 0.0339813
step 14590, training accuracy 0.918153
step 14590, cost 161.876
step 14590, change in cost 0.0341187
step 14600, training accuracy 0.918153
step 14600, cost 161.842
step 14600, change in cost 0.0340576
step 14610, training accuracy 0.918153
step 14610, cost 161.808
step 14610, change in cost 0.0339813
step 14620, training accuracy 0.918153
step 14620, cost 161.774
step 14620, change in cost 0.0340118
step 14630, training accuracy 0.918153
step 14630, cost 161.74
step 14630, change in cost 0.0339355
step 14640, training accuracy 0.918153
step 14640, cost 161.706
step 14640, change in cost 0.0339508
step 14650, training accuracy 0.918153
step 14650, cost 161.672
step 14650, change in cost 0.0339813
step 14660, training accuracy 0.918153
step 14660, cost 161.638
step 14660, change in cost 0.0339355
step 14670, training accuracy 0.918153
step 14670, cost 161.604
step 14670, change in cost 0.0339661
step 14680, training accuracy 0.918153
step 14680, cost 161.571
step 14680, change in cost 0.033844
step 14690, training accuracy 0.918153
step 14690, cost 161.537
step 14690, change in cost 0.0339203
step 14700, training accuracy 0.918153
step 14700, cost 161.503
step 14700, change in cost 0.033844
step 14710, training accuracy 0.918153
step 14710, cost 161.469
step 14710, change in cost 0.0338745
step 14720, training accuracy 0.918153
step 14720, cost 161.435
step 14720, change in cost 0.0337982
step 14730, training accuracy 0.918153
step 14730, cost 161.401
step 14730, change in cost 0.033844
step 14740, training accuracy 0.918153
step 14740, cost 161.367
step 14740, change in cost 0.033844
step 14750, training accuracy 0.918153
step 14750, cost 161.334
step 14750, change in cost 0.0337524
step 14760, training accuracy 0.918153
step 14760, cost 161.3
step 14760, change in cost 0.0338135
step 14770, training accuracy 0.918153
step 14770, cost 161.266
step 14770, change in cost 0.0337219
step 14780, training accuracy 0.918153
step 14780, cost 161.232
step 14780, change in cost 0.0337372
step 14790, training accuracy 0.918153
step 14790, cost 161.199
step 14790, change in cost 0.0336761
step 14800, training accuracy 0.918153
step 14800, cost 161.165
step 14800, change in cost 0.0337372
step 14810, training accuracy 0.918153
step 14810, cost 161.131
step 14810, change in cost 0.0337219
step 14820, training accuracy 0.918153
step 14820, cost 161.098
step 14820, change in cost 0.0336914
step 14830, training accuracy 0.919203
step 14830, cost 161.064
step 14830, change in cost 0.0336609
step 14840, training accuracy 0.919203
step 14840, cost 161.03
step 14840, change in cost 0.0336456
step 14850, training accuracy 0.919203
step 14850, cost 160.997
step 14850, change in cost 0.0335541
step 14860, training accuracy 0.919203
step 14860, cost 160.963
step 14860, change in cost 0.0336761
step 14870, training accuracy 0.919203
step 14870, cost 160.93
step 14870, change in cost 0.0335693
step 14880, training accuracy 0.919203
step 14880, cost 160.896
step 14880, change in cost 0.0336151
step 14890, training accuracy 0.919203
step 14890, cost 160.862
step 14890, change in cost 0.0335846
step 14900, training accuracy 0.919203
step 14900, cost 160.829
step 14900, change in cost 0.033493
step 14910, training accuracy 0.919203
step 14910, cost 160.795
step 14910, change in cost 0.0335846
step 14920, training accuracy 0.919203
step 14920, cost 160.762
step 14920, change in cost 0.0335083
step 14930, training accuracy 0.919203
step 14930, cost 160.728
step 14930, change in cost 0.0334778
step 14940, training accuracy 0.919203
step 14940, cost 160.695
step 14940, change in cost 0.033493
step 14950, training accuracy 0.919203
step 14950, cost 160.661
step 14950, change in cost 0.0334778
step 14960, training accuracy 0.919203
step 14960, cost 160.628
step 14960, change in cost 0.0334625
step 14970, training accuracy 0.919203
step 14970, cost 160.594
step 14970, change in cost 0.0334625
step 14980, training accuracy 0.919203
step 14980, cost 160.561
step 14980, change in cost 0.0334167
step 14990, training accuracy 0.919203
step 14990, cost 160.527
step 14990, change in cost 0.0334473
step 15000, training accuracy 0.919203
step 15000, cost 160.494
step 15000, change in cost 0.033371
step 15010, training accuracy 0.919203
step 15010, cost 160.461
step 15010, change in cost 0.0333099
step 15020, training accuracy 0.919203
step 15020, cost 160.427
step 15020, change in cost 0.0333557
step 15030, training accuracy 0.919203
step 15030, cost 160.394
step 15030, change in cost 0.0334015
step 15040, training accuracy 0.919203
step 15040, cost 160.361
step 15040, change in cost 0.0332947
step 15050, training accuracy 0.919203
step 15050, cost 160.327
step 15050, change in cost 0.0333557
step 15060, training accuracy 0.919203
step 15060, cost 160.294
step 15060, change in cost 0.0332642
step 15070, training accuracy 0.919203
step 15070, cost 160.261
step 15070, change in cost 0.0333099
step 15080, training accuracy 0.919203
step 15080, cost 160.228
step 15080, change in cost 0.0332794
step 15090, training accuracy 0.919203
step 15090, cost 160.194
step 15090, change in cost 0.0332642
step 15100, training accuracy 0.919203
step 15100, cost 160.161
step 15100, change in cost 0.0331879
step 15110, training accuracy 0.919203
step 15110, cost 160.128
step 15110, change in cost 0.0331879
step 15120, training accuracy 0.919203
step 15120, cost 160.095
step 15120, change in cost 0.0332794
step 15130, training accuracy 0.919203
step 15130, cost 160.061
step 15130, change in cost 0.0331879
step 15140, training accuracy 0.919203
step 15140, cost 160.028
step 15140, change in cost 0.0332031
step 15150, training accuracy 0.919203
step 15150, cost 159.995
step 15150, change in cost 0.0330811
step 15160, training accuracy 0.919203
step 15160, cost 159.962
step 15160, change in cost 0.0331726
step 15170, training accuracy 0.919203
step 15170, cost 159.929
step 15170, change in cost 0.0331726
step 15180, training accuracy 0.919203
step 15180, cost 159.896
step 15180, change in cost 0.0330811
step 15190, training accuracy 0.919203
step 15190, cost 159.863
step 15190, change in cost 0.0331421
step 15200, training accuracy 0.919203
step 15200, cost 159.83
step 15200, change in cost 0.0330353
step 15210, training accuracy 0.919203
step 15210, cost 159.796
step 15210, change in cost 0.0331268
step 15220, training accuracy 0.919203
step 15220, cost 159.763
step 15220, change in cost 0.0330353
step 15230, training accuracy 0.919203
step 15230, cost 159.73
step 15230, change in cost 0.0330505
step 15240, training accuracy 0.919203
step 15240, cost 159.697
step 15240, change in cost 0.0330048
step 15250, training accuracy 0.919203
step 15250, cost 159.664
step 15250, change in cost 0.0330048
step 15260, training accuracy 0.919203
step 15260, cost 159.631
step 15260, change in cost 0.032959
step 15270, training accuracy 0.919203
step 15270, cost 159.598
step 15270, change in cost 0.0329895
step 15280, training accuracy 0.919203
step 15280, cost 159.565
step 15280, change in cost 0.0329742
step 15290, training accuracy 0.919203
step 15290, cost 159.532
step 15290, change in cost 0.032959
step 15300, training accuracy 0.919203
step 15300, cost 159.5
step 15300, change in cost 0.0328979
step 15310, training accuracy 0.919203
step 15310, cost 159.467
step 15310, change in cost 0.0329132
step 15320, training accuracy 0.919203
step 15320, cost 159.434
step 15320, change in cost 0.0328979
step 15330, training accuracy 0.919203
step 15330, cost 159.401
step 15330, change in cost 0.0328217
step 15340, training accuracy 0.919203
step 15340, cost 159.368
step 15340, change in cost 0.032959
step 15350, training accuracy 0.919203
step 15350, cost 159.335
step 15350, change in cost 0.0328522
step 15360, training accuracy 0.919203
step 15360, cost 159.302
step 15360, change in cost 0.0328369
step 15370, training accuracy 0.919203
step 15370, cost 159.269
step 15370, change in cost 0.0328217
step 15380, training accuracy 0.919203
step 15380, cost 159.237
step 15380, change in cost 0.0328217
step 15390, training accuracy 0.919203
step 15390, cost 159.204
step 15390, change in cost 0.0327911
step 15400, training accuracy 0.919203
step 15400, cost 159.171
step 15400, change in cost 0.0327454
step 15410, training accuracy 0.920252
step 15410, cost 159.138
step 15410, change in cost 0.0327454
step 15420, training accuracy 0.920252
step 15420, cost 159.106
step 15420, change in cost 0.0328064
step 15430, training accuracy 0.920252
step 15430, cost 159.073
step 15430, change in cost 0.0327148
step 15440, training accuracy 0.921301
step 15440, cost 159.04
step 15440, change in cost 0.0327148
step 15450, training accuracy 0.921301
step 15450, cost 159.007
step 15450, change in cost 0.0326843
step 15460, training accuracy 0.921301
step 15460, cost 158.975
step 15460, change in cost 0.0327301
step 15470, training accuracy 0.921301
step 15470, cost 158.942
step 15470, change in cost 0.0325775
step 15480, training accuracy 0.921301
step 15480, cost 158.909
step 15480, change in cost 0.0326843
step 15490, training accuracy 0.921301
step 15490, cost 158.877
step 15490, change in cost 0.0326691
step 15500, training accuracy 0.921301
step 15500, cost 158.844
step 15500, change in cost 0.032608
step 15510, training accuracy 0.921301
step 15510, cost 158.812
step 15510, change in cost 0.0326233
step 15520, training accuracy 0.921301
step 15520, cost 158.779
step 15520, change in cost 0.0325317
step 15530, training accuracy 0.921301
step 15530, cost 158.746
step 15530, change in cost 0.0326385
step 15540, training accuracy 0.921301
step 15540, cost 158.714
step 15540, change in cost 0.0325928
step 15550, training accuracy 0.921301
step 15550, cost 158.681
step 15550, change in cost 0.0325317
step 15560, training accuracy 0.921301
step 15560, cost 158.649
step 15560, change in cost 0.0325623
step 15570, training accuracy 0.921301
step 15570, cost 158.616
step 15570, change in cost 0.0324097
step 15580, training accuracy 0.921301
step 15580, cost 158.584
step 15580, change in cost 0.0325623
step 15590, training accuracy 0.921301
step 15590, cost 158.551
step 15590, change in cost 0.0325165
step 15600, training accuracy 0.921301
step 15600, cost 158.519
step 15600, change in cost 0.0325012
step 15610, training accuracy 0.92235
step 15610, cost 158.486
step 15610, change in cost 0.0324402
step 15620, training accuracy 0.92235
step 15620, cost 158.454
step 15620, change in cost 0.0324554
step 15630, training accuracy 0.92235
step 15630, cost 158.421
step 15630, change in cost 0.0323792
step 15640, training accuracy 0.92235
step 15640, cost 158.389
step 15640, change in cost 0.0324097
step 15650, training accuracy 0.92235
step 15650, cost 158.357
step 15650, change in cost 0.0324249
step 15660, training accuracy 0.92235
step 15660, cost 158.324
step 15660, change in cost 0.0324097
step 15670, training accuracy 0.92235
step 15670, cost 158.292
step 15670, change in cost 0.0323486
step 15680, training accuracy 0.92235
step 15680, cost 158.259
step 15680, change in cost 0.0323334
step 15690, training accuracy 0.92235
step 15690, cost 158.227
step 15690, change in cost 0.0323181
step 15700, training accuracy 0.92235
step 15700, cost 158.195
step 15700, change in cost 0.0323944
step 15710, training accuracy 0.92235
step 15710, cost 158.163
step 15710, change in cost 0.0322571
step 15720, training accuracy 0.92235
step 15720, cost 158.13
step 15720, change in cost 0.0323181
step 15730, training accuracy 0.92235
step 15730, cost 158.098
step 15730, change in cost 0.0322418
step 15740, training accuracy 0.92235
step 15740, cost 158.066
step 15740, change in cost 0.0322876
step 15750, training accuracy 0.92235
step 15750, cost 158.033
step 15750, change in cost 0.0322571
step 15760, training accuracy 0.92235
step 15760, cost 158.001
step 15760, change in cost 0.0322418
step 15770, training accuracy 0.92235
step 15770, cost 157.969
step 15770, change in cost 0.0322113
step 15780, training accuracy 0.92235
step 15780, cost 157.937
step 15780, change in cost 0.0321655
step 15790, training accuracy 0.92235
step 15790, cost 157.905
step 15790, change in cost 0.0322571
step 15800, training accuracy 0.92235
step 15800, cost 157.872
step 15800, change in cost 0.0321503
step 15810, training accuracy 0.92235
step 15810, cost 157.84
step 15810, change in cost 0.032196
step 15820, training accuracy 0.92235
step 15820, cost 157.808
step 15820, change in cost 0.0321655
step 15830, training accuracy 0.92235
step 15830, cost 157.776
step 15830, change in cost 0.0320892
step 15840, training accuracy 0.92235
step 15840, cost 157.744
step 15840, change in cost 0.0321503
step 15850, training accuracy 0.92235
step 15850, cost 157.712
step 15850, change in cost 0.0320587
step 15860, training accuracy 0.92235
step 15860, cost 157.68
step 15860, change in cost 0.032074
step 15870, training accuracy 0.92235
step 15870, cost 157.648
step 15870, change in cost 0.0320892
step 15880, training accuracy 0.92235
step 15880, cost 157.616
step 15880, change in cost 0.0320435
step 15890, training accuracy 0.92235
step 15890, cost 157.583
step 15890, change in cost 0.0321045
step 15900, training accuracy 0.92235
step 15900, cost 157.551
step 15900, change in cost 0.0320129
step 15910, training accuracy 0.92235
step 15910, cost 157.519
step 15910, change in cost 0.0320129
step 15920, training accuracy 0.92235
step 15920, cost 157.487
step 15920, change in cost 0.0319672
step 15930, training accuracy 0.92235
step 15930, cost 157.455
step 15930, change in cost 0.0320435
step 15940, training accuracy 0.92235
step 15940, cost 157.423
step 15940, change in cost 0.0319214
step 15950, training accuracy 0.92235
step 15950, cost 157.391
step 15950, change in cost 0.0319824
step 15960, training accuracy 0.92235
step 15960, cost 157.36
step 15960, change in cost 0.0319214
step 15970, training accuracy 0.9234
step 15970, cost 157.328
step 15970, change in cost 0.0319214
step 15980, training accuracy 0.9234
step 15980, cost 157.296
step 15980, change in cost 0.0319672
step 15990, training accuracy 0.9234
step 15990, cost 157.264
step 15990, change in cost 0.0318604
step 16000, training accuracy 0.9234
step 16000, cost 157.232
step 16000, change in cost 0.0318604
step 16010, training accuracy 0.9234
step 16010, cost 157.2
step 16010, change in cost 0.0318756
step 16020, training accuracy 0.9234
step 16020, cost 157.168
step 16020, change in cost 0.0318298
step 16030, training accuracy 0.9234
step 16030, cost 157.136
step 16030, change in cost 0.0318909
step 16040, training accuracy 0.9234
step 16040, cost 157.104
step 16040, change in cost 0.0318451
step 16050, training accuracy 0.9234
step 16050, cost 157.073
step 16050, change in cost 0.0317383
step 16060, training accuracy 0.9234
step 16060, cost 157.041
step 16060, change in cost 0.0317993
step 16070, training accuracy 0.9234
step 16070, cost 157.009
step 16070, change in cost 0.0317993
step 16080, training accuracy 0.9234
step 16080, cost 156.977
step 16080, change in cost 0.0317688
step 16090, training accuracy 0.9234
step 16090, cost 156.946
step 16090, change in cost 0.0317841
step 16100, training accuracy 0.9234
step 16100, cost 156.914
step 16100, change in cost 0.0317078
step 16110, training accuracy 0.9234
step 16110, cost 156.882
step 16110, change in cost 0.0317688
step 16120, training accuracy 0.9234
step 16120, cost 156.85
step 16120, change in cost 0.0316467
step 16130, training accuracy 0.9234
step 16130, cost 156.819
step 16130, change in cost 0.0317078
step 16140, training accuracy 0.9234
step 16140, cost 156.787
step 16140, change in cost 0.0316772
step 16150, training accuracy 0.9234
step 16150, cost 156.755
step 16150, change in cost 0.0316925
step 16160, training accuracy 0.9234
step 16160, cost 156.724
step 16160, change in cost 0.0316315
step 16170, training accuracy 0.9234
step 16170, cost 156.692
step 16170, change in cost 0.0316467
step 16180, training accuracy 0.9234
step 16180, cost 156.661
step 16180, change in cost 0.031601
step 16190, training accuracy 0.9234
step 16190, cost 156.629
step 16190, change in cost 0.0316162
step 16200, training accuracy 0.9234
step 16200, cost 156.597
step 16200, change in cost 0.031601
step 16210, training accuracy 0.924449
step 16210, cost 156.566
step 16210, change in cost 0.0315857
step 16220, training accuracy 0.924449
step 16220, cost 156.534
step 16220, change in cost 0.0315704
step 16230, training accuracy 0.924449
step 16230, cost 156.503
step 16230, change in cost 0.0315247
step 16240, training accuracy 0.925498
step 16240, cost 156.471
step 16240, change in cost 0.0315857
step 16250, training accuracy 0.925498
step 16250, cost 156.44
step 16250, change in cost 0.0314941
step 16260, training accuracy 0.925498
step 16260, cost 156.408
step 16260, change in cost 0.0314636
step 16270, training accuracy 0.925498
step 16270, cost 156.377
step 16270, change in cost 0.0315094
step 16280, training accuracy 0.925498
step 16280, cost 156.345
step 16280, change in cost 0.0314636
step 16290, training accuracy 0.925498
step 16290, cost 156.314
step 16290, change in cost 0.0315094
step 16300, training accuracy 0.925498
step 16300, cost 156.282
step 16300, change in cost 0.0314026
step 16310, training accuracy 0.925498
step 16310, cost 156.251
step 16310, change in cost 0.0313873
step 16320, training accuracy 0.925498
step 16320, cost 156.219
step 16320, change in cost 0.0314789
step 16330, training accuracy 0.925498
step 16330, cost 156.188
step 16330, change in cost 0.0314026
step 16340, training accuracy 0.925498
step 16340, cost 156.157
step 16340, change in cost 0.0313568
step 16350, training accuracy 0.925498
step 16350, cost 156.125
step 16350, change in cost 0.0314484
step 16360, training accuracy 0.925498
step 16360, cost 156.094
