In [1]:
import tensorflow as tf
import numpy as np
import gym
from keras.models import Sequential
from keras.layers import Dense
from keras.optimizers import Adam
from collections import deque


Using TensorFlow backend.

In [3]:
env = gym.make('CartPole-v0')


[2017-07-12 23:06:44,506] Making new env: CartPole-v0

In [4]:
state = env.reset()
for i in range(500):
    next_state, reward, done, _ = env.step(env.action_space.sample())
    print(reward)
    if done:
        print(i)
        break


1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
15

In [19]:
lr = 0.001
mem_size = 2000
episodes = 1000
steps = 2000

epsilon = 1
epsilon_min = 0.01
epsilon_decay = 0.995
gamma = 0.95
batch_size = 32

In [20]:
bot = Sequential()
bot.add(Dense(units=24, input_dim=4, activation='relu'))
bot.add(Dense(units=24, activation='relu'))
bot.add(Dense(units=2, activation='linear'))
bot.compile(loss='mse', optimizer=Adam(lr))

memory = deque(maxlen=mem_size)

In [21]:
reward_list = []
for episode in range(episodes):
    state = env.reset()
    total_rewards = 0
    for step in range(steps):
        if epsilon >= np.random.rand():
            action = env.action_space.sample()
        else:
            action = np.argmax(bot.predict(np.reshape(state, (1, 4))))
        next_state, reward, done, _ = env.step(action)
        # reward = reward if not done else -10
        memory.append((state, action, next_state, reward, done))
        total_rewards = total_rewards + reward
        state = next_state
        if done:
            reward_list.append(total_rewards)
            print('episode:{} steps:{} total_reward:{} epsilon:{}'.format(episode+1, step+1, total_rewards, epsilon))
            break
    
    minibatch = [memory[ii] for ii in np.random.choice(range(len(memory)), batch_size)]
    for state, action, next_state, reward, done in minibatch:
        target = reward
        if not done:
            target = reward + gamma * np.amax(bot.predict(np.reshape(next_state, (1, 4))))
        target_act = bot.predict(np.reshape(state, (1, 4)))
        target_act[0][action] = target
        bot.fit( np.reshape(state, (1, 4)), target_act, epochs=1, verbose=0)
    if epsilon > epsilon_min:
        epsilon = epsilon * epsilon_decay


episode:1 steps:20 total_reward:20.0 epsilon:1
episode:2 steps:11 total_reward:11.0 epsilon:0.995
episode:3 steps:89 total_reward:89.0 epsilon:0.990025
episode:4 steps:18 total_reward:18.0 epsilon:0.985074875
episode:5 steps:12 total_reward:12.0 epsilon:0.9801495006250001
episode:6 steps:26 total_reward:26.0 epsilon:0.9752487531218751
episode:7 steps:13 total_reward:13.0 epsilon:0.9703725093562657
episode:8 steps:24 total_reward:24.0 epsilon:0.9655206468094844
episode:9 steps:15 total_reward:15.0 epsilon:0.960693043575437
episode:10 steps:55 total_reward:55.0 epsilon:0.9558895783575597
episode:11 steps:13 total_reward:13.0 epsilon:0.9511101304657719
episode:12 steps:12 total_reward:12.0 epsilon:0.946354579813443
episode:13 steps:14 total_reward:14.0 epsilon:0.9416228069143757
episode:14 steps:18 total_reward:18.0 epsilon:0.9369146928798039
episode:15 steps:28 total_reward:28.0 epsilon:0.9322301194154049
episode:16 steps:13 total_reward:13.0 epsilon:0.9275689688183278
episode:17 steps:9 total_reward:9.0 epsilon:0.9229311239742362
episode:18 steps:14 total_reward:14.0 epsilon:0.918316468354365
episode:19 steps:19 total_reward:19.0 epsilon:0.9137248860125932
episode:20 steps:23 total_reward:23.0 epsilon:0.9091562615825302
episode:21 steps:17 total_reward:17.0 epsilon:0.9046104802746175
episode:22 steps:36 total_reward:36.0 epsilon:0.9000874278732445
episode:23 steps:19 total_reward:19.0 epsilon:0.8955869907338783
episode:24 steps:24 total_reward:24.0 epsilon:0.8911090557802088
episode:25 steps:17 total_reward:17.0 epsilon:0.8866535105013078
episode:26 steps:12 total_reward:12.0 epsilon:0.8822202429488013
episode:27 steps:19 total_reward:19.0 epsilon:0.8778091417340573
episode:28 steps:22 total_reward:22.0 epsilon:0.8734200960253871
episode:29 steps:22 total_reward:22.0 epsilon:0.8690529955452602
episode:30 steps:14 total_reward:14.0 epsilon:0.8647077305675338
episode:31 steps:26 total_reward:26.0 epsilon:0.8603841919146962
episode:32 steps:19 total_reward:19.0 epsilon:0.8560822709551227
episode:33 steps:38 total_reward:38.0 epsilon:0.851801859600347
episode:34 steps:19 total_reward:19.0 epsilon:0.8475428503023453
episode:35 steps:8 total_reward:8.0 epsilon:0.8433051360508336
episode:36 steps:26 total_reward:26.0 epsilon:0.8390886103705794
episode:37 steps:11 total_reward:11.0 epsilon:0.8348931673187264
episode:38 steps:17 total_reward:17.0 epsilon:0.8307187014821328
episode:39 steps:23 total_reward:23.0 epsilon:0.8265651079747222
episode:40 steps:15 total_reward:15.0 epsilon:0.8224322824348486
episode:41 steps:25 total_reward:25.0 epsilon:0.8183201210226743
episode:42 steps:14 total_reward:14.0 epsilon:0.8142285204175609
episode:43 steps:16 total_reward:16.0 epsilon:0.810157377815473
episode:44 steps:10 total_reward:10.0 epsilon:0.8061065909263957
episode:45 steps:23 total_reward:23.0 epsilon:0.8020760579717637
episode:46 steps:39 total_reward:39.0 epsilon:0.798065677681905
episode:47 steps:28 total_reward:28.0 epsilon:0.7940753492934954
episode:48 steps:16 total_reward:16.0 epsilon:0.7901049725470279
episode:49 steps:12 total_reward:12.0 epsilon:0.7861544476842928
episode:50 steps:15 total_reward:15.0 epsilon:0.7822236754458713
episode:51 steps:53 total_reward:53.0 epsilon:0.778312557068642
episode:52 steps:38 total_reward:38.0 epsilon:0.7744209942832988
episode:53 steps:20 total_reward:20.0 epsilon:0.7705488893118823
episode:54 steps:16 total_reward:16.0 epsilon:0.7666961448653229
episode:55 steps:11 total_reward:11.0 epsilon:0.7628626641409962
episode:56 steps:16 total_reward:16.0 epsilon:0.7590483508202912
episode:57 steps:19 total_reward:19.0 epsilon:0.7552531090661897
episode:58 steps:15 total_reward:15.0 epsilon:0.7514768435208588
episode:59 steps:20 total_reward:20.0 epsilon:0.7477194593032545
episode:60 steps:13 total_reward:13.0 epsilon:0.7439808620067382
episode:61 steps:22 total_reward:22.0 epsilon:0.7402609576967045
episode:62 steps:10 total_reward:10.0 epsilon:0.736559652908221
episode:63 steps:15 total_reward:15.0 epsilon:0.7328768546436799
episode:64 steps:15 total_reward:15.0 epsilon:0.7292124703704616
episode:65 steps:14 total_reward:14.0 epsilon:0.7255664080186093
episode:66 steps:11 total_reward:11.0 epsilon:0.7219385759785162
episode:67 steps:17 total_reward:17.0 epsilon:0.7183288830986236
episode:68 steps:14 total_reward:14.0 epsilon:0.7147372386831305
episode:69 steps:17 total_reward:17.0 epsilon:0.7111635524897149
episode:70 steps:12 total_reward:12.0 epsilon:0.7076077347272662
episode:71 steps:15 total_reward:15.0 epsilon:0.7040696960536299
episode:72 steps:18 total_reward:18.0 epsilon:0.7005493475733617
episode:73 steps:16 total_reward:16.0 epsilon:0.697046600835495
episode:74 steps:12 total_reward:12.0 epsilon:0.6935613678313175
episode:75 steps:11 total_reward:11.0 epsilon:0.6900935609921609
episode:76 steps:41 total_reward:41.0 epsilon:0.6866430931872001
episode:77 steps:14 total_reward:14.0 epsilon:0.6832098777212641
episode:78 steps:12 total_reward:12.0 epsilon:0.6797938283326578
episode:79 steps:11 total_reward:11.0 epsilon:0.6763948591909945
episode:80 steps:12 total_reward:12.0 epsilon:0.6730128848950395
episode:81 steps:16 total_reward:16.0 epsilon:0.6696478204705644
episode:82 steps:22 total_reward:22.0 epsilon:0.6662995813682115
episode:83 steps:33 total_reward:33.0 epsilon:0.6629680834613705
episode:84 steps:15 total_reward:15.0 epsilon:0.6596532430440636
episode:85 steps:17 total_reward:17.0 epsilon:0.6563549768288433
episode:86 steps:10 total_reward:10.0 epsilon:0.653073201944699
episode:87 steps:18 total_reward:18.0 epsilon:0.6498078359349755
episode:88 steps:18 total_reward:18.0 epsilon:0.6465587967553006
episode:89 steps:11 total_reward:11.0 epsilon:0.6433260027715241
episode:90 steps:17 total_reward:17.0 epsilon:0.6401093727576664
episode:91 steps:12 total_reward:12.0 epsilon:0.6369088258938781
episode:92 steps:12 total_reward:12.0 epsilon:0.6337242817644086
episode:93 steps:13 total_reward:13.0 epsilon:0.6305556603555866
episode:94 steps:23 total_reward:23.0 epsilon:0.6274028820538087
episode:95 steps:13 total_reward:13.0 epsilon:0.6242658676435396
episode:96 steps:25 total_reward:25.0 epsilon:0.6211445383053219
episode:97 steps:21 total_reward:21.0 epsilon:0.6180388156137953
episode:98 steps:16 total_reward:16.0 epsilon:0.6149486215357263
episode:99 steps:11 total_reward:11.0 epsilon:0.6118738784280476
episode:100 steps:17 total_reward:17.0 epsilon:0.6088145090359074
episode:101 steps:13 total_reward:13.0 epsilon:0.6057704364907278
episode:102 steps:14 total_reward:14.0 epsilon:0.6027415843082742
episode:103 steps:19 total_reward:19.0 epsilon:0.5997278763867329
episode:104 steps:14 total_reward:14.0 epsilon:0.5967292370047992
episode:105 steps:13 total_reward:13.0 epsilon:0.5937455908197752
episode:106 steps:12 total_reward:12.0 epsilon:0.5907768628656763
episode:107 steps:11 total_reward:11.0 epsilon:0.5878229785513479
episode:108 steps:11 total_reward:11.0 epsilon:0.5848838636585911
episode:109 steps:41 total_reward:41.0 epsilon:0.5819594443402982
episode:110 steps:14 total_reward:14.0 epsilon:0.5790496471185967
episode:111 steps:15 total_reward:15.0 epsilon:0.5761543988830038
episode:112 steps:11 total_reward:11.0 epsilon:0.5732736268885887
episode:113 steps:23 total_reward:23.0 epsilon:0.5704072587541458
episode:114 steps:70 total_reward:70.0 epsilon:0.567555222460375
episode:115 steps:38 total_reward:38.0 epsilon:0.5647174463480732
episode:116 steps:20 total_reward:20.0 epsilon:0.5618938591163328
episode:117 steps:18 total_reward:18.0 epsilon:0.5590843898207511
episode:118 steps:28 total_reward:28.0 epsilon:0.5562889678716474
episode:119 steps:12 total_reward:12.0 epsilon:0.5535075230322891
episode:120 steps:13 total_reward:13.0 epsilon:0.5507399854171277
episode:121 steps:12 total_reward:12.0 epsilon:0.547986285490042
episode:122 steps:16 total_reward:16.0 epsilon:0.5452463540625918
episode:123 steps:19 total_reward:19.0 epsilon:0.5425201222922789
episode:124 steps:12 total_reward:12.0 epsilon:0.5398075216808175
episode:125 steps:10 total_reward:10.0 epsilon:0.5371084840724134
episode:126 steps:13 total_reward:13.0 epsilon:0.5344229416520513
episode:127 steps:10 total_reward:10.0 epsilon:0.531750826943791
episode:128 steps:9 total_reward:9.0 epsilon:0.5290920728090721
episode:129 steps:21 total_reward:21.0 epsilon:0.5264466124450268
episode:130 steps:14 total_reward:14.0 epsilon:0.5238143793828016
episode:131 steps:13 total_reward:13.0 epsilon:0.5211953074858876
episode:132 steps:18 total_reward:18.0 epsilon:0.5185893309484582
episode:133 steps:22 total_reward:22.0 epsilon:0.5159963842937159
episode:134 steps:54 total_reward:54.0 epsilon:0.5134164023722473
episode:135 steps:42 total_reward:42.0 epsilon:0.510849320360386
episode:136 steps:61 total_reward:61.0 epsilon:0.5082950737585841
episode:137 steps:56 total_reward:56.0 epsilon:0.5057535983897912
episode:138 steps:51 total_reward:51.0 epsilon:0.5032248303978422
episode:139 steps:59 total_reward:59.0 epsilon:0.500708706245853
episode:140 steps:16 total_reward:16.0 epsilon:0.4982051627146237
episode:141 steps:18 total_reward:18.0 epsilon:0.49571413690105054
episode:142 steps:19 total_reward:19.0 epsilon:0.4932355662165453
episode:143 steps:22 total_reward:22.0 epsilon:0.4907693883854626
episode:144 steps:47 total_reward:47.0 epsilon:0.4883155414435353
episode:145 steps:73 total_reward:73.0 epsilon:0.4858739637363176
episode:146 steps:28 total_reward:28.0 epsilon:0.483444593917636
episode:147 steps:48 total_reward:48.0 epsilon:0.4810273709480478
episode:148 steps:74 total_reward:74.0 epsilon:0.47862223409330756
episode:149 steps:52 total_reward:52.0 epsilon:0.47622912292284103
episode:150 steps:20 total_reward:20.0 epsilon:0.4738479773082268