step 16360, change in cost 0.0313263
step 16370, training accuracy 0.925498
step 16370, cost 156.062
step 16370, change in cost 0.0313721
step 16380, training accuracy 0.925498
step 16380, cost 156.031
step 16380, change in cost 0.0313568
step 16390, training accuracy 0.925498
step 16390, cost 156
step 16390, change in cost 0.0312653
step 16400, training accuracy 0.925498
step 16400, cost 155.968
step 16400, change in cost 0.0313721
step 16410, training accuracy 0.925498
step 16410, cost 155.937
step 16410, change in cost 0.0312653
step 16420, training accuracy 0.925498
step 16420, cost 155.906
step 16420, change in cost 0.0312653
step 16430, training accuracy 0.925498
step 16430, cost 155.875
step 16430, change in cost 0.0312347
step 16440, training accuracy 0.925498
step 16440, cost 155.843
step 16440, change in cost 0.031311
step 16450, training accuracy 0.925498
step 16450, cost 155.812
step 16450, change in cost 0.0311737
step 16460, training accuracy 0.925498
step 16460, cost 155.781
step 16460, change in cost 0.0312347
step 16470, training accuracy 0.925498
step 16470, cost 155.75
step 16470, change in cost 0.0312195
step 16480, training accuracy 0.925498
step 16480, cost 155.719
step 16480, change in cost 0.031189
step 16490, training accuracy 0.925498
step 16490, cost 155.687
step 16490, change in cost 0.0312195
step 16500, training accuracy 0.925498
step 16500, cost 155.656
step 16500, change in cost 0.0311432
step 16510, training accuracy 0.925498
step 16510, cost 155.625
step 16510, change in cost 0.0311279
step 16520, training accuracy 0.925498
step 16520, cost 155.594
step 16520, change in cost 0.0311432
step 16530, training accuracy 0.925498
step 16530, cost 155.563
step 16530, change in cost 0.0310974
step 16540, training accuracy 0.925498
step 16540, cost 155.532
step 16540, change in cost 0.0311127
step 16550, training accuracy 0.925498
step 16550, cost 155.501
step 16550, change in cost 0.0310822
step 16560, training accuracy 0.925498
step 16560, cost 155.469
step 16560, change in cost 0.0311279
step 16570, training accuracy 0.925498
step 16570, cost 155.438
step 16570, change in cost 0.0310211
step 16580, training accuracy 0.925498
step 16580, cost 155.407
step 16580, change in cost 0.0310669
step 16590, training accuracy 0.925498
step 16590, cost 155.376
step 16590, change in cost 0.0310516
step 16600, training accuracy 0.925498
step 16600, cost 155.345
step 16600, change in cost 0.0310516
step 16610, training accuracy 0.925498
step 16610, cost 155.314
step 16610, change in cost 0.0309753
step 16620, training accuracy 0.925498
step 16620, cost 155.283
step 16620, change in cost 0.0310059
step 16630, training accuracy 0.925498
step 16630, cost 155.252
step 16630, change in cost 0.0309601
step 16640, training accuracy 0.925498
step 16640, cost 155.221
step 16640, change in cost 0.0309753
step 16650, training accuracy 0.925498
step 16650, cost 155.19
step 16650, change in cost 0.0309601
step 16660, training accuracy 0.926548
step 16660, cost 155.159
step 16660, change in cost 0.0309448
step 16670, training accuracy 0.926548
step 16670, cost 155.129
step 16670, change in cost 0.0309296
step 16680, training accuracy 0.926548
step 16680, cost 155.098
step 16680, change in cost 0.0309143
step 16690, training accuracy 0.926548
step 16690, cost 155.067
step 16690, change in cost 0.0309143
step 16700, training accuracy 0.926548
step 16700, cost 155.036
step 16700, change in cost 0.0308533
step 16710, training accuracy 0.926548
step 16710, cost 155.005
step 16710, change in cost 0.0308685
step 16720, training accuracy 0.926548
step 16720, cost 154.974
step 16720, change in cost 0.0308533
step 16730, training accuracy 0.927597
step 16730, cost 154.943
step 16730, change in cost 0.0308533
step 16740, training accuracy 0.927597
step 16740, cost 154.912
step 16740, change in cost 0.0308533
step 16750, training accuracy 0.927597
step 16750, cost 154.882
step 16750, change in cost 0.030777
step 16760, training accuracy 0.927597
step 16760, cost 154.851
step 16760, change in cost 0.0308228
step 16770, training accuracy 0.927597
step 16770, cost 154.82
step 16770, change in cost 0.0307617
step 16780, training accuracy 0.927597
step 16780, cost 154.789
step 16780, change in cost 0.0307617
step 16790, training accuracy 0.927597
step 16790, cost 154.758
step 16790, change in cost 0.030838
step 16800, training accuracy 0.927597
step 16800, cost 154.728
step 16800, change in cost 0.0307465
step 16810, training accuracy 0.927597
step 16810, cost 154.697
step 16810, change in cost 0.0306854
step 16820, training accuracy 0.927597
step 16820, cost 154.666
step 16820, change in cost 0.0307617
step 16830, training accuracy 0.927597
step 16830, cost 154.636
step 16830, change in cost 0.0306549
step 16840, training accuracy 0.927597
step 16840, cost 154.605
step 16840, change in cost 0.0307617
step 16850, training accuracy 0.927597
step 16850, cost 154.574
step 16850, change in cost 0.0305634
step 16860, training accuracy 0.927597
step 16860, cost 154.544
step 16860, change in cost 0.0307159
step 16870, training accuracy 0.927597
step 16870, cost 154.513
step 16870, change in cost 0.0306549
step 16880, training accuracy 0.927597
step 16880, cost 154.482
step 16880, change in cost 0.0306091
step 16890, training accuracy 0.927597
step 16890, cost 154.452
step 16890, change in cost 0.0305786
step 16900, training accuracy 0.927597
step 16900, cost 154.421
step 16900, change in cost 0.0305939
step 16910, training accuracy 0.927597
step 16910, cost 154.39
step 16910, change in cost 0.0306549
step 16920, training accuracy 0.928646
step 16920, cost 154.36
step 16920, change in cost 0.0306396
step 16930, training accuracy 0.929696
step 16930, cost 154.329
step 16930, change in cost 0.0305176
step 16940, training accuracy 0.929696
step 16940, cost 154.299
step 16940, change in cost 0.0305939
step 16950, training accuracy 0.929696
step 16950, cost 154.268
step 16950, change in cost 0.0305176
step 16960, training accuracy 0.929696
step 16960, cost 154.238
step 16960, change in cost 0.0305023
step 16970, training accuracy 0.929696
step 16970, cost 154.207
step 16970, change in cost 0.0304718
step 16980, training accuracy 0.929696
step 16980, cost 154.177
step 16980, change in cost 0.0305481
step 16990, training accuracy 0.929696
step 16990, cost 154.146
step 16990, change in cost 0.0304413
step 17000, training accuracy 0.929696
step 17000, cost 154.116
step 17000, change in cost 0.0305023
step 17010, training accuracy 0.929696
step 17010, cost 154.085
step 17010, change in cost 0.0304718
step 17020, training accuracy 0.929696
step 17020, cost 154.055
step 17020, change in cost 0.0304108
step 17030, training accuracy 0.929696
step 17030, cost 154.024
step 17030, change in cost 0.0304413
step 17040, training accuracy 0.929696
step 17040, cost 153.994
step 17040, change in cost 0.0304108
step 17050, training accuracy 0.929696
step 17050, cost 153.964
step 17050, change in cost 0.0304871
step 17060, training accuracy 0.929696
step 17060, cost 153.933
step 17060, change in cost 0.0303192
step 17070, training accuracy 0.929696
step 17070, cost 153.903
step 17070, change in cost 0.030426
step 17080, training accuracy 0.929696
step 17080, cost 153.872
step 17080, change in cost 0.030365
step 17090, training accuracy 0.929696
step 17090, cost 153.842
step 17090, change in cost 0.030304
step 17100, training accuracy 0.929696
step 17100, cost 153.812
step 17100, change in cost 0.0303345
step 17110, training accuracy 0.929696
step 17110, cost 153.781
step 17110, change in cost 0.0303497
step 17120, training accuracy 0.929696
step 17120, cost 153.751
step 17120, change in cost 0.0301971
step 17130, training accuracy 0.929696
step 17130, cost 153.721
step 17130, change in cost 0.0303955
step 17140, training accuracy 0.929696
step 17140, cost 153.691
step 17140, change in cost 0.0302429
step 17150, training accuracy 0.929696
step 17150, cost 153.66
step 17150, change in cost 0.0302887
step 17160, training accuracy 0.929696
step 17160, cost 153.63
step 17160, change in cost 0.0301819
step 17170, training accuracy 0.929696
step 17170, cost 153.6
step 17170, change in cost 0.0303192
step 17180, training accuracy 0.929696
step 17180, cost 153.57
step 17180, change in cost 0.0301971
step 17190, training accuracy 0.929696
step 17190, cost 153.539
step 17190, change in cost 0.0302124
step 17200, training accuracy 0.929696
step 17200, cost 153.509
step 17200, change in cost 0.0302277
step 17210, training accuracy 0.929696
step 17210, cost 153.479
step 17210, change in cost 0.0301819
step 17220, training accuracy 0.929696
step 17220, cost 153.449
step 17220, change in cost 0.0302277
step 17230, training accuracy 0.929696
step 17230, cost 153.419
step 17230, change in cost 0.0301208
step 17240, training accuracy 0.929696
step 17240, cost 153.389
step 17240, change in cost 0.0300903
step 17250, training accuracy 0.929696
step 17250, cost 153.358
step 17250, change in cost 0.0302277
step 17260, training accuracy 0.929696
step 17260, cost 153.328
step 17260, change in cost 0.0300598
step 17270, training accuracy 0.929696
step 17270, cost 153.298
step 17270, change in cost 0.0301361
step 17280, training accuracy 0.929696
step 17280, cost 153.268
step 17280, change in cost 0.0300903
step 17290, training accuracy 0.929696
step 17290, cost 153.238
step 17290, change in cost 0.0300598
step 17300, training accuracy 0.930745
step 17300, cost 153.208
step 17300, change in cost 0.0300751
step 17310, training accuracy 0.930745
step 17310, cost 153.178
step 17310, change in cost 0.0300293
step 17320, training accuracy 0.930745
step 17320, cost 153.148
step 17320, change in cost 0.0300598
step 17330, training accuracy 0.930745
step 17330, cost 153.118
step 17330, change in cost 0.0299988
step 17340, training accuracy 0.930745
step 17340, cost 153.088
step 17340, change in cost 0.0300446
step 17350, training accuracy 0.930745
step 17350, cost 153.058
step 17350, change in cost 0.0299683
step 17360, training accuracy 0.930745
step 17360, cost 153.028
step 17360, change in cost 0.0300293
step 17370, training accuracy 0.930745
step 17370, cost 152.998
step 17370, change in cost 0.029953
step 17380, training accuracy 0.930745
step 17380, cost 152.968
step 17380, change in cost 0.0299835
step 17390, training accuracy 0.930745
step 17390, cost 152.938
step 17390, change in cost 0.029892
step 17400, training accuracy 0.930745
step 17400, cost 152.908
step 17400, change in cost 0.0299835
step 17410, training accuracy 0.930745
step 17410, cost 152.878
step 17410, change in cost 0.0299072
step 17420, training accuracy 0.930745
step 17420, cost 152.848
step 17420, change in cost 0.0299072
step 17430, training accuracy 0.930745
step 17430, cost 152.818
step 17430, change in cost 0.029892
step 17440, training accuracy 0.930745
step 17440, cost 152.788
step 17440, change in cost 0.029892
step 17450, training accuracy 0.930745
step 17450, cost 152.759
step 17450, change in cost 0.0298004
step 17460, training accuracy 0.930745
step 17460, cost 152.729
step 17460, change in cost 0.029892
step 17470, training accuracy 0.930745
step 17470, cost 152.699
step 17470, change in cost 0.0298157
step 17480, training accuracy 0.930745
step 17480, cost 152.669
step 17480, change in cost 0.0298615
step 17490, training accuracy 0.930745
step 17490, cost 152.639
step 17490, change in cost 0.0297546
step 17500, training accuracy 0.930745
step 17500, cost 152.609
step 17500, change in cost 0.0298462
step 17510, training accuracy 0.930745
step 17510, cost 152.58
step 17510, change in cost 0.0298157
step 17520, training accuracy 0.930745
step 17520, cost 152.55
step 17520, change in cost 0.0297546
step 17530, training accuracy 0.930745
step 17530, cost 152.52
step 17530, change in cost 0.0297699
step 17540, training accuracy 0.930745
step 17540, cost 152.49
step 17540, change in cost 0.0297852
step 17550, training accuracy 0.930745
step 17550, cost 152.461
step 17550, change in cost 0.0297089
step 17560, training accuracy 0.930745
step 17560, cost 152.431
step 17560, change in cost 0.0297241
step 17570, training accuracy 0.930745
step 17570, cost 152.401
step 17570, change in cost 0.0297089
step 17580, training accuracy 0.930745
step 17580, cost 152.371
step 17580, change in cost 0.0296631
step 17590, training accuracy 0.930745
step 17590, cost 152.342
step 17590, change in cost 0.0296631
step 17600, training accuracy 0.930745
step 17600, cost 152.312
step 17600, change in cost 0.0297241
step 17610, training accuracy 0.930745
step 17610, cost 152.282
step 17610, change in cost 0.0296631
step 17620, training accuracy 0.930745
step 17620, cost 152.253
step 17620, change in cost 0.0296478
step 17630, training accuracy 0.930745
step 17630, cost 152.223
step 17630, change in cost 0.0296021
step 17640, training accuracy 0.930745
step 17640, cost 152.194
step 17640, change in cost 0.0296021
step 17650, training accuracy 0.930745
step 17650, cost 152.164
step 17650, change in cost 0.0296173
step 17660, training accuracy 0.930745
step 17660, cost 152.134
step 17660, change in cost 0.0296326
step 17670, training accuracy 0.931794
step 17670, cost 152.105
step 17670, change in cost 0.0294952
step 17680, training accuracy 0.931794
step 17680, cost 152.075
step 17680, change in cost 0.0295715
step 17690, training accuracy 0.931794
step 17690, cost 152.046
step 17690, change in cost 0.0296021
step 17700, training accuracy 0.931794
step 17700, cost 152.016
step 17700, change in cost 0.0295105
step 17710, training accuracy 0.931794
step 17710, cost 151.987
step 17710, change in cost 0.0295258
step 17720, training accuracy 0.931794
step 17720, cost 151.957
step 17720, change in cost 0.02948
step 17730, training accuracy 0.931794
step 17730, cost 151.928
step 17730, change in cost 0.0295105
step 17740, training accuracy 0.931794
step 17740, cost 151.898
step 17740, change in cost 0.0295563
step 17750, training accuracy 0.931794
step 17750, cost 151.869
step 17750, change in cost 0.0294495
step 17760, training accuracy 0.931794
step 17760, cost 151.839
step 17760, change in cost 0.0294495
step 17770, training accuracy 0.931794
step 17770, cost 151.81
step 17770, change in cost 0.0294647
step 17780, training accuracy 0.931794
step 17780, cost 151.78
step 17780, change in cost 0.0294037
step 17790, training accuracy 0.931794
step 17790, cost 151.751
step 17790, change in cost 0.0294037
step 17800, training accuracy 0.931794
step 17800, cost 151.721
step 17800, change in cost 0.0294037
step 17810, training accuracy 0.931794
step 17810, cost 151.692
step 17810, change in cost 0.0293884
step 17820, training accuracy 0.931794
step 17820, cost 151.663
step 17820, change in cost 0.0294189
step 17830, training accuracy 0.931794
step 17830, cost 151.633
step 17830, change in cost 0.0294037
step 17840, training accuracy 0.931794
step 17840, cost 151.604
step 17840, change in cost 0.0293274
step 17850, training accuracy 0.931794
step 17850, cost 151.575
step 17850, change in cost 0.0292969
step 17860, training accuracy 0.931794
step 17860, cost 151.545
step 17860, change in cost 0.0293884
step 17870, training accuracy 0.931794
step 17870, cost 151.516
step 17870, change in cost 0.0293121
step 17880, training accuracy 0.931794
step 17880, cost 151.487
step 17880, change in cost 0.0293121
step 17890, training accuracy 0.931794
step 17890, cost 151.457
step 17890, change in cost 0.0292664
step 17900, training accuracy 0.931794
step 17900, cost 151.428
step 17900, change in cost 0.0292511
step 17910, training accuracy 0.931794
step 17910, cost 151.399
step 17910, change in cost 0.0292816
step 17920, training accuracy 0.931794
step 17920, cost 151.37
step 17920, change in cost 0.0292664
step 17930, training accuracy 0.931794
step 17930, cost 151.34
step 17930, change in cost 0.0292206
step 17940, training accuracy 0.931794
step 17940, cost 151.311
step 17940, change in cost 0.0292816
step 17950, training accuracy 0.931794
step 17950, cost 151.282
step 17950, change in cost 0.0292206
step 17960, training accuracy 0.931794
step 17960, cost 151.253
step 17960, change in cost 0.0291901
step 17970, training accuracy 0.931794
step 17970, cost 151.223
step 17970, change in cost 0.0291595
step 17980, training accuracy 0.931794
step 17980, cost 151.194
step 17980, change in cost 0.0291595
step 17990, training accuracy 0.931794
step 17990, cost 151.165
step 17990, change in cost 0.0291901
step 18000, training accuracy 0.931794
step 18000, cost 151.136
step 18000, change in cost 0.0291901
step 18010, training accuracy 0.931794
step 18010, cost 151.107
step 18010, change in cost 0.0290527
step 18020, training accuracy 0.931794
step 18020, cost 151.078
step 18020, change in cost 0.0291595
step 18030, training accuracy 0.931794
step 18030, cost 151.049
step 18030, change in cost 0.029129
step 18040, training accuracy 0.931794
step 18040, cost 151.019
step 18040, change in cost 0.029129
step 18050, training accuracy 0.931794
step 18050, cost 150.99
step 18050, change in cost 0.029068
step 18060, training accuracy 0.931794
step 18060, cost 150.961
step 18060, change in cost 0.0290527
step 18070, training accuracy 0.931794
step 18070, cost 150.932
step 18070, change in cost 0.029068
step 18080, training accuracy 0.931794
step 18080, cost 150.903
step 18080, change in cost 0.0290833
step 18090, training accuracy 0.931794
step 18090, cost 150.874
step 18090, change in cost 0.0290375
step 18100, training accuracy 0.931794
step 18100, cost 150.845
step 18100, change in cost 0.0290222
step 18110, training accuracy 0.931794
step 18110, cost 150.816
step 18110, change in cost 0.0290375
step 18120, training accuracy 0.931794
step 18120, cost 150.787
step 18120, change in cost 0.0289154
step 18130, training accuracy 0.932844
step 18130, cost 150.758