episode:151 steps:24 total_reward:24.0 epsilon:0.47147873742168567
episode:152 steps:46 total_reward:46.0 epsilon:0.46912134373457726
episode:153 steps:18 total_reward:18.0 epsilon:0.46677573701590436
episode:154 steps:21 total_reward:21.0 epsilon:0.46444185833082485
episode:155 steps:53 total_reward:53.0 epsilon:0.46211964903917074
episode:156 steps:15 total_reward:15.0 epsilon:0.4598090507939749
episode:157 steps:11 total_reward:11.0 epsilon:0.457510005540005
episode:158 steps:14 total_reward:14.0 epsilon:0.45522245551230495
episode:159 steps:12 total_reward:12.0 epsilon:0.4529463432347434
episode:160 steps:17 total_reward:17.0 epsilon:0.4506816115185697
episode:161 steps:15 total_reward:15.0 epsilon:0.4484282034609769
episode:162 steps:43 total_reward:43.0 epsilon:0.446186062443672
episode:163 steps:51 total_reward:51.0 epsilon:0.4439551321314536
episode:164 steps:75 total_reward:75.0 epsilon:0.4417353564707963
episode:165 steps:72 total_reward:72.0 epsilon:0.43952667968844233
episode:166 steps:81 total_reward:81.0 epsilon:0.43732904629000013
episode:167 steps:15 total_reward:15.0 epsilon:0.4351424010585501
episode:168 steps:27 total_reward:27.0 epsilon:0.43296668905325736
episode:169 steps:127 total_reward:127.0 epsilon:0.43080185560799106
episode:170 steps:76 total_reward:76.0 epsilon:0.4286478463299511
episode:171 steps:23 total_reward:23.0 epsilon:0.42650460709830135
episode:172 steps:66 total_reward:66.0 epsilon:0.42437208406280985
episode:173 steps:67 total_reward:67.0 epsilon:0.4222502236424958
episode:174 steps:10 total_reward:10.0 epsilon:0.42013897252428334
episode:175 steps:44 total_reward:44.0 epsilon:0.4180382776616619
episode:176 steps:12 total_reward:12.0 epsilon:0.4159480862733536
episode:177 steps:67 total_reward:67.0 epsilon:0.41386834584198684
episode:178 steps:177 total_reward:177.0 epsilon:0.4117990041127769
episode:179 steps:24 total_reward:24.0 epsilon:0.40974000909221303
episode:180 steps:76 total_reward:76.0 epsilon:0.40769130904675194
episode:181 steps:20 total_reward:20.0 epsilon:0.40565285250151817
episode:182 steps:19 total_reward:19.0 epsilon:0.4036245882390106
episode:183 steps:41 total_reward:41.0 epsilon:0.4016064652978155
episode:184 steps:20 total_reward:20.0 epsilon:0.3995984329713264
episode:185 steps:57 total_reward:57.0 epsilon:0.3976004408064698
episode:186 steps:20 total_reward:20.0 epsilon:0.39561243860243744
episode:187 steps:18 total_reward:18.0 epsilon:0.3936343764094253
episode:188 steps:21 total_reward:21.0 epsilon:0.39166620452737816
episode:189 steps:81 total_reward:81.0 epsilon:0.3897078735047413
episode:190 steps:83 total_reward:83.0 epsilon:0.3877593341372176
episode:191 steps:18 total_reward:18.0 epsilon:0.3858205374665315
episode:192 steps:12 total_reward:12.0 epsilon:0.38389143477919885
episode:193 steps:13 total_reward:13.0 epsilon:0.3819719776053028
episode:194 steps:28 total_reward:28.0 epsilon:0.3800621177172763
episode:195 steps:20 total_reward:20.0 epsilon:0.37816180712868996
episode:196 steps:19 total_reward:19.0 epsilon:0.37627099809304654
episode:197 steps:13 total_reward:13.0 epsilon:0.3743896431025813
episode:198 steps:15 total_reward:15.0 epsilon:0.37251769488706843
episode:199 steps:9 total_reward:9.0 epsilon:0.3706551064126331
episode:200 steps:14 total_reward:14.0 epsilon:0.36880183088056995
episode:201 steps:141 total_reward:141.0 epsilon:0.3669578217261671
episode:202 steps:25 total_reward:25.0 epsilon:0.36512303261753626
episode:203 steps:64 total_reward:64.0 epsilon:0.3632974174544486
episode:204 steps:54 total_reward:54.0 epsilon:0.3614809303671764
episode:205 steps:19 total_reward:19.0 epsilon:0.3596735257153405
episode:206 steps:22 total_reward:22.0 epsilon:0.3578751580867638
episode:207 steps:13 total_reward:13.0 epsilon:0.35608578229633
episode:208 steps:17 total_reward:17.0 epsilon:0.3543053533848483
episode:209 steps:26 total_reward:26.0 epsilon:0.35253382661792404
episode:210 steps:16 total_reward:16.0 epsilon:0.3507711574848344
episode:211 steps:28 total_reward:28.0 epsilon:0.34901730169741024
episode:212 steps:81 total_reward:81.0 epsilon:0.3472722151889232
episode:213 steps:28 total_reward:28.0 epsilon:0.3455358541129786
episode:214 steps:24 total_reward:24.0 epsilon:0.3438081748424137
episode:215 steps:27 total_reward:27.0 epsilon:0.3420891339682016
episode:216 steps:20 total_reward:20.0 epsilon:0.3403786882983606
episode:217 steps:41 total_reward:41.0 epsilon:0.3386767948568688
episode:218 steps:25 total_reward:25.0 epsilon:0.33698341088258443
episode:219 steps:13 total_reward:13.0 epsilon:0.3352984938281715
episode:220 steps:18 total_reward:18.0 epsilon:0.33362200135903064
episode:221 steps:16 total_reward:16.0 epsilon:0.33195389135223546
episode:222 steps:23 total_reward:23.0 epsilon:0.3302941218954743
episode:223 steps:39 total_reward:39.0 epsilon:0.32864265128599696
episode:224 steps:18 total_reward:18.0 epsilon:0.326999438029567
episode:225 steps:24 total_reward:24.0 epsilon:0.3253644408394192
episode:226 steps:43 total_reward:43.0 epsilon:0.3237376186352221
episode:227 steps:25 total_reward:25.0 epsilon:0.322118930542046
episode:228 steps:200 total_reward:200.0 epsilon:0.32050833588933575
episode:229 steps:40 total_reward:40.0 epsilon:0.31890579420988907
episode:230 steps:40 total_reward:40.0 epsilon:0.3173112652388396
episode:231 steps:58 total_reward:58.0 epsilon:0.3157247089126454
episode:232 steps:29 total_reward:29.0 epsilon:0.3141460853680822
episode:233 steps:29 total_reward:29.0 epsilon:0.3125753549412418
episode:234 steps:39 total_reward:39.0 epsilon:0.31101247816653554
episode:235 steps:44 total_reward:44.0 epsilon:0.30945741577570285
episode:236 steps:116 total_reward:116.0 epsilon:0.3079101286968243
episode:237 steps:161 total_reward:161.0 epsilon:0.3063705780533402
episode:238 steps:67 total_reward:67.0 epsilon:0.30483872516307353
episode:239 steps:47 total_reward:47.0 epsilon:0.3033145315372582
episode:240 steps:32 total_reward:32.0 epsilon:0.3017979588795719
episode:241 steps:99 total_reward:99.0 epsilon:0.30028896908517405
episode:242 steps:18 total_reward:18.0 epsilon:0.2987875242397482
episode:243 steps:28 total_reward:28.0 epsilon:0.29729358661854943
episode:244 steps:36 total_reward:36.0 epsilon:0.29580711868545667
episode:245 steps:48 total_reward:48.0 epsilon:0.2943280830920294
episode:246 steps:29 total_reward:29.0 epsilon:0.29285644267656924
episode:247 steps:62 total_reward:62.0 epsilon:0.2913921604631864
episode:248 steps:48 total_reward:48.0 epsilon:0.28993519966087045
episode:249 steps:25 total_reward:25.0 epsilon:0.2884855236625661
episode:250 steps:46 total_reward:46.0 epsilon:0.28704309604425327
episode:251 steps:59 total_reward:59.0 epsilon:0.285607880564032
episode:252 steps:30 total_reward:30.0 epsilon:0.28417984116121187
episode:253 steps:122 total_reward:122.0 epsilon:0.2827589419554058
episode:254 steps:31 total_reward:31.0 epsilon:0.28134514724562876
episode:255 steps:19 total_reward:19.0 epsilon:0.2799384215094006
episode:256 steps:24 total_reward:24.0 epsilon:0.27853872940185365
episode:257 steps:41 total_reward:41.0 epsilon:0.27714603575484437
episode:258 steps:42 total_reward:42.0 epsilon:0.2757603055760701
episode:259 steps:100 total_reward:100.0 epsilon:0.2743815040481898
episode:260 steps:59 total_reward:59.0 epsilon:0.2730095965279488
episode:261 steps:27 total_reward:27.0 epsilon:0.27164454854530906
episode:262 steps:21 total_reward:21.0 epsilon:0.2702863258025825
episode:263 steps:43 total_reward:43.0 epsilon:0.2689348941735696
episode:264 steps:42 total_reward:42.0 epsilon:0.26759021970270175
episode:265 steps:38 total_reward:38.0 epsilon:0.2662522686041882
episode:266 steps:81 total_reward:81.0 epsilon:0.2649210072611673
episode:267 steps:53 total_reward:53.0 epsilon:0.26359640222486147
episode:268 steps:58 total_reward:58.0 epsilon:0.26227842021373715
episode:269 steps:40 total_reward:40.0 epsilon:0.2609670281126685
episode:270 steps:51 total_reward:51.0 epsilon:0.25966219297210513
episode:271 steps:35 total_reward:35.0 epsilon:0.2583638820072446
episode:272 steps:32 total_reward:32.0 epsilon:0.2570720625972084
episode:273 steps:59 total_reward:59.0 epsilon:0.25578670228422234
episode:274 steps:35 total_reward:35.0 epsilon:0.25450776877280124
episode:275 steps:30 total_reward:30.0 epsilon:0.2532352299289372
episode:276 steps:48 total_reward:48.0 epsilon:0.2519690537792925
episode:277 steps:53 total_reward:53.0 epsilon:0.2507092085103961
episode:278 steps:22 total_reward:22.0 epsilon:0.2494556624678441
episode:279 steps:26 total_reward:26.0 epsilon:0.24820838415550486
episode:280 steps:38 total_reward:38.0 epsilon:0.24696734223472733
episode:281 steps:128 total_reward:128.0 epsilon:0.2457325055235537
episode:282 steps:23 total_reward:23.0 epsilon:0.24450384299593592
episode:283 steps:40 total_reward:40.0 epsilon:0.24328132378095624
episode:284 steps:24 total_reward:24.0 epsilon:0.24206491716205145
episode:285 steps:19 total_reward:19.0 epsilon:0.2408545925762412
episode:286 steps:24 total_reward:24.0 epsilon:0.23965031961336
episode:287 steps:30 total_reward:30.0 epsilon:0.2384520680152932
episode:288 steps:47 total_reward:47.0 epsilon:0.23725980767521673
episode:289 steps:67 total_reward:67.0 epsilon:0.23607350863684065
episode:290 steps:77 total_reward:77.0 epsilon:0.23489314109365644
episode:291 steps:12 total_reward:12.0 epsilon:0.23371867538818816
episode:292 steps:10 total_reward:10.0 epsilon:0.23255008201124722
episode:293 steps:16 total_reward:16.0 epsilon:0.231387331601191
episode:294 steps:19 total_reward:19.0 epsilon:0.23023039494318503
episode:295 steps:21 total_reward:21.0 epsilon:0.2290792429684691
episode:296 steps:24 total_reward:24.0 epsilon:0.22793384675362674
episode:297 steps:90 total_reward:90.0 epsilon:0.22679417751985861