step 18130, change in cost 0.0289917
step 18140, training accuracy 0.932844
step 18140, cost 150.729
step 18140, change in cost 0.0289764
step 18150, training accuracy 0.932844
step 18150, cost 150.7
step 18150, change in cost 0.0289764
step 18160, training accuracy 0.932844
step 18160, cost 150.671
step 18160, change in cost 0.0289764
step 18170, training accuracy 0.932844
step 18170, cost 150.642
step 18170, change in cost 0.0289001
step 18180, training accuracy 0.932844
step 18180, cost 150.613
step 18180, change in cost 0.0289459
step 18190, training accuracy 0.932844
step 18190, cost 150.585
step 18190, change in cost 0.0288696
step 18200, training accuracy 0.932844
step 18200, cost 150.556
step 18200, change in cost 0.0289001
step 18210, training accuracy 0.932844
step 18210, cost 150.527
step 18210, change in cost 0.0289154
step 18220, training accuracy 0.932844
step 18220, cost 150.498
step 18220, change in cost 0.0288544
step 18230, training accuracy 0.932844
step 18230, cost 150.469
step 18230, change in cost 0.0289154
step 18240, training accuracy 0.932844
step 18240, cost 150.44
step 18240, change in cost 0.0287781
step 18250, training accuracy 0.933893
step 18250, cost 150.411
step 18250, change in cost 0.0288544
step 18260, training accuracy 0.933893
step 18260, cost 150.383
step 18260, change in cost 0.0288239
step 18270, training accuracy 0.933893
step 18270, cost 150.354
step 18270, change in cost 0.0287933
step 18280, training accuracy 0.933893
step 18280, cost 150.325
step 18280, change in cost 0.0288544
step 18290, training accuracy 0.933893
step 18290, cost 150.296
step 18290, change in cost 0.028717
step 18300, training accuracy 0.933893
step 18300, cost 150.267
step 18300, change in cost 0.0288239
step 18310, training accuracy 0.933893
step 18310, cost 150.239
step 18310, change in cost 0.028717
step 18320, training accuracy 0.933893
step 18320, cost 150.21
step 18320, change in cost 0.0287018
step 18330, training accuracy 0.933893
step 18330, cost 150.181
step 18330, change in cost 0.0288086
step 18340, training accuracy 0.933893
step 18340, cost 150.152
step 18340, change in cost 0.0287018
step 18350, training accuracy 0.934942
step 18350, cost 150.124
step 18350, change in cost 0.0287323
step 18360, training accuracy 0.934942
step 18360, cost 150.095
step 18360, change in cost 0.0287018
step 18370, training accuracy 0.934942
step 18370, cost 150.066
step 18370, change in cost 0.028656
step 18380, training accuracy 0.934942
step 18380, cost 150.038
step 18380, change in cost 0.0287018
step 18390, training accuracy 0.934942
step 18390, cost 150.009
step 18390, change in cost 0.0286407
step 18400, training accuracy 0.934942
step 18400, cost 149.98
step 18400, change in cost 0.0286102
step 18410, training accuracy 0.934942
step 18410, cost 149.952
step 18410, change in cost 0.0287018
step 18420, training accuracy 0.934942
step 18420, cost 149.923
step 18420, change in cost 0.0285492
step 18430, training accuracy 0.935992
step 18430, cost 149.894
step 18430, change in cost 0.0286255
step 18440, training accuracy 0.935992
step 18440, cost 149.866
step 18440, change in cost 0.0286255
step 18450, training accuracy 0.935992
step 18450, cost 149.837
step 18450, change in cost 0.0285339
step 18460, training accuracy 0.935992
step 18460, cost 149.809
step 18460, change in cost 0.0286102
step 18470, training accuracy 0.935992
step 18470, cost 149.78
step 18470, change in cost 0.0285339
step 18480, training accuracy 0.935992
step 18480, cost 149.752
step 18480, change in cost 0.0285492
step 18490, training accuracy 0.935992
step 18490, cost 149.723
step 18490, change in cost 0.0285645
step 18500, training accuracy 0.935992
step 18500, cost 149.695
step 18500, change in cost 0.0284729
step 18510, training accuracy 0.937041
step 18510, cost 149.666
step 18510, change in cost 0.0285339
step 18520, training accuracy 0.93809
step 18520, cost 149.638
step 18520, change in cost 0.0285339
step 18530, training accuracy 0.93809
step 18530, cost 149.609
step 18530, change in cost 0.0284729
step 18540, training accuracy 0.93809
step 18540, cost 149.581
step 18540, change in cost 0.0285187
step 18550, training accuracy 0.93809
step 18550, cost 149.552
step 18550, change in cost 0.0284119
step 18560, training accuracy 0.93809
step 18560, cost 149.524
step 18560, change in cost 0.0284576
step 18570, training accuracy 0.93809
step 18570, cost 149.495
step 18570, change in cost 0.0284271
step 18580, training accuracy 0.93809
step 18580, cost 149.467
step 18580, change in cost 0.0284119
step 18590, training accuracy 0.93809
step 18590, cost 149.438
step 18590, change in cost 0.0284882
step 18600, training accuracy 0.93809
step 18600, cost 149.41
step 18600, change in cost 0.0283051
step 18610, training accuracy 0.93809
step 18610, cost 149.382
step 18610, change in cost 0.0284424
step 18620, training accuracy 0.93809
step 18620, cost 149.353
step 18620, change in cost 0.0283661
step 18630, training accuracy 0.93809
step 18630, cost 149.325
step 18630, change in cost 0.0283661
step 18640, training accuracy 0.93809
step 18640, cost 149.297
step 18640, change in cost 0.0283356
step 18650, training accuracy 0.93809
step 18650, cost 149.268
step 18650, change in cost 0.0283508
step 18660, training accuracy 0.93809
step 18660, cost 149.24
step 18660, change in cost 0.0283661
step 18670, training accuracy 0.93809
step 18670, cost 149.212
step 18670, change in cost 0.028244
step 18680, training accuracy 0.93809
step 18680, cost 149.183
step 18680, change in cost 0.0283203
step 18690, training accuracy 0.93809
step 18690, cost 149.155
step 18690, change in cost 0.0282593
step 18700, training accuracy 0.93809
step 18700, cost 149.127
step 18700, change in cost 0.0282745
step 18710, training accuracy 0.93809
step 18710, cost 149.098
step 18710, change in cost 0.0282745
step 18720, training accuracy 0.93809
step 18720, cost 149.07
step 18720, change in cost 0.0282593
step 18730, training accuracy 0.93809
step 18730, cost 149.042
step 18730, change in cost 0.0282288
step 18740, training accuracy 0.93809
step 18740, cost 149.014
step 18740, change in cost 0.0282745
step 18750, training accuracy 0.93809
step 18750, cost 148.985
step 18750, change in cost 0.0282135
step 18760, training accuracy 0.93809
step 18760, cost 148.957
step 18760, change in cost 0.0281677
step 18770, training accuracy 0.93809
step 18770, cost 148.929
step 18770, change in cost 0.028183
step 18780, training accuracy 0.93809
step 18780, cost 148.901
step 18780, change in cost 0.0281982
step 18790, training accuracy 0.93809
step 18790, cost 148.873
step 18790, change in cost 0.0281372
step 18800, training accuracy 0.93809
step 18800, cost 148.845
step 18800, change in cost 0.028183
step 18810, training accuracy 0.93809
step 18810, cost 148.816
step 18810, change in cost 0.028183
step 18820, training accuracy 0.93809
step 18820, cost 148.788
step 18820, change in cost 0.0281219
step 18830, training accuracy 0.93809
step 18830, cost 148.76
step 18830, change in cost 0.0281219
step 18840, training accuracy 0.93809
step 18840, cost 148.732
step 18840, change in cost 0.0280609
step 18850, training accuracy 0.93809
step 18850, cost 148.704
step 18850, change in cost 0.0281067
step 18860, training accuracy 0.93809
step 18860, cost 148.676
step 18860, change in cost 0.0281067
step 18870, training accuracy 0.93809
step 18870, cost 148.648
step 18870, change in cost 0.0280762
step 18880, training accuracy 0.93809
step 18880, cost 148.62
step 18880, change in cost 0.0280151
step 18890, training accuracy 0.93809
step 18890, cost 148.592
step 18890, change in cost 0.0280914
step 18900, training accuracy 0.93809
step 18900, cost 148.564
step 18900, change in cost 0.0280457
step 18910, training accuracy 0.93809
step 18910, cost 148.536
step 18910, change in cost 0.0280151
step 18920, training accuracy 0.93809
step 18920, cost 148.508
step 18920, change in cost 0.0280151
step 18930, training accuracy 0.93809
step 18930, cost 148.48
step 18930, change in cost 0.0279999
step 18940, training accuracy 0.93809
step 18940, cost 148.452
step 18940, change in cost 0.0279999
step 18950, training accuracy 0.93809
step 18950, cost 148.424
step 18950, change in cost 0.0279236
step 18960, training accuracy 0.93809
step 18960, cost 148.396
step 18960, change in cost 0.0279846
step 18970, training accuracy 0.93809
step 18970, cost 148.368
step 18970, change in cost 0.0279694
step 18980, training accuracy 0.93809
step 18980, cost 148.34
step 18980, change in cost 0.0278778
step 18990, training accuracy 0.93809
step 18990, cost 148.312
step 18990, change in cost 0.0279541
step 19000, training accuracy 0.93809
step 19000, cost 148.284
step 19000, change in cost 0.0279083
step 19010, training accuracy 0.93809
step 19010, cost 148.256
step 19010, change in cost 0.0278931
step 19020, training accuracy 0.93809
step 19020, cost 148.228
step 19020, change in cost 0.0279236
step 19030, training accuracy 0.93809
step 19030, cost 148.2
step 19030, change in cost 0.0278778
step 19040, training accuracy 0.93809
step 19040, cost 148.172
step 19040, change in cost 0.0278931
step 19050, training accuracy 0.93809
step 19050, cost 148.145
step 19050, change in cost 0.0278168
step 19060, training accuracy 0.93809
step 19060, cost 148.117
step 19060, change in cost 0.0278625
step 19070, training accuracy 0.93809
step 19070, cost 148.089
step 19070, change in cost 0.027832
step 19080, training accuracy 0.93809
step 19080, cost 148.061
step 19080, change in cost 0.0278168
step 19090, training accuracy 0.93809
step 19090, cost 148.033
step 19090, change in cost 0.0278625
step 19100, training accuracy 0.93809
step 19100, cost 148.006
step 19100, change in cost 0.0277252
step 19110, training accuracy 0.93809
step 19110, cost 147.978
step 19110, change in cost 0.027771
step 19120, training accuracy 0.93809
step 19120, cost 147.95
step 19120, change in cost 0.0278473
step 19130, training accuracy 0.93809
step 19130, cost 147.922
step 19130, change in cost 0.0276794
step 19140, training accuracy 0.93809
step 19140, cost 147.894
step 19140, change in cost 0.0277863
step 19150, training accuracy 0.93809
step 19150, cost 147.867
step 19150, change in cost 0.0277252
step 19160, training accuracy 0.93809
step 19160, cost 147.839
step 19160, change in cost 0.0277405
step 19170, training accuracy 0.93809
step 19170, cost 147.811
step 19170, change in cost 0.0276489
step 19180, training accuracy 0.93809
step 19180, cost 147.784
step 19180, change in cost 0.027771
step 19190, training accuracy 0.93809
step 19190, cost 147.756
step 19190, change in cost 0.0276489
step 19200, training accuracy 0.93809
step 19200, cost 147.728
step 19200, change in cost 0.0277252
step 19210, training accuracy 0.93914
step 19210, cost 147.701
step 19210, change in cost 0.0276031
step 19220, training accuracy 0.93914
step 19220, cost 147.673
step 19220, change in cost 0.02771
step 19230, training accuracy 0.93914
step 19230, cost 147.645
step 19230, change in cost 0.0276031
step 19240, training accuracy 0.93914
step 19240, cost 147.618
step 19240, change in cost 0.0276794
step 19250, training accuracy 0.93914
step 19250, cost 147.59
step 19250, change in cost 0.0275879
step 19260, training accuracy 0.93914
step 19260, cost 147.562
step 19260, change in cost 0.0276184
step 19270, training accuracy 0.93914
step 19270, cost 147.535
step 19270, change in cost 0.0275574
step 19280, training accuracy 0.93914
step 19280, cost 147.507
step 19280, change in cost 0.0276337
step 19290, training accuracy 0.93914
step 19290, cost 147.48
step 19290, change in cost 0.0275574
step 19300, training accuracy 0.93914
step 19300, cost 147.452
step 19300, change in cost 0.0275879
step 19310, training accuracy 0.93914
step 19310, cost 147.424
step 19310, change in cost 0.0275269
step 19320, training accuracy 0.93914
step 19320, cost 147.397
step 19320, change in cost 0.0275269
step 19330, training accuracy 0.93914
step 19330, cost 147.369
step 19330, change in cost 0.0274963
step 19340, training accuracy 0.93914
step 19340, cost 147.342
step 19340, change in cost 0.0275726
step 19350, training accuracy 0.93914
step 19350, cost 147.314
step 19350, change in cost 0.0274658
step 19360, training accuracy 0.93914
step 19360, cost 147.287
step 19360, change in cost 0.0275116
step 19370, training accuracy 0.93914
step 19370, cost 147.259
step 19370, change in cost 0.0274506
step 19380, training accuracy 0.93914
step 19380, cost 147.232
step 19380, change in cost 0.0274811
step 19390, training accuracy 0.93914
step 19390, cost 147.205
step 19390, change in cost 0.02742
step 19400, training accuracy 0.93914
step 19400, cost 147.177
step 19400, change in cost 0.0275116
step 19410, training accuracy 0.93914
step 19410, cost 147.15
step 19410, change in cost 0.0274506
step 19420, training accuracy 0.93914
step 19420, cost 147.122
step 19420, change in cost 0.0273743
step 19430, training accuracy 0.93914
step 19430, cost 147.095
step 19430, change in cost 0.0274506
step 19440, training accuracy 0.93914
step 19440, cost 147.067
step 19440, change in cost 0.0273895
step 19450, training accuracy 0.93914
step 19450, cost 147.04
step 19450, change in cost 0.027359
step 19460, training accuracy 0.93914
step 19460, cost 147.013
step 19460, change in cost 0.0273132
step 19470, training accuracy 0.93914
step 19470, cost 146.985
step 19470, change in cost 0.02742
step 19480, training accuracy 0.93914
step 19480, cost 146.958
step 19480, change in cost 0.0273438
step 19490, training accuracy 0.93914
step 19490, cost 146.931
step 19490, change in cost 0.0273285
step 19500, training accuracy 0.93914
step 19500, cost 146.903
step 19500, change in cost 0.027359
step 19510, training accuracy 0.93914
step 19510, cost 146.876
step 19510, change in cost 0.0273132
step 19520, training accuracy 0.93914
step 19520, cost 146.849
step 19520, change in cost 0.027298
step 19530, training accuracy 0.93914
step 19530, cost 146.821
step 19530, change in cost 0.0272522
step 19540, training accuracy 0.93914
step 19540, cost 146.794
step 19540, change in cost 0.027298
step 19550, training accuracy 0.93914
step 19550, cost 146.767
step 19550, change in cost 0.0272522
step 19560, training accuracy 0.93914
step 19560, cost 146.74
step 19560, change in cost 0.0272827
step 19570, training accuracy 0.93914
step 19570, cost 146.712
step 19570, change in cost 0.0272369
step 19580, training accuracy 0.93914
step 19580, cost 146.685
step 19580, change in cost 0.0272522
step 19590, training accuracy 0.93914
step 19590, cost 146.658
step 19590, change in cost 0.0272064
step 19600, training accuracy 0.93914
step 19600, cost 146.631
step 19600, change in cost 0.0272217
step 19610, training accuracy 0.93914
step 19610, cost 146.603
step 19610, change in cost 0.0271759
step 19620, training accuracy 0.93914
step 19620, cost 146.576
step 19620, change in cost 0.0272064
step 19630, training accuracy 0.93914
step 19630, cost 146.549
step 19630, change in cost 0.0271301
step 19640, training accuracy 0.93914
step 19640, cost 146.522
step 19640, change in cost 0.0272064
step 19650, training accuracy 0.93914
step 19650, cost 146.495
step 19650, change in cost 0.0271912
step 19660, training accuracy 0.93914
step 19660, cost 146.468
step 19660, change in cost 0.0270996
step 19670, training accuracy 0.93914
step 19670, cost 146.441
step 19670, change in cost 0.0270844
step 19680, training accuracy 0.93914
step 19680, cost 146.413
step 19680, change in cost 0.0271606
step 19690, training accuracy 0.93914
step 19690, cost 146.386
step 19690, change in cost 0.0270996
step 19700, training accuracy 0.93914
step 19700, cost 146.359
step 19700, change in cost 0.0270844
step 19710, training accuracy 0.93914
step 19710, cost 146.332
step 19710, change in cost 0.0271606
step 19720, training accuracy 0.93914
step 19720, cost 146.305
step 19720, change in cost 0.0270233
step 19730, training accuracy 0.93914
step 19730, cost 146.278
step 19730, change in cost 0.0270844
step 19740, training accuracy 0.93914
step 19740, cost 146.251
step 19740, change in cost 0.0270691
step 19750, training accuracy 0.93914
step 19750, cost 146.224
step 19750, change in cost 0.0270386
step 19760, training accuracy 0.93914
step 19760, cost 146.197
step 19760, change in cost 0.0270233
step 19770, training accuracy 0.93914
step 19770, cost 146.17
step 19770, change in cost 0.0269775
step 19780, training accuracy 0.93914
step 19780, cost 146.143
step 19780, change in cost 0.0270081
step 19790, training accuracy 0.93914
step 19790, cost 146.116
step 19790, change in cost 0.0270233
step 19800, training accuracy 0.93914
step 19800, cost 146.089
step 19800, change in cost 0.0269165
step 19810, training accuracy 0.93914
step 19810, cost 146.062
step 19810, change in cost 0.0270691
step 19820, training accuracy 0.93914
step 19820, cost 146.035
step 19820, change in cost 0.026947
step 19830, training accuracy 0.93914
step 19830, cost 146.008
step 19830, change in cost 0.026947
step 19840, training accuracy 0.93914
step 19840, cost 145.981
step 19840, change in cost 0.0269012
step 19850, training accuracy 0.93914
step 19850, cost 145.954
step 19850, change in cost 0.026947
step 19860, training accuracy 0.93914
step 19860, cost 145.927
step 19860, change in cost 0.0269012
step 19870, training accuracy 0.93914
step 19870, cost 145.9
step 19870, change in cost 0.0269012
step 19880, training accuracy 0.93914
step 19880, cost 145.873
step 19880, change in cost 0.0268555
step 19890, training accuracy 0.93914
step 19890, cost 145.846
step 19890, change in cost 0.026947
step 19900, training accuracy 0.93914
step 19900, cost 145.82
step 19900, change in cost 0.0268402
step 19910, training accuracy 0.93914
step 19910, cost 145.793
step 19910, change in cost 0.0268555