episode:298 steps:41 total_reward:41.0 epsilon:0.22566020663225933
episode:299 steps:28 total_reward:28.0 epsilon:0.22453190559909803
episode:300 steps:158 total_reward:158.0 epsilon:0.22340924607110255
episode:301 steps:200 total_reward:200.0 epsilon:0.22229219984074702
episode:302 steps:71 total_reward:71.0 epsilon:0.2211807388415433
episode:303 steps:26 total_reward:26.0 epsilon:0.22007483514733558
episode:304 steps:110 total_reward:110.0 epsilon:0.2189744609715989
episode:305 steps:31 total_reward:31.0 epsilon:0.2178795886667409
episode:306 steps:160 total_reward:160.0 epsilon:0.2167901907234072
episode:307 steps:141 total_reward:141.0 epsilon:0.21570623976979014
episode:308 steps:127 total_reward:127.0 epsilon:0.21462770857094118
episode:309 steps:101 total_reward:101.0 epsilon:0.21355457002808648
episode:310 steps:111 total_reward:111.0 epsilon:0.21248679717794605
episode:311 steps:104 total_reward:104.0 epsilon:0.21142436319205632
episode:312 steps:55 total_reward:55.0 epsilon:0.21036724137609603
episode:313 steps:35 total_reward:35.0 epsilon:0.20931540516921554
episode:314 steps:142 total_reward:142.0 epsilon:0.20826882814336947
episode:315 steps:73 total_reward:73.0 epsilon:0.20722748400265262
episode:316 steps:49 total_reward:49.0 epsilon:0.20619134658263935
episode:317 steps:98 total_reward:98.0 epsilon:0.20516038984972615
episode:318 steps:126 total_reward:126.0 epsilon:0.2041345879004775
episode:319 steps:140 total_reward:140.0 epsilon:0.2031139149609751
episode:320 steps:110 total_reward:110.0 epsilon:0.20209834538617025
episode:321 steps:102 total_reward:102.0 epsilon:0.2010878536592394
episode:322 steps:45 total_reward:45.0 epsilon:0.2000824143909432
episode:323 steps:51 total_reward:51.0 epsilon:0.19908200231898848
episode:324 steps:53 total_reward:53.0 epsilon:0.19808659230739353
episode:325 steps:16 total_reward:16.0 epsilon:0.19709615934585656
episode:326 steps:21 total_reward:21.0 epsilon:0.19611067854912728
episode:327 steps:46 total_reward:46.0 epsilon:0.19513012515638165
episode:328 steps:36 total_reward:36.0 epsilon:0.19415447453059972
episode:329 steps:44 total_reward:44.0 epsilon:0.19318370215794672
episode:330 steps:23 total_reward:23.0 epsilon:0.192217783647157
episode:331 steps:31 total_reward:31.0 epsilon:0.1912566947289212
episode:332 steps:52 total_reward:52.0 epsilon:0.1903004112552766
episode:333 steps:21 total_reward:21.0 epsilon:0.18934890919900021
episode:334 steps:19 total_reward:19.0 epsilon:0.18840216465300522
episode:335 steps:36 total_reward:36.0 epsilon:0.18746015382974018
episode:336 steps:23 total_reward:23.0 epsilon:0.1865228530605915
episode:337 steps:26 total_reward:26.0 epsilon:0.18559023879528855
episode:338 steps:63 total_reward:63.0 epsilon:0.1846622876013121
episode:339 steps:23 total_reward:23.0 epsilon:0.18373897616330553
episode:340 steps:57 total_reward:57.0 epsilon:0.182820281282489
episode:341 steps:85 total_reward:85.0 epsilon:0.18190617987607657
episode:342 steps:33 total_reward:33.0 epsilon:0.18099664897669618
episode:343 steps:28 total_reward:28.0 epsilon:0.1800916657318127
episode:344 steps:41 total_reward:41.0 epsilon:0.17919120740315364
episode:345 steps:22 total_reward:22.0 epsilon:0.17829525136613786
episode:346 steps:55 total_reward:55.0 epsilon:0.17740377510930716
episode:347 steps:27 total_reward:27.0 epsilon:0.17651675623376062
episode:348 steps:65 total_reward:65.0 epsilon:0.1756341724525918
episode:349 steps:182 total_reward:182.0 epsilon:0.17475600159032884
episode:350 steps:76 total_reward:76.0 epsilon:0.17388222158237718
episode:351 steps:32 total_reward:32.0 epsilon:0.1730128104744653
episode:352 steps:66 total_reward:66.0 epsilon:0.17214774642209296
episode:353 steps:128 total_reward:128.0 epsilon:0.1712870076899825
episode:354 steps:37 total_reward:37.0 epsilon:0.17043057265153258
episode:355 steps:27 total_reward:27.0 epsilon:0.16957841978827493
episode:356 steps:41 total_reward:41.0 epsilon:0.16873052768933355
episode:357 steps:44 total_reward:44.0 epsilon:0.1678868750508869
episode:358 steps:60 total_reward:60.0 epsilon:0.16704744067563246
episode:359 steps:47 total_reward:47.0 epsilon:0.1662122034722543
episode:360 steps:111 total_reward:111.0 epsilon:0.16538114245489302
episode:361 steps:38 total_reward:38.0 epsilon:0.16455423674261854
episode:362 steps:82 total_reward:82.0 epsilon:0.16373146555890544
episode:363 steps:22 total_reward:22.0 epsilon:0.16291280823111093
episode:364 steps:37 total_reward:37.0 epsilon:0.16209824418995536
episode:365 steps:41 total_reward:41.0 epsilon:0.16128775296900558
episode:366 steps:115 total_reward:115.0 epsilon:0.16048131420416054
episode:367 steps:60 total_reward:60.0 epsilon:0.15967890763313974
episode:368 steps:73 total_reward:73.0 epsilon:0.15888051309497406
episode:369 steps:39 total_reward:39.0 epsilon:0.1580861105294992
episode:370 steps:18 total_reward:18.0 epsilon:0.1572956799768517
episode:371 steps:52 total_reward:52.0 epsilon:0.15650920157696743
episode:372 steps:107 total_reward:107.0 epsilon:0.1557266555690826
episode:373 steps:109 total_reward:109.0 epsilon:0.1549480222912372
episode:374 steps:109 total_reward:109.0 epsilon:0.15417328217978102
episode:375 steps:83 total_reward:83.0 epsilon:0.1534024157688821
episode:376 steps:72 total_reward:72.0 epsilon:0.1526354036900377
episode:377 steps:98 total_reward:98.0 epsilon:0.1518722266715875
episode:378 steps:47 total_reward:47.0 epsilon:0.15111286553822956
episode:379 steps:53 total_reward:53.0 epsilon:0.15035730121053842
episode:380 steps:60 total_reward:60.0 epsilon:0.14960551470448571
episode:381 steps:108 total_reward:108.0 epsilon:0.14885748713096328
episode:382 steps:172 total_reward:172.0 epsilon:0.14811319969530845
episode:383 steps:24 total_reward:24.0 epsilon:0.1473726336968319
episode:384 steps:25 total_reward:25.0 epsilon:0.14663577052834775
episode:385 steps:81 total_reward:81.0 epsilon:0.14590259167570602
episode:386 steps:99 total_reward:99.0 epsilon:0.1451730787173275
episode:387 steps:104 total_reward:104.0 epsilon:0.14444721332374086
episode:388 steps:200 total_reward:200.0 epsilon:0.14372497725712216
episode:389 steps:108 total_reward:108.0 epsilon:0.14300635237083656
episode:390 steps:52 total_reward:52.0 epsilon:0.14229132060898236
episode:391 steps:40 total_reward:40.0 epsilon:0.14157986400593744
episode:392 steps:74 total_reward:74.0 epsilon:0.14087196468590776
episode:393 steps:88 total_reward:88.0 epsilon:0.14016760486247823
episode:394 steps:125 total_reward:125.0 epsilon:0.13946676683816583
episode:395 steps:102 total_reward:102.0 epsilon:0.138769433003975
episode:396 steps:187 total_reward:187.0 epsilon:0.13807558583895513
episode:397 steps:172 total_reward:172.0 epsilon:0.13738520790976036
episode:398 steps:53 total_reward:53.0 epsilon:0.13669828187021155
episode:399 steps:55 total_reward:55.0 epsilon:0.13601479046086049
episode:400 steps:138 total_reward:138.0 epsilon:0.1353347165085562
episode:401 steps:112 total_reward:112.0 epsilon:0.1346580429260134
episode:402 steps:90 total_reward:90.0 epsilon:0.13398475271138335
episode:403 steps:135 total_reward:135.0 epsilon:0.13331482894782642
episode:404 steps:140 total_reward:140.0 epsilon:0.13264825480308728
episode:405 steps:48 total_reward:48.0 epsilon:0.13198501352907185
episode:406 steps:100 total_reward:100.0 epsilon:0.1313250884614265
episode:407 steps:189 total_reward:189.0 epsilon:0.13066846301911936
episode:408 steps:200 total_reward:200.0 epsilon:0.13001512070402377
episode:409 steps:51 total_reward:51.0 epsilon:0.12936504510050365
episode:410 steps:59 total_reward:59.0 epsilon:0.12871821987500112
episode:411 steps:79 total_reward:79.0 epsilon:0.12807462877562611
episode:412 steps:123 total_reward:123.0 epsilon:0.12743425563174798
episode:413 steps:123 total_reward:123.0 epsilon:0.12679708435358925
episode:414 steps:114 total_reward:114.0 epsilon:0.1261630989318213
episode:415 steps:25 total_reward:25.0 epsilon:0.1255322834371622
episode:416 steps:48 total_reward:48.0 epsilon:0.12490462201997637
episode:417 steps:43 total_reward:43.0 epsilon:0.1242800989098765
episode:418 steps:110 total_reward:110.0 epsilon:0.12365869841532712
episode:419 steps:138 total_reward:138.0 epsilon:0.12304040492325048
episode:420 steps:165 total_reward:165.0 epsilon:0.12242520289863423
episode:421 steps:130 total_reward:130.0 epsilon:0.12181307688414106
episode:422 steps:141 total_reward:141.0 epsilon:0.12120401149972035
episode:423 steps:147 total_reward:147.0 epsilon:0.12059799144222175
episode:424 steps:191 total_reward:191.0 epsilon:0.11999500148501063
episode:425 steps:200 total_reward:200.0 epsilon:0.11939502647758558
episode:426 steps:200 total_reward:200.0 epsilon:0.11879805134519765
episode:427 steps:181 total_reward:181.0 epsilon:0.11820406108847166
episode:428 steps:175 total_reward:175.0 epsilon:0.1176130407830293
episode:429 steps:130 total_reward:130.0 epsilon:0.11702497557911415
episode:430 steps:178 total_reward:178.0 epsilon:0.11643985070121858
episode:431 steps:197 total_reward:197.0 epsilon:0.11585765144771248
episode:432 steps:29 total_reward:29.0 epsilon:0.11527836319047392
episode:433 steps:107 total_reward:107.0 epsilon:0.11470197137452155
episode:434 steps:126 total_reward:126.0 epsilon:0.11412846151764894
episode:435 steps:111 total_reward:111.0 epsilon:0.1135578192100607
episode:436 steps:161 total_reward:161.0 epsilon:0.11299003011401039
episode:437 steps:174 total_reward:174.0 epsilon:0.11242507996344034
episode:438 steps:127 total_reward:127.0 epsilon:0.11186295456362313
episode:439 steps:200 total_reward:200.0 epsilon:0.11130363979080501
episode:440 steps:200 total_reward:200.0 epsilon:0.11074712159185099
episode:441 steps:146 total_reward:146.0 epsilon:0.11019338598389174
episode:442 steps:185 total_reward:185.0 epsilon:0.10964241905397228
episode:443 steps:31 total_reward:31.0 epsilon:0.10909420695870241
episode:444 steps:200 total_reward:200.0 epsilon:0.1085487359239089
episode:445 steps:200 total_reward:200.0 epsilon:0.10800599224428936
episode:446 steps:200 total_reward:200.0 epsilon:0.10746596228306791
episode:447 steps:200 total_reward:200.0 epsilon:0.10692863247165257
episode:448 steps:173 total_reward:173.0 epsilon:0.1063939893092943
episode:449 steps:200 total_reward:200.0 epsilon:0.10586201936274783
episode:450 steps:200 total_reward:200.0 epsilon:0.10533270926593409
episode:451 steps:200 total_reward:200.0 epsilon:0.10480604571960442
episode:452 steps:200 total_reward:200.0 epsilon:0.1042820154910064
episode:453 steps:200 total_reward:200.0 epsilon:0.10376060541355137
episode:454 steps:200 total_reward:200.0 epsilon:0.1032418023864836
episode:455 steps:200 total_reward:200.0 epsilon:0.10272559337455119
episode:456 steps:200 total_reward:200.0 epsilon:0.10221196540767843
episode:457 steps:200 total_reward:200.0 epsilon:0.10170090558064004
episode:458 steps:200 total_reward:200.0 epsilon:0.10119240105273684
episode:459 steps:200 total_reward:200.0 epsilon:0.10068643904747315
episode:460 steps:200 total_reward:200.0 epsilon:0.10018300685223579
episode:461 steps:200 total_reward:200.0 epsilon:0.0996820918179746
episode:462 steps:200 total_reward:200.0 epsilon:0.09918368135888474
episode:463 steps:200 total_reward:200.0 epsilon:0.09868776295209031
episode:464 steps:200 total_reward:200.0 epsilon:0.09819432413732986
episode:465 steps:129 total_reward:129.0 epsilon:0.09770335251664321
episode:466 steps:189 total_reward:189.0 epsilon:0.09721483575406
episode:467 steps:155 total_reward:155.0 epsilon:0.09672876157528969
episode:468 steps:33 total_reward:33.0 epsilon:0.09624511776741324
episode:469 steps:196 total_reward:196.0 epsilon:0.09576389217857617
episode:470 steps:33 total_reward:33.0 epsilon:0.09528507271768329
episode:471 steps:200 total_reward:200.0 epsilon:0.09480864735409487
episode:472 steps:158 total_reward:158.0 epsilon:0.0943346041173244
episode:473 steps:120 total_reward:120.0 epsilon:0.09386293109673778
episode:474 steps:38 total_reward:38.0 epsilon:0.09339361644125409
episode:475 steps:114 total_reward:114.0 epsilon:0.09292664835904782
episode:476 steps:98 total_reward:98.0 epsilon:0.09246201511725258
episode:477 steps:134 total_reward:134.0 epsilon:0.09199970504166631
episode:478 steps:200 total_reward:200.0 epsilon:0.09153970651645797
episode:479 steps:126 total_reward:126.0 epsilon:0.09108200798387568
episode:480 steps:200 total_reward:200.0 epsilon:0.0906265979439563
episode:481 steps:200 total_reward:200.0 epsilon:0.09017346495423652
episode:482 steps:141 total_reward:141.0 epsilon:0.08972259762946533
episode:483 steps:200 total_reward:200.0 epsilon:0.089273984641318
episode:484 steps:84 total_reward:84.0 epsilon:0.0888276147181114
episode:485 steps:107 total_reward:107.0 epsilon:0.08838347664452084
episode:486 steps:154 total_reward:154.0 epsilon:0.08794155926129824
episode:487 steps:141 total_reward:141.0 epsilon:0.08750185146499175
episode:488 steps:200 total_reward:200.0 epsilon:0.08706434220766679
episode:489 steps:185 total_reward:185.0 epsilon:0.08662902049662846
episode:490 steps:107 total_reward:107.0 epsilon:0.08619587539414532
episode:491 steps:190 total_reward:190.0 epsilon:0.08576489601717459
episode:492 steps:44 total_reward:44.0 epsilon:0.08533607153708872
episode:493 steps:112 total_reward:112.0 epsilon:0.08490939117940327
episode:494 steps:116 total_reward:116.0 epsilon:0.08448484422350626
episode:495 steps:48 total_reward:48.0 epsilon:0.08406242000238873
episode:496 steps:200 total_reward:200.0 epsilon:0.08364210790237678
episode:497 steps:111 total_reward:111.0 epsilon:0.0832238973628649
episode:498 steps:117 total_reward:117.0 epsilon:0.08280777787605056
episode:499 steps:119 total_reward:119.0 epsilon:0.08239373898667031
episode:500 steps:200 total_reward:200.0 epsilon:0.08198177029173696
episode:501 steps:200 total_reward:200.0 epsilon:0.08157186144027828
episode:502 steps:109 total_reward:109.0 epsilon:0.0811640021330769
episode:503 steps:68 total_reward:68.0 epsilon:0.08075818212241151
episode:504 steps:110 total_reward:110.0 epsilon:0.08035439121179945
episode:505 steps:118 total_reward:118.0 epsilon:0.07995261925574046
episode:506 steps:200 total_reward:200.0 epsilon:0.07955285615946175
episode:507 steps:128 total_reward:128.0 epsilon:0.07915509187866444
episode:508 steps:200 total_reward:200.0 epsilon:0.07875931641927113
episode:509 steps:153 total_reward:153.0 epsilon:0.07836551983717477
episode:510 steps:200 total_reward:200.0 epsilon:0.07797369223798889
episode:511 steps:200 total_reward:200.0 epsilon:0.07758382377679894
episode:512 steps:121 total_reward:121.0 epsilon:0.07719590465791494
episode:513 steps:200 total_reward:200.0 epsilon:0.07680992513462537
episode:514 steps:102 total_reward:102.0 epsilon:0.07642587550895225
episode:515 steps:164 total_reward:164.0 epsilon:0.07604374613140748
episode:516 steps:116 total_reward:116.0 epsilon:0.07566352740075044
episode:517 steps:200 total_reward:200.0 epsilon:0.07528520976374668
episode:518 steps:129 total_reward:129.0 epsilon:0.07490878371492794
episode:519 steps:134 total_reward:134.0 epsilon:0.0745342397963533
episode:520 steps:110 total_reward:110.0 epsilon:0.07416156859737154
episode:521 steps:171 total_reward:171.0 epsilon:0.07379076075438468
episode:522 steps:117 total_reward:117.0 epsilon:0.07342180695061275
episode:523 steps:200 total_reward:200.0 epsilon:0.07305469791585968
episode:524 steps:200 total_reward:200.0 epsilon:0.07268942442628039
episode:525 steps:200 total_reward:200.0 epsilon:0.07232597730414898
episode:526 steps:156 total_reward:156.0 epsilon:0.07196434741762824
episode:527 steps:200 total_reward:200.0 epsilon:0.0716045256805401
episode:528 steps:39 total_reward:39.0 epsilon:0.0712465030521374
episode:529 steps:95 total_reward:95.0 epsilon:0.0708902705368767
episode:530 steps:122 total_reward:122.0 epsilon:0.07053581918419231
episode:531 steps:39 total_reward:39.0 epsilon:0.07018314008827135
episode:532 steps:131 total_reward:131.0 epsilon:0.06983222438783
episode:533 steps:137 total_reward:137.0 epsilon:0.06948306326589085
episode:534 steps:200 total_reward:200.0 epsilon:0.0691356479495614
episode:535 steps:200 total_reward:200.0 epsilon:0.06878996970981359
episode:536 steps:200 total_reward:200.0 epsilon:0.06844601986126451
episode:537 steps:148 total_reward:148.0 epsilon:0.06810378976195819
episode:538 steps:200 total_reward:200.0 epsilon:0.0677632708131484
episode:539 steps:150 total_reward:150.0 epsilon:0.06742445445908266
episode:540 steps:180 total_reward:180.0 epsilon:0.06708733218678724
episode:541 steps:200 total_reward:200.0 epsilon:0.0667518955258533
episode:542 steps:143 total_reward:143.0 epsilon:0.06641813604822402
episode:543 steps:94 total_reward:94.0 epsilon:0.0660860453679829
episode:544 steps:108 total_reward:108.0 epsilon:0.06575561514114299
episode:545 steps:164 total_reward:164.0 epsilon:0.06542683706543727
episode:546 steps:111 total_reward:111.0 epsilon:0.06509970288011008
episode:547 steps:200 total_reward:200.0 epsilon:0.06477420436570952
episode:548 steps:191 total_reward:191.0 epsilon:0.06445033334388098
episode:549 steps:200 total_reward:200.0 epsilon:0.06412808167716157
episode:550 steps:200 total_reward:200.0 epsilon:0.06380744126877576
episode:551 steps:200 total_reward:200.0 epsilon:0.06348840406243188
episode:552 steps:143 total_reward:143.0 epsilon:0.06317096204211972
episode:553 steps:200 total_reward:200.0 epsilon:0.06285510723190912
episode:554 steps:200 total_reward:200.0 epsilon:0.06254083169574957
episode:555 steps:126 total_reward:126.0 epsilon:0.062228127537270826
episode:556 steps:142 total_reward:142.0 epsilon:0.06191698689958447
episode:557 steps:127 total_reward:127.0 epsilon:0.061607401965086545
episode:558 steps:200 total_reward:200.0 epsilon:0.06129936495526111
episode:559 steps:200 total_reward:200.0 epsilon:0.0609928681304848
episode:560 steps:179 total_reward:179.0 epsilon:0.060687903789832374
episode:561 steps:128 total_reward:128.0 epsilon:0.06038446427088321
episode:562 steps:172 total_reward:172.0 epsilon:0.06008254194952879
episode:563 steps:142 total_reward:142.0 epsilon:0.05978212923978115
episode:564 steps:200 total_reward:200.0 epsilon:0.05948321859358224
episode:565 steps:200 total_reward:200.0 epsilon:0.05918580250061433
episode:566 steps:135 total_reward:135.0 epsilon:0.058889873488111255
episode:567 steps:164 total_reward:164.0 epsilon:0.058595424120670696
episode:568 steps:50 total_reward:50.0 epsilon:0.05830244700006734
episode:569 steps:154 total_reward:154.0 epsilon:0.058010934765067
episode:570 steps:118 total_reward:118.0 epsilon:0.05772088009124167
episode:571 steps:150 total_reward:150.0 epsilon:0.05743227569078546
episode:572 steps:200 total_reward:200.0 epsilon:0.05714511431233153
episode:573 steps:200 total_reward:200.0 epsilon:0.05685938874076987
episode:574 steps:200 total_reward:200.0 epsilon:0.056575091797066025
episode:575 steps:200 total_reward:200.0 epsilon:0.056292216338080694
episode:576 steps:61 total_reward:61.0 epsilon:0.05601075525639029
episode:577 steps:189 total_reward:189.0 epsilon:0.05573070148010834
episode:578 steps:101 total_reward:101.0 epsilon:0.0554520479727078
episode:579 steps:200 total_reward:200.0 epsilon:0.05517478773284426
episode:580 steps:187 total_reward:187.0 epsilon:0.05489891379418004
episode:581 steps:200 total_reward:200.0 epsilon:0.05462441922520914
episode:582 steps:200 total_reward:200.0 epsilon:0.0543512971290831
episode:583 steps:200 total_reward:200.0 epsilon:0.05407954064343768