step 19920, training accuracy 0.93914
step 19920, cost 145.766
step 19920, change in cost 0.0268402
step 19930, training accuracy 0.93914
step 19930, cost 145.739
step 19930, change in cost 0.0268555
step 19940, training accuracy 0.93914
step 19940, cost 145.712
step 19940, change in cost 0.0267639
step 19950, training accuracy 0.93914
step 19950, cost 145.686
step 19950, change in cost 0.026825
step 19960, training accuracy 0.93914
step 19960, cost 145.659
step 19960, change in cost 0.0268707
step 19970, training accuracy 0.93914
step 19970, cost 145.632
step 19970, change in cost 0.0267487
step 19980, training accuracy 0.93914
step 19980, cost 145.605
step 19980, change in cost 0.0267944
step 19990, training accuracy 0.93914
step 19990, cost 145.578
step 19990, change in cost 0.0267334
step 20000, training accuracy 0.93914
step 20000, cost 145.552
step 20000, change in cost 0.0267792
step 20010, training accuracy 0.93914
step 20010, cost 145.525
step 20010, change in cost 0.0267487
step 20020, training accuracy 0.93914
step 20020, cost 145.498
step 20020, change in cost 0.0267334
step 20030, training accuracy 0.93914
step 20030, cost 145.471
step 20030, change in cost 0.0267181
step 20040, training accuracy 0.93914
step 20040, cost 145.445
step 20040, change in cost 0.0267029
step 20050, training accuracy 0.93914
step 20050, cost 145.418
step 20050, change in cost 0.0267181
step 20060, training accuracy 0.93914
step 20060, cost 145.391
step 20060, change in cost 0.0267029
step 20070, training accuracy 0.93914
step 20070, cost 145.365
step 20070, change in cost 0.0266876
step 20080, training accuracy 0.93914
step 20080, cost 145.338
step 20080, change in cost 0.0266418
step 20090, training accuracy 0.93914
step 20090, cost 145.311
step 20090, change in cost 0.0266724
step 20100, training accuracy 0.93914
step 20100, cost 145.285
step 20100, change in cost 0.0266266
step 20110, training accuracy 0.93914
step 20110, cost 145.258
step 20110, change in cost 0.0266571
step 20120, training accuracy 0.93914
step 20120, cost 145.231
step 20120, change in cost 0.0265808
step 20130, training accuracy 0.93914
step 20130, cost 145.205
step 20130, change in cost 0.0265961
step 20140, training accuracy 0.93914
step 20140, cost 145.178
step 20140, change in cost 0.0266266
step 20150, training accuracy 0.940189
step 20150, cost 145.152
step 20150, change in cost 0.0265503
step 20160, training accuracy 0.940189
step 20160, cost 145.125
step 20160, change in cost 0.0266113
step 20170, training accuracy 0.940189
step 20170, cost 145.098
step 20170, change in cost 0.0265808
step 20180, training accuracy 0.940189
step 20180, cost 145.072
step 20180, change in cost 0.0265808
step 20190, training accuracy 0.940189
step 20190, cost 145.045
step 20190, change in cost 0.0265198
step 20200, training accuracy 0.940189
step 20200, cost 145.019
step 20200, change in cost 0.0265808
step 20210, training accuracy 0.940189
step 20210, cost 144.992
step 20210, change in cost 0.026474
step 20220, training accuracy 0.940189
step 20220, cost 144.966
step 20220, change in cost 0.0265198
step 20230, training accuracy 0.940189
step 20230, cost 144.939
step 20230, change in cost 0.0265198
step 20240, training accuracy 0.940189
step 20240, cost 144.913
step 20240, change in cost 0.026474
step 20250, training accuracy 0.940189
step 20250, cost 144.886
step 20250, change in cost 0.0265045
step 20260, training accuracy 0.940189
step 20260, cost 144.86
step 20260, change in cost 0.0264435
step 20270, training accuracy 0.940189
step 20270, cost 144.833
step 20270, change in cost 0.0264282
step 20280, training accuracy 0.940189
step 20280, cost 144.807
step 20280, change in cost 0.0265198
step 20290, training accuracy 0.940189
step 20290, cost 144.78
step 20290, change in cost 0.0264282
step 20300, training accuracy 0.940189
step 20300, cost 144.754
step 20300, change in cost 0.0263824
step 20310, training accuracy 0.940189
step 20310, cost 144.728
step 20310, change in cost 0.026413
step 20320, training accuracy 0.940189
step 20320, cost 144.701
step 20320, change in cost 0.0263977
step 20330, training accuracy 0.940189
step 20330, cost 144.675
step 20330, change in cost 0.0263672
step 20340, training accuracy 0.940189
step 20340, cost 144.648
step 20340, change in cost 0.026413
step 20350, training accuracy 0.940189
step 20350, cost 144.622
step 20350, change in cost 0.0263214
step 20360, training accuracy 0.940189
step 20360, cost 144.596
step 20360, change in cost 0.0263977
step 20370, training accuracy 0.940189
step 20370, cost 144.569
step 20370, change in cost 0.0263214
step 20380, training accuracy 0.940189
step 20380, cost 144.543
step 20380, change in cost 0.0263519
step 20390, training accuracy 0.940189
step 20390, cost 144.517
step 20390, change in cost 0.0263062
step 20400, training accuracy 0.940189
step 20400, cost 144.49
step 20400, change in cost 0.0263672
step 20410, training accuracy 0.940189
step 20410, cost 144.464
step 20410, change in cost 0.0262756
step 20420, training accuracy 0.940189
step 20420, cost 144.438
step 20420, change in cost 0.0262909
step 20430, training accuracy 0.940189
step 20430, cost 144.412
step 20430, change in cost 0.0262299
step 20440, training accuracy 0.940189
step 20440, cost 144.385
step 20440, change in cost 0.0263062
step 20450, training accuracy 0.940189
step 20450, cost 144.359
step 20450, change in cost 0.0262604
step 20460, training accuracy 0.940189
step 20460, cost 144.333
step 20460, change in cost 0.0262756
step 20470, training accuracy 0.940189
step 20470, cost 144.307
step 20470, change in cost 0.0262299
step 20480, training accuracy 0.940189
step 20480, cost 144.28
step 20480, change in cost 0.0262146
step 20490, training accuracy 0.940189
step 20490, cost 144.254
step 20490, change in cost 0.0262299
step 20500, training accuracy 0.940189
step 20500, cost 144.228
step 20500, change in cost 0.0262146
step 20510, training accuracy 0.940189
step 20510, cost 144.202
step 20510, change in cost 0.0261688
step 20520, training accuracy 0.940189
step 20520, cost 144.175
step 20520, change in cost 0.0262299
step 20530, training accuracy 0.940189
step 20530, cost 144.149
step 20530, change in cost 0.0261688
step 20540, training accuracy 0.940189
step 20540, cost 144.123
step 20540, change in cost 0.0261078
step 20550, training accuracy 0.940189
step 20550, cost 144.097
step 20550, change in cost 0.0261688
step 20560, training accuracy 0.940189
step 20560, cost 144.071
step 20560, change in cost 0.0261688
step 20570, training accuracy 0.940189
step 20570, cost 144.045
step 20570, change in cost 0.0260925
step 20580, training accuracy 0.940189
step 20580, cost 144.019
step 20580, change in cost 0.0260925
step 20590, training accuracy 0.940189
step 20590, cost 143.993
step 20590, change in cost 0.0261383
step 20600, training accuracy 0.940189
step 20600, cost 143.966
step 20600, change in cost 0.0261078
step 20610, training accuracy 0.940189
step 20610, cost 143.94
step 20610, change in cost 0.026001
step 20620, training accuracy 0.940189
step 20620, cost 143.914
step 20620, change in cost 0.0261383
step 20630, training accuracy 0.940189
step 20630, cost 143.888
step 20630, change in cost 0.0260468
step 20640, training accuracy 0.940189
step 20640, cost 143.862
step 20640, change in cost 0.0260925
step 20650, training accuracy 0.940189
step 20650, cost 143.836
step 20650, change in cost 0.026062
step 20660, training accuracy 0.940189
step 20660, cost 143.81
step 20660, change in cost 0.0259552
step 20670, training accuracy 0.940189
step 20670, cost 143.784
step 20670, change in cost 0.0260925
step 20680, training accuracy 0.940189
step 20680, cost 143.758
step 20680, change in cost 0.0259705
step 20690, training accuracy 0.940189
step 20690, cost 143.732
step 20690, change in cost 0.0259857
step 20700, training accuracy 0.940189
step 20700, cost 143.706
step 20700, change in cost 0.0260162
step 20710, training accuracy 0.940189
step 20710, cost 143.68
step 20710, change in cost 0.0259705
step 20720, training accuracy 0.940189
step 20720, cost 143.654
step 20720, change in cost 0.0259399
step 20730, training accuracy 0.940189
step 20730, cost 143.628
step 20730, change in cost 0.0259552
step 20740, training accuracy 0.940189
step 20740, cost 143.602
step 20740, change in cost 0.0259552
step 20750, training accuracy 0.940189
step 20750, cost 143.576
step 20750, change in cost 0.0259857
step 20760, training accuracy 0.940189
step 20760, cost 143.55
step 20760, change in cost 0.0259094
step 20770, training accuracy 0.940189
step 20770, cost 143.524
step 20770, change in cost 0.0259094
step 20780, training accuracy 0.940189
step 20780, cost 143.499
step 20780, change in cost 0.0258789
step 20790, training accuracy 0.940189
step 20790, cost 143.473
step 20790, change in cost 0.0258942
step 20800, training accuracy 0.940189
step 20800, cost 143.447
step 20800, change in cost 0.0258942
step 20810, training accuracy 0.940189
step 20810, cost 143.421
step 20810, change in cost 0.0258636
step 20820, training accuracy 0.940189
step 20820, cost 143.395
step 20820, change in cost 0.0258179
step 20830, training accuracy 0.940189
step 20830, cost 143.369
step 20830, change in cost 0.0258636
step 20840, training accuracy 0.940189
step 20840, cost 143.343
step 20840, change in cost 0.0259094
step 20850, training accuracy 0.940189
step 20850, cost 143.318
step 20850, change in cost 0.0257721
step 20860, training accuracy 0.940189
step 20860, cost 143.292
step 20860, change in cost 0.0258636
step 20870, training accuracy 0.940189
step 20870, cost 143.266
step 20870, change in cost 0.0258026
step 20880, training accuracy 0.940189
step 20880, cost 143.24
step 20880, change in cost 0.0258179
step 20890, training accuracy 0.940189
step 20890, cost 143.214
step 20890, change in cost 0.0257263
step 20900, training accuracy 0.940189
step 20900, cost 143.189
step 20900, change in cost 0.0257721
step 20910, training accuracy 0.940189
step 20910, cost 143.163
step 20910, change in cost 0.0257721
step 20920, training accuracy 0.940189
step 20920, cost 143.137
step 20920, change in cost 0.0257874
step 20930, training accuracy 0.940189
step 20930, cost 143.111
step 20930, change in cost 0.0257416
step 20940, training accuracy 0.940189
step 20940, cost 143.086
step 20940, change in cost 0.0257263
step 20950, training accuracy 0.941238
step 20950, cost 143.06
step 20950, change in cost 0.0256958
step 20960, training accuracy 0.941238
step 20960, cost 143.034
step 20960, change in cost 0.0257111
step 20970, training accuracy 0.941238
step 20970, cost 143.008
step 20970, change in cost 0.0256958
step 20980, training accuracy 0.941238
step 20980, cost 142.983
step 20980, change in cost 0.0256805
step 20990, training accuracy 0.941238
step 20990, cost 142.957
step 20990, change in cost 0.0257111
step 21000, training accuracy 0.941238
step 21000, cost 142.931
step 21000, change in cost 0.0256348
step 21010, training accuracy 0.941238
step 21010, cost 142.906
step 21010, change in cost 0.0256958
step 21020, training accuracy 0.941238
step 21020, cost 142.88
step 21020, change in cost 0.0256195
step 21030, training accuracy 0.941238
step 21030, cost 142.854
step 21030, change in cost 0.02565
step 21040, training accuracy 0.941238
step 21040, cost 142.829
step 21040, change in cost 0.0256042
step 21050, training accuracy 0.941238
step 21050, cost 142.803
step 21050, change in cost 0.0256805
step 21060, training accuracy 0.941238
step 21060, cost 142.778
step 21060, change in cost 0.0255737
step 21070, training accuracy 0.941238
step 21070, cost 142.752
step 21070, change in cost 0.0256042
step 21080, training accuracy 0.941238
step 21080, cost 142.726
step 21080, change in cost 0.0255737
step 21090, training accuracy 0.941238
step 21090, cost 142.701
step 21090, change in cost 0.025589
step 21100, training accuracy 0.941238
step 21100, cost 142.675
step 21100, change in cost 0.0255127
step 21110, training accuracy 0.941238
step 21110, cost 142.65
step 21110, change in cost 0.0256042
step 21120, training accuracy 0.941238
step 21120, cost 142.624
step 21120, change in cost 0.0255432
step 21130, training accuracy 0.941238
step 21130, cost 142.599
step 21130, change in cost 0.0255585
step 21140, training accuracy 0.941238
step 21140, cost 142.573
step 21140, change in cost 0.0254974
step 21150, training accuracy 0.941238
step 21150, cost 142.548
step 21150, change in cost 0.0255432
step 21160, training accuracy 0.941238
step 21160, cost 142.522
step 21160, change in cost 0.0254669
step 21170, training accuracy 0.941238
step 21170, cost 142.497
step 21170, change in cost 0.0254822
step 21180, training accuracy 0.941238
step 21180, cost 142.471
step 21180, change in cost 0.0255127
step 21190, training accuracy 0.941238
step 21190, cost 142.446
step 21190, change in cost 0.0254517
step 21200, training accuracy 0.941238
step 21200, cost 142.42
step 21200, change in cost 0.0253906
step 21210, training accuracy 0.941238
step 21210, cost 142.395
step 21210, change in cost 0.0254974
step 21220, training accuracy 0.941238
step 21220, cost 142.369
step 21220, change in cost 0.0254517
step 21230, training accuracy 0.941238
step 21230, cost 142.344
step 21230, change in cost 0.0254517
step 21240, training accuracy 0.941238
step 21240, cost 142.318
step 21240, change in cost 0.0254364
step 21250, training accuracy 0.941238
step 21250, cost 142.293
step 21250, change in cost 0.0253601
step 21260, training accuracy 0.941238
step 21260, cost 142.268
step 21260, change in cost 0.0253754
step 21270, training accuracy 0.941238
step 21270, cost 142.242
step 21270, change in cost 0.0254517
step 21280, training accuracy 0.941238
step 21280, cost 142.217
step 21280, change in cost 0.0253143
step 21290, training accuracy 0.941238
step 21290, cost 142.192
step 21290, change in cost 0.0253754
step 21300, training accuracy 0.941238
step 21300, cost 142.166
step 21300, change in cost 0.0253601
step 21310, training accuracy 0.941238
step 21310, cost 142.141
step 21310, change in cost 0.0253143
step 21320, training accuracy 0.941238
step 21320, cost 142.115
step 21320, change in cost 0.0254059
step 21330, training accuracy 0.941238
step 21330, cost 142.09
step 21330, change in cost 0.0252686
step 21340, training accuracy 0.941238
step 21340, cost 142.065
step 21340, change in cost 0.0253448
step 21350, training accuracy 0.941238
step 21350, cost 142.04
step 21350, change in cost 0.0252991
step 21360, training accuracy 0.941238
step 21360, cost 142.014
step 21360, change in cost 0.0253143
step 21370, training accuracy 0.941238
step 21370, cost 141.989
step 21370, change in cost 0.0252838
step 21380, training accuracy 0.941238
step 21380, cost 141.964
step 21380, change in cost 0.0252838
step 21390, training accuracy 0.941238
step 21390, cost 141.938
step 21390, change in cost 0.0252686
step 21400, training accuracy 0.941238
step 21400, cost 141.913
step 21400, change in cost 0.0252686
step 21410, training accuracy 0.941238
step 21410, cost 141.888
step 21410, change in cost 0.025238
step 21420, training accuracy 0.941238
step 21420, cost 141.863
step 21420, change in cost 0.0251923
step 21430, training accuracy 0.941238
step 21430, cost 141.837
step 21430, change in cost 0.0252533
step 21440, training accuracy 0.941238
step 21440, cost 141.812
step 21440, change in cost 0.025177
step 21450, training accuracy 0.941238
step 21450, cost 141.787
step 21450, change in cost 0.025238
step 21460, training accuracy 0.941238
step 21460, cost 141.762
step 21460, change in cost 0.0251617
step 21470, training accuracy 0.941238
step 21470, cost 141.737
step 21470, change in cost 0.0251617
step 21480, training accuracy 0.941238
step 21480, cost 141.711
step 21480, change in cost 0.0252228
step 21490, training accuracy 0.941238
step 21490, cost 141.686
step 21490, change in cost 0.025116
step 21500, training accuracy 0.941238
step 21500, cost 141.661
step 21500, change in cost 0.0251007
step 21510, training accuracy 0.941238
step 21510, cost 141.636
step 21510, change in cost 0.0252228
step 21520, training accuracy 0.941238
step 21520, cost 141.611
step 21520, change in cost 0.0251312
step 21530, training accuracy 0.941238
step 21530, cost 141.586
step 21530, change in cost 0.0251007
step 21540, training accuracy 0.941238
step 21540, cost 141.561
step 21540, change in cost 0.0251007
step 21550, training accuracy 0.941238
step 21550, cost 141.536
step 21550, change in cost 0.0251007
step 21560, training accuracy 0.941238
step 21560, cost 141.511
step 21560, change in cost 0.0250702
step 21570, training accuracy 0.941238
step 21570, cost 141.485
step 21570, change in cost 0.0251312
step 21580, training accuracy 0.941238
step 21580, cost 141.46
step 21580, change in cost 0.0250397
step 21590, training accuracy 0.941238
step 21590, cost 141.435
step 21590, change in cost 0.0250397
step 21600, training accuracy 0.941238
step 21600, cost 141.41
step 21600, change in cost 0.0250549
step 21610, training accuracy 0.941238
step 21610, cost 141.385
step 21610, change in cost 0.0250854
step 21620, training accuracy 0.941238
step 21620, cost 141.36
step 21620, change in cost 0.0250092
step 21630, training accuracy 0.941238
step 21630, cost 141.335
step 21630, change in cost 0.0249939
step 21640, training accuracy 0.941238
step 21640, cost 141.31
step 21640, change in cost 0.0250244
step 21650, training accuracy 0.941238
step 21650, cost 141.285
step 21650, change in cost 0.0249634
step 21660, training accuracy 0.941238
step 21660, cost 141.26
step 21660, change in cost 0.0249939
step 21670, training accuracy 0.941238
step 21670, cost 141.235
step 21670, change in cost 0.0250092
step 21680, training accuracy 0.941238
step 21680, cost 141.21
step 21680, change in cost 0.0249481
step 21690, training accuracy 0.941238