episode:584 steps:158 total_reward:158.0 epsilon:0.05380914294022049
episode:585 steps:200 total_reward:200.0 epsilon:0.05354009722551939
episode:586 steps:144 total_reward:144.0 epsilon:0.05327239673939179
episode:587 steps:126 total_reward:126.0 epsilon:0.053006034755694834
episode:588 steps:162 total_reward:162.0 epsilon:0.052741004581916356
episode:589 steps:168 total_reward:168.0 epsilon:0.052477299559006776
episode:590 steps:131 total_reward:131.0 epsilon:0.052214913061211746
episode:591 steps:188 total_reward:188.0 epsilon:0.05195383849590569
episode:592 steps:200 total_reward:200.0 epsilon:0.05169406930342616
episode:593 steps:156 total_reward:156.0 epsilon:0.05143559895690903
episode:594 steps:180 total_reward:180.0 epsilon:0.051178420962124486
episode:595 steps:195 total_reward:195.0 epsilon:0.05092252885731386
episode:596 steps:200 total_reward:200.0 epsilon:0.05066791621302729
episode:597 steps:200 total_reward:200.0 epsilon:0.05041457663196215
episode:598 steps:145 total_reward:145.0 epsilon:0.050162503748802344
episode:599 steps:200 total_reward:200.0 epsilon:0.049911691230058335
episode:600 steps:200 total_reward:200.0 epsilon:0.04966213277390804
episode:601 steps:172 total_reward:172.0 epsilon:0.0494138221100385
episode:602 steps:112 total_reward:112.0 epsilon:0.04916675299948831
episode:603 steps:144 total_reward:144.0 epsilon:0.04892091923449087
episode:604 steps:200 total_reward:200.0 epsilon:0.04867631463831842
episode:605 steps:200 total_reward:200.0 epsilon:0.048432933065126825
episode:606 steps:143 total_reward:143.0 epsilon:0.048190768399801194
episode:607 steps:46 total_reward:46.0 epsilon:0.04794981455780219
episode:608 steps:177 total_reward:177.0 epsilon:0.04771006548501318
episode:609 steps:163 total_reward:163.0 epsilon:0.047471515157588115
episode:610 steps:200 total_reward:200.0 epsilon:0.047234157581800176
episode:611 steps:200 total_reward:200.0 epsilon:0.046997986793891174
episode:612 steps:107 total_reward:107.0 epsilon:0.04676299685992172
episode:613 steps:60 total_reward:60.0 epsilon:0.04652918187562211
episode:614 steps:185 total_reward:185.0 epsilon:0.046296535966244
episode:615 steps:44 total_reward:44.0 epsilon:0.046065053286412784
episode:616 steps:37 total_reward:37.0 epsilon:0.04583472801998072
episode:617 steps:200 total_reward:200.0 epsilon:0.045605554379880814
episode:618 steps:41 total_reward:41.0 epsilon:0.04537752660798141
episode:619 steps:92 total_reward:92.0 epsilon:0.0451506389749415
episode:620 steps:170 total_reward:170.0 epsilon:0.044924885780066794
episode:621 steps:165 total_reward:165.0 epsilon:0.04470026135116646
episode:622 steps:200 total_reward:200.0 epsilon:0.04447676004441063
episode:623 steps:161 total_reward:161.0 epsilon:0.04425437624418858
episode:624 steps:174 total_reward:174.0 epsilon:0.04403310436296763
episode:625 steps:200 total_reward:200.0 epsilon:0.043812938841152796
episode:626 steps:145 total_reward:145.0 epsilon:0.04359387414694703
episode:627 steps:153 total_reward:153.0 epsilon:0.043375904776212296
episode:628 steps:200 total_reward:200.0 epsilon:0.043159025252331236
episode:629 steps:192 total_reward:192.0 epsilon:0.04294323012606958
episode:630 steps:200 total_reward:200.0 epsilon:0.04272851397543923
episode:631 steps:70 total_reward:70.0 epsilon:0.04251487140556204
episode:632 steps:139 total_reward:139.0 epsilon:0.04230229704853423
episode:633 steps:200 total_reward:200.0 epsilon:0.04209078556329156
episode:634 steps:164 total_reward:164.0 epsilon:0.0418803316354751
episode:635 steps:151 total_reward:151.0 epsilon:0.041670929977297724
episode:636 steps:142 total_reward:142.0 epsilon:0.04146257532741124
episode:637 steps:171 total_reward:171.0 epsilon:0.04125526245077418
episode:638 steps:196 total_reward:196.0 epsilon:0.04104898613852031
episode:639 steps:74 total_reward:74.0 epsilon:0.04084374120782771
episode:640 steps:131 total_reward:131.0 epsilon:0.04063952250178857
episode:641 steps:117 total_reward:117.0 epsilon:0.04043632488927963
episode:642 steps:146 total_reward:146.0 epsilon:0.04023414326483323
episode:643 steps:127 total_reward:127.0 epsilon:0.040032972548509065
episode:644 steps:200 total_reward:200.0 epsilon:0.03983280768576652
episode:645 steps:200 total_reward:200.0 epsilon:0.03963364364733769
episode:646 steps:200 total_reward:200.0 epsilon:0.039435475429100995
episode:647 steps:200 total_reward:200.0 epsilon:0.03923829805195549
episode:648 steps:200 total_reward:200.0 epsilon:0.03904210656169572
episode:649 steps:200 total_reward:200.0 epsilon:0.03884689602888724
episode:650 steps:200 total_reward:200.0 epsilon:0.0386526615487428
episode:651 steps:200 total_reward:200.0 epsilon:0.03845939824099909
episode:652 steps:200 total_reward:200.0 epsilon:0.03826710124979409
episode:653 steps:148 total_reward:148.0 epsilon:0.038075765743545126
episode:654 steps:138 total_reward:138.0 epsilon:0.0378853869148274
episode:655 steps:200 total_reward:200.0 epsilon:0.03769595998025326
episode:656 steps:118 total_reward:118.0 epsilon:0.03750748018035199
episode:657 steps:200 total_reward:200.0 epsilon:0.037319942779450235
episode:658 steps:200 total_reward:200.0 epsilon:0.037133343065552986
episode:659 steps:104 total_reward:104.0 epsilon:0.03694767635022522
episode:660 steps:200 total_reward:200.0 epsilon:0.036762937968474095
episode:661 steps:200 total_reward:200.0 epsilon:0.03657912327863173
episode:662 steps:198 total_reward:198.0 epsilon:0.036396227662238566
episode:663 steps:186 total_reward:186.0 epsilon:0.03621424652392737
episode:664 steps:182 total_reward:182.0 epsilon:0.036033175291307735
episode:665 steps:163 total_reward:163.0 epsilon:0.03585300941485119
episode:666 steps:200 total_reward:200.0 epsilon:0.035673744367776934
episode:667 steps:148 total_reward:148.0 epsilon:0.03549537564593805
episode:668 steps:200 total_reward:200.0 epsilon:0.035317898767708356
episode:669 steps:200 total_reward:200.0 epsilon:0.03514130927386981
episode:670 steps:200 total_reward:200.0 epsilon:0.03496560272750046
episode:671 steps:198 total_reward:198.0 epsilon:0.03479077471386296
episode:672 steps:181 total_reward:181.0 epsilon:0.03461682084029365
episode:673 steps:200 total_reward:200.0 epsilon:0.034443736736092176
episode:674 steps:50 total_reward:50.0 epsilon:0.034271518052411715
episode:675 steps:36 total_reward:36.0 epsilon:0.034100160462149656
episode:676 steps:147 total_reward:147.0 epsilon:0.03392965965983891
episode:677 steps:129 total_reward:129.0 epsilon:0.033760011361539714
episode:678 steps:122 total_reward:122.0 epsilon:0.03359121130473201
episode:679 steps:153 total_reward:153.0 epsilon:0.033423255248208356
episode:680 steps:126 total_reward:126.0 epsilon:0.03325613897196732
episode:681 steps:133 total_reward:133.0 epsilon:0.03308985827710748
episode:682 steps:23 total_reward:23.0 epsilon:0.032924408985721944
episode:683 steps:119 total_reward:119.0 epsilon:0.03275978694079333
episode:684 steps:200 total_reward:200.0 epsilon:0.032595988006089364
episode:685 steps:200 total_reward:200.0 epsilon:0.032433008066058915
episode:686 steps:173 total_reward:173.0 epsilon:0.03227084302572862
episode:687 steps:200 total_reward:200.0 epsilon:0.032109488810599975
episode:688 steps:195 total_reward:195.0 epsilon:0.031948941366546975
episode:689 steps:185 total_reward:185.0 epsilon:0.03178919665971424
episode:690 steps:139 total_reward:139.0 epsilon:0.03163025067641567
episode:691 steps:169 total_reward:169.0 epsilon:0.03147209942303359
episode:692 steps:160 total_reward:160.0 epsilon:0.03131473892591842
episode:693 steps:76 total_reward:76.0 epsilon:0.031158165231288826
episode:694 steps:90 total_reward:90.0 epsilon:0.03100237440513238
episode:695 steps:99 total_reward:99.0 epsilon:0.030847362533106718
episode:696 steps:136 total_reward:136.0 epsilon:0.030693125720441184
episode:697 steps:151 total_reward:151.0 epsilon:0.030539660091838977
episode:698 steps:149 total_reward:149.0 epsilon:0.03038696179137978
episode:699 steps:102 total_reward:102.0 epsilon:0.030235026982422884
episode:700 steps:200 total_reward:200.0 epsilon:0.030083851847510768
episode:701 steps:200 total_reward:200.0 epsilon:0.029933432588273214
episode:702 steps:107 total_reward:107.0 epsilon:0.029783765425331846
episode:703 steps:159 total_reward:159.0 epsilon:0.029634846598205186
episode:704 steps:197 total_reward:197.0 epsilon:0.02948667236521416
episode:705 steps:140 total_reward:140.0 epsilon:0.029339239003388088
episode:706 steps:200 total_reward:200.0 epsilon:0.029192542808371146
episode:707 steps:28 total_reward:28.0 epsilon:0.02904658009432929
episode:708 steps:154 total_reward:154.0 epsilon:0.028901347193857643
episode:709 steps:149 total_reward:149.0 epsilon:0.028756840457888354
episode:710 steps:166 total_reward:166.0 epsilon:0.02861305625559891
episode:711 steps:144 total_reward:144.0 epsilon:0.028469990974320916
episode:712 steps:153 total_reward:153.0 epsilon:0.02832764101944931
episode:713 steps:189 total_reward:189.0 epsilon:0.028186002814352063
episode:714 steps:200 total_reward:200.0 epsilon:0.0280450728002803
episode:715 steps:187 total_reward:187.0 epsilon:0.0279048474362789
episode:716 steps:200 total_reward:200.0 epsilon:0.027765323199097504
episode:717 steps:156 total_reward:156.0 epsilon:0.027626496583102015
episode:718 steps:200 total_reward:200.0 epsilon:0.027488364100186506
episode:719 steps:200 total_reward:200.0 epsilon:0.027350922279685573
episode:720 steps:200 total_reward:200.0 epsilon:0.027214167668287145
episode:721 steps:70 total_reward:70.0 epsilon:0.02707809682994571
episode:722 steps:174 total_reward:174.0 epsilon:0.02694270634579598
episode:723 steps:180 total_reward:180.0 epsilon:0.026807992814067
episode:724 steps:200 total_reward:200.0 epsilon:0.026673952849996664