step 21690, cost 141.185
step 21690, change in cost 0.0249634
step 21700, training accuracy 0.941238
step 21700, cost 141.16
step 21700, change in cost 0.0249176
step 21710, training accuracy 0.941238
step 21710, cost 141.135
step 21710, change in cost 0.0249634
step 21720, training accuracy 0.941238
step 21720, cost 141.11
step 21720, change in cost 0.0249481
step 21730, training accuracy 0.941238
step 21730, cost 141.086
step 21730, change in cost 0.0249023
step 21740, training accuracy 0.941238
step 21740, cost 141.061
step 21740, change in cost 0.0249176
step 21750, training accuracy 0.941238
step 21750, cost 141.036
step 21750, change in cost 0.0248566
step 21760, training accuracy 0.941238
step 21760, cost 141.011
step 21760, change in cost 0.0248871
step 21770, training accuracy 0.941238
step 21770, cost 140.986
step 21770, change in cost 0.0249023
step 21780, training accuracy 0.941238
step 21780, cost 140.961
step 21780, change in cost 0.0248566
step 21790, training accuracy 0.941238
step 21790, cost 140.936
step 21790, change in cost 0.0248413
step 21800, training accuracy 0.941238
step 21800, cost 140.911
step 21800, change in cost 0.0248566
step 21810, training accuracy 0.941238
step 21810, cost 140.887
step 21810, change in cost 0.0248566
step 21820, training accuracy 0.941238
step 21820, cost 140.862
step 21820, change in cost 0.0247803
step 21830, training accuracy 0.941238
step 21830, cost 140.837
step 21830, change in cost 0.0248566
step 21840, training accuracy 0.941238
step 21840, cost 140.812
step 21840, change in cost 0.0248413
step 21850, training accuracy 0.941238
step 21850, cost 140.787
step 21850, change in cost 0.0247803
step 21860, training accuracy 0.941238
step 21860, cost 140.763
step 21860, change in cost 0.0247955
step 21870, training accuracy 0.941238
step 21870, cost 140.738
step 21870, change in cost 0.0247803
step 21880, training accuracy 0.941238
step 21880, cost 140.713
step 21880, change in cost 0.0247192
step 21890, training accuracy 0.941238
step 21890, cost 140.688
step 21890, change in cost 0.024765
step 21900, training accuracy 0.941238
step 21900, cost 140.663
step 21900, change in cost 0.024765
step 21910, training accuracy 0.941238
step 21910, cost 140.639
step 21910, change in cost 0.0247192
step 21920, training accuracy 0.941238
step 21920, cost 140.614
step 21920, change in cost 0.0247192
step 21930, training accuracy 0.941238
step 21930, cost 140.589
step 21930, change in cost 0.0247345
step 21940, training accuracy 0.941238
step 21940, cost 140.565
step 21940, change in cost 0.0247192
step 21950, training accuracy 0.941238
step 21950, cost 140.54
step 21950, change in cost 0.0247192
step 21960, training accuracy 0.941238
step 21960, cost 140.515
step 21960, change in cost 0.0246429
step 21970, training accuracy 0.941238
step 21970, cost 140.491
step 21970, change in cost 0.024704
step 21980, training accuracy 0.941238
step 21980, cost 140.466
step 21980, change in cost 0.0246582
step 21990, training accuracy 0.941238
step 21990, cost 140.441
step 21990, change in cost 0.0246277
step 22000, training accuracy 0.941238
step 22000, cost 140.417
step 22000, change in cost 0.024704
step 22010, training accuracy 0.941238
step 22010, cost 140.392
step 22010, change in cost 0.0245819
step 22020, training accuracy 0.941238
step 22020, cost 140.367
step 22020, change in cost 0.0246582
step 22030, training accuracy 0.941238
step 22030, cost 140.343
step 22030, change in cost 0.0246124
step 22040, training accuracy 0.941238
step 22040, cost 140.318
step 22040, change in cost 0.0246277
step 22050, training accuracy 0.941238
step 22050, cost 140.293
step 22050, change in cost 0.0245819
step 22060, training accuracy 0.941238
step 22060, cost 140.269
step 22060, change in cost 0.0245667
step 22070, training accuracy 0.941238
step 22070, cost 140.244
step 22070, change in cost 0.0246277
step 22080, training accuracy 0.941238
step 22080, cost 140.22
step 22080, change in cost 0.0245209
step 22090, training accuracy 0.941238
step 22090, cost 140.195
step 22090, change in cost 0.0245819
step 22100, training accuracy 0.941238
step 22100, cost 140.171
step 22100, change in cost 0.0245209
step 22110, training accuracy 0.941238
step 22110, cost 140.146
step 22110, change in cost 0.0245514
step 22120, training accuracy 0.941238
step 22120, cost 140.122
step 22120, change in cost 0.0245056
step 22130, training accuracy 0.941238
step 22130, cost 140.097
step 22130, change in cost 0.0245514
step 22140, training accuracy 0.941238
step 22140, cost 140.072
step 22140, change in cost 0.0245514
step 22150, training accuracy 0.941238
step 22150, cost 140.048
step 22150, change in cost 0.0245056
step 22160, training accuracy 0.941238
step 22160, cost 140.024
step 22160, change in cost 0.0244598
step 22170, training accuracy 0.941238
step 22170, cost 139.999
step 22170, change in cost 0.0244751
step 22180, training accuracy 0.941238
step 22180, cost 139.975
step 22180, change in cost 0.0244293
step 22190, training accuracy 0.941238
step 22190, cost 139.95
step 22190, change in cost 0.0244751
step 22200, training accuracy 0.941238
step 22200, cost 139.926
step 22200, change in cost 0.0244751
step 22210, training accuracy 0.941238
step 22210, cost 139.901
step 22210, change in cost 0.0244446
step 22220, training accuracy 0.941238
step 22220, cost 139.877
step 22220, change in cost 0.0243988
step 22230, training accuracy 0.941238
step 22230, cost 139.852
step 22230, change in cost 0.0244293
step 22240, training accuracy 0.941238
step 22240, cost 139.828
step 22240, change in cost 0.0244293
step 22250, training accuracy 0.941238
step 22250, cost 139.804
step 22250, change in cost 0.0244141
step 22260, training accuracy 0.941238
step 22260, cost 139.779
step 22260, change in cost 0.0243835
step 22270, training accuracy 0.941238
step 22270, cost 139.755
step 22270, change in cost 0.0243683
step 22280, training accuracy 0.941238
step 22280, cost 139.73
step 22280, change in cost 0.0244141
step 22290, training accuracy 0.941238
step 22290, cost 139.706
step 22290, change in cost 0.0243225
step 22300, training accuracy 0.941238
step 22300, cost 139.682
step 22300, change in cost 0.0243835
step 22310, training accuracy 0.941238
step 22310, cost 139.657
step 22310, change in cost 0.0243225
step 22320, training accuracy 0.941238
step 22320, cost 139.633
step 22320, change in cost 0.024353
step 22330, training accuracy 0.941238
step 22330, cost 139.609
step 22330, change in cost 0.0243073
step 22340, training accuracy 0.942288
step 22340, cost 139.584
step 22340, change in cost 0.0242615
step 22350, training accuracy 0.942288
step 22350, cost 139.56
step 22350, change in cost 0.0243683
step 22360, training accuracy 0.942288
step 22360, cost 139.536
step 22360, change in cost 0.0242615
step 22370, training accuracy 0.942288
step 22370, cost 139.512
step 22370, change in cost 0.024292
step 22380, training accuracy 0.942288
step 22380, cost 139.487
step 22380, change in cost 0.0243225
step 22390, training accuracy 0.942288
step 22390, cost 139.463
step 22390, change in cost 0.0241852
step 22400, training accuracy 0.942288
step 22400, cost 139.439
step 22400, change in cost 0.024292
step 22410, training accuracy 0.942288
step 22410, cost 139.414
step 22410, change in cost 0.0242615
step 22420, training accuracy 0.942288
step 22420, cost 139.39
step 22420, change in cost 0.0241852
step 22430, training accuracy 0.942288
step 22430, cost 139.366
step 22430, change in cost 0.024231
step 22440, training accuracy 0.942288
step 22440, cost 139.342
step 22440, change in cost 0.0242462
step 22450, training accuracy 0.942288
step 22450, cost 139.318
step 22450, change in cost 0.0242462
step 22460, training accuracy 0.942288
step 22460, cost 139.293
step 22460, change in cost 0.0241547
step 22470, training accuracy 0.942288
step 22470, cost 139.269
step 22470, change in cost 0.0241852
step 22480, training accuracy 0.942288
step 22480, cost 139.245
step 22480, change in cost 0.0241852
step 22490, training accuracy 0.942288
step 22490, cost 139.221
step 22490, change in cost 0.0241547
step 22500, training accuracy 0.942288
step 22500, cost 139.197
step 22500, change in cost 0.0242004
step 22510, training accuracy 0.942288
step 22510, cost 139.173
step 22510, change in cost 0.0241394
step 22520, training accuracy 0.942288
step 22520, cost 139.148
step 22520, change in cost 0.0241394
step 22530, training accuracy 0.942288
step 22530, cost 139.124
step 22530, change in cost 0.0241241
step 22540, training accuracy 0.942288
step 22540, cost 139.1
step 22540, change in cost 0.0241089
step 22550, training accuracy 0.942288
step 22550, cost 139.076
step 22550, change in cost 0.0241394
step 22560, training accuracy 0.942288
step 22560, cost 139.052
step 22560, change in cost 0.0241089
step 22570, training accuracy 0.942288
step 22570, cost 139.028
step 22570, change in cost 0.0241089
step 22580, training accuracy 0.942288
step 22580, cost 139.004
step 22580, change in cost 0.0240631
step 22590, training accuracy 0.942288
step 22590, cost 138.98
step 22590, change in cost 0.0240784
step 22600, training accuracy 0.942288
step 22600, cost 138.956
step 22600, change in cost 0.0240631
step 22610, training accuracy 0.942288
step 22610, cost 138.932
step 22610, change in cost 0.0240173
step 22620, training accuracy 0.942288
step 22620, cost 138.907
step 22620, change in cost 0.0240936
step 22630, training accuracy 0.942288
step 22630, cost 138.883
step 22630, change in cost 0.0240326
step 22640, training accuracy 0.942288
step 22640, cost 138.859
step 22640, change in cost 0.0240326
step 22650, training accuracy 0.942288
step 22650, cost 138.835
step 22650, change in cost 0.0240173
step 22660, training accuracy 0.942288
step 22660, cost 138.811
step 22660, change in cost 0.0240021
step 22670, training accuracy 0.942288
step 22670, cost 138.787
step 22670, change in cost 0.0239868
step 22680, training accuracy 0.942288
step 22680, cost 138.763
step 22680, change in cost 0.0240021
step 22690, training accuracy 0.942288
step 22690, cost 138.739
step 22690, change in cost 0.023941
step 22700, training accuracy 0.942288
step 22700, cost 138.715
step 22700, change in cost 0.0240173
step 22710, training accuracy 0.942288
step 22710, cost 138.692
step 22710, change in cost 0.02388
step 22720, training accuracy 0.942288
step 22720, cost 138.668
step 22720, change in cost 0.0240021
step 22730, training accuracy 0.942288
step 22730, cost 138.644
step 22730, change in cost 0.0239258
step 22740, training accuracy 0.942288
step 22740, cost 138.62
step 22740, change in cost 0.0239258
step 22750, training accuracy 0.942288
step 22750, cost 138.596
step 22750, change in cost 0.0238953
step 22760, training accuracy 0.942288
step 22760, cost 138.572
step 22760, change in cost 0.0239258
step 22770, training accuracy 0.942288
step 22770, cost 138.548
step 22770, change in cost 0.0239563
step 22780, training accuracy 0.942288
step 22780, cost 138.524
step 22780, change in cost 0.02388
step 22790, training accuracy 0.942288
step 22790, cost 138.5
step 22790, change in cost 0.0238953
step 22800, training accuracy 0.942288
step 22800, cost 138.476
step 22800, change in cost 0.0237885
step 22810, training accuracy 0.942288
step 22810, cost 138.452
step 22810, change in cost 0.0239258
step 22820, training accuracy 0.942288
step 22820, cost 138.429
step 22820, change in cost 0.0238647
step 22830, training accuracy 0.942288
step 22830, cost 138.405
step 22830, change in cost 0.0238495
step 22840, training accuracy 0.942288
step 22840, cost 138.381
step 22840, change in cost 0.0238342
step 22850, training accuracy 0.942288
step 22850, cost 138.357
step 22850, change in cost 0.023819
step 22860, training accuracy 0.942288
step 22860, cost 138.333
step 22860, change in cost 0.0238037
step 22870, training accuracy 0.942288
step 22870, cost 138.309
step 22870, change in cost 0.023819
step 22880, training accuracy 0.942288
step 22880, cost 138.286
step 22880, change in cost 0.0238495
step 22890, training accuracy 0.942288
step 22890, cost 138.262
step 22890, change in cost 0.0237427
step 22900, training accuracy 0.942288
step 22900, cost 138.238
step 22900, change in cost 0.0238037
step 22910, training accuracy 0.942288
step 22910, cost 138.214
step 22910, change in cost 0.0237427
step 22920, training accuracy 0.942288
step 22920, cost 138.191
step 22920, change in cost 0.0237427
step 22930, training accuracy 0.942288
step 22930, cost 138.167
step 22930, change in cost 0.0237885
step 22940, training accuracy 0.942288
step 22940, cost 138.143
step 22940, change in cost 0.0237274
step 22950, training accuracy 0.942288
step 22950, cost 138.119
step 22950, change in cost 0.0237122
step 22960, training accuracy 0.942288
step 22960, cost 138.096
step 22960, change in cost 0.0237274
step 22970, training accuracy 0.942288
step 22970, cost 138.072
step 22970, change in cost 0.0237732
step 22980, training accuracy 0.942288
step 22980, cost 138.048
step 22980, change in cost 0.0236969
step 22990, training accuracy 0.942288
step 22990, cost 138.024
step 22990, change in cost 0.0236664
step 23000, training accuracy 0.943337
step 23000, cost 138.001
step 23000, change in cost 0.0236816
step 23010, training accuracy 0.943337
step 23010, cost 137.977
step 23010, change in cost 0.0236816
step 23020, training accuracy 0.943337
step 23020, cost 137.953
step 23020, change in cost 0.0236816
step 23030, training accuracy 0.943337
step 23030, cost 137.93
step 23030, change in cost 0.0236206
step 23040, training accuracy 0.943337
step 23040, cost 137.906
step 23040, change in cost 0.0236816
step 23050, training accuracy 0.943337
step 23050, cost 137.882
step 23050, change in cost 0.0236511
step 23060, training accuracy 0.943337
step 23060, cost 137.859
step 23060, change in cost 0.0236206
step 23070, training accuracy 0.943337
step 23070, cost 137.835
step 23070, change in cost 0.0236053
step 23080, training accuracy 0.943337
step 23080, cost 137.812
step 23080, change in cost 0.0236664
step 23090, training accuracy 0.943337
step 23090, cost 137.788
step 23090, change in cost 0.0235596
step 23100, training accuracy 0.943337
step 23100, cost 137.764
step 23100, change in cost 0.0236053
step 23110, training accuracy 0.943337
step 23110, cost 137.741
step 23110, change in cost 0.0236206
step 23120, training accuracy 0.943337
step 23120, cost 137.717
step 23120, change in cost 0.0235138
step 23130, training accuracy 0.943337
step 23130, cost 137.694
step 23130, change in cost 0.0235901
step 23140, training accuracy 0.943337
step 23140, cost 137.67
step 23140, change in cost 0.0235443
step 23150, training accuracy 0.943337
step 23150, cost 137.647
step 23150, change in cost 0.0235596
step 23160, training accuracy 0.943337
step 23160, cost 137.623
step 23160, change in cost 0.0235748
step 23170, training accuracy 0.943337
step 23170, cost 137.6
step 23170, change in cost 0.0235138
step 23180, training accuracy 0.943337
step 23180, cost 137.576
step 23180, change in cost 0.0235138
step 23190, training accuracy 0.943337
step 23190, cost 137.552
step 23190, change in cost 0.0235291
step 23200, training accuracy 0.943337
step 23200, cost 137.529
step 23200, change in cost 0.0234528
step 23210, training accuracy 0.943337
step 23210, cost 137.506
step 23210, change in cost 0.0235138
step 23220, training accuracy 0.943337
step 23220, cost 137.482
step 23220, change in cost 0.023468
step 23230, training accuracy 0.943337
step 23230, cost 137.459
step 23230, change in cost 0.023468
step 23240, training accuracy 0.943337
step 23240, cost 137.435
step 23240, change in cost 0.0234985
step 23250, training accuracy 0.943337
step 23250, cost 137.412
step 23250, change in cost 0.0234222
step 23260, training accuracy 0.943337
step 23260, cost 137.388
step 23260, change in cost 0.0234528
step 23270, training accuracy 0.943337
step 23270, cost 137.365
step 23270, change in cost 0.023407
step 23280, training accuracy 0.943337
step 23280, cost 137.341
step 23280, change in cost 0.0234528
step 23290, training accuracy 0.943337
step 23290, cost 137.318
step 23290, change in cost 0.0233917
step 23300, training accuracy 0.943337
step 23300, cost 137.295
step 23300, change in cost 0.0233917
step 23310, training accuracy 0.943337
step 23310, cost 137.271
step 23310, change in cost 0.0234222
step 23320, training accuracy 0.943337
step 23320, cost 137.248
step 23320, change in cost 0.0233917
step 23330, training accuracy 0.943337
step 23330, cost 137.224
step 23330, change in cost 0.0233612
step 23340, training accuracy 0.943337
step 23340, cost 137.201
step 23340, change in cost 0.0234222
step 23350, training accuracy 0.943337
step 23350, cost 137.178
step 23350, change in cost 0.0233154
step 23360, training accuracy 0.943337
step 23360, cost 137.154
step 23360, change in cost 0.0234222
step 23370, training accuracy 0.943337
step 23370, cost 137.131
step 23370, change in cost 0.0233154
step 23380, training accuracy 0.943337
step 23380, cost 137.108
step 23380, change in cost 0.0233154
step 23390, training accuracy 0.943337
step 23390, cost 137.084
step 23390, change in cost 0.0233765
step 23400, training accuracy 0.943337
step 23400, cost 137.061
step 23400, change in cost 0.0232544
step 23410, training accuracy 0.943337
step 23410, cost 137.038
step 23410, change in cost 0.0233307
step 23420, training accuracy 0.943337
step 23420, cost 137.014
step 23420, change in cost 0.0232849
step 23430, training accuracy 0.943337
step 23430, cost 136.991
step 23430, change in cost 0.0233612
step 23440, training accuracy 0.943337
step 23440, cost 136.968
step 23440, change in cost 0.0232239
step 23450, training accuracy 0.943337
step 23450, cost 136.944
step 23450, change in cost 0.0232697
step 23460, training accuracy 0.943337
step 23460, cost 136.921
step 23460, change in cost 0.0233002
step 23470, training accuracy 0.943337
step 23470, cost 136.898
step 23470, change in cost 0.0232086
step 23480, training accuracy 0.944386
step 23480, cost 136.875
step 23480, change in cost 0.0232697
step 23490, training accuracy 0.944386
step 23490, cost 136.852
step 23490, change in cost 0.0231934
step 23500, training accuracy 0.944386
step 23500, cost 136.828
step 23500, change in cost 0.0232849
step 23510, training accuracy 0.944386
step 23510, cost 136.805
step 23510, change in cost 0.0231781
step 23520, training accuracy 0.944386
step 23520, cost 136.782
step 23520, change in cost 0.0232391
step 23530, training accuracy 0.944386
step 23530, cost 136.759
step 23530, change in cost 0.0232086
step 23540, training accuracy 0.944386
step 23540, cost 136.735
step 23540, change in cost 0.0231476
step 23550, training accuracy 0.944386
step 23550, cost 136.712
step 23550, change in cost 0.0231934
step 23560, training accuracy 0.944386
step 23560, cost 136.689
step 23560, change in cost 0.0231628
step 23570, training accuracy 0.944386
step 23570, cost 136.666
step 23570, change in cost 0.0231781
step 23580, training accuracy 0.944386
step 23580, cost 136.643
step 23580, change in cost 0.0231476
step 23590, training accuracy 0.944386
step 23590, cost 136.62
step 23590, change in cost 0.0231781
step 23600, training accuracy 0.944386
step 23600, cost 136.597
step 23600, change in cost 0.0230865
step 23610, training accuracy 0.944386
step 23610, cost 136.573
step 23610, change in cost 0.0231628
step 23620, training accuracy 0.944386
step 23620, cost 136.55
step 23620, change in cost 0.0231018
step 23630, training accuracy 0.944386
step 23630, cost 136.527
step 23630, change in cost 0.0231476
step 23640, training accuracy 0.944386
step 23640, cost 136.504
step 23640, change in cost 0.0230255
step 23650, training accuracy 0.944386
step 23650, cost 136.481
step 23650, change in cost 0.0231628
step 23660, training accuracy 0.944386
step 23660, cost 136.458
step 23660, change in cost 0.023056
step 23670, training accuracy 0.944386
step 23670, cost 136.435
step 23670, change in cost 0.023056
step 23680, training accuracy 0.944386
step 23680, cost 136.412
step 23680, change in cost 0.0231171
step 23690, training accuracy 0.944386
step 23690, cost 136.389
step 23690, change in cost 0.0230408
step 23700, training accuracy 0.944386
step 23700, cost 136.366
step 23700, change in cost 0.0230408
step 23710, training accuracy 0.944386
step 23710, cost 136.343
step 23710, change in cost 0.0230408
step 23720, training accuracy 0.944386
step 23720, cost 136.32
step 23720, change in cost 0.0230103
step 23730, training accuracy 0.944386
step 23730, cost 136.297
step 23730, change in cost 0.0230255
step 23740, training accuracy 0.944386
step 23740, cost 136.274
step 23740, change in cost 0.0230103
step 23750, training accuracy 0.944386
step 23750, cost 136.251
step 23750, change in cost 0.022995
step 23760, training accuracy 0.944386
step 23760, cost 136.228
step 23760, change in cost 0.0230103
step 23770, training accuracy 0.944386
step 23770, cost 136.205
step 23770, change in cost 0.0229797
step 23780, training accuracy 0.944386
step 23780, cost 136.182
step 23780, change in cost 0.0229492
step 23790, training accuracy 0.944386
step 23790, cost 136.159
step 23790, change in cost 0.022995
step 23800, training accuracy 0.944386
step 23800, cost 136.136
step 23800, change in cost 0.0229645
step 23810, training accuracy 0.944386
step 23810, cost 136.113
step 23810, change in cost 0.022934
step 23820, training accuracy 0.944386
step 23820, cost 136.09
step 23820, change in cost 0.0229492
step 23830, training accuracy 0.944386
step 23830, cost 136.067
step 23830, change in cost 0.0229492
step 23840, training accuracy 0.944386
step 23840, cost 136.044
step 23840, change in cost 0.0228882
step 23850, training accuracy 0.944386
step 23850, cost 136.021
step 23850, change in cost 0.0229797
step 23860, training accuracy 0.944386
step 23860, cost 135.998
step 23860, change in cost 0.0228271
step 23870, training accuracy 0.944386
step 23870, cost 135.975
step 23870, change in cost 0.022934
step 23880, training accuracy 0.944386
step 23880, cost 135.952
step 23880, change in cost 0.0228729
step 23890, training accuracy 0.944386
step 23890, cost 135.929
step 23890, change in cost 0.0229034
step 23900, training accuracy 0.944386
step 23900, cost 135.907
step 23900, change in cost 0.0228271
step 23910, training accuracy 0.944386
step 23910, cost 135.884
step 23910, change in cost 0.0228729
step 23920, training accuracy 0.944386
step 23920, cost 135.861
step 23920, change in cost 0.0228882
step 23930, training accuracy 0.944386
step 23930, cost 135.838
step 23930, change in cost 0.0228271
step 23940, training accuracy 0.944386
step 23940, cost 135.815
step 23940, change in cost 0.0227966
step 23950, training accuracy 0.944386
step 23950, cost 135.792
step 23950, change in cost 0.0228577
step 23960, training accuracy 0.944386
step 23960, cost 135.77
step 23960, change in cost 0.0228119
step 23970, training accuracy 0.944386
step 23970, cost 135.747
step 23970, change in cost 0.0228119
step 23980, training accuracy 0.944386
step 23980, cost 135.724
step 23980, change in cost 0.0227966
step 23990, training accuracy 0.944386
step 23990, cost 135.701
step 23990, change in cost 0.0228119
step 24000, training accuracy 0.944386
step 24000, cost 135.678
step 24000, change in cost 0.0227356
step 24010, training accuracy 0.944386
step 24010, cost 135.656
step 24010, change in cost 0.0227814
step 24020, training accuracy 0.944386
step 24020, cost 135.633
step 24020, change in cost 0.0227814
step 24030, training accuracy 0.944386
step 24030, cost 135.61
step 24030, change in cost 0.0227661
step 24040, training accuracy 0.944386
step 24040, cost 135.587
step 24040, change in cost 0.0227661
step 24050, training accuracy 0.944386
step 24050, cost 135.565
step 24050, change in cost 0.0227356
step 24060, training accuracy 0.944386
step 24060, cost 135.542
step 24060, change in cost 0.0226746
step 24070, training accuracy 0.944386
step 24070, cost 135.519
step 24070, change in cost 0.0227356
step 24080, training accuracy 0.944386
step 24080, cost 135.496
step 24080, change in cost 0.0227356
step 24090, training accuracy 0.944386
step 24090, cost 135.474
step 24090, change in cost 0.0226746
step 24100, training accuracy 0.944386
step 24100, cost 135.451
step 24100, change in cost 0.0227356
step 24110, training accuracy 0.944386
step 24110, cost 135.428
step 24110, change in cost 0.022644
step 24120, training accuracy 0.944386
step 24120, cost 135.406
step 24120, change in cost 0.0227051
step 24130, training accuracy 0.944386
step 24130, cost 135.383
step 24130, change in cost 0.0226135
step 24140, training accuracy 0.944386
step 24140, cost 135.36
step 24140, change in cost 0.0227051
step 24150, training accuracy 0.944386
step 24150, cost 135.338
step 24150, change in cost 0.0226288
step 24160, training accuracy 0.944386
step 24160, cost 135.315
step 24160, change in cost 0.0226593
step 24170, training accuracy 0.944386
step 24170, cost 135.292
step 24170, change in cost 0.0226593
step 24180, training accuracy 0.944386
step 24180, cost 135.27
step 24180, change in cost 0.0226135
step 24190, training accuracy 0.944386
step 24190, cost 135.247
step 24190, change in cost 0.022583
step 24200, training accuracy 0.944386
step 24200, cost 135.225
step 24200, change in cost 0.0226135
step 24210, training accuracy 0.944386
step 24210, cost 135.202
step 24210, change in cost 0.0226135
step 24220, training accuracy 0.944386
step 24220, cost 135.179
step 24220, change in cost 0.0225677
step 24230, training accuracy 0.944386
step 24230, cost 135.157
step 24230, change in cost 0.0226135
step 24240, training accuracy 0.944386
step 24240, cost 135.134
step 24240, change in cost 0.0225677
step 24250, training accuracy 0.944386
step 24250, cost 135.112
step 24250, change in cost 0.0225372
step 24260, training accuracy 0.944386
step 24260, cost 135.089
step 24260, change in cost 0.022583
step 24270, training accuracy 0.944386
step 24270, cost 135.067
step 24270, change in cost 0.022522
step 24280, training accuracy 0.945435
step 24280, cost 135.044
step 24280, change in cost 0.022522
step 24290, training accuracy 0.945435
step 24290, cost 135.021
step 24290, change in cost 0.0225525
step 24300, training accuracy 0.945435
step 24300, cost 134.999
step 24300, change in cost 0.0225372
step 24310, training accuracy 0.945435
step 24310, cost 134.976
step 24310, change in cost 0.022522
step 24320, training accuracy 0.945435
step 24320, cost 134.954
step 24320, change in cost 0.0224609
step 24330, training accuracy 0.945435
step 24330, cost 134.931
step 24330, change in cost 0.022522
step 24340, training accuracy 0.945435
step 24340, cost 134.909
step 24340, change in cost 0.0224609
step 24350, training accuracy 0.945435
step 24350, cost 134.886
step 24350, change in cost 0.022522
step 24360, training accuracy 0.945435
step 24360, cost 134.864
step 24360, change in cost 0.0224152
step 24370, training accuracy 0.945435
step 24370, cost 134.842
step 24370, change in cost 0.0224762
step 24380, training accuracy 0.945435
step 24380, cost 134.819
step 24380, change in cost 0.0224304
step 24390, training accuracy 0.946485
step 24390, cost 134.797
step 24390, change in cost 0.0224609
step 24400, training accuracy 0.946485
step 24400, cost 134.774
step 24400, change in cost 0.0224152
step 24410, training accuracy 0.946485
step 24410, cost 134.752
step 24410, change in cost 0.0224762
step 24420, training accuracy 0.946485
step 24420, cost 134.729
step 24420, change in cost 0.0224152
step 24430, training accuracy 0.946485
step 24430, cost 134.707
step 24430, change in cost 0.0223999
step 24440, training accuracy 0.946485
step 24440, cost 134.685
step 24440, change in cost 0.0223541
step 24450, training accuracy 0.946485
step 24450, cost 134.662
step 24450, change in cost 0.0224457
step 24460, training accuracy 0.946485
step 24460, cost 134.64
step 24460, change in cost 0.0223846
step 24470, training accuracy 0.946485
step 24470, cost 134.617
step 24470, change in cost 0.0223694
step 24480, training accuracy 0.946485
step 24480, cost 134.595
step 24480, change in cost 0.0223083
step 24490, training accuracy 0.946485
step 24490, cost 134.573
step 24490, change in cost 0.0223999
step 24500, training accuracy 0.946485
step 24500, cost 134.55
step 24500, change in cost 0.0223389
step 24510, training accuracy 0.946485
step 24510, cost 134.528
step 24510, change in cost 0.0223389
step 24520, training accuracy 0.946485
step 24520, cost 134.506
step 24520, change in cost 0.0223236
step 24530, training accuracy 0.946485
step 24530, cost 134.483
step 24530, change in cost 0.0223236
step 24540, training accuracy 0.946485
step 24540, cost 134.461
step 24540, change in cost 0.0222778
step 24550, training accuracy 0.946485
step 24550, cost 134.439
step 24550, change in cost 0.0223389
step 24560, training accuracy 0.946485
step 24560, cost 134.416
step 24560, change in cost 0.0222778
step 24570, training accuracy 0.946485
step 24570, cost 134.394
step 24570, change in cost 0.0222931
step 24580, training accuracy 0.946485
step 24580, cost 134.372
step 24580, change in cost 0.0222778
step 24590, training accuracy 0.946485
step 24590, cost 134.35
step 24590, change in cost 0.0222626
step 24600, training accuracy 0.946485
step 24600, cost 134.327
step 24600, change in cost 0.0222473
step 24610, training accuracy 0.946485
step 24610, cost 134.305
step 24610, change in cost 0.0222778
step 24620, training accuracy 0.946485
step 24620, cost 134.283
step 24620, change in cost 0.0222778
step 24630, training accuracy 0.946485
step 24630, cost 134.261
step 24630, change in cost 0.0222321
step 24640, training accuracy 0.946485
step 24640, cost 134.238
step 24640, change in cost 0.0222626
step 24650, training accuracy 0.946485
step 24650, cost 134.216
step 24650, change in cost 0.022171
step 24660, training accuracy 0.946485
step 24660, cost 134.194
step 24660, change in cost 0.0222168
step 24670, training accuracy 0.946485
step 24670, cost 134.172
step 24670, change in cost 0.022171
step 24680, training accuracy 0.946485
step 24680, cost 134.15
step 24680, change in cost 0.0222321
step 24690, training accuracy 0.946485
step 24690, cost 134.127
step 24690, change in cost 0.022171
step 24700, training accuracy 0.946485
step 24700, cost 134.105
step 24700, change in cost 0.0221558
step 24710, training accuracy 0.946485
step 24710, cost 134.083
step 24710, change in cost 0.0222321
step 24720, training accuracy 0.946485
step 24720, cost 134.061
step 24720, change in cost 0.022171
step 24730, training accuracy 0.946485
step 24730, cost 134.039
step 24730, change in cost 0.0221252
step 24740, training accuracy 0.946485
step 24740, cost 134.017
step 24740, change in cost 0.02211
step 24750, training accuracy 0.946485
step 24750, cost 133.994
step 24750, change in cost 0.0221405
step 24760, training accuracy 0.946485
step 24760, cost 133.972
step 24760, change in cost 0.0221558
step 24770, training accuracy 0.946485
step 24770, cost 133.95
step 24770, change in cost 0.02211
step 24780, training accuracy 0.946485
step 24780, cost 133.928
step 24780, change in cost 0.0221252
step 24790, training accuracy 0.946485
step 24790, cost 133.906
step 24790, change in cost 0.0220337
step 24800, training accuracy 0.946485
step 24800, cost 133.884
step 24800, change in cost 0.0221252
step 24810, training accuracy 0.946485
step 24810, cost 133.862
step 24810, change in cost 0.02211
step 24820, training accuracy 0.946485
step 24820, cost 133.84
step 24820, change in cost 0.02211
step 24830, training accuracy 0.946485
step 24830, cost 133.818
step 24830, change in cost 0.0220337
step 24840, training accuracy 0.946485
step 24840, cost 133.796
step 24840, change in cost 0.0220795
step 24850, training accuracy 0.946485
step 24850, cost 133.773
step 24850, change in cost 0.0220642
step 24860, training accuracy 0.946485
step 24860, cost 133.751
step 24860, change in cost 0.0220032
step 24870, training accuracy 0.946485
step 24870, cost 133.729
step 24870, change in cost 0.0220337
step 24880, training accuracy 0.946485
step 24880, cost 133.707
step 24880, change in cost 0.0220337
step 24890, training accuracy 0.946485
step 24890, cost 133.685
step 24890, change in cost 0.0220184
step 24900, training accuracy 0.946485
step 24900, cost 133.663
step 24900, change in cost 0.0220184
step 24910, training accuracy 0.946485
step 24910, cost 133.641
step 24910, change in cost 0.0219574
step 24920, training accuracy 0.946485
step 24920, cost 133.619
step 24920, change in cost 0.022049
step 24930, training accuracy 0.946485
step 24930, cost 133.597
step 24930, change in cost 0.0220032
step 24940, training accuracy 0.946485
step 24940, cost 133.575
step 24940, change in cost 0.0219421
step 24950, training accuracy 0.946485
step 24950, cost 133.553
step 24950, change in cost 0.0219727
step 24960, training accuracy 0.946485
step 24960, cost 133.531
step 24960, change in cost 0.0219574
step 24970, training accuracy 0.946485
step 24970, cost 133.51
step 24970, change in cost 0.0219421
step 24980, training accuracy 0.946485
step 24980, cost 133.488
step 24980, change in cost 0.0219879
step 24990, training accuracy 0.946485
step 24990, cost 133.466
step 24990, change in cost 0.0218811
step 25000, training accuracy 0.946485
step 25000, cost 133.444
step 25000, change in cost 0.0219421
step 25010, training accuracy 0.946485
step 25010, cost 133.422
step 25010, change in cost 0.0219421
step 25020, training accuracy 0.946485
step 25020, cost 133.4
step 25020, change in cost 0.0219116
step 25030, training accuracy 0.946485
step 25030, cost 133.378
step 25030, change in cost 0.0218811
step 25040, training accuracy 0.946485
step 25040, cost 133.356
step 25040, change in cost 0.0218964
step 25050, training accuracy 0.946485
step 25050, cost 133.334
step 25050, change in cost 0.0218811
step 25060, training accuracy 0.946485
step 25060, cost 133.312
step 25060, change in cost 0.0218964
step 25070, training accuracy 0.946485
step 25070, cost 133.29
step 25070, change in cost 0.0218506
step 25080, training accuracy 0.946485
step 25080, cost 133.269
step 25080, change in cost 0.0218811
step 25090, training accuracy 0.946485
step 25090, cost 133.247
step 25090, change in cost 0.0218353
step 25100, training accuracy 0.946485
step 25100, cost 133.225
step 25100, change in cost 0.0218506
step 25110, training accuracy 0.946485
step 25110, cost 133.203
step 25110, change in cost 0.0218506
step 25120, training accuracy 0.946485
step 25120, cost 133.181
step 25120, change in cost 0.0218353
step 25130, training accuracy 0.946485
step 25130, cost 133.159
step 25130, change in cost 0.0217743
step 25140, training accuracy 0.946485
step 25140, cost 133.138
step 25140, change in cost 0.0218506
step 25150, training accuracy 0.946485
step 25150, cost 133.116
step 25150, change in cost 0.0217743
step 25160, training accuracy 0.946485
step 25160, cost 133.094
step 25160, change in cost 0.0218048
step 25170, training accuracy 0.946485
step 25170, cost 133.072
step 25170, change in cost 0.0217896
step 25180, training accuracy 0.946485
step 25180, cost 133.05
step 25180, change in cost 0.0217896
step 25190, training accuracy 0.946485
step 25190, cost 133.029
step 25190, change in cost 0.0217896
step 25200, training accuracy 0.946485
step 25200, cost 133.007
step 25200, change in cost 0.0217438
step 25210, training accuracy 0.946485
step 25210, cost 132.985
step 25210, change in cost 0.021759
step 25220, training accuracy 0.946485
step 25220, cost 132.963
step 25220, change in cost 0.0217133
step 25230, training accuracy 0.946485
step 25230, cost 132.942
step 25230, change in cost 0.0217133
step 25240, training accuracy 0.946485
step 25240, cost 132.92
step 25240, change in cost 0.0217743
step 25250, training accuracy 0.946485
step 25250, cost 132.898
step 25250, change in cost 0.021698
step 25260, training accuracy 0.946485
step 25260, cost 132.877
step 25260, change in cost 0.0217438
step 25270, training accuracy 0.946485
step 25270, cost 132.855
step 25270, change in cost 0.0217133
step 25280, training accuracy 0.946485
step 25280, cost 132.833
step 25280, change in cost 0.021637
step 25290, training accuracy 0.946485
step 25290, cost 132.811
step 25290, change in cost 0.0217743
step 25300, training accuracy 0.946485
step 25300, cost 132.79
step 25300, change in cost 0.0216217
step 25310, training accuracy 0.946485
step 25310, cost 132.768
step 25310, change in cost 0.0216522
step 25320, training accuracy 0.946485
step 25320, cost 132.746
step 25320, change in cost 0.0217438
step 25330, training accuracy 0.946485
step 25330, cost 132.725
step 25330, change in cost 0.021637
step 25340, training accuracy 0.946485
step 25340, cost 132.703
step 25340, change in cost 0.021637
step 25350, training accuracy 0.946485
step 25350, cost 132.681
step 25350, change in cost 0.0216064
step 25360, training accuracy 0.946485
step 25360, cost 132.66
step 25360, change in cost 0.0216827
step 25370, training accuracy 0.946485
step 25370, cost 132.638
step 25370, change in cost 0.021637
step 25380, training accuracy 0.946485
step 25380, cost 132.617
step 25380, change in cost 0.0215912
step 25390, training accuracy 0.946485
step 25390, cost 132.595
step 25390, change in cost 0.0215912
step 25400, training accuracy 0.946485
step 25400, cost 132.573
step 25400, change in cost 0.0216217
step 25410, training accuracy 0.946485
step 25410, cost 132.552
step 25410, change in cost 0.0215454
step 25420, training accuracy 0.946485
step 25420, cost 132.53
step 25420, change in cost 0.0215912
step 25430, training accuracy 0.946485
step 25430, cost 132.509
step 25430, change in cost 0.0215759
step 25440, training accuracy 0.946485
step 25440, cost 132.487
step 25440, change in cost 0.0215454
step 25450, training accuracy 0.946485
step 25450, cost 132.466
step 25450, change in cost 0.0215454
step 25460, training accuracy 0.946485
step 25460, cost 132.444
step 25460, change in cost 0.0215607
step 25470, training accuracy 0.946485
step 25470, cost 132.422
step 25470, change in cost 0.0215759
step 25480, training accuracy 0.946485
step 25480, cost 132.401
step 25480, change in cost 0.0215454
step 25490, training accuracy 0.946485
step 25490, cost 132.379
step 25490, change in cost 0.0215149
step 25500, training accuracy 0.946485
step 25500, cost 132.358
step 25500, change in cost 0.0215149
step 25510, training accuracy 0.946485
step 25510, cost 132.336
step 25510, change in cost 0.0215454
step 25520, training accuracy 0.946485
step 25520, cost 132.315
step 25520, change in cost 0.0214844
step 25530, training accuracy 0.946485
step 25530, cost 132.293
step 25530, change in cost 0.0214996
step 25540, training accuracy 0.946485
step 25540, cost 132.272
step 25540, change in cost 0.0214996
step 25550, training accuracy 0.946485
step 25550, cost 132.25
step 25550, change in cost 0.0214844
step 25560, training accuracy 0.946485
step 25560, cost 132.229
step 25560, change in cost 0.0214539
step 25570, training accuracy 0.946485
step 25570, cost 132.207
step 25570, change in cost 0.0214233
step 25580, training accuracy 0.946485
step 25580, cost 132.186
step 25580, change in cost 0.0214844
step 25590, training accuracy 0.946485
step 25590, cost 132.165
step 25590, change in cost 0.0214233
step 25600, training accuracy 0.946485
step 25600, cost 132.143
step 25600, change in cost 0.0214539
step 25610, training accuracy 0.946485
step 25610, cost 132.122
step 25610, change in cost 0.0214539
step 25620, training accuracy 0.946485
step 25620, cost 132.1
step 25620, change in cost 0.0214081
step 25630, training accuracy 0.946485
step 25630, cost 132.079
step 25630, change in cost 0.0214233
step 25640, training accuracy 0.946485
step 25640, cost 132.057
step 25640, change in cost 0.0213776
step 25650, training accuracy 0.946485
step 25650, cost 132.036
step 25650, change in cost 0.0213928
step 25660, training accuracy 0.946485
step 25660, cost 132.015
step 25660, change in cost 0.0214233
step 25670, training accuracy 0.946485
step 25670, cost 131.993
step 25670, change in cost 0.0213623
step 25680, training accuracy 0.946485
step 25680, cost 131.972
step 25680, change in cost 0.0213623
step 25690, training accuracy 0.946485
step 25690, cost 131.951
step 25690, change in cost 0.0213776
step 25700, training accuracy 0.946485
step 25700, cost 131.929
step 25700, change in cost 0.0213776
step 25710, training accuracy 0.946485
step 25710, cost 131.908
step 25710, change in cost 0.0213623
step 25720, training accuracy 0.946485
step 25720, cost 131.886
step 25720, change in cost 0.021347
step 25730, training accuracy 0.946485
step 25730, cost 131.865
step 25730, change in cost 0.021286
step 25740, training accuracy 0.946485
step 25740, cost 131.844
step 25740, change in cost 0.0213318
step 25750, training accuracy 0.946485
step 25750, cost 131.822
step 25750, change in cost 0.0213318
step 25760, training accuracy 0.946485
step 25760, cost 131.801
step 25760, change in cost 0.021286
step 25770, training accuracy 0.947534
step 25770, cost 131.78
step 25770, change in cost 0.0213013
step 25780, training accuracy 0.947534
step 25780, cost 131.759
step 25780, change in cost 0.0213013
step 25790, training accuracy 0.947534
step 25790, cost 131.737
step 25790, change in cost 0.021286
step 25800, training accuracy 0.947534
step 25800, cost 131.716
step 25800, change in cost 0.0213013
step 25810, training accuracy 0.947534
step 25810, cost 131.695
step 25810, change in cost 0.021225
step 25820, training accuracy 0.947534
step 25820, cost 131.673
step 25820, change in cost 0.021286
step 25830, training accuracy 0.948583
step 25830, cost 131.652
step 25830, change in cost 0.0212555
step 25840, training accuracy 0.948583
step 25840, cost 131.631
step 25840, change in cost 0.0212402
step 25850, training accuracy 0.948583
step 25850, cost 131.61
step 25850, change in cost 0.0212708
step 25860, training accuracy 0.948583
step 25860, cost 131.589
step 25860, change in cost 0.0212097
step 25870, training accuracy 0.948583
step 25870, cost 131.567
step 25870, change in cost 0.0212555
step 25880, training accuracy 0.948583
step 25880, cost 131.546
step 25880, change in cost 0.0211945
step 25890, training accuracy 0.948583
step 25890, cost 131.525
step 25890, change in cost 0.021225
step 25900, training accuracy 0.948583
step 25900, cost 131.504
step 25900, change in cost 0.0212402
step 25910, training accuracy 0.948583
step 25910, cost 131.482
step 25910, change in cost 0.0211487
step 25920, training accuracy 0.948583
step 25920, cost 131.461
step 25920, change in cost 0.0211639
step 25930, training accuracy 0.948583
step 25930, cost 131.44
step 25930, change in cost 0.0211945
step 25940, training accuracy 0.948583
step 25940, cost 131.419
step 25940, change in cost 0.0211487
step 25950, training accuracy 0.948583
step 25950, cost 131.398
step 25950, change in cost 0.0211639
step 25960, training accuracy 0.948583
step 25960, cost 131.377
step 25960, change in cost 0.0211029
step 25970, training accuracy 0.948583
step 25970, cost 131.356
step 25970, change in cost 0.0211487
step 25980, training accuracy 0.948583
step 25980, cost 131.334
step 25980, change in cost 0.0211792
step 25990, training accuracy 0.948583
step 25990, cost 131.313
step 25990, change in cost 0.0211182
step 26000, training accuracy 0.948583
step 26000, cost 131.292
step 26000, change in cost 0.0210876
step 26010, training accuracy 0.948583
step 26010, cost 131.271
step 26010, change in cost 0.0211334
step 26020, training accuracy 0.948583
step 26020, cost 131.25
step 26020, change in cost 0.0211182
step 26030, training accuracy 0.948583
step 26030, cost 131.229
step 26030, change in cost 0.0211182
step 26040, training accuracy 0.948583
step 26040, cost 131.208
step 26040, change in cost 0.0210876
step 26050, training accuracy 0.948583
step 26050, cost 131.187
step 26050, change in cost 0.0210419
step 26060, training accuracy 0.948583
step 26060, cost 131.166
step 26060, change in cost 0.0210876
step 26070, training accuracy 0.948583
step 26070, cost 131.145
step 26070, change in cost 0.0210571
step 26080, training accuracy 0.948583
step 26080, cost 131.123
step 26080, change in cost 0.0210876
step 26090, training accuracy 0.948583
step 26090, cost 131.102
step 26090, change in cost 0.0210419
step 26100, training accuracy 0.948583
step 26100, cost 131.081
step 26100, change in cost 0.0210419
step 26110, training accuracy 0.948583
step 26110, cost 131.06
step 26110, change in cost 0.0210571
step 26120, training accuracy 0.948583
step 26120, cost 131.039
step 26120, change in cost 0.0210266
step 26130, training accuracy 0.948583
step 26130, cost 131.018
step 26130, change in cost 0.0210724
step 26140, training accuracy 0.948583
step 26140, cost 130.997
step 26140, change in cost 0.0209503
step 26150, training accuracy 0.948583
step 26150, cost 130.976
step 26150, change in cost 0.0210114
step 26160, training accuracy 0.948583
step 26160, cost 130.955
step 26160, change in cost 0.0209808
step 26170, training accuracy 0.948583
step 26170, cost 130.934
step 26170, change in cost 0.0209961
step 26180, training accuracy 0.948583
step 26180, cost 130.913
step 26180, change in cost 0.0209808
step 26190, training accuracy 0.948583
step 26190, cost 130.892
step 26190, change in cost 0.0209503
step 26200, training accuracy 0.948583
step 26200, cost 130.871
step 26200, change in cost 0.0209808
step 26210, training accuracy 0.948583
step 26210, cost 130.85
step 26210, change in cost 0.0209198
step 26220, training accuracy 0.948583
step 26220, cost 130.829
step 26220, change in cost 0.0209961
step 26230, training accuracy 0.948583
step 26230, cost 130.808
step 26230, change in cost 0.0209198
step 26240, training accuracy 0.948583
step 26240, cost 130.788
step 26240, change in cost 0.0209198
step 26250, training accuracy 0.948583
step 26250, cost 130.767
step 26250, change in cost 0.0209808
step 26260, training accuracy 0.948583
step 26260, cost 130.746
step 26260, change in cost 0.0209045
step 26270, training accuracy 0.948583
step 26270, cost 130.725
step 26270, change in cost 0.0209045
step 26280, training accuracy 0.948583
step 26280, cost 130.704
step 26280, change in cost 0.0209198
step 26290, training accuracy 0.948583
step 26290, cost 130.683
step 26290, change in cost 0.0208588
step 26300, training accuracy 0.948583
step 26300, cost 130.662
step 26300, change in cost 0.0209198
step 26310, training accuracy 0.948583
step 26310, cost 130.641
step 26310, change in cost 0.0208588
step 26320, training accuracy 0.948583
step 26320, cost 130.62
step 26320, change in cost 0.0208893
step 26330, training accuracy 0.948583
step 26330, cost 130.599
step 26330, change in cost 0.0208588
step 26340, training accuracy 0.948583
step 26340, cost 130.579
step 26340, change in cost 0.0208588
step 26350, training accuracy 0.948583
step 26350, cost 130.558
step 26350, change in cost 0.020874
step 26360, training accuracy 0.948583
step 26360, cost 130.537
step 26360, change in cost 0.020813
step 26370, training accuracy 0.948583
step 26370, cost 130.516
step 26370, change in cost 0.0208435
step 26380, training accuracy 0.948583
step 26380, cost 130.495
step 26380, change in cost 0.0208282
step 26390, training accuracy 0.948583
step 26390, cost 130.474
step 26390, change in cost 0.020813
step 26400, training accuracy 0.948583
step 26400, cost 130.454
step 26400, change in cost 0.0208282
step 26410, training accuracy 0.948583
step 26410, cost 130.433
step 26410, change in cost 0.0207977
step 26420, training accuracy 0.948583
step 26420, cost 130.412
step 26420, change in cost 0.0208282
step 26430, training accuracy 0.948583
step 26430, cost 130.391
step 26430, change in cost 0.0207672
step 26440, training accuracy 0.948583
step 26440, cost 130.37
step 26440, change in cost 0.0207672
step 26450, training accuracy 0.948583
step 26450, cost 130.35
step 26450, change in cost 0.0207825
step 26460, training accuracy 0.948583
step 26460, cost 130.329
step 26460, change in cost 0.0207672
step 26470, training accuracy 0.948583
step 26470, cost 130.308
step 26470, change in cost 0.0207214
step 26480, training accuracy 0.948583
step 26480, cost 130.287
step 26480, change in cost 0.0207977
step 26490, training accuracy 0.948583
step 26490, cost 130.267
step 26490, change in cost 0.0207062
step 26500, training accuracy 0.948583
step 26500, cost 130.246
step 26500, change in cost 0.0207367
step 26510, training accuracy 0.948583
step 26510, cost 130.225
step 26510, change in cost 0.0207062
step 26520, training accuracy 0.948583
step 26520, cost 130.204
step 26520, change in cost 0.0207672
step 26530, training accuracy 0.948583
step 26530, cost 130.184
step 26530, change in cost 0.0206909
step 26540, training accuracy 0.948583
step 26540, cost 130.163
step 26540, change in cost 0.0207062
step 26550, training accuracy 0.948583
step 26550, cost 130.142
step 26550, change in cost 0.0207062
step 26560, training accuracy 0.948583
step 26560, cost 130.122
step 26560, change in cost 0.0206757
step 26570, training accuracy 0.948583
step 26570, cost 130.101
step 26570, change in cost 0.0206909
step 26580, training accuracy 0.948583
step 26580, cost 130.08
step 26580, change in cost 0.0206757
step 26590, training accuracy 0.948583
step 26590, cost 130.06
step 26590, change in cost 0.0206451
step 26600, training accuracy 0.948583
step 26600, cost 130.039
step 26600, change in cost 0.0206146
step 26610, training accuracy 0.948583
step 26610, cost 130.018
step 26610, change in cost 0.0206909
step 26620, training accuracy 0.948583
step 26620, cost 129.998
step 26620, change in cost 0.0206757
step 26630, training accuracy 0.948583
step 26630, cost 129.977
step 26630, change in cost 0.0206146
step 26640, training accuracy 0.949633
step 26640, cost 129.956
step 26640, change in cost 0.0206451
step 26650, training accuracy 0.949633
step 26650, cost 129.936
step 26650, change in cost 0.0205688
step 26660, training accuracy 0.949633
step 26660, cost 129.915
step 26660, change in cost 0.0206451
step 26670, training accuracy 0.950682
step 26670, cost 129.895
step 26670, change in cost 0.0206146
step 26680, training accuracy 0.950682
step 26680, cost 129.874
step 26680, change in cost 0.0206146
step 26690, training accuracy 0.950682
step 26690, cost 129.853
step 26690, change in cost 0.0205536
step 26700, training accuracy 0.950682
step 26700, cost 129.833
step 26700, change in cost 0.0206604
step 26710, training accuracy 0.950682
step 26710, cost 129.812
step 26710, change in cost 0.0205383
step 26720, training accuracy 0.950682
step 26720, cost 129.792
step 26720, change in cost 0.0205383
step 26730, training accuracy 0.950682
step 26730, cost 129.771
step 26730, change in cost 0.0205841
step 26740, training accuracy 0.950682
step 26740, cost 129.751
step 26740, change in cost 0.0205688
step 26750, training accuracy 0.950682
step 26750, cost 129.73
step 26750, change in cost 0.0205536
step 26760, training accuracy 0.950682
step 26760, cost 129.71
step 26760, change in cost 0.020462
step 26770, training accuracy 0.950682
step 26770, cost 129.689
step 26770, change in cost 0.0205994
step 26780, training accuracy 0.950682
step 26780, cost 129.668
step 26780, change in cost 0.0205231
step 26790, training accuracy 0.950682
step 26790, cost 129.648
step 26790, change in cost 0.0204926
step 26800, training accuracy 0.950682
step 26800, cost 129.627
step 26800, change in cost 0.0205231
step 26810, training accuracy 0.950682
step 26810, cost 129.607
step 26810, change in cost 0.0204468
step 26820, training accuracy 0.950682
step 26820, cost 129.586
step 26820, change in cost 0.0205383
step 26830, training accuracy 0.950682
step 26830, cost 129.566
step 26830, change in cost 0.0204773
step 26840, training accuracy 0.950682
step 26840, cost 129.545
step 26840, change in cost 0.020462
step 26850, training accuracy 0.950682
step 26850, cost 129.525
step 26850, change in cost 0.020462
step 26860, training accuracy 0.950682
step 26860, cost 129.505
step 26860, change in cost 0.020462
step 26870, training accuracy 0.950682
step 26870, cost 129.484
step 26870, change in cost 0.0204315
step 26880, training accuracy 0.950682
step 26880, cost 129.464
step 26880, change in cost 0.0204773
step 26890, training accuracy 0.950682
step 26890, cost 129.443
step 26890, change in cost 0.0204315
step 26900, training accuracy 0.950682
step 26900, cost 129.423
step 26900, change in cost 0.020401
step 26910, training accuracy 0.950682
step 26910, cost 129.402
step 26910, change in cost 0.020462
step 26920, training accuracy 0.950682
step 26920, cost 129.382
step 26920, change in cost 0.0204468
step 26930, training accuracy 0.950682
step 26930, cost 129.361
step 26930, change in cost 0.0204315
step 26940, training accuracy 0.950682
step 26940, cost 129.341
step 26940, change in cost 0.0203857
step 26950, training accuracy 0.950682
step 26950, cost 129.321
step 26950, change in cost 0.0203857
step 26960, training accuracy 0.950682
step 26960, cost 129.3
step 26960, change in cost 0.020401
step 26970, training accuracy 0.950682
step 26970, cost 129.28
step 26970, change in cost 0.0203857
step 26980, training accuracy 0.950682
step 26980, cost 129.26
step 26980, change in cost 0.0203705
step 26990, training accuracy 0.950682
step 26990, cost 129.239
step 26990, change in cost 0.0203857
final accuracy on test set: 0.904762

저작자표시 비영리 변경금지

'Text Mining/Python' 관련 글

[텍스트마이닝] SVM 텍스트 분류

Date : 2017.12.29

[텍스트마이닝]한국어 Word2vec+tsne

Date : 2017.12.29

[형태소 분석] Konlpy 사용하기

Date : 2017.02.17

[nltk Books] Chapter 5. Categorizing and Tagging Words

Date : 2016.10.27

Admin

04-29 22:41

Contact Us

Address
경기도 수원시 영통구 원천동 산5번지 아주대학교 다산관 429호

E-mail
textminings@gmail.com

Phone
031-219-2910

[Classification] e-mail 분류하기

DATA import

학습 파라미터 세팅

TensorFlow Placeholders

TensorFlow variables

오퍼레이션 : 가중치곱 / 바이어스합/ 시그모이드

'Text Mining/Python' 관련 글

[텍스트마이닝] SVM 텍스트 분류

[텍스트마이닝]한국어 Word2vec+tsne

[형태소 분석] Konlpy 사용하기

[nltk Books] Chapter 5. Categorizing and Tagging Words

Category

Recent

Archives

Links

Admin

Contact Us

Tags

Calendar

Copyright © All Rights Reserved

Designed by CMSFactory.NET

티스토리툴바

« 2024/04 »
일	월	화	수	목	금	토
	1	2	3	4	5	6
7	8	9	10	11	12	13
14	15	16	17	18	19	20
21	22	23	24	25	26	27
28	29	30