episode:725 steps:21 total_reward:21.0 epsilon:0.02654058308574668
episode:726 steps:152 total_reward:152.0 epsilon:0.026407880170317945
episode:727 steps:200 total_reward:200.0 epsilon:0.026275840769466357
episode:728 steps:200 total_reward:200.0 epsilon:0.026144461565619025
episode:729 steps:200 total_reward:200.0 epsilon:0.02601373925779093
episode:730 steps:148 total_reward:148.0 epsilon:0.025883670561501974
episode:731 steps:200 total_reward:200.0 epsilon:0.025754252208694463
episode:732 steps:200 total_reward:200.0 epsilon:0.02562548094765099
episode:733 steps:200 total_reward:200.0 epsilon:0.025497353542912736
episode:734 steps:200 total_reward:200.0 epsilon:0.02536986677519817
episode:735 steps:200 total_reward:200.0 epsilon:0.02524301744132218
episode:736 steps:200 total_reward:200.0 epsilon:0.025116802354115567
episode:737 steps:200 total_reward:200.0 epsilon:0.024991218342344988
episode:738 steps:200 total_reward:200.0 epsilon:0.024866262250633264
episode:739 steps:48 total_reward:48.0 epsilon:0.024741930939380097
episode:740 steps:198 total_reward:198.0 epsilon:0.024618221284683196
episode:741 steps:200 total_reward:200.0 epsilon:0.02449513017825978
episode:742 steps:200 total_reward:200.0 epsilon:0.02437265452736848
episode:743 steps:200 total_reward:200.0 epsilon:0.024250791254731636
episode:744 steps:200 total_reward:200.0 epsilon:0.024129537298457977
episode:745 steps:22 total_reward:22.0 epsilon:0.024008889611965685
episode:746 steps:127 total_reward:127.0 epsilon:0.023888845163905856
episode:747 steps:200 total_reward:200.0 epsilon:0.023769400938086327
episode:748 steps:200 total_reward:200.0 epsilon:0.023650553933395897
episode:749 steps:200 total_reward:200.0 epsilon:0.023532301163728918
episode:750 steps:131 total_reward:131.0 epsilon:0.023414639657910272
episode:751 steps:139 total_reward:139.0 epsilon:0.023297566459620722
episode:752 steps:200 total_reward:200.0 epsilon:0.023181078627322618
episode:753 steps:200 total_reward:200.0 epsilon:0.023065173234186005
episode:754 steps:200 total_reward:200.0 epsilon:0.022949847368015076
episode:755 steps:200 total_reward:200.0 epsilon:0.022835098131175
episode:756 steps:200 total_reward:200.0 epsilon:0.022720922640519125
episode:757 steps:200 total_reward:200.0 epsilon:0.02260731802731653
episode:758 steps:200 total_reward:200.0 epsilon:0.022494281437179946
episode:759 steps:200 total_reward:200.0 epsilon:0.022381810029994047
episode:760 steps:200 total_reward:200.0 epsilon:0.022269900979844076
episode:761 steps:200 total_reward:200.0 epsilon:0.022158551474944856
episode:762 steps:200 total_reward:200.0 epsilon:0.022047758717570132
episode:763 steps:200 total_reward:200.0 epsilon:0.021937519923982282
episode:764 steps:200 total_reward:200.0 epsilon:0.021827832324362372
episode:765 steps:200 total_reward:200.0 epsilon:0.02171869316274056
episode:766 steps:200 total_reward:200.0 epsilon:0.021610099696926857
episode:767 steps:200 total_reward:200.0 epsilon:0.021502049198442223
episode:768 steps:200 total_reward:200.0 epsilon:0.021394538952450012
episode:769 steps:200 total_reward:200.0 epsilon:0.02128756625768776
episode:770 steps:200 total_reward:200.0 epsilon:0.021181128426399323
episode:771 steps:200 total_reward:200.0 epsilon:0.021075222784267326
episode:772 steps:200 total_reward:200.0 epsilon:0.020969846670345987
episode:773 steps:190 total_reward:190.0 epsilon:0.020864997436994256
episode:774 steps:200 total_reward:200.0 epsilon:0.020760672449809284
episode:775 steps:200 total_reward:200.0 epsilon:0.020656869087560238
episode:776 steps:200 total_reward:200.0 epsilon:0.020553584742122436
episode:777 steps:200 total_reward:200.0 epsilon:0.020450816818411825
episode:778 steps:200 total_reward:200.0 epsilon:0.020348562734319765
episode:779 steps:200 total_reward:200.0 epsilon:0.020246819920648168
episode:780 steps:200 total_reward:200.0 epsilon:0.020145585821044927
episode:781 steps:200 total_reward:200.0 epsilon:0.020044857891939702
episode:782 steps:200 total_reward:200.0 epsilon:0.019944633602480003
episode:783 steps:200 total_reward:200.0 epsilon:0.019844910434467605
episode:784 steps:200 total_reward:200.0 epsilon:0.019745685882295267
episode:785 steps:200 total_reward:200.0 epsilon:0.01964695745288379
episode:786 steps:200 total_reward:200.0 epsilon:0.01954872266561937
episode:787 steps:200 total_reward:200.0 epsilon:0.019450979052291272
episode:788 steps:200 total_reward:200.0 epsilon:0.019353724157029815
episode:789 steps:200 total_reward:200.0 epsilon:0.019256955536244666
episode:790 steps:200 total_reward:200.0 epsilon:0.019160670758563442
episode:791 steps:200 total_reward:200.0 epsilon:0.019064867404770626
episode:792 steps:200 total_reward:200.0 epsilon:0.018969543067746772
episode:793 steps:200 total_reward:200.0 epsilon:0.018874695352408037
episode:794 steps:200 total_reward:200.0 epsilon:0.018780321875645996
episode:795 steps:200 total_reward:200.0 epsilon:0.018686420266267767
episode:796 steps:200 total_reward:200.0 epsilon:0.018592988164936427
episode:797 steps:200 total_reward:200.0 epsilon:0.018500023224111744
episode:798 steps:200 total_reward:200.0 epsilon:0.018407523107991184
episode:799 steps:200 total_reward:200.0 epsilon:0.01831548549245123
episode:800 steps:200 total_reward:200.0 epsilon:0.018223908064988973
episode:801 steps:200 total_reward:200.0 epsilon:0.018132788524664028
episode:802 steps:200 total_reward:200.0 epsilon:0.018042124582040707
episode:803 steps:200 total_reward:200.0 epsilon:0.017951913959130504
episode:804 steps:200 total_reward:200.0 epsilon:0.01786215438933485
episode:805 steps:200 total_reward:200.0 epsilon:0.017772843617388175
episode:806 steps:105 total_reward:105.0 epsilon:0.017683979399301233
episode:807 steps:200 total_reward:200.0 epsilon:0.017595559502304726
episode:808 steps:200 total_reward:200.0 epsilon:0.0175075817047932
episode:809 steps:200 total_reward:200.0 epsilon:0.017420043796269234
episode:810 steps:200 total_reward:200.0 epsilon:0.017332943577287888
episode:811 steps:200 total_reward:200.0 epsilon:0.01724627885940145
episode:812 steps:200 total_reward:200.0 epsilon:0.017160047465104442
episode:813 steps:199 total_reward:199.0 epsilon:0.01707424722777892
episode:814 steps:182 total_reward:182.0 epsilon:0.016988875991640028
episode:815 steps:200 total_reward:200.0 epsilon:0.016903931611681827
episode:816 steps:200 total_reward:200.0 epsilon:0.01681941195362342
episode:817 steps:185 total_reward:185.0 epsilon:0.016735314893855303
episode:818 steps:179 total_reward:179.0 epsilon:0.016651638319386028
episode:819 steps:200 total_reward:200.0 epsilon:0.0165683801277891
episode:820 steps:200 total_reward:200.0 epsilon:0.016485538227150154
episode:821 steps:198 total_reward:198.0 epsilon:0.0164031105360144
episode:822 steps:179 total_reward:179.0 epsilon:0.01632109498333433
episode:823 steps:194 total_reward:194.0 epsilon:0.016239489508417658
episode:824 steps:200 total_reward:200.0 epsilon:0.01615829206087557
episode:825 steps:170 total_reward:170.0 epsilon:0.01607750060057119
episode:826 steps:170 total_reward:170.0 epsilon:0.015997113097568336
episode:827 steps:174 total_reward:174.0 epsilon:0.015917127532080494
episode:828 steps:179 total_reward:179.0 epsilon:0.01583754189442009
episode:829 steps:19 total_reward:19.0 epsilon:0.01575835418494799
episode:830 steps:38 total_reward:38.0 epsilon:0.01567956241402325
episode:831 steps:200 total_reward:200.0 epsilon:0.015601164601953134
episode:832 steps:31 total_reward:31.0 epsilon:0.015523158778943369
episode:833 steps:22 total_reward:22.0 epsilon:0.015445542985048652
episode:834 steps:12 total_reward:12.0 epsilon:0.015368315270123408
episode:835 steps:23 total_reward:23.0 epsilon:0.01529147369377279
episode:836 steps:18 total_reward:18.0 epsilon:0.015215016325303928
episode:837 steps:35 total_reward:35.0 epsilon:0.015138941243677408
episode:838 steps:30 total_reward:30.0 epsilon:0.01506324653745902
episode:839 steps:27 total_reward:27.0 epsilon:0.014987930304771725
episode:840 steps:22 total_reward:22.0 epsilon:0.014912990653247866
episode:841 steps:107 total_reward:107.0 epsilon:0.014838425699981627
episode:842 steps:194 total_reward:194.0 epsilon:0.01476423357148172
episode:843 steps:26 total_reward:26.0 epsilon:0.014690412403624311
episode:844 steps:143 total_reward:143.0 epsilon:0.01461696034160619
episode:845 steps:200 total_reward:200.0 epsilon:0.014543875539898159
episode:846 steps:200 total_reward:200.0 epsilon:0.014471156162198668
episode:847 steps:23 total_reward:23.0 epsilon:0.014398800381387675
episode:848 steps:27 total_reward:27.0 epsilon:0.014326806379480736
episode:849 steps:142 total_reward:142.0 epsilon:0.014255172347583332
episode:850 steps:200 total_reward:200.0 epsilon:0.014183896485845416
episode:851 steps:200 total_reward:200.0 epsilon:0.014112977003416188
episode:852 steps:200 total_reward:200.0 epsilon:0.014042412118399107
episode:853 steps:200 total_reward:200.0 epsilon:0.013972200057807112
episode:854 steps:200 total_reward:200.0 epsilon:0.013902339057518077
episode:855 steps:194 total_reward:194.0 epsilon:0.013832827362230486
episode:856 steps:200 total_reward:200.0 epsilon:0.013763663225419333
episode:857 steps:176 total_reward:176.0 epsilon:0.013694844909292236
episode:858 steps:200 total_reward:200.0 epsilon:0.013626370684745774
episode:859 steps:200 total_reward:200.0 epsilon:0.013558238831322046
episode:860 steps:200 total_reward:200.0 epsilon:0.013490447637165436
episode:861 steps:200 total_reward:200.0 epsilon:0.013422995398979608
episode:862 steps:200 total_reward:200.0 epsilon:0.01335588042198471
episode:863 steps:200 total_reward:200.0 epsilon:0.013289101019874787
episode:864 steps:200 total_reward:200.0 epsilon:0.013222655514775413
episode:865 steps:200 total_reward:200.0 epsilon:0.013156542237201536
episode:866 steps:200 total_reward:200.0 epsilon:0.013090759526015528
episode:867 steps:102 total_reward:102.0 epsilon:0.01302530572838545
episode:868 steps:200 total_reward:200.0 epsilon:0.012960179199743523
episode:869 steps:200 total_reward:200.0 epsilon:0.012895378303744804
episode:870 steps:200 total_reward:200.0 epsilon:0.01283090141222608
episode:871 steps:181 total_reward:181.0 epsilon:0.012766746905164949
episode:872 steps:187 total_reward:187.0 epsilon:0.012702913170639124
episode:873 steps:191 total_reward:191.0 epsilon:0.012639398604785928
episode:874 steps:190 total_reward:190.0 epsilon:0.012576201611761997
episode:875 steps:161 total_reward:161.0 epsilon:0.012513320603703188
episode:876 steps:181 total_reward:181.0 epsilon:0.012450754000684672
episode:877 steps:150 total_reward:150.0 epsilon:0.012388500230681249
episode:878 steps:33 total_reward:33.0 epsilon:0.012326557729527843
episode:879 steps:121 total_reward:121.0 epsilon:0.012264924940880204
episode:880 steps:159 total_reward:159.0 epsilon:0.012203600316175803
episode:881 steps:136 total_reward:136.0 epsilon:0.012142582314594924
episode:882 steps:178 total_reward:178.0 epsilon:0.01208186940302195
episode:883 steps:167 total_reward:167.0 epsilon:0.01202146005600684
episode:884 steps:138 total_reward:138.0 epsilon:0.011961352755726806
episode:885 steps:176 total_reward:176.0 epsilon:0.01190154599194817
episode:886 steps:159 total_reward:159.0 epsilon:0.01184203826198843
episode:887 steps:200 total_reward:200.0 epsilon:0.011782828070678488
episode:888 steps:128 total_reward:128.0 epsilon:0.011723913930325095
episode:889 steps:200 total_reward:200.0 epsilon:0.01166529436067347
episode:890 steps:142 total_reward:142.0 epsilon:0.011606967888870102
episode:891 steps:109 total_reward:109.0 epsilon:0.01154893304942575
episode:892 steps:113 total_reward:113.0 epsilon:0.011491188384178622
episode:893 steps:120 total_reward:120.0 epsilon:0.011433732442257729
episode:894 steps:200 total_reward:200.0 epsilon:0.01137656378004644
episode:895 steps:135 total_reward:135.0 epsilon:0.011319680961146208
episode:896 steps:33 total_reward:33.0 epsilon:0.011263082556340478
episode:897 steps:113 total_reward:113.0 epsilon:0.011206767143558775
episode:898 steps:110 total_reward:110.0 epsilon:0.011150733307840981
episode:899 steps:119 total_reward:119.0 epsilon:0.011094979641301777
episode:900 steps:122 total_reward:122.0 epsilon:0.011039504743095268
episode:901 steps:20 total_reward:20.0 epsilon:0.01098430721937979
episode:902 steps:29 total_reward:29.0 epsilon:0.010929385683282892
episode:903 steps:27 total_reward:27.0 epsilon:0.010874738754866477
episode:904 steps:32 total_reward:32.0 epsilon:0.010820365061092144
episode:905 steps:22 total_reward:22.0 epsilon:0.010766263235786683
episode:906 steps:134 total_reward:134.0 epsilon:0.01071243191960775
episode:907 steps:153 total_reward:153.0 epsilon:0.010658869760009713
episode:908 steps:136 total_reward:136.0 epsilon:0.010605575411209664
episode:909 steps:125 total_reward:125.0 epsilon:0.010552547534153616
episode:910 steps:128 total_reward:128.0 epsilon:0.010499784796482848
episode:911 steps:143 total_reward:143.0 epsilon:0.010447285872500434
episode:912 steps:27 total_reward:27.0 epsilon:0.01039504944313793
episode:913 steps:128 total_reward:128.0 epsilon:0.010343074195922241
episode:914 steps:124 total_reward:124.0 epsilon:0.01029135882494263
episode:915 steps:124 total_reward:124.0 epsilon:0.010239902030817916
episode:916 steps:14 total_reward:14.0 epsilon:0.010188702520663827
episode:917 steps:25 total_reward:25.0 epsilon:0.010137759008060509
episode:918 steps:186 total_reward:186.0 epsilon:0.010087070213020206
episode:919 steps:200 total_reward:200.0 epsilon:0.010036634861955105
episode:920 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:921 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:922 steps:53 total_reward:53.0 epsilon:0.00998645168764533
episode:923 steps:113 total_reward:113.0 epsilon:0.00998645168764533
episode:924 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:925 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:926 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:927 steps:105 total_reward:105.0 epsilon:0.00998645168764533
episode:928 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:929 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:930 steps:196 total_reward:196.0 epsilon:0.00998645168764533
episode:931 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:932 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:933 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:934 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:935 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:936 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:937 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:938 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:939 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:940 steps:44 total_reward:44.0 epsilon:0.00998645168764533
episode:941 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:942 steps:194 total_reward:194.0 epsilon:0.00998645168764533
episode:943 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:944 steps:181 total_reward:181.0 epsilon:0.00998645168764533
episode:945 steps:184 total_reward:184.0 epsilon:0.00998645168764533
episode:946 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:947 steps:164 total_reward:164.0 epsilon:0.00998645168764533
episode:948 steps:164 total_reward:164.0 epsilon:0.00998645168764533
episode:949 steps:188 total_reward:188.0 epsilon:0.00998645168764533
episode:950 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:951 steps:180 total_reward:180.0 epsilon:0.00998645168764533
episode:952 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:953 steps:183 total_reward:183.0 epsilon:0.00998645168764533
episode:954 steps:198 total_reward:198.0 epsilon:0.00998645168764533
episode:955 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:956 steps:191 total_reward:191.0 epsilon:0.00998645168764533
episode:957 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:958 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:959 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:960 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:961 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:962 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:963 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:964 steps:88 total_reward:88.0 epsilon:0.00998645168764533
episode:965 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:966 steps:186 total_reward:186.0 epsilon:0.00998645168764533
episode:967 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:968 steps:132 total_reward:132.0 epsilon:0.00998645168764533
episode:969 steps:159 total_reward:159.0 epsilon:0.00998645168764533
episode:970 steps:122 total_reward:122.0 epsilon:0.00998645168764533
episode:971 steps:139 total_reward:139.0 epsilon:0.00998645168764533
episode:972 steps:134 total_reward:134.0 epsilon:0.00998645168764533
episode:973 steps:154 total_reward:154.0 epsilon:0.00998645168764533
episode:974 steps:142 total_reward:142.0 epsilon:0.00998645168764533
episode:975 steps:20 total_reward:20.0 epsilon:0.00998645168764533
episode:976 steps:129 total_reward:129.0 epsilon:0.00998645168764533
episode:977 steps:117 total_reward:117.0 epsilon:0.00998645168764533
episode:978 steps:107 total_reward:107.0 epsilon:0.00998645168764533
episode:979 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:980 steps:167 total_reward:167.0 epsilon:0.00998645168764533
episode:981 steps:111 total_reward:111.0 epsilon:0.00998645168764533
episode:982 steps:146 total_reward:146.0 epsilon:0.00998645168764533
episode:983 steps:156 total_reward:156.0 epsilon:0.00998645168764533
episode:984 steps:150 total_reward:150.0 epsilon:0.00998645168764533
episode:985 steps:128 total_reward:128.0 epsilon:0.00998645168764533
episode:986 steps:123 total_reward:123.0 epsilon:0.00998645168764533
episode:987 steps:126 total_reward:126.0 epsilon:0.00998645168764533
episode:988 steps:114 total_reward:114.0 epsilon:0.00998645168764533
episode:989 steps:120 total_reward:120.0 epsilon:0.00998645168764533
episode:990 steps:148 total_reward:148.0 epsilon:0.00998645168764533
episode:991 steps:176 total_reward:176.0 epsilon:0.00998645168764533
episode:992 steps:128 total_reward:128.0 epsilon:0.00998645168764533
episode:993 steps:200 total_reward:200.0 epsilon:0.00998645168764533
episode:994 steps:167 total_reward:167.0 epsilon:0.00998645168764533
episode:995 steps:145 total_reward:145.0 epsilon:0.00998645168764533
episode:996 steps:141 total_reward:141.0 epsilon:0.00998645168764533
episode:997 steps:41 total_reward:41.0 epsilon:0.00998645168764533
episode:998 steps:122 total_reward:122.0 epsilon:0.00998645168764533
episode:999 steps:199 total_reward:199.0 epsilon:0.00998645168764533
episode:1000 steps:113 total_reward:113.0 epsilon:0.00998645168764533

In [22]:
import matplotlib.pyplot as plt
%matplotlib inline
def running_mean(x, N):
    cumsum = np.cumsum(np.insert(x, 0, 0)) 
    return (cumsum[N:] - cumsum[:-N]) / N 

smoothed_rews = running_mean(reward_list, 10)

plt.plot(reward_list, color='grey', alpha=0.3)
plt.plot(smoothed_rews)


Out[22]:
[<matplotlib.lines.Line2D at 0x239e49e27b8>]

In [28]:
state = env.reset()
total_reward = 0
for i in range(500):
    env.render()
    next_state, reward, done, _ = env.step(np.argmax(bot.predict(np.reshape(state, (1, 4)))))
    total_reward = total_reward + reward
    state = next_state
    if done:
        print(total_reward)
        env.close()
        break


200.0

In [ ]: