{
  "trx_id": "e717a4097afcd25b65fc680bc72208c03dad0fc8",
  "block": 106151231,
  "trx_in_block": 0,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2026-05-18T06:46:00",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "7939.637324 VESTS"
    }
  ]
}

{
  "trx_id": "a99d2a46b4f8de473838e4c445fe73362fb23ab6",
  "block": 106007441,
  "trx_in_block": 1,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2026-05-13T06:20:39",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "5227.426919 VESTS"
    }
  ]
}

{
  "trx_id": "8db02b4532c39ff0e3c96e0278503ed0ea1d44a0",
  "block": 105518702,
  "trx_in_block": 0,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2026-04-26T05:57:06",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "7952.153080 VESTS"
    }
  ]
}

{
  "trx_id": "ec41978a884516a7710af3473d7b1a04dd31cee6",
  "block": 102872834,
  "trx_in_block": 3,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2026-01-24T01:21:09",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "5268.973738 VESTS"
    }
  ]
}

{
  "trx_id": "302f47c440b2992c00cfcce87653acb557c20f2d",
  "block": 91319043,
  "trx_in_block": 5,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2024-12-17T20:30:48",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "5433.192935 VESTS"
    }
  ]
}

{
  "trx_id": "1a3ccc9f39e548ad3186979783b84f44025ece9d",
  "block": 79873175,
  "trx_in_block": 1,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2023-11-14T12:11:18",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "5602.326467 VESTS"
    }
  ]
}

{
  "trx_id": "f8e427896f7ab217bef24f0297b3d42abc00faee",
  "block": 78363558,
  "trx_in_block": 7,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2023-09-22T10:58:15",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "8539.235253 VESTS"
    }
  ]
}

{
  "trx_id": "2e829da1cc7309a6a043582014924a7267cd3c04",
  "block": 69121206,
  "trx_in_block": 5,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2022-11-03T18:21:21",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "8761.286691 VESTS"
    }
  ]
}

{
  "trx_id": "0036a9745379251c1a2c1414f9d12532a18f6aeb",
  "block": 60824398,
  "trx_in_block": 23,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2022-01-17T23:30:18",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "8981.394292 VESTS"
    }
  ]
}

{
  "trx_id": "6df352f11ce3821352b94ac546bafa1ab4cbf2b9",
  "block": 54614693,
  "trx_in_block": 0,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2021-06-14T06:39:48",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "9165.588580 VESTS"
    }
  ]
}

{
  "trx_id": "0b64d4c4f6c1601b4c4ef4206d36e101b1988c1c",
  "block": 49361940,
  "trx_in_block": 2,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2020-12-11T16:51:36",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "9353.010554 VESTS"
    }
  ]
}

{
  "trx_id": "df36572f909a53c50d05d8972548a62802d78139",
  "block": 49213455,
  "trx_in_block": 0,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2020-12-06T10:27:03",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "1912.543513 VESTS"
    }
  ]
}

{
  "trx_id": "d76a7b92ad5a84eec06c478c5010bbd888bfca7c",
  "block": 49197026,
  "trx_in_block": 2,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2020-12-05T20:29:24",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "9359.218408 VESTS"
    }
  ]
}

{
  "trx_id": "cb5f33c14ebaa772eb1e8abe84d566dd92bb70c5",
  "block": 48271827,
  "trx_in_block": 2,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2020-11-03T03:33:12",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "1920.017158 VESTS"
    }
  ]
}

{
  "trx_id": "5bb3888fbd06486816e30acef3636db89ef84623",
  "block": 43223792,
  "trx_in_block": 4,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2020-05-09T11:30:39",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "9562.023767 VESTS"
    }
  ]
}

{
  "trx_id": "53db273c5e1c5d2145489cf112b21351af56f90d",
  "block": 43200906,
  "trx_in_block": 18,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2020-05-08T15:58:24",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "1953.311140 VESTS"
    }
  ]
}

{
  "trx_id": "dd3ea8060093d000c276574ad3d0b228437dfdcb",
  "block": 35952069,
  "trx_in_block": 20,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2019-08-28T20:17:51",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "9706.645218 VESTS"
    }
  ]
}

{
  "trx_id": "db7ed8e2f731ecc01c807642abed0ec0e0044e05",
  "block": 31513820,
  "trx_in_block": 5,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2019-03-27T07:13:39",
  "op": [
    "comment",
    {
      "parent_author": "south-man",
      "parent_permlink": "image-of-airplane-car-bicycle-deep-learning-source",
      "author": "steemitboard",
      "permlink": "steemitboard-notify-south-man-20190327t071338000z",
      "title": "",
      "body": "Congratulations @south-man! You received a personal award!\n\n<table><tr><td>https://steemitimages.com/70x70/http://steemitboard.com/@south-man/birthday1.png</td><td>Happy Birthday! - You are on the Steem blockchain for 1 year!</td></tr></table>\n\n<sub>_You can view [your badges on your Steem Board](https://steemitboard.com/@south-man) and compare to others on the [Steem Ranking](http://steemitboard.com/ranking/index.php?name=south-man)_</sub>\n\n\n**Do not miss the last post from @steemitboard:**\n<table><tr><td><a href=\"https://steemit.com/steem/@steemitboard/3-years-on-steem-happy-birthday-the-distribution-of-commemorative-badges-has-begun\"><img src=\"https://steemitimages.com/64x128/http://u.cubeupload.com/arcange/BG6u6k.png\"></a></td><td><a href=\"https://steemit.com/steem/@steemitboard/3-years-on-steem-happy-birthday-the-distribution-of-commemorative-badges-has-begun\">3 years on Steem - The distribution of commemorative badges has begun!</a></td></tr><tr><td><a href=\"https://steemit.com/steem/@steemitboard/happy-birthday-the-steem-blockchain-is-running-for-3-years\"><img src=\"https://steemitimages.com/64x128/http://u.cubeupload.com/arcange/BG6u6k.png\"></a></td><td><a href=\"https://steemit.com/steem/@steemitboard/happy-birthday-the-steem-blockchain-is-running-for-3-years\">Happy Birthday! The Steem blockchain is running for 3 years.</a></td></tr></table>\n\n###### [Vote for @Steemitboard as a witness](https://v2.steemconnect.com/sign/account-witness-vote?witness=steemitboard&approve=1) to get one more award and increased upvotes!",
      "json_metadata": "{\"image\":[\"https://steemitboard.com/img/notify.png\"]}"
    }
  ]
}

{
  "trx_id": "585e556b59a46690af16e656930f23671d14c27b",
  "block": 25956752,
  "trx_in_block": 16,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-09-14T15:51:36",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "9905.410270 VESTS"
    }
  ]
}

{
  "trx_id": "3c4a14ea2ea6c73f87e4735b255c9011d322388b",
  "block": 25572351,
  "trx_in_block": 4,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-09-01T07:22:45",
  "op": [
    "vote",
    {
      "voter": "hyunhonoh",
      "author": "south-man",
      "permlink": "4invru",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "81588f0ef44a6353d9f7f0ea4317255059991848",
  "block": 25572349,
  "trx_in_block": 13,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-09-01T07:22:39",
  "op": [
    "vote",
    {
      "voter": "hyunhonoh",
      "author": "south-man",
      "permlink": "image-of-airplane-car-bicycle-deep-learning-source",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "fa9b70f8df36a9b72bc87d6d6f3eae7b8f157714",
  "block": 24785151,
  "trx_in_block": 55,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-08-04T23:06:24",
  "op": [
    "delegate_vesting_shares",
    {
      "delegator": "steem",
      "delegatee": "south-man",
      "vesting_shares": "30190.715309 VESTS"
    }
  ]
}

{
  "trx_id": "e0f2a4487fc80467243a8b8cea8cf9b01bea7dcd",
  "block": 24742500,
  "trx_in_block": 31,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-08-03T11:33:15",
  "op": [
    "transfer",
    {
      "from": "merlin7",
      "to": "south-man",
      "amount": "0.001 SBD",
      "memo": "Hi I am lady Merlin...You are awesome.I need your friendship,i am following you, kindly follow me .I can get you FREE UPVOTES JUST FOR FRIENDSHIP..Thank you"
    }
  ]
}

{
  "trx_id": "bbb534801677d3d278d9ac80117568419c6b564f",
  "block": 24359808,
  "trx_in_block": 10,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-07-21T04:04:42",
  "op": [
    "vote",
    {
      "voter": "hyunhonoh",
      "author": "south-man",
      "permlink": "5cmvsm",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "5fbd804dade42235a7c10c1b844fdce9dee6adf1",
  "block": 23653925,
  "trx_in_block": 10,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-26T07:22:36",
  "op": [
    "vote",
    {
      "voter": "eee7407",
      "author": "south-man",
      "permlink": "75bqdz",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "6ea59a43dcf7a5bf88a60bd4f82b3619ba9c021f",
  "block": 23345704,
  "trx_in_block": 0,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-15T14:27:12",
  "op": [
    "account_update",
    {
      "account": "south-man",
      "memo_key": "STM74HZ4PYpZrMBbQmXEfiKG4hRan1oe9iNx8jfBPVPaGwcYrMFA5",
      "json_metadata": "{\"profile\":{\"cover_image\":\"https://img.esteem.ws/o2olfpohru.jpg\",\"profile_image\":\"https://cdn.steemitimages.com/DQmS8RURRt5EPd8K3PaMjT15G4tgJy1Q688AKuAYmmJBRBU/KakaoTalk_20180615_231733775%20(2).jpg\",\"name\":\"southman\",\"location\":\"Seoul in korea\"}}"
    }
  ]
}

{
  "trx_id": "8f5dbef224f45968b5e342beeb3cf87985decb4e",
  "block": 23287575,
  "trx_in_block": 77,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-13T14:00:03",
  "op": [
    "custom_json",
    {
      "required_auths": [],
      "required_posting_auths": [
        "south-man"
      ],
      "id": "follow",
      "json": "[\"follow\",{\"follower\":\"south-man\",\"following\":\"youngogmarqs\",\"what\":[\"blog\"]}]"
    }
  ]
}

{
  "trx_id": "a6031d7ddb1e05ed0e4f20889ec8427f327b1d7e",
  "block": 23240880,
  "trx_in_block": 27,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-11T23:04:21",
  "op": [
    "vote",
    {
      "voter": "youngogmarqs",
      "author": "south-man",
      "permlink": "5cmvsm",
      "weight": 2
    }
  ]
}

{
  "trx_id": "65f49712ab0429ce123dd0d7cf97b4de11f04ca5",
  "block": 23240865,
  "trx_in_block": 44,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-11T23:03:33",
  "op": [
    "vote",
    {
      "voter": "youngogmarqs",
      "author": "south-man",
      "permlink": "4invru",
      "weight": 2
    }
  ]
}

{
  "trx_id": "e6d39c8eb3dd40a9fb8fa38fed7515dfaa5af82d",
  "block": 23240859,
  "trx_in_block": 8,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-11T23:03:12",
  "op": [
    "comment",
    {
      "parent_author": "",
      "parent_permlink": "python",
      "author": "south-man",
      "permlink": "5cmvsm",
      "title": "구글이미지 다운로드하는 파이썬 라이브러리",
      "body": "구글 이미지 다운로드가 필요해서 검색하는 중 찾은 라이브러리 입니다. 넘 편하고 좋네요. 속이 후련^^\nhttps://github.com/hardikvasa/google-images-download\n\n\n![](https://cdn.steemitimages.com/DQmPekZRMJbNb1SRUFc46UzheXSQCXPFT1TU6969j2hukcq/image.png)",
      "json_metadata": "{\"tags\":[\"python\",\"kr-dev\",\"google\",\"image\",\"download\"],\"links\":[\"https://github.com/hardikvasa/google-images-download\"],\"app\":\"steemkr/0.1\",\"format\":\"markdown\",\"image\":[\"https://cdn.steemitimages.com/DQmPekZRMJbNb1SRUFc46UzheXSQCXPFT1TU6969j2hukcq/image.png\"]}"
    }
  ]
}

{
  "trx_id": "25e3981b8b4883d51f8a4e5ec252bf47a8e562e3",
  "block": 23240846,
  "trx_in_block": 16,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-11T23:02:33",
  "op": [
    "comment",
    {
      "parent_author": "",
      "parent_permlink": "deep-learning",
      "author": "south-man",
      "permlink": "4invru",
      "title": "비행기, 자동차, 자전거의 이미지 딥러닝 소스",
      "body": "비행기, 자동차, 자전거의 google이미지를 받아서 딥러닝으로 학습한 후에 사진을 분류하는 소스입니다.\n사진을 많이 수집해서 하면 좀더 높은 정확도가 나올 것 같은데 사진을 추려내는 데 시간이 많이 걸려서 각 400개 정도의 사진만으로 학습을 시켰기 때문에 정확도가 그리 높게 나오진 않았어요^^\n\n깃 허브에도 동일한 소스가 있습니다.\nhttps://github.com/llejo3/deep-learning/blob/master/vehicle_classification_learning.md\n\n\n```python\nfrom google_images_download import google_images_download\n```\n\n\n```python\nresponse = google_images_download.googleimagesdownload() \n```\n\n구글 사진을 검색어로 가져오는데, 검색어와 맞지 않는 사진은 삭제하는 작업이 좀 필요합니다.\n```python\n#arguments = {\"keywords\":\"forsythia,cherry blossoms,magnolia,azalea,tulip\",\narguments = {\"keywords\":\"Planes, Bicycles, Cars\",\n             \"limit\":600,\n             \"print_urls\":True,\n             format: \"jpg,png\",\n             \"chromedriver\" : \"./chromedriver.exe\",\n             \"exact_size\":\"200,200\"}\npaths = response.download(arguments) \n```\n\n\n\n```python\nfrom keras.layers import Conv2D, Dense, Flatten, MaxPooling2D, Dropout\nfrom keras.layers.normalization import BatchNormalization\nfrom keras.models import Sequential\nfrom keras.preprocessing.image import ImageDataGenerator\nfrom keras.optimizers import Adam, RMSprop, SGD\nfrom keras.callbacks import EarlyStopping, TensorBoard, ModelCheckpoint\n \n```\n\n\n```python\ntrain_datagen = ImageDataGenerator(\n        rescale=1./255,\n        rotation_range=40,      # 40도까지 회전\n        width_shift_range=0.2,  # 20%까지 좌우 이동\n        height_shift_range=0.2, # 20%까지 상하 이동\n        shear_range=0.2,        # 20%까지 기울임\n        zoom_range=0.2,         # 20%까지 확대\n        horizontal_flip=True   # 좌우 뒤집기\n)\n```\n\n\n```python\ntrain = train_datagen.flow_from_directory(\n    'vehicles/train',\n    target_size=(200, 200),\n    #batch_size=32,\n    class_mode='categorical')\n\nvalid = ImageDataGenerator(rescale=1.0/255).flow_from_directory(\n    'vehicles/validation',\n    target_size=(200, 200),\n    #batch_size=32,\n    class_mode='categorical',\n    shuffle=False)\n```\n\n    Found 884 images belonging to 3 classes.\n    Found 298 images belonging to 3 classes.\n    \n\n\n```python\nmodel1 = Sequential()\nmodel1.add(Conv2D(32, (3, 3), activation='relu', input_shape=(200, 200, 3)))\nmodel1.add(MaxPooling2D((2, 2)))\nmodel1.add(Flatten())\nmodel1.add(Dense(3, activation='softmax'))\n```\n\n\n```python\nmodel1.summary()\n```\n\n    _________________________________________________________________\n    Layer (type)                 Output Shape              Param #   \n    =================================================================\n    conv2d_8 (Conv2D)            (None, 198, 198, 32)      896       \n    _________________________________________________________________\n    max_pooling2d_8 (MaxPooling2 (None, 99, 99, 32)        0         \n    _________________________________________________________________\n    flatten_7 (Flatten)          (None, 313632)            0         \n    _________________________________________________________________\n    dense_7 (Dense)              (None, 3)                 940899    \n    =================================================================\n    Total params: 941,795\n    Trainable params: 941,795\n    Non-trainable params: 0\n    _________________________________________________________________\n    \n\n\n```python\nsgd = SGD(lr=0.01, decay=1e-6, momentum=0.9, nesterov=True)\nmodel1.compile(loss='categorical_crossentropy', metrics=['accuracy'], optimizer=sgd)\n```\n\n\n```python\nhistory1 = model1.fit_generator(\n    train, validation_data=valid, epochs=30,\n    callbacks=[\n        EarlyStopping(monitor = \"val_loss\", patience=2),\n        TensorBoard(log_dir='tensorboard_logs/log_model1')\n    ])\n```\n\n    Epoch 1/30\n     1/28 [>.............................] - ETA: 46s - loss: 1.1692 - acc: 0.3438\n\n    C:\\Users\\south\\Anaconda3\\lib\\site-packages\\PIL\\Image.py:918: UserWarning: Palette images with Transparency   expressed in bytes should be converted to RGBA images\n      'to RGBA images')\n    \n\n    28/28 [==============================] - 36s 1s/step - loss: 3.7392 - acc: 0.3803 - val_loss: 1.0593 - val_acc: 0.4564\n    Epoch 2/30\n    28/28 [==============================] - 35s 1s/step - loss: 1.1686 - acc: 0.4723 - val_loss: 1.0099 - val_acc: 0.5839\n    Epoch 3/30\n    28/28 [==============================] - 36s 1s/step - loss: 0.9614 - acc: 0.5998 - val_loss: 0.9786 - val_acc: 0.5772\n    Epoch 4/30\n    28/28 [==============================] - 36s 1s/step - loss: 0.9075 - acc: 0.6209 - val_loss: 0.9295 - val_acc: 0.5805\n    Epoch 5/30\n    28/28 [==============================] - 36s 1s/step - loss: 0.8156 - acc: 0.6610 - val_loss: 0.8961 - val_acc: 0.6242\n    Epoch 6/30\n    28/28 [==============================] - 36s 1s/step - loss: 0.7874 - acc: 0.6738 - val_loss: 0.9063 - val_acc: 0.5973\n    Epoch 7/30\n    28/28 [==============================] - 37s 1s/step - loss: 0.7597 - acc: 0.6724 - val_loss: 0.8911 - val_acc: 0.6074\n    Epoch 8/30\n    28/28 [==============================] - 37s 1s/step - loss: 0.7685 - acc: 0.6638 - val_loss: 0.9621 - val_acc: 0.5772\n    Epoch 9/30\n    28/28 [==============================] - 34s 1s/step - loss: 0.7566 - acc: 0.6674 - val_loss: 0.8563 - val_acc: 0.6376\n    Epoch 10/30\n    28/28 [==============================] - 34s 1s/step - loss: 0.6990 - acc: 0.7087 - val_loss: 0.9243 - val_acc: 0.6074\n    Epoch 11/30\n    28/28 [==============================] - 35s 1s/step - loss: 0.7083 - acc: 0.7040 - val_loss: 0.9144 - val_acc: 0.6208\n    \n\n\n```python\nmodel2 = Sequential()\nmodel2.add(Conv2D(32, (3, 3), activation='relu', input_shape=(200, 200, 3)))\nmodel2.add(MaxPooling2D((2, 2)))\nmodel2.add(Conv2D(32, (3, 3), activation='relu'))\nmodel2.add(MaxPooling2D((2, 2)))\nmodel2.add(Flatten())\nmodel2.add(Dense(3, activation='softmax'))\n```\n\n\n```python\nmodel2.summary()\n```\n\n    _________________________________________________________________\n    Layer (type)                 Output Shape              Param #   \n    =================================================================\n    conv2d_11 (Conv2D)           (None, 198, 198, 32)      896       \n    _________________________________________________________________\n    max_pooling2d_11 (MaxPooling (None, 99, 99, 32)        0         \n    _________________________________________________________________\n    conv2d_12 (Conv2D)           (None, 97, 97, 32)        9248      \n    _________________________________________________________________\n    max_pooling2d_12 (MaxPooling (None, 48, 48, 32)        0         \n    _________________________________________________________________\n    flatten_9 (Flatten)          (None, 73728)             0         \n    _________________________________________________________________\n    dense_9 (Dense)              (None, 3)                 221187    \n    =================================================================\n    Total params: 231,331\n    Trainable params: 231,331\n    Non-trainable params: 0\n    _________________________________________________________________\n    \n\n\n```python\nsgd = SGD(lr=0.001, decay=1e-6, momentum=0.9, nesterov=True)\nmodel2.compile(loss='categorical_crossentropy', metrics=['accuracy'], optimizer=sgd)\n```\n\n\n```python\nhistory2 = model2.fit_generator(\n    train, validation_data=valid, epochs=100,\n    callbacks=[\n        EarlyStopping(monitor = \"val_loss\", patience=2),\n        TensorBoard(log_dir='tensorboard_logs/log_model2')\n    ])\n```\n\n    Epoch 1/30\n     4/28 [===>..........................] - ETA: 44s - loss: 1.4224 - acc: 0.3391\n\n    C:\\Users\\south\\Anaconda3\\lib\\site-packages\\PIL\\Image.py:918: UserWarning: Palette images with Transparency   expressed in bytes should be converted to RGBA images\n      'to RGBA images')\n    \n\n    28/28 [==============================] - 57s 2s/step - loss: 1.1169 - acc: 0.4312 - val_loss: 1.3070 - val_acc: 0.3658\n    Epoch 2/30\n    28/28 [==============================] - 53s 2s/step - loss: 1.0442 - acc: 0.4781 - val_loss: 1.0760 - val_acc: 0.4262\n    Epoch 3/30\n    28/28 [==============================] - 55s 2s/step - loss: 0.9853 - acc: 0.5179 - val_loss: 0.9727 - val_acc: 0.5503\n    Epoch 4/30\n    28/28 [==============================] - 55s 2s/step - loss: 0.9722 - acc: 0.5405 - val_loss: 1.0100 - val_acc: 0.6107\n    Epoch 5/30\n    28/28 [==============================] - 54s 2s/step - loss: 0.9162 - acc: 0.5844 - val_loss: 0.9376 - val_acc: 0.6040\n    Epoch 6/30\n    28/28 [==============================] - 59s 2s/step - loss: 0.7842 - acc: 0.6691 - val_loss: 1.0358 - val_acc: 0.6107\n    Epoch 7/30\n    28/28 [==============================] - 55s 2s/step - loss: 0.7229 - acc: 0.6908 - val_loss: 0.7254 - val_acc: 0.7148\n    Epoch 8/30\n    28/28 [==============================] - 55s 2s/step - loss: 0.7381 - acc: 0.6953 - val_loss: 0.8614 - val_acc: 0.6342\n    Epoch 9/30\n    28/28 [==============================] - 54s 2s/step - loss: 0.6778 - acc: 0.7174 - val_loss: 0.7262 - val_acc: 0.6980\n    \n\n\n```python\nmodel3 = Sequential()\nmodel3.add(Conv2D(32, (3, 3), activation='relu', input_shape=(200, 200, 3)))\nmodel3.add(BatchNormalization())\nmodel3.add(Conv2D(32, (3, 3), activation='relu'))\nmodel3.add(MaxPooling2D(pool_size=(2, 2)))\nmodel3.add(Dropout(0.25))\n\nmodel3.add(Conv2D(64, (3, 3), activation='relu'))\nmodel3.add(BatchNormalization())\nmodel3.add(Conv2D(64, (3, 3), activation='relu'))\nmodel3.add(MaxPooling2D(pool_size=(2, 2)))\nmodel3.add(Dropout(0.25))\n\nmodel3.add(Flatten())\nmodel3.add(Dense(256, activation='relu'))\nmodel3.add(Dropout(0.5))\nmodel3.add(Dense(3, activation='softmax'))\n```\n\n\n```python\nmodel3.summary()\n```\n\n    _________________________________________________________________\n    Layer (type)                 Output Shape              Param #   \n    =================================================================\n    conv2d_19 (Conv2D)           (None, 198, 198, 32)      896       \n    _________________________________________________________________\n    max_pooling2d_16 (MaxPooling (None, 99, 99, 32)        0         \n    _________________________________________________________________\n    dropout_4 (Dropout)          (None, 99, 99, 32)        0         \n    _________________________________________________________________\n    conv2d_20 (Conv2D)           (None, 97, 97, 64)        18496     \n    _________________________________________________________________\n    max_pooling2d_17 (MaxPooling (None, 48, 48, 64)        0         \n    _________________________________________________________________\n    dropout_5 (Dropout)          (None, 48, 48, 64)        0         \n    _________________________________________________________________\n    flatten_11 (Flatten)         (None, 147456)            0         \n    _________________________________________________________________\n    dense_12 (Dense)             (None, 256)               37748992  \n    _________________________________________________________________\n    dropout_6 (Dropout)          (None, 256)               0         \n    _________________________________________________________________\n    dense_13 (Dense)             (None, 3)                 771       \n    =================================================================\n    Total params: 37,769,155\n    Trainable params: 37,769,155\n    Non-trainable params: 0\n    _________________________________________________________________\n    \n\n\n```python\nsgd = SGD(lr=0.01, decay=1e-6, momentum=0.9, nesterov=True)\nmodel3.compile(loss='categorical_crossentropy', metrics=['accuracy'], optimizer=sgd)\n```\n\n\n```python\nhistory3 = model3.fit_generator(\n    train, validation_data=valid, epochs=30,\n    callbacks=[\n        EarlyStopping(monitor = \"val_loss\", patience=2),\n        ModelCheckpoint('model3-{epoch:02d}.hdf5', save_best_only=True),\n        TensorBoard(log_dir='tensorboard_logs/log_model3')\n    ])\n```\n\n    Epoch 1/30\n     1/28 [>.............................] - ETA: 1:34 - loss: 1.1490 - acc: 0.2500\n\n    C:\\Users\\south\\Anaconda3\\lib\\site-packages\\PIL\\Image.py:918: UserWarning: Palette images with Transparency   expressed in bytes should be converted to RGBA images\n      'to RGBA images')\n    \n\n    28/28 [==============================] - 85s 3s/step - loss: 1.2178 - acc: 0.3502 - val_loss: 1.0796 - val_acc: 0.3893\n    Epoch 2/30\n    28/28 [==============================] - 86s 3s/step - loss: 1.0463 - acc: 0.4851 - val_loss: 1.0564 - val_acc: 0.4899\n    Epoch 3/30\n    28/28 [==============================] - 82s 3s/step - loss: 0.9830 - acc: 0.5198 - val_loss: 1.0233 - val_acc: 0.5839\n    Epoch 4/30\n    28/28 [==============================] - 86s 3s/step - loss: 0.8876 - acc: 0.6224 - val_loss: 0.9437 - val_acc: 0.5336\n    Epoch 5/30\n    28/28 [==============================] - 84s 3s/step - loss: 0.7544 - acc: 0.6600 - val_loss: 0.8263 - val_acc: 0.6577\n    Epoch 6/30\n    28/28 [==============================] - 86s 3s/step - loss: 0.7302 - acc: 0.7052 - val_loss: 0.7034 - val_acc: 0.6879\n    Epoch 7/30\n    28/28 [==============================] - 89s 3s/step - loss: 0.6840 - acc: 0.7004 - val_loss: 0.6810 - val_acc: 0.7047\n    Epoch 8/30\n    28/28 [==============================] - 89s 3s/step - loss: 0.6499 - acc: 0.7214 - val_loss: 0.6697 - val_acc: 0.7248\n    Epoch 9/30\n    28/28 [==============================] - 87s 3s/step - loss: 0.6163 - acc: 0.7403 - val_loss: 0.6561 - val_acc: 0.7383\n    Epoch 10/30\n    28/28 [==============================] - 107s 4s/step - loss: 0.6387 - acc: 0.7238 - val_loss: 0.6992 - val_acc: 0.6846\n    Epoch 11/30\n    28/28 [==============================] - 90s 3s/step - loss: 0.5976 - acc: 0.7457 - val_loss: 0.6568 - val_acc: 0.7114\n\n\n![deepLearning.PNG](https://cdn.steemitimages.com/DQmaqW4SM7dpaZaKFAxqUDM7oMCYFN4qbH3TU3cLmzXzNKs/deepLearning.PNG)",
      "json_metadata": "{\"tags\":[\"kr-dev\",\"deep-learning\",\"keras\",\"vehicle\",\"python\"],\"image\":[\"https://cdn.steemitimages.com/DQmaqW4SM7dpaZaKFAxqUDM7oMCYFN4qbH3TU3cLmzXzNKs/deepLearning.PNG\"],\"links\":[\"https://github.com/llejo3/deep-learning/blob/master/vehicle_classification_learning.md\"],\"app\":\"steemkr/0.1\",\"format\":\"markdown\"}"
    }
  ]
}

{
  "trx_id": "2a3e4f58f9428cbb8f0ab8d3e2b0f2182a29e58b",
  "block": 23129914,
  "trx_in_block": 3,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-08T02:07:09",
  "op": [
    "custom_json",
    {
      "required_auths": [],
      "required_posting_auths": [
        "south-man"
      ],
      "id": "follow",
      "json": "[\"follow\",{\"follower\":\"south-man\",\"following\":\"wonsama\",\"what\":[\"blog\"]}]"
    }
  ]
}

{
  "trx_id": "dc0a3a7a9d07dfaf47bb5909240b629717b5c8b3",
  "block": 23116187,
  "trx_in_block": 5,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-07T14:40:09",
  "op": [
    "vote",
    {
      "voter": "wonsama",
      "author": "south-man",
      "permlink": "4invru",
      "weight": 1000
    }
  ]
}

{
  "trx_id": "3637634a1a696ef807df26700be5b446f9a12870",
  "block": 23114640,
  "trx_in_block": 5,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-07T13:22:42",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "south-man",
      "permlink": "image-of-airplane-car-bicycle-deep-learning-source",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "2b03cdd50476575c51abed1d015a19683a4fb624",
  "block": 23114639,
  "trx_in_block": 18,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-07T13:22:39",
  "op": [
    "comment",
    {
      "parent_author": "south-man",
      "parent_permlink": "image-of-airplane-car-bicycle-deep-learning-source",
      "author": "allnatural",
      "permlink": "re-south-man-image-of-airplane-car-bicycle-deep-learning-source-20180607t132238802z",
      "title": "",
      "body": "Get your post resteemed to 72,000 followers. Go here https://steemit.com/@a-a-a",
      "json_metadata": "{\"tags\":[\"deep-learning\"],\"links\":[\"https://steemit.com/@a-a-a\"],\"app\":\"steemit/0.1\"}"
    }
  ]
}

{
  "trx_id": "9319ca22e0ba1850b9cdfe519a53064e84121d91",
  "block": 23114637,
  "trx_in_block": 33,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-07T13:22:33",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "south-man",
      "permlink": "4invru",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "bfca12501ad93a34cd3cec8e1b8265fbf26ca306",
  "block": 23114630,
  "trx_in_block": 40,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-07T13:22:12",
  "op": [
    "comment",
    {
      "parent_author": "",
      "parent_permlink": "deep-learning",
      "author": "south-man",
      "permlink": "image-of-airplane-car-bicycle-deep-learning-source",
      "title": "Image of airplane, car, bicycle Deep learning source",
      "body": "It is a source to sort pictures after learning google image of airplane, car, bicycle and learn by deep learning.\nIf you collect a lot of photos, you will get higher accuracy. It took a lot of time to sort out the pictures, so I learned about 400 pictures only, so the accuracy did not come out so much .\n\nThe feather hub has the same source.\nhttps://github.com/llejo3/deep-learning/blob/master/vehicle_classification_learning.md\n\n\n```python\nfrom google_images_download import google_images_download\n```\n\n\n```python\nresponse = google_images_download.googleimagesdownload() \n```\n\nWhen I import a Google photo as a search query, I need to delete some photos that do not match my query.\n```python\n#arguments = {\"keywords\":\"forsythia,cherry blossoms,magnolia,azalea,tulip\",\narguments = {\"keywords\":\"Planes, Bicycles, Cars\",\n             \"limit\":600,\n             \"print_urls\":True,\n             format: \"jpg,png\",\n             \"chromedriver\" : \"./chromedriver.exe\",\n             \"exact_size\":\"200,200\"}\npaths = response.download(arguments) \n```\n\n\n\n```python\nfrom keras.layers import Conv2D, Dense, Flatten, MaxPooling2D, Dropout\nfrom keras.layers.normalization import BatchNormalization\nfrom keras.models import Sequential\nfrom keras.preprocessing.image import ImageDataGenerator\nfrom keras.optimizers import Adam, RMSprop, SGD\nfrom keras.callbacks import EarlyStopping, TensorBoard, ModelCheckpoint\n \n```\n\n\n```python\ntrain_datagen = ImageDataGenerator(\n        rescale=1./255,\n        rotation_range=40,      # 40도까지 회전\n        width_shift_range=0.2,  # 20%까지 좌우 이동\n        height_shift_range=0.2, # 20%까지 상하 이동\n        shear_range=0.2,        # 20%까지 기울임\n        zoom_range=0.2,         # 20%까지 확대\n        horizontal_flip=True   # 좌우 뒤집기\n)\n```\n\n\n```python\ntrain = train_datagen.flow_from_directory(\n    'vehicles/train',\n    target_size=(200, 200),\n    #batch_size=32,\n    class_mode='categorical')\n\nvalid = ImageDataGenerator(rescale=1.0/255).flow_from_directory(\n    'vehicles/validation',\n    target_size=(200, 200),\n    #batch_size=32,\n    class_mode='categorical',\n    shuffle=False)\n```\n\n    Found 884 images belonging to 3 classes.\n    Found 298 images belonging to 3 classes.\n    \n\n\n```python\nmodel1 = Sequential()\nmodel1.add(Conv2D(32, (3, 3), activation='relu', input_shape=(200, 200, 3)))\nmodel1.add(MaxPooling2D((2, 2)))\nmodel1.add(Flatten())\nmodel1.add(Dense(3, activation='softmax'))\n```\n\n\n```python\nmodel1.summary()\n```\n\n    _________________________________________________________________\n    Layer (type)                 Output Shape              Param #   \n    =================================================================\n    conv2d_8 (Conv2D)            (None, 198, 198, 32)      896       \n    _________________________________________________________________\n    max_pooling2d_8 (MaxPooling2 (None, 99, 99, 32)        0         \n    _________________________________________________________________\n    flatten_7 (Flatten)          (None, 313632)            0         \n    _________________________________________________________________\n    dense_7 (Dense)              (None, 3)                 940899    \n    =================================================================\n    Total params: 941,795\n    Trainable params: 941,795\n    Non-trainable params: 0\n    _________________________________________________________________\n    \n\n\n```python\nsgd = SGD(lr=0.01, decay=1e-6, momentum=0.9, nesterov=True)\nmodel1.compile(loss='categorical_crossentropy', metrics=['accuracy'], optimizer=sgd)\n```\n\n\n```python\nhistory1 = model1.fit_generator(\n    train, validation_data=valid, epochs=30,\n    callbacks=[\n        EarlyStopping(monitor = \"val_loss\", patience=2),\n        TensorBoard(log_dir='tensorboard_logs/log_model1')\n    ])\n```\n\n    Epoch 1/30\n     1/28 [>.............................] - ETA: 46s - loss: 1.1692 - acc: 0.3438\n\n    C:\\Users\\south\\Anaconda3\\lib\\site-packages\\PIL\\Image.py:918: UserWarning: Palette images with Transparency   expressed in bytes should be converted to RGBA images\n      'to RGBA images')\n    \n\n    28/28 [==============================] - 36s 1s/step - loss: 3.7392 - acc: 0.3803 - val_loss: 1.0593 - val_acc: 0.4564\n    Epoch 2/30\n    28/28 [==============================] - 35s 1s/step - loss: 1.1686 - acc: 0.4723 - val_loss: 1.0099 - val_acc: 0.5839\n    Epoch 3/30\n    28/28 [==============================] - 36s 1s/step - loss: 0.9614 - acc: 0.5998 - val_loss: 0.9786 - val_acc: 0.5772\n    Epoch 4/30\n    28/28 [==============================] - 36s 1s/step - loss: 0.9075 - acc: 0.6209 - val_loss: 0.9295 - val_acc: 0.5805\n    Epoch 5/30\n    28/28 [==============================] - 36s 1s/step - loss: 0.8156 - acc: 0.6610 - val_loss: 0.8961 - val_acc: 0.6242\n    Epoch 6/30\n    28/28 [==============================] - 36s 1s/step - loss: 0.7874 - acc: 0.6738 - val_loss: 0.9063 - val_acc: 0.5973\n    Epoch 7/30\n    28/28 [==============================] - 37s 1s/step - loss: 0.7597 - acc: 0.6724 - val_loss: 0.8911 - val_acc: 0.6074\n    Epoch 8/30\n    28/28 [==============================] - 37s 1s/step - loss: 0.7685 - acc: 0.6638 - val_loss: 0.9621 - val_acc: 0.5772\n    Epoch 9/30\n    28/28 [==============================] - 34s 1s/step - loss: 0.7566 - acc: 0.6674 - val_loss: 0.8563 - val_acc: 0.6376\n    Epoch 10/30\n    28/28 [==============================] - 34s 1s/step - loss: 0.6990 - acc: 0.7087 - val_loss: 0.9243 - val_acc: 0.6074\n    Epoch 11/30\n    28/28 [==============================] - 35s 1s/step - loss: 0.7083 - acc: 0.7040 - val_loss: 0.9144 - val_acc: 0.6208\n    \n\n\n```python\nmodel2 = Sequential()\nmodel2.add(Conv2D(32, (3, 3), activation='relu', input_shape=(200, 200, 3)))\nmodel2.add(MaxPooling2D((2, 2)))\nmodel2.add(Conv2D(32, (3, 3), activation='relu'))\nmodel2.add(MaxPooling2D((2, 2)))\nmodel2.add(Flatten())\nmodel2.add(Dense(3, activation='softmax'))\n```\n\n\n```python\nmodel2.summary()\n```\n\n    _________________________________________________________________\n    Layer (type)                 Output Shape              Param #   \n    =================================================================\n    conv2d_11 (Conv2D)           (None, 198, 198, 32)      896       \n    _________________________________________________________________\n    max_pooling2d_11 (MaxPooling (None, 99, 99, 32)        0         \n    _________________________________________________________________\n    conv2d_12 (Conv2D)           (None, 97, 97, 32)        9248      \n    _________________________________________________________________\n    max_pooling2d_12 (MaxPooling (None, 48, 48, 32)        0         \n    _________________________________________________________________\n    flatten_9 (Flatten)          (None, 73728)             0         \n    _________________________________________________________________\n    dense_9 (Dense)              (None, 3)                 221187    \n    =================================================================\n    Total params: 231,331\n    Trainable params: 231,331\n    Non-trainable params: 0\n    _________________________________________________________________\n    \n\n\n```python\nsgd = SGD(lr=0.001, decay=1e-6, momentum=0.9, nesterov=True)\nmodel2.compile(loss='categorical_crossentropy', metrics=['accuracy'], optimizer=sgd)\n```\n\n\n```python\nhistory2 = model2.fit_generator(\n    train, validation_data=valid, epochs=100,\n    callbacks=[\n        EarlyStopping(monitor = \"val_loss\", patience=2),\n        TensorBoard(log_dir='tensorboard_logs/log_model2')\n    ])\n```\n\n    Epoch 1/30\n     4/28 [===>..........................] - ETA: 44s - loss: 1.4224 - acc: 0.3391\n\n    C:\\Users\\south\\Anaconda3\\lib\\site-packages\\PIL\\Image.py:918: UserWarning: Palette images with Transparency   expressed in bytes should be converted to RGBA images\n      'to RGBA images')\n    \n\n    28/28 [==============================] - 57s 2s/step - loss: 1.1169 - acc: 0.4312 - val_loss: 1.3070 - val_acc: 0.3658\n    Epoch 2/30\n    28/28 [==============================] - 53s 2s/step - loss: 1.0442 - acc: 0.4781 - val_loss: 1.0760 - val_acc: 0.4262\n    Epoch 3/30\n    28/28 [==============================] - 55s 2s/step - loss: 0.9853 - acc: 0.5179 - val_loss: 0.9727 - val_acc: 0.5503\n    Epoch 4/30\n    28/28 [==============================] - 55s 2s/step - loss: 0.9722 - acc: 0.5405 - val_loss: 1.0100 - val_acc: 0.6107\n    Epoch 5/30\n    28/28 [==============================] - 54s 2s/step - loss: 0.9162 - acc: 0.5844 - val_loss: 0.9376 - val_acc: 0.6040\n    Epoch 6/30\n    28/28 [==============================] - 59s 2s/step - loss: 0.7842 - acc: 0.6691 - val_loss: 1.0358 - val_acc: 0.6107\n    Epoch 7/30\n    28/28 [==============================] - 55s 2s/step - loss: 0.7229 - acc: 0.6908 - val_loss: 0.7254 - val_acc: 0.7148\n    Epoch 8/30\n    28/28 [==============================] - 55s 2s/step - loss: 0.7381 - acc: 0.6953 - val_loss: 0.8614 - val_acc: 0.6342\n    Epoch 9/30\n    28/28 [==============================] - 54s 2s/step - loss: 0.6778 - acc: 0.7174 - val_loss: 0.7262 - val_acc: 0.6980\n    \n\n\n```python\nmodel3 = Sequential()\nmodel3.add(Conv2D(32, (3, 3), activation='relu', input_shape=(200, 200, 3)))\nmodel3.add(BatchNormalization())\nmodel3.add(Conv2D(32, (3, 3), activation='relu'))\nmodel3.add(MaxPooling2D(pool_size=(2, 2)))\nmodel3.add(Dropout(0.25))\n\nmodel3.add(Conv2D(64, (3, 3), activation='relu'))\nmodel3.add(BatchNormalization())\nmodel3.add(Conv2D(64, (3, 3), activation='relu'))\nmodel3.add(MaxPooling2D(pool_size=(2, 2)))\nmodel3.add(Dropout(0.25))\n\nmodel3.add(Flatten())\nmodel3.add(Dense(256, activation='relu'))\nmodel3.add(Dropout(0.5))\nmodel3.add(Dense(3, activation='softmax'))\n```\n\n\n```python\nmodel3.summary()\n```\n\n    _________________________________________________________________\n    Layer (type)                 Output Shape              Param #   \n    =================================================================\n    conv2d_19 (Conv2D)           (None, 198, 198, 32)      896       \n    _________________________________________________________________\n    max_pooling2d_16 (MaxPooling (None, 99, 99, 32)        0         \n    _________________________________________________________________\n    dropout_4 (Dropout)          (None, 99, 99, 32)        0         \n    _________________________________________________________________\n    conv2d_20 (Conv2D)           (None, 97, 97, 64)        18496     \n    _________________________________________________________________\n    max_pooling2d_17 (MaxPooling (None, 48, 48, 64)        0         \n    _________________________________________________________________\n    dropout_5 (Dropout)          (None, 48, 48, 64)        0         \n    _________________________________________________________________\n    flatten_11 (Flatten)         (None, 147456)            0         \n    _________________________________________________________________\n    dense_12 (Dense)             (None, 256)               37748992  \n    _________________________________________________________________\n    dropout_6 (Dropout)          (None, 256)               0         \n    _________________________________________________________________\n    dense_13 (Dense)             (None, 3)                 771       \n    =================================================================\n    Total params: 37,769,155\n    Trainable params: 37,769,155\n    Non-trainable params: 0\n    _________________________________________________________________\n    \n\n\n```python\nsgd = SGD(lr=0.01, decay=1e-6, momentum=0.9, nesterov=True)\nmodel3.compile(loss='categorical_crossentropy', metrics=['accuracy'], optimizer=sgd)\n```\n\n\n```python\nhistory3 = model3.fit_generator(\n    train, validation_data=valid, epochs=30,\n    callbacks=[\n        EarlyStopping(monitor = \"val_loss\", patience=2),\n        ModelCheckpoint('model3-{epoch:02d}.hdf5', save_best_only=True),\n        TensorBoard(log_dir='tensorboard_logs/log_model3')\n    ])\n```\n\n    Epoch 1/30\n     1/28 [>.............................] - ETA: 1:34 - loss: 1.1490 - acc: 0.2500\n\n    C:\\Users\\south\\Anaconda3\\lib\\site-packages\\PIL\\Image.py:918: UserWarning: Palette images with Transparency   expressed in bytes should be converted to RGBA images\n      'to RGBA images')\n    \n\n    28/28 [==============================] - 85s 3s/step - loss: 1.2178 - acc: 0.3502 - val_loss: 1.0796 - val_acc: 0.3893\n    Epoch 2/30\n    28/28 [==============================] - 86s 3s/step - loss: 1.0463 - acc: 0.4851 - val_loss: 1.0564 - val_acc: 0.4899\n    Epoch 3/30\n    28/28 [==============================] - 82s 3s/step - loss: 0.9830 - acc: 0.5198 - val_loss: 1.0233 - val_acc: 0.5839\n    Epoch 4/30\n    28/28 [==============================] - 86s 3s/step - loss: 0.8876 - acc: 0.6224 - val_loss: 0.9437 - val_acc: 0.5336\n    Epoch 5/30\n    28/28 [==============================] - 84s 3s/step - loss: 0.7544 - acc: 0.6600 - val_loss: 0.8263 - val_acc: 0.6577\n    Epoch 6/30\n    28/28 [==============================] - 86s 3s/step - loss: 0.7302 - acc: 0.7052 - val_loss: 0.7034 - val_acc: 0.6879\n    Epoch 7/30\n    28/28 [==============================] - 89s 3s/step - loss: 0.6840 - acc: 0.7004 - val_loss: 0.6810 - val_acc: 0.7047\n    Epoch 8/30\n    28/28 [==============================] - 89s 3s/step - loss: 0.6499 - acc: 0.7214 - val_loss: 0.6697 - val_acc: 0.7248\n    Epoch 9/30\n    28/28 [==============================] - 87s 3s/step - loss: 0.6163 - acc: 0.7403 - val_loss: 0.6561 - val_acc: 0.7383\n    Epoch 10/30\n    28/28 [==============================] - 107s 4s/step - loss: 0.6387 - acc: 0.7238 - val_loss: 0.6992 - val_acc: 0.6846\n    Epoch 11/30\n    28/28 [==============================] - 90s 3s/step - loss: 0.5976 - acc: 0.7457 - val_loss: 0.6568 - val_acc: 0.7114\n\n\n![deepLearning.PNG](https://cdn.steemitimages.com/DQmaqW4SM7dpaZaKFAxqUDM7oMCYFN4qbH3TU3cLmzXzNKs/deepLearning.PNG)",
      "json_metadata": "{\"tags\":[\"deep-learning\",\"keras\",\"vehicle\",\"python\"],\"image\":[\"https://cdn.steemitimages.com/DQmaqW4SM7dpaZaKFAxqUDM7oMCYFN4qbH3TU3cLmzXzNKs/deepLearning.PNG\"],\"links\":[\"https://github.com/llejo3/deep-learning/blob/master/vehicle_classification_learning.md\"],\"app\":\"steemit/0.1\",\"format\":\"markdown\"}"
    }
  ]
}

{
  "trx_id": "46d0e969de5db3860faf18be38351d33ffff56b6",
  "block": 23114554,
  "trx_in_block": 24,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-07T13:18:24",
  "op": [
    "vote",
    {
      "voter": "raise-me-up",
      "author": "south-man",
      "permlink": "4invru",
      "weight": 60
    }
  ]
}

{
  "trx_id": "0529e3e49acacdd585ad2a5872bb192e27e2e197",
  "block": 23114529,
  "trx_in_block": 26,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-07T13:17:09",
  "op": [
    "comment",
    {
      "parent_author": "",
      "parent_permlink": "deep-learning",
      "author": "south-man",
      "permlink": "4invru",
      "title": "비행기, 자동차, 자전거의 이미지 딥러닝 소스",
      "body": "비행기, 자동차, 자전거의 google이미지를 받아서 딥러닝으로 학습한 후에 사진을 분류하는 소스입니다.\n사진을 많이 수집해서 하면 좀더 높은 정확도가 나올 것 같은데 사진을 추려내는 데 시간이 많이 걸려서 각 400개 정도의 사진만으로 학습을 시켰기 때문에 정확도가 그리 높게 나오진 않았어요^^\n\n깃 허브에도 동일한 소스가 있습니다.\nhttps://github.com/llejo3/deep-learning/blob/master/vehicle_classification_learning.md\n\n\n```python\nfrom google_images_download import google_images_download\n```\n\n\n```python\nresponse = google_images_download.googleimagesdownload() \n```\n\n구글 사진을 검색어로 가져오는데, 검색어와 맞지 않는 사진은 삭제하는 작업이 좀 필요합니다.\n```python\n#arguments = {\"keywords\":\"forsythia,cherry blossoms,magnolia,azalea,tulip\",\narguments = {\"keywords\":\"Planes, Bicycles, Cars\",\n             \"limit\":600,\n             \"print_urls\":True,\n             format: \"jpg,png\",\n             \"chromedriver\" : \"./chromedriver.exe\",\n             \"exact_size\":\"200,200\"}\npaths = response.download(arguments) \n```\n\n\n\n```python\nfrom keras.layers import Conv2D, Dense, Flatten, MaxPooling2D, Dropout\nfrom keras.layers.normalization import BatchNormalization\nfrom keras.models import Sequential\nfrom keras.preprocessing.image import ImageDataGenerator\nfrom keras.optimizers import Adam, RMSprop, SGD\nfrom keras.callbacks import EarlyStopping, TensorBoard, ModelCheckpoint\n \n```\n\n\n```python\ntrain_datagen = ImageDataGenerator(\n        rescale=1./255,\n        rotation_range=40,      # 40도까지 회전\n        width_shift_range=0.2,  # 20%까지 좌우 이동\n        height_shift_range=0.2, # 20%까지 상하 이동\n        shear_range=0.2,        # 20%까지 기울임\n        zoom_range=0.2,         # 20%까지 확대\n        horizontal_flip=True   # 좌우 뒤집기\n)\n```\n\n\n```python\ntrain = train_datagen.flow_from_directory(\n    'vehicles/train',\n    target_size=(200, 200),\n    #batch_size=32,\n    class_mode='categorical')\n\nvalid = ImageDataGenerator(rescale=1.0/255).flow_from_directory(\n    'vehicles/validation',\n    target_size=(200, 200),\n    #batch_size=32,\n    class_mode='categorical',\n    shuffle=False)\n```\n\n    Found 884 images belonging to 3 classes.\n    Found 298 images belonging to 3 classes.\n    \n\n\n```python\nmodel1 = Sequential()\nmodel1.add(Conv2D(32, (3, 3), activation='relu', input_shape=(200, 200, 3)))\nmodel1.add(MaxPooling2D((2, 2)))\nmodel1.add(Flatten())\nmodel1.add(Dense(3, activation='softmax'))\n```\n\n\n```python\nmodel1.summary()\n```\n\n    _________________________________________________________________\n    Layer (type)                 Output Shape              Param #   \n    =================================================================\n    conv2d_8 (Conv2D)            (None, 198, 198, 32)      896       \n    _________________________________________________________________\n    max_pooling2d_8 (MaxPooling2 (None, 99, 99, 32)        0         \n    _________________________________________________________________\n    flatten_7 (Flatten)          (None, 313632)            0         \n    _________________________________________________________________\n    dense_7 (Dense)              (None, 3)                 940899    \n    =================================================================\n    Total params: 941,795\n    Trainable params: 941,795\n    Non-trainable params: 0\n    _________________________________________________________________\n    \n\n\n```python\nsgd = SGD(lr=0.01, decay=1e-6, momentum=0.9, nesterov=True)\nmodel1.compile(loss='categorical_crossentropy', metrics=['accuracy'], optimizer=sgd)\n```\n\n\n```python\nhistory1 = model1.fit_generator(\n    train, validation_data=valid, epochs=30,\n    callbacks=[\n        EarlyStopping(monitor = \"val_loss\", patience=2),\n        TensorBoard(log_dir='tensorboard_logs/log_model1')\n    ])\n```\n\n    Epoch 1/30\n     1/28 [>.............................] - ETA: 46s - loss: 1.1692 - acc: 0.3438\n\n    C:\\Users\\south\\Anaconda3\\lib\\site-packages\\PIL\\Image.py:918: UserWarning: Palette images with Transparency   expressed in bytes should be converted to RGBA images\n      'to RGBA images')\n    \n\n    28/28 [==============================] - 36s 1s/step - loss: 3.7392 - acc: 0.3803 - val_loss: 1.0593 - val_acc: 0.4564\n    Epoch 2/30\n    28/28 [==============================] - 35s 1s/step - loss: 1.1686 - acc: 0.4723 - val_loss: 1.0099 - val_acc: 0.5839\n    Epoch 3/30\n    28/28 [==============================] - 36s 1s/step - loss: 0.9614 - acc: 0.5998 - val_loss: 0.9786 - val_acc: 0.5772\n    Epoch 4/30\n    28/28 [==============================] - 36s 1s/step - loss: 0.9075 - acc: 0.6209 - val_loss: 0.9295 - val_acc: 0.5805\n    Epoch 5/30\n    28/28 [==============================] - 36s 1s/step - loss: 0.8156 - acc: 0.6610 - val_loss: 0.8961 - val_acc: 0.6242\n    Epoch 6/30\n    28/28 [==============================] - 36s 1s/step - loss: 0.7874 - acc: 0.6738 - val_loss: 0.9063 - val_acc: 0.5973\n    Epoch 7/30\n    28/28 [==============================] - 37s 1s/step - loss: 0.7597 - acc: 0.6724 - val_loss: 0.8911 - val_acc: 0.6074\n    Epoch 8/30\n    28/28 [==============================] - 37s 1s/step - loss: 0.7685 - acc: 0.6638 - val_loss: 0.9621 - val_acc: 0.5772\n    Epoch 9/30\n    28/28 [==============================] - 34s 1s/step - loss: 0.7566 - acc: 0.6674 - val_loss: 0.8563 - val_acc: 0.6376\n    Epoch 10/30\n    28/28 [==============================] - 34s 1s/step - loss: 0.6990 - acc: 0.7087 - val_loss: 0.9243 - val_acc: 0.6074\n    Epoch 11/30\n    28/28 [==============================] - 35s 1s/step - loss: 0.7083 - acc: 0.7040 - val_loss: 0.9144 - val_acc: 0.6208\n    \n\n\n```python\nmodel2 = Sequential()\nmodel2.add(Conv2D(32, (3, 3), activation='relu', input_shape=(200, 200, 3)))\nmodel2.add(MaxPooling2D((2, 2)))\nmodel2.add(Conv2D(32, (3, 3), activation='relu'))\nmodel2.add(MaxPooling2D((2, 2)))\nmodel2.add(Flatten())\nmodel2.add(Dense(3, activation='softmax'))\n```\n\n\n```python\nmodel2.summary()\n```\n\n    _________________________________________________________________\n    Layer (type)                 Output Shape              Param #   \n    =================================================================\n    conv2d_11 (Conv2D)           (None, 198, 198, 32)      896       \n    _________________________________________________________________\n    max_pooling2d_11 (MaxPooling (None, 99, 99, 32)        0         \n    _________________________________________________________________\n    conv2d_12 (Conv2D)           (None, 97, 97, 32)        9248      \n    _________________________________________________________________\n    max_pooling2d_12 (MaxPooling (None, 48, 48, 32)        0         \n    _________________________________________________________________\n    flatten_9 (Flatten)          (None, 73728)             0         \n    _________________________________________________________________\n    dense_9 (Dense)              (None, 3)                 221187    \n    =================================================================\n    Total params: 231,331\n    Trainable params: 231,331\n    Non-trainable params: 0\n    _________________________________________________________________\n    \n\n\n```python\nsgd = SGD(lr=0.001, decay=1e-6, momentum=0.9, nesterov=True)\nmodel2.compile(loss='categorical_crossentropy', metrics=['accuracy'], optimizer=sgd)\n```\n\n\n```python\nhistory2 = model2.fit_generator(\n    train, validation_data=valid, epochs=100,\n    callbacks=[\n        EarlyStopping(monitor = \"val_loss\", patience=2),\n        TensorBoard(log_dir='tensorboard_logs/log_model2')\n    ])\n```\n\n    Epoch 1/30\n     4/28 [===>..........................] - ETA: 44s - loss: 1.4224 - acc: 0.3391\n\n    C:\\Users\\south\\Anaconda3\\lib\\site-packages\\PIL\\Image.py:918: UserWarning: Palette images with Transparency   expressed in bytes should be converted to RGBA images\n      'to RGBA images')\n    \n\n    28/28 [==============================] - 57s 2s/step - loss: 1.1169 - acc: 0.4312 - val_loss: 1.3070 - val_acc: 0.3658\n    Epoch 2/30\n    28/28 [==============================] - 53s 2s/step - loss: 1.0442 - acc: 0.4781 - val_loss: 1.0760 - val_acc: 0.4262\n    Epoch 3/30\n    28/28 [==============================] - 55s 2s/step - loss: 0.9853 - acc: 0.5179 - val_loss: 0.9727 - val_acc: 0.5503\n    Epoch 4/30\n    28/28 [==============================] - 55s 2s/step - loss: 0.9722 - acc: 0.5405 - val_loss: 1.0100 - val_acc: 0.6107\n    Epoch 5/30\n    28/28 [==============================] - 54s 2s/step - loss: 0.9162 - acc: 0.5844 - val_loss: 0.9376 - val_acc: 0.6040\n    Epoch 6/30\n    28/28 [==============================] - 59s 2s/step - loss: 0.7842 - acc: 0.6691 - val_loss: 1.0358 - val_acc: 0.6107\n    Epoch 7/30\n    28/28 [==============================] - 55s 2s/step - loss: 0.7229 - acc: 0.6908 - val_loss: 0.7254 - val_acc: 0.7148\n    Epoch 8/30\n    28/28 [==============================] - 55s 2s/step - loss: 0.7381 - acc: 0.6953 - val_loss: 0.8614 - val_acc: 0.6342\n    Epoch 9/30\n    28/28 [==============================] - 54s 2s/step - loss: 0.6778 - acc: 0.7174 - val_loss: 0.7262 - val_acc: 0.6980\n    \n\n\n```python\nmodel3 = Sequential()\nmodel3.add(Conv2D(32, (3, 3), activation='relu', input_shape=(200, 200, 3)))\nmodel3.add(BatchNormalization())\nmodel3.add(Conv2D(32, (3, 3), activation='relu'))\nmodel3.add(MaxPooling2D(pool_size=(2, 2)))\nmodel3.add(Dropout(0.25))\n\nmodel3.add(Conv2D(64, (3, 3), activation='relu'))\nmodel3.add(BatchNormalization())\nmodel3.add(Conv2D(64, (3, 3), activation='relu'))\nmodel3.add(MaxPooling2D(pool_size=(2, 2)))\nmodel3.add(Dropout(0.25))\n\nmodel3.add(Flatten())\nmodel3.add(Dense(256, activation='relu'))\nmodel3.add(Dropout(0.5))\nmodel3.add(Dense(3, activation='softmax'))\n```\n\n\n```python\nmodel3.summary()\n```\n\n    _________________________________________________________________\n    Layer (type)                 Output Shape              Param #   \n    =================================================================\n    conv2d_19 (Conv2D)           (None, 198, 198, 32)      896       \n    _________________________________________________________________\n    max_pooling2d_16 (MaxPooling (None, 99, 99, 32)        0         \n    _________________________________________________________________\n    dropout_4 (Dropout)          (None, 99, 99, 32)        0         \n    _________________________________________________________________\n    conv2d_20 (Conv2D)           (None, 97, 97, 64)        18496     \n    _________________________________________________________________\n    max_pooling2d_17 (MaxPooling (None, 48, 48, 64)        0         \n    _________________________________________________________________\n    dropout_5 (Dropout)          (None, 48, 48, 64)        0         \n    _________________________________________________________________\n    flatten_11 (Flatten)         (None, 147456)            0         \n    _________________________________________________________________\n    dense_12 (Dense)             (None, 256)               37748992  \n    _________________________________________________________________\n    dropout_6 (Dropout)          (None, 256)               0         \n    _________________________________________________________________\n    dense_13 (Dense)             (None, 3)                 771       \n    =================================================================\n    Total params: 37,769,155\n    Trainable params: 37,769,155\n    Non-trainable params: 0\n    _________________________________________________________________\n    \n\n\n```python\nsgd = SGD(lr=0.01, decay=1e-6, momentum=0.9, nesterov=True)\nmodel3.compile(loss='categorical_crossentropy', metrics=['accuracy'], optimizer=sgd)\n```\n\n\n```python\nhistory3 = model3.fit_generator(\n    train, validation_data=valid, epochs=30,\n    callbacks=[\n        EarlyStopping(monitor = \"val_loss\", patience=2),\n        ModelCheckpoint('model3-{epoch:02d}.hdf5', save_best_only=True),\n        TensorBoard(log_dir='tensorboard_logs/log_model3')\n    ])\n```\n\n    Epoch 1/30\n     1/28 [>.............................] - ETA: 1:34 - loss: 1.1490 - acc: 0.2500\n\n    C:\\Users\\south\\Anaconda3\\lib\\site-packages\\PIL\\Image.py:918: UserWarning: Palette images with Transparency   expressed in bytes should be converted to RGBA images\n      'to RGBA images')\n    \n\n    28/28 [==============================] - 85s 3s/step - loss: 1.2178 - acc: 0.3502 - val_loss: 1.0796 - val_acc: 0.3893\n    Epoch 2/30\n    28/28 [==============================] - 86s 3s/step - loss: 1.0463 - acc: 0.4851 - val_loss: 1.0564 - val_acc: 0.4899\n    Epoch 3/30\n    28/28 [==============================] - 82s 3s/step - loss: 0.9830 - acc: 0.5198 - val_loss: 1.0233 - val_acc: 0.5839\n    Epoch 4/30\n    28/28 [==============================] - 86s 3s/step - loss: 0.8876 - acc: 0.6224 - val_loss: 0.9437 - val_acc: 0.5336\n    Epoch 5/30\n    28/28 [==============================] - 84s 3s/step - loss: 0.7544 - acc: 0.6600 - val_loss: 0.8263 - val_acc: 0.6577\n    Epoch 6/30\n    28/28 [==============================] - 86s 3s/step - loss: 0.7302 - acc: 0.7052 - val_loss: 0.7034 - val_acc: 0.6879\n    Epoch 7/30\n    28/28 [==============================] - 89s 3s/step - loss: 0.6840 - acc: 0.7004 - val_loss: 0.6810 - val_acc: 0.7047\n    Epoch 8/30\n    28/28 [==============================] - 89s 3s/step - loss: 0.6499 - acc: 0.7214 - val_loss: 0.6697 - val_acc: 0.7248\n    Epoch 9/30\n    28/28 [==============================] - 87s 3s/step - loss: 0.6163 - acc: 0.7403 - val_loss: 0.6561 - val_acc: 0.7383\n    Epoch 10/30\n    28/28 [==============================] - 107s 4s/step - loss: 0.6387 - acc: 0.7238 - val_loss: 0.6992 - val_acc: 0.6846\n    Epoch 11/30\n    28/28 [==============================] - 90s 3s/step - loss: 0.5976 - acc: 0.7457 - val_loss: 0.6568 - val_acc: 0.7114\n\n\n![deepLearning.PNG](https://cdn.steemitimages.com/DQmaqW4SM7dpaZaKFAxqUDM7oMCYFN4qbH3TU3cLmzXzNKs/deepLearning.PNG)",
      "json_metadata": "{\"tags\":[\"deep-learning\",\"kr\",\"keras\",\"vehicle\",\"python\"],\"image\":[\"https://cdn.steemitimages.com/DQmaqW4SM7dpaZaKFAxqUDM7oMCYFN4qbH3TU3cLmzXzNKs/deepLearning.PNG\"],\"links\":[\"https://github.com/llejo3/deep-learning/blob/master/vehicle_classification_learning.md\"],\"app\":\"steemit/0.1\",\"format\":\"markdown\"}"
    }
  ]
}

{
  "trx_id": "36493916d0ce6ce8b8270e4c5a883d7b9cec55da",
  "block": 23105941,
  "trx_in_block": 4,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-07T06:07:36",
  "op": [
    "comment",
    {
      "parent_author": "",
      "parent_permlink": "python",
      "author": "south-man",
      "permlink": "5cmvsm",
      "title": "구글이미지 다운로드하는 파이썬 라이브러리",
      "body": "@@ -22,17 +22,17 @@\n %EB%8A%94 %EC%A4%91 %EC%B0%BE%EC%9D%80 %EB%9D%BC\n-%EB%A6%AC\n+%EC%9D%B4\n %EB%B8%8C%EB%9F%AC%EB%A6%AC %EC%9E%85%EB%8B%88%EB%8B%A4.\n",
      "json_metadata": "{\"tags\":[\"python\",\"kr\",\"google\",\"image\",\"download\"],\"links\":[\"https://github.com/hardikvasa/google-images-download\"],\"app\":\"steemit/0.1\",\"format\":\"markdown\",\"image\":[\"https://cdn.steemitimages.com/DQmPekZRMJbNb1SRUFc46UzheXSQCXPFT1TU6969j2hukcq/image.png\"]}"
    }
  ]
}

{
  "trx_id": "45ce6657caf1dcaef24d20138a092e02ca007b82",
  "block": 23075279,
  "trx_in_block": 32,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-06T04:34:18",
  "op": [
    "comment",
    {
      "parent_author": "",
      "parent_permlink": "python",
      "author": "south-man",
      "permlink": "5cmvsm",
      "title": "구글이미지 다운로드하는 파이썬 라이브러리",
      "body": "@@ -103,8 +103,103 @@\n download\n+%0A%0A%0A!%5B%5D(https://cdn.steemitimages.com/DQmPekZRMJbNb1SRUFc46UzheXSQCXPFT1TU6969j2hukcq/image.png)\n",
      "json_metadata": "{\"tags\":[\"python\",\"kr\",\"google\",\"image\",\"download\"],\"links\":[\"https://github.com/hardikvasa/google-images-download\"],\"app\":\"steemit/0.1\",\"format\":\"markdown\",\"image\":[\"https://cdn.steemitimages.com/DQmPekZRMJbNb1SRUFc46UzheXSQCXPFT1TU6969j2hukcq/image.png\"]}"
    }
  ]
}

{
  "trx_id": "a1f8c638fb7f282d2f01acfbf77e0a4367a3ddde",
  "block": 23075257,
  "trx_in_block": 53,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-06T04:33:12",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "south-man",
      "permlink": "5cmvsm",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "22f22f2dcaa3112355eb495dba338ebb971314f5",
  "block": 23075250,
  "trx_in_block": 54,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-06T04:32:51",
  "op": [
    "comment",
    {
      "parent_author": "",
      "parent_permlink": "python",
      "author": "south-man",
      "permlink": "5cmvsm",
      "title": "구글이미지 다운로드하는 파이썬 라이브러리",
      "body": "구글 이미지 다운로드가 필요해서 검색하는 중 찾은 라리브러리 입니다. 넘 편하고 좋네요. 속이 후련^^\nhttps://github.com/hardikvasa/google-images-download",
      "json_metadata": "{\"tags\":[\"python\",\"kr\",\"google\",\"image\",\"download\"],\"links\":[\"https://github.com/hardikvasa/google-images-download\"],\"app\":\"steemit/0.1\",\"format\":\"markdown\"}"
    }
  ]
}

{
  "trx_id": "16018dd84d856ab8d36c04417bd1002772203ce5",
  "block": 23037064,
  "trx_in_block": 107,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-04T20:42:57",
  "op": [
    "custom_json",
    {
      "required_auths": [],
      "required_posting_auths": [
        "south-man"
      ],
      "id": "follow",
      "json": "[\"reblog\",{\"account\":\"south-man\",\"author\":\"gmreza\",\"permlink\":\"nice-flower-click-on-this-red\"}]"
    }
  ]
}

{
  "trx_id": "96f9c382545995b870cd0ad84eda2370fa6e580a",
  "block": 23037059,
  "trx_in_block": 26,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-06-04T20:42:42",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "gmreza",
      "permlink": "nice-flower-click-on-this-red",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "18054b13b7d310f2c8e2eeadc19cea15ec95dcba",
  "block": 22882477,
  "trx_in_block": 40,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-30T11:50:33",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "south-man",
      "permlink": "6kq5ue",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "82945493cf710a35403dab5a184ef6da5e53fcd1",
  "block": 22882466,
  "trx_in_block": 30,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-30T11:50:00",
  "op": [
    "comment",
    {
      "parent_author": "",
      "parent_permlink": "monitor",
      "author": "south-man",
      "permlink": "6kq5ue",
      "title": "프로그램 감시하다가 다운되면 다시 살리는 쉘 스크립트",
      "body": "깃 허브에도 올라가 있는데, 이전에 소개드린 바이넨스 자동 매매 프로그램이 죽었을 때 다시 살리는 쉘 스크립트입니다.\nhttps://github.com/llejo3/binance-trading/blob/master/tradingMonitor.sh\n\n```\n#!/bin/bash\n\nlog=./logs/monitor.log\nSLEEP_SECONDS=300     # 대기 시간\nBASIC_GAP_SECONDS=300 # 비교를 위한 기준 간격 시간\n\nwhile [ 1 ] \ndo\n  # 로그가 1분정도마다 남기게 되어 있어서 INFO로 남긴 데이터 중에서 마지막 시간을 가져온다.\n  last_date=`tail -200 ./logs/process.log | grep INFO | tail -1 | cut -c 2-20`\n  diff_seconds=$(($(date -d now +%s) - $(date -d \"$last_date\" +%s)))    \t\n  if [ $diff_seconds -gt $BASIC_GAP_SECONDS ]  # 로그가 없으면 다시 실행              \n  then\n    # process ID를 구한다음 프로세스를 종료시킨 다음 다시 시작한다.\n    PROCESS=`ps -ef|grep process|grep -v grep | awk '{print $2}'`\n\tkill -9 $PROCESS\n\twait\n\tnow_date=`date '+%Y-%m-%d %H:%M:%S'`\n\techo \"$now_date : Process restarted.\" >> $log\n\t./startProcess.sh\n\twait\n  fi\n  sleep $SLEEP_SECONDS\ndone\n```\n\n![](https://cdn.steemitimages.com/DQmNdmJKRFtDQr5RvEUZZEaPESa2ghFc8MMN1gm88WsM2xf/image.png)",
      "json_metadata": "{\"tags\":[\"monitor\",\"shell\",\"re-live\",\"kr\",\"unix\"],\"image\":[\"https://cdn.steemitimages.com/DQmNdmJKRFtDQr5RvEUZZEaPESa2ghFc8MMN1gm88WsM2xf/image.png\"],\"links\":[\"https://github.com/llejo3/binance-trading/blob/master/tradingMonitor.sh\"],\"app\":\"steemit/0.1\",\"format\":\"markdown\"}"
    }
  ]
}

{
  "trx_id": "2068dab49100668195dee472ea44dac27cc64e18",
  "block": 22882083,
  "trx_in_block": 12,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-30T11:30:51",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "shinhancard",
      "permlink": "2pny1t",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "4246eef754e8a11b37a7b6c059e1bc620b7e3ec2",
  "block": 22869551,
  "trx_in_block": 32,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-30T01:04:09",
  "op": [
    "custom_json",
    {
      "required_auths": [],
      "required_posting_auths": [
        "south-man"
      ],
      "id": "follow",
      "json": "[\"follow\",{\"follower\":\"south-man\",\"following\":\"aslan1\",\"what\":[\"blog\"]}]"
    }
  ]
}

{
  "trx_id": "af09a46649f09c3ba25cffe2f43eadbe32b0265a",
  "block": 22869550,
  "trx_in_block": 8,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-30T01:04:06",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "aslan1",
      "permlink": "spring-flower-on-steemit",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "2a676bb0e54fa30445db912bd71e504a736b3da3",
  "block": 22869548,
  "trx_in_block": 47,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-30T01:04:00",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "aslan1",
      "permlink": "flowers-view",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "5927e2d234a8fa3170294a829c25ea468a877c68",
  "block": 22869548,
  "trx_in_block": 29,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-30T01:04:00",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "aslan1",
      "permlink": "flower-mountain-day",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "4a851ad8dc950eefffaab5a06fff665ac4622022",
  "block": 22869507,
  "trx_in_block": 45,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-30T01:01:57",
  "op": [
    "custom_json",
    {
      "required_auths": [],
      "required_posting_auths": [
        "south-man"
      ],
      "id": "follow",
      "json": "[\"follow\",{\"follower\":\"south-man\",\"following\":\"sd974201\",\"what\":[\"blog\"]}]"
    }
  ]
}

{
  "trx_id": "82826ae8ceaf64af51e47902c26075c84e046616",
  "block": 22869496,
  "trx_in_block": 50,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-30T01:01:24",
  "op": [
    "custom_json",
    {
      "required_auths": [],
      "required_posting_auths": [
        "south-man"
      ],
      "id": "follow",
      "json": "[\"follow\",{\"follower\":\"south-man\",\"following\":\"anomaly\",\"what\":[\"blog\"]}]"
    }
  ]
}

{
  "trx_id": "4d2d1939f2ae6d9c29b857be41a104228512b89c",
  "block": 22868041,
  "trx_in_block": 6,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-29T23:48:39",
  "op": [
    "vote",
    {
      "voter": "anomaly",
      "author": "south-man",
      "permlink": "tensorflow-source-to-classify-cifar-10-data",
      "weight": 100
    }
  ]
}

{
  "trx_id": "f31c094ef3be139b1dbd54078b1a4e3ba64ddcf1",
  "block": 22867424,
  "trx_in_block": 56,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-29T23:17:48",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "south-man",
      "permlink": "tensorflow-source-to-classify-cifar-10-data",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "39f02c74a75d7ed032fc5b503f9ca7bcebab0da2",
  "block": 22867400,
  "trx_in_block": 42,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-29T23:16:36",
  "op": [
    "comment",
    {
      "parent_author": "",
      "parent_permlink": "tensorflow",
      "author": "south-man",
      "permlink": "tensorflow-source-to-classify-cifar-10-data",
      "title": "Tensorflow source to classify CIFAR 10 data",
      "body": "Source is also on github.\nhttps://github.com/llejo3/deep-learning/blob/master/CIFAR10_tensorflow.md\n\n\n```python\nimport tensorflow as tf\n```\n\n\n```python\n# Helper functions\n\ndef weight_variable(shape):\n    initial = tf.truncated_normal(shape, stddev=0.1)\n    return tf.Variable(initial)\n\n\ndef bias_variable(shape):\n    initial = tf.constant(0.1, shape=shape)\n    return tf.Variable(initial)\n\n\ndef conv2d(x, W):\n    return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')\n\n\ndef max_pool_2x2(x):\n    return tf.nn.max_pool(x, ksize=[1, 2, 2, 1],\n                          strides=[1, 2, 2, 1], padding='SAME')\n\n\ndef conv_layer(input, shape):\n    W = weight_variable(shape)\n    b = bias_variable([shape[3]])\n    return tf.nn.relu(conv2d(input, W) + b)\n\ndef conv_norm_layer(input, shape, phase):\n    W = weight_variable(shape)\n    b = bias_variable([shape[3]])\n    return tf.nn.relu( batch_norm_wrapper( conv2d(input, W) + b, phase))\n\ndef full_layer(input, size):\n    in_size = int(input.get_shape()[1])\n    W = weight_variable([in_size, size])\n    b = bias_variable([size])\n    return tf.matmul(input, W) + b\n\n# Batch Nomalization \ndef batch_norm_wrapper(inputs, is_training, decay = 0.999):\n    scale = tf.Variable(tf.ones([inputs.get_shape()[-1]]))\n    beta = tf.Variable(tf.zeros([inputs.get_shape()[-1]]))\n    pop_mean = tf.Variable(tf.zeros([inputs.get_shape()[-1]]), trainable=False)\n    pop_var = tf.Variable(tf.ones([inputs.get_shape()[-1]]), trainable=False)\n\n    if is_training == True:\n        batch_mean, batch_var = tf.nn.moments(inputs,[0])\n        train_mean = tf.assign(pop_mean, pop_mean * decay + batch_mean * (1 - decay))\n        train_var = tf.assign(pop_var, pop_var * decay + batch_var * (1 - decay))\n        with tf.control_dependencies([train_mean, train_var]):\n            return tf.nn.batch_normalization(inputs,\n                batch_mean, batch_var, beta, scale, epsilon)\n    else:\n        return tf.nn.batch_normalization(inputs, pop_mean, pop_var, beta, scale, epsilon)\n```\n\n\n```python\nimport pickle\nimport os\nimport numpy as np\nimport matplotlib.pyplot as plt\nimport tensorflow as tf\n\n# CIFAR 10 data path\nDATA_PATH = \"./cifar-10-batches-py\"\nBATCH_SIZE = 50\nSTEPS = 500000\nepsilon = 1e-3\n\ndef one_hot(vec, vals=10):\n    n = len(vec)\n    out = np.zeros((n, vals))\n    out[range(n), vec] = 1\n    return out\n\n\ndef unpickle(file):\n    with open(os.path.join(DATA_PATH, file), 'rb') as fo:\n        u = pickle._Unpickler(fo)\n        u.encoding = 'latin1'\n        dict = u.load()\n    return dict\n\n\ndef display_cifar(images, size):\n    n = len(images)\n    plt.figure()\n    plt.gca().set_axis_off()\n    im = np.vstack([np.hstack([images[np.random.choice(n)] for i in range(size)])\n                    for i in range(size)])\n    plt.imshow(im)\n    plt.show()\n\n\nclass CifarLoader(object):\n    \"\"\"\n    Load and mange the CIFAR dataset.\n    (for any practical use there is no reason not to use the built-in dataset handler instead)\n    \"\"\"\n    def __init__(self, source_files):\n        self._source = source_files\n        self._i = 0\n        self.images = None\n        self.labels = None\n\n    def load(self):\n        data = [unpickle(f) for f in self._source]\n        images = np.vstack([d[\"data\"] for d in data])\n        n = len(images)\n        self.images = images.reshape(n, 3, 32, 32).transpose(0, 2, 3, 1).astype(float) / 255\n        self.labels = one_hot(np.hstack([d[\"labels\"] for d in data]), 10)\n        return self\n\n    def next_batch(self, batch_size):\n        x, y = self.images[self._i:self._i+batch_size], self.labels[self._i:self._i+batch_size]\n        self._i = (self._i + batch_size) % len(self.images)\n        return x, y\n\n    def random_batch(self, batch_size):\n        n = len(self.images)\n        ix = np.random.choice(n, batch_size)\n        return self.images[ix], self.labels[ix]\n\nclass CifarDataManager(object):\n    def __init__(self):\n        self.train = CifarLoader([\"data_batch_{}\".format(i) for i in range(1, 6)]).load()\n        self.test = CifarLoader([\"test_batch\"]).load()\n\n\ndef run_simple_net():\n    cifar = CifarDataManager()\n    x = tf.placeholder(tf.float32, shape=[None, 32, 32, 3])\n    y_ = tf.placeholder(tf.float32, shape=[None, 10])\n    keep_prob = tf.placeholder(tf.float32)\n    phase = tf.placeholder(tf.bool) \n\n    conv1 = conv_norm_layer(x, [5, 5, 3, 32], phase)\n    conv1_pool = max_pool_2x2(conv1)\n\n    conv2 = conv_norm_layer(conv1_pool, [5, 5, 32, 64], phase)\n    conv2_pool = max_pool_2x2(conv2)\n\n    conv3 = conv_norm_layer(conv2_pool, [5, 5, 64, 128], phase)\n    conv3_pool = max_pool_2x2(conv3)\n    conv3_flat = tf.reshape(conv3_pool, [-1, 4 * 4 * 128])\n    conv3_drop = tf.nn.dropout(conv3_flat, keep_prob=keep_prob)\n\n    full_1 = tf.nn.relu(full_layer(conv3_drop, 512))\n    full1_drop = tf.nn.dropout(full_1, keep_prob=keep_prob)\n\n    y_conv = full_layer(full1_drop, 10)\n\n    cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=y_conv, labels=y_))\n    train_step = tf.train.AdamOptimizer(1e-3).minimize(cross_entropy)\n\n    correct_prediction = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1))\n    accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))\n\n    def test(sess):\n        X = cifar.test.images.reshape(10, 1000, 32, 32, 3)\n        Y = cifar.test.labels.reshape(10, 1000, 10)\n        acc = np.mean([sess.run(accuracy, feed_dict={x: X[i], y_: Y[i], phase: False, keep_prob: 1.0})\n                       for i in range(10)])\n        print(\"Accuracy: {:.4}%\".format(acc * 100))\n\n    with tf.Session() as sess:\n        sess.run(tf.global_variables_initializer())\n\n        for i in range(STEPS):\n            batch = cifar.train.next_batch(BATCH_SIZE)\n            sess.run(train_step, feed_dict={x: batch[0], y_: batch[1], phase: True,  keep_prob: 0.5})\n\n            if i % 500 == 0:\n                test(sess)\n\n        test(sess)\n\n\ndef build_second_net():\n    cifar = CifarDataManager()\n    x = tf.placeholder(tf.float32, shape=[None, 32, 32, 3])\n    y_ = tf.placeholder(tf.float32, shape=[None, 10])\n    keep_prob = tf.placeholder(tf.float32)\n    phase = tf.placeholder(tf.bool) \n\n    C1, C2, C3 = 32, 64, 128\n    F1 = 600\n\n    conv1_1 = conv_norm_layer(x, [3, 3, 3, C1], phase)\n    conv1_2 = conv_norm_layer(conv1_1, [3, 3, C1, C1], phase)\n    conv1_3 = conv_norm_layer(conv1_2, [3, 3, C1, C1], phase)\n    conv1_pool = max_pool_2x2(conv1_3)\n    conv1_drop = tf.nn.dropout(conv1_pool, keep_prob=keep_prob)\n\n    conv2_1 = conv_norm_layer(conv1_drop, [3, 3, C1, C2], phase)\n    conv2_2 = conv_norm_layer(conv2_1, [3, 3, C2, C2], phase)\n    conv2_3 = conv_norm_layer(conv2_2, [3, 3, C2, C2], phase)\n    conv2_pool = max_pool_2x2(conv2_3)\n    conv2_drop = tf.nn.dropout(conv2_pool, keep_prob=keep_prob)\n\n    conv3_1 = conv_norm_layer(conv2_drop, [3, 3, C2, C3], phase)\n    conv3_2 = conv_norm_layer(conv3_1, [3, 3, C3, C3], phase)\n    conv3_3 = conv_norm_layer(conv3_2, [3, 3, C3, C3], phase)\n    conv3_pool = tf.nn.max_pool(conv3_3, ksize=[1, 8, 8, 1], strides=[1, 8, 8, 1], padding='SAME')\n    conv3_flat = tf.reshape(conv3_pool, [-1, C3])\n    conv3_drop = tf.nn.dropout(conv3_flat, keep_prob=keep_prob)\n\n    full1 = tf.nn.relu(full_layer(conv3_flat, F1))\n    full1_drop = tf.nn.dropout(full1, keep_prob=keep_prob)\n\n    y_conv = full_layer(full1_drop, 10)\n\n    cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=y_conv, labels=y_))\n    train_step = tf.train.AdamOptimizer(5e-4).minimize(cross_entropy)\n\n    correct_prediction = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1))\n    accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))\n\n    def test(sess):\n        X = cifar.test.images.reshape(10, 1000, 32, 32, 3)\n        Y = cifar.test.labels.reshape(10, 1000, 10)\n        acc = np.mean([sess.run(accuracy, feed_dict={x: X[i], y_: Y[i], phase: False , keep_prob: 1.0})\n                       for i in range(10)])\n        print(\"Accuracy: {:.4}%\".format(acc * 100))\n\n    with tf.Session() as sess:\n        sess.run(tf.global_variables_initializer())\n\n        for i in range(STEPS):\n            batch = cifar.train.next_batch(BATCH_SIZE)\n            sess.run(train_step, feed_dict={x: batch[0], y_: batch[1], phase: True, keep_prob: 0.5})\n\n            if i % 500 == 0:\n                test(sess)\n\n        test(sess)\n\n\ndef create_cifar_image():\n    d = CifarDataManager()\n    print(\"Number of train images: {}\".format(len(d.train.images)))\n    print(\"Number of train labels: {}\".format(len(d.train.labels)))\n    print(\"Number of test images: {}\".format(len(d.test.images)))\n    print(\"Number of test images: {}\".format(len(d.test.labels)))\n    images = d.train.images\n    display_cifar(images, 10)\n```\n\n\n```python\ncreate_cifar_image()\n```\n\n    Number of train images: 50000\n    Number of train labels: 50000\n    Number of test images: 10000\n    Number of test images: 10000\n\n\n\n![](https://cdn.steemitimages.com/DQmcEnsxs3tT7HneBPDSio2rBGGFPpVP4ax2gQDR87B1eGv/image.png)\n\n\n\n```python\nrun_simple_net()\n```\n\n    Accuracy: 11.14%\n    Accuracy: 43.33%\n    Accuracy: 47.72%\n    Accuracy: 49.42%\n    Accuracy: 55.05%\n    Accuracy: 58.3%\n    Accuracy: 61.28%\n    Accuracy: 57.95%\n    Accuracy: 65.16%\n    Accuracy: 63.02%\n    Accuracy: 67.24%\n    Accuracy: 63.56%\n    Accuracy: 67.67%\n    Accuracy: 65.49%\n    Accuracy: 69.49%\n    Accuracy: 69.16%\n    Accuracy: 69.66%\n    Accuracy: 70.72%\n    Accuracy: 71.93%\n    Accuracy: 69.34%\n    Accuracy: 70.39%\n    Accuracy: 71.15%\n    Accuracy: 72.94%\n    Accuracy: 71.53%\n    Accuracy: 73.49%\n    Accuracy: 71.79%\n    Accuracy: 71.93%\n    Accuracy: 71.12%\n    Accuracy: 73.47%\n    Accuracy: 72.49%\n    Accuracy: 74.03%\n    Accuracy: 70.01%\n    Accuracy: 74.21%\n    Accuracy: 72.42%\n    Accuracy: 73.63%\n    Accuracy: 72.65%\n    Accuracy: 74.18%\n    Accuracy: 73.05%\n    Accuracy: 74.97%\n    Accuracy: 74.02%\n    Accuracy: 73.46%\n    Accuracy: 73.5%\n    Accuracy: 74.29%\n    Accuracy: 73.23%\n    Accuracy: 75.01%\n    Accuracy: 73.64%\n    Accuracy: 74.37%\n    Accuracy: 75.41%\n    Accuracy: 75.28%\n    Accuracy: 75.45%\n    Accuracy: 75.35%\n    Accuracy: 75.24%\n    Accuracy: 75.59%\n    Accuracy: 75.92%\n    Accuracy: 75.43%\n    Accuracy: 75.02%\n    Accuracy: 75.24%\n    Accuracy: 75.87%\n    Accuracy: 75.38%\n    Accuracy: 76.03%\n    Accuracy: 75.42%\n    Accuracy: 75.64%\n    Accuracy: 75.83%\n    Accuracy: 75.49%\n    Accuracy: 75.36%\n    Accuracy: 75.14%\n    Accuracy: 75.94%\n    Accuracy: 75.73%\n    Accuracy: 76.33%\n    Accuracy: 75.53%\n    Accuracy: 75.77%\n    Accuracy: 75.8%\n    Accuracy: 75.4%\n    Accuracy: 76.26%\n    Accuracy: 75.9%\n    Accuracy: 73.55%\n    Accuracy: 75.29%\n    Accuracy: 76.37%\n    Accuracy: 76.49%\n    Accuracy: 75.92%\n    Accuracy: 75.24%\n    Accuracy: 76.35%\n    Accuracy: 75.96%\n    Accuracy: 76.49%\n    Accuracy: 76.67%\n    Accuracy: 76.46%\n    Accuracy: 75.63%\n    Accuracy: 76.93%\n    Accuracy: 75.73%\n    Accuracy: 76.22%\n    Accuracy: 74.97%\n    Accuracy: 75.99%\n    Accuracy: 76.77%\n    Accuracy: 76.16%\n    Accuracy: 75.58%\n    Accuracy: 77.01%\n    Accuracy: 76.35%\n    Accuracy: 76.84%\n    Accuracy: 76.02%\n    Accuracy: 76.27%\n    Accuracy: 76.15%\n    Accuracy: 77.02%\n    Accuracy: 75.18%\n    Accuracy: 76.91%\n    Accuracy: 76.39%\n    Accuracy: 76.95%\n    Accuracy: 75.67%\n    Accuracy: 76.51%\n    Accuracy: 76.46%\n    Accuracy: 76.47%\n    Accuracy: 76.34%\n    Accuracy: 76.74%\n    Accuracy: 76.62%\n    Accuracy: 76.04%\n    Accuracy: 76.66%\n    Accuracy: 76.6%\n    Accuracy: 76.47%\n    Accuracy: 76.9%\n    Accuracy: 76.29%\n    Accuracy: 76.89%\n    Accuracy: 76.45%\n    Accuracy: 77.04%\n    Accuracy: 77.04%\n    Accuracy: 77.44%\n    Accuracy: 75.54%\n    Accuracy: 77.28%\n    Accuracy: 76.51%\n    Accuracy: 77.3%\n    Accuracy: 76.69%\n    Accuracy: 77.1%\n    Accuracy: 76.69%\n    Accuracy: 77.18%\n    Accuracy: 76.99%\n    Accuracy: 77.11%\n    Accuracy: 76.37%\n    Accuracy: 77.45%\n    Accuracy: 76.04%\n    Accuracy: 77.34%\n    Accuracy: 76.5%\n    Accuracy: 76.94%\n    Accuracy: 76.32%\n    Accuracy: 77.07%\n    Accuracy: 76.9%\n    Accuracy: 77.57%\n    Accuracy: 77.33%\n    Accuracy: 77.48%\n    Accuracy: 77.0%\n    Accuracy: 77.14%\n    Accuracy: 77.25%\n    Accuracy: 77.75%\n    Accuracy: 76.36%\n    Accuracy: 77.1%\n    Accuracy: 76.81%\n    Accuracy: 76.96%\n    Accuracy: 76.85%\n    Accuracy: 77.19%\n    Accuracy: 77.33%\n    Accuracy: 76.7%\n    Accuracy: 76.71%\n    Accuracy: 77.5%\n    Accuracy: 77.19%\n    Accuracy: 77.37%\n    Accuracy: 76.71%\n    Accuracy: 77.27%\n    Accuracy: 76.44%\n    Accuracy: 77.3%\n    Accuracy: 76.75%\n    Accuracy: 76.26%\n    Accuracy: 76.87%\n    Accuracy: 77.38%\n    Accuracy: 77.18%\n    Accuracy: 77.28%\n    Accuracy: 76.74%\n    Accuracy: 77.01%\n    Accuracy: 76.84%\n    Accuracy: 76.87%\n    Accuracy: 77.38%\n    Accuracy: 77.01%\n    Accuracy: 77.42%\n    Accuracy: 77.11%\n    Accuracy: 76.76%\n    Accuracy: 77.19%\n    Accuracy: 76.77%\n    Accuracy: 77.01%\n    Accuracy: 76.07%\n    Accuracy: 76.98%\n    Accuracy: 77.22%\n    Accuracy: 77.37%\n    Accuracy: 76.72%\n    Accuracy: 77.0%\n    Accuracy: 77.46%\n    Accuracy: 77.38%\n    Accuracy: 76.91%\n    Accuracy: 77.34%\n    Accuracy: 77.33%\n    Accuracy: 76.82%\n    Accuracy: 76.82%\n    Accuracy: 77.2%\n    Accuracy: 76.29%\n    Accuracy: 77.04%\n    Accuracy: 76.87%\n    Accuracy: 77.89%\n    Accuracy: 76.4%\n    Accuracy: 77.35%\n    Accuracy: 75.91%\n    Accuracy: 77.28%\n    Accuracy: 76.6%\n    Accuracy: 77.57%\n    Accuracy: 77.42%\n    Accuracy: 77.21%\n    Accuracy: 77.37%\n    Accuracy: 76.98%\n    Accuracy: 77.08%\n    Accuracy: 77.56%\n    Accuracy: 77.3%\n    Accuracy: 77.62%\n    Accuracy: 77.05%\n    Accuracy: 77.71%\n    Accuracy: 77.2%\n    Accuracy: 77.39%\n    Accuracy: 77.51%\n    Accuracy: 76.78%\n    Accuracy: 77.15%\n    Accuracy: 77.02%\n    Accuracy: 77.48%\n    Accuracy: 77.4%\n    Accuracy: 76.12%\n    Accuracy: 77.32%\n    Accuracy: 75.95%\n    Accuracy: 76.92%\n    Accuracy: 76.84%\n    Accuracy: 76.95%\n    Accuracy: 76.45%\n    Accuracy: 76.9%\n    Accuracy: 77.29%\n    Accuracy: 77.23%\n    Accuracy: 76.67%\n    Accuracy: 77.18%\n    Accuracy: 76.35%\n    Accuracy: 77.66%\n    Accuracy: 77.03%\n    Accuracy: 77.06%\n    Accuracy: 77.15%\n    Accuracy: 77.92%\n    Accuracy: 76.58%\n    Accuracy: 77.41%\n    Accuracy: 77.02%\n    Accuracy: 77.51%\n    Accuracy: 76.09%\n    Accuracy: 77.84%\n    Accuracy: 76.7%\n    Accuracy: 77.56%\n    Accuracy: 77.17%\n    Accuracy: 77.19%\n    Accuracy: 77.46%\n    Accuracy: 77.01%\n    Accuracy: 77.58%\n    Accuracy: 77.48%\n    Accuracy: 77.78%\n    Accuracy: 77.03%\n    Accuracy: 76.44%\n    Accuracy: 77.08%\n    Accuracy: 77.07%\n    Accuracy: 77.61%\n    Accuracy: 77.09%\n    Accuracy: 77.76%\n    Accuracy: 76.98%\n    Accuracy: 77.39%\n    Accuracy: 76.83%\n    Accuracy: 77.11%\n    Accuracy: 76.37%\n    Accuracy: 76.82%\n    Accuracy: 77.36%\n    Accuracy: 77.73%\n    Accuracy: 76.12%\n    Accuracy: 77.52%\n    Accuracy: 76.76%\n    Accuracy: 77.23%\n    Accuracy: 77.2%\n    Accuracy: 77.39%\n    Accuracy: 76.92%\n    Accuracy: 77.51%\n    Accuracy: 77.15%\n    Accuracy: 77.75%\n    Accuracy: 77.15%\n    Accuracy: 77.25%\n    Accuracy: 77.18%\n    Accuracy: 77.07%\n    Accuracy: 77.16%\n    Accuracy: 76.77%\n    Accuracy: 76.06%\n    Accuracy: 77.52%\n    Accuracy: 77.36%\n    Accuracy: 77.24%\n    Accuracy: 77.42%\n    Accuracy: 77.7%\n    Accuracy: 76.77%\n    Accuracy: 77.25%\n    Accuracy: 76.95%\n    Accuracy: 76.48%\n    Accuracy: 76.45%\n    Accuracy: 77.2%\n    Accuracy: 76.58%\n    Accuracy: 77.43%\n    Accuracy: 75.77%\n    Accuracy: 76.97%\n    Accuracy: 77.11%\n    Accuracy: 78.03%\n    Accuracy: 76.67%\n    Accuracy: 76.86%\n    Accuracy: 76.88%\n    Accuracy: 77.09%\n    Accuracy: 77.69%\n    Accuracy: 77.47%\n    Accuracy: 77.01%\n    Accuracy: 77.6%\n    Accuracy: 77.46%\n    Accuracy: 77.03%\n    Accuracy: 77.32%\n    Accuracy: 77.34%\n    Accuracy: 77.13%\n    Accuracy: 77.71%\n    Accuracy: 77.46%\n    Accuracy: 77.29%\n    Accuracy: 76.58%\n    Accuracy: 78.04%\n    Accuracy: 77.09%\n    Accuracy: 77.65%\n    Accuracy: 77.26%\n    Accuracy: 77.55%\n    Accuracy: 77.89%\n    Accuracy: 77.48%\n    Accuracy: 76.85%\n    Accuracy: 76.91%\n    Accuracy: 77.39%\n    Accuracy: 77.59%\n    Accuracy: 77.11%\n    Accuracy: 77.29%\n    Accuracy: 77.9%\n    Accuracy: 77.54%\n    Accuracy: 76.59%\n    Accuracy: 77.66%\n    Accuracy: 77.07%\n    Accuracy: 77.03%\n    Accuracy: 76.92%\n    Accuracy: 77.39%\n    Accuracy: 77.16%\n    Accuracy: 77.39%\n    Accuracy: 76.8%\n    Accuracy: 77.3%\n    Accuracy: 77.65%\n    Accuracy: 77.41%\n    Accuracy: 77.11%\n    Accuracy: 78.0%\n    Accuracy: 77.34%\n    Accuracy: 77.9%\n    Accuracy: 76.84%\n    Accuracy: 77.41%\n    Accuracy: 77.19%\n    Accuracy: 77.61%\n    Accuracy: 77.16%\n    Accuracy: 77.89%\n    Accuracy: 77.14%\n    Accuracy: 77.59%\n    Accuracy: 76.52%\n    Accuracy: 77.85%\n    Accuracy: 77.05%\n    Accuracy: 77.56%\n    Accuracy: 77.59%\n    Accuracy: 76.75%\n    Accuracy: 77.67%\n    Accuracy: 77.74%\n    Accuracy: 76.49%\n    Accuracy: 77.43%\n    Accuracy: 77.49%\n    Accuracy: 76.95%\n    Accuracy: 77.5%\n    Accuracy: 77.31%\n    Accuracy: 77.06%\n    Accuracy: 77.73%\n    Accuracy: 77.42%\n    Accuracy: 77.35%\n    Accuracy: 77.2%\n    Accuracy: 77.65%\n    Accuracy: 77.14%\n    Accuracy: 77.02%\n    Accuracy: 77.09%\n    Accuracy: 77.31%\n    Accuracy: 77.54%\n    Accuracy: 77.37%\n    Accuracy: 77.19%\n    Accuracy: 77.58%\n    Accuracy: 77.18%\n    Accuracy: 78.04%\n    Accuracy: 77.18%\n    Accuracy: 78.09%\n    Accuracy: 76.67%\n    Accuracy: 78.05%\n    Accuracy: 77.47%\n    Accuracy: 77.51%\n    Accuracy: 77.78%\n    Accuracy: 76.92%\n    Accuracy: 77.21%\n    Accuracy: 77.65%\n    Accuracy: 77.1%\n    Accuracy: 78.08%\n    Accuracy: 77.36%\n    Accuracy: 77.07%\n    Accuracy: 77.34%\n    Accuracy: 77.86%\n    Accuracy: 76.9%\n    Accuracy: 77.7%\n    Accuracy: 77.91%\n    Accuracy: 77.1%\n    Accuracy: 77.32%\n    Accuracy: 77.53%\n    Accuracy: 77.59%\n    Accuracy: 77.22%\n    Accuracy: 77.32%\n    Accuracy: 77.48%\n    Accuracy: 77.35%\n    Accuracy: 77.29%\n    Accuracy: 77.45%\n    Accuracy: 77.5%\n    Accuracy: 76.78%\n    Accuracy: 77.8%\n    Accuracy: 77.14%\n    Accuracy: 77.39%\n    Accuracy: 76.39%\n    Accuracy: 77.81%\n    Accuracy: 76.83%\n    Accuracy: 77.12%\n    Accuracy: 76.96%\n    Accuracy: 77.72%\n    Accuracy: 77.18%\n    Accuracy: 77.66%\n    Accuracy: 77.59%\n    Accuracy: 77.98%\n    Accuracy: 76.89%\n    Accuracy: 77.33%\n    Accuracy: 76.9%\n    Accuracy: 77.57%\n    Accuracy: 77.87%\n    Accuracy: 77.16%\n    Accuracy: 77.58%\n    Accuracy: 78.44%\n    Accuracy: 77.33%\n    Accuracy: 77.4%\n    Accuracy: 77.49%\n    Accuracy: 77.69%\n    Accuracy: 76.34%\n    Accuracy: 77.7%\n    Accuracy: 77.42%\n    Accuracy: 77.62%\n    Accuracy: 77.3%\n    Accuracy: 77.13%\n    Accuracy: 77.11%\n    Accuracy: 77.97%\n    Accuracy: 77.76%\n    Accuracy: 78.15%\n    Accuracy: 77.69%\n    Accuracy: 78.06%\n    Accuracy: 77.78%\n    Accuracy: 77.52%\n    Accuracy: 77.64%\n    Accuracy: 77.16%\n    Accuracy: 77.25%\n    Accuracy: 77.95%\n    Accuracy: 76.6%\n    Accuracy: 77.89%\n    Accuracy: 77.38%\n    Accuracy: 77.46%\n    Accuracy: 77.68%\n    Accuracy: 76.93%\n    Accuracy: 77.07%\n    Accuracy: 77.93%\n    Accuracy: 78.11%\n    Accuracy: 77.95%\n    Accuracy: 78.18%\n    Accuracy: 77.4%\n    Accuracy: 76.7%\n    Accuracy: 77.65%\n    Accuracy: 77.51%\n    Accuracy: 77.04%\n    Accuracy: 77.11%\n    Accuracy: 77.69%\n    Accuracy: 77.25%\n    Accuracy: 77.3%\n    Accuracy: 77.31%\n    Accuracy: 77.53%\n    Accuracy: 77.56%\n    Accuracy: 77.13%\n    Accuracy: 77.47%\n    Accuracy: 77.63%\n    Accuracy: 76.85%\n    Accuracy: 77.42%\n    Accuracy: 77.49%\n    Accuracy: 76.93%\n    Accuracy: 76.89%\n    Accuracy: 77.7%\n    Accuracy: 77.32%\n    Accuracy: 77.73%\n    Accuracy: 77.42%\n    Accuracy: 77.5%\n    Accuracy: 76.96%\n    Accuracy: 77.39%\n    Accuracy: 77.46%\n    Accuracy: 77.9%\n    Accuracy: 76.7%\n    Accuracy: 77.43%\n    Accuracy: 77.5%\n    Accuracy: 77.71%\n    Accuracy: 77.63%\n    Accuracy: 77.65%\n    Accuracy: 78.0%\n    Accuracy: 77.16%\n    Accuracy: 76.95%\n    Accuracy: 77.47%\n    Accuracy: 77.5%\n    Accuracy: 77.64%\n    Accuracy: 77.96%\n    Accuracy: 77.29%\n    Accuracy: 77.24%\n    Accuracy: 77.24%\n    Accuracy: 77.33%\n    Accuracy: 77.23%\n    Accuracy: 77.2%\n    Accuracy: 77.46%\n    Accuracy: 76.82%\n    Accuracy: 77.33%\n    Accuracy: 77.74%\n    Accuracy: 77.97%\n    Accuracy: 77.0%\n    Accuracy: 77.36%\n    Accuracy: 77.59%\n    Accuracy: 77.24%\n    Accuracy: 76.67%\n    Accuracy: 77.55%\n    Accuracy: 76.98%\n    Accuracy: 77.41%\n    Accuracy: 77.08%\n    Accuracy: 77.37%\n    Accuracy: 77.31%\n    Accuracy: 77.52%\n    Accuracy: 77.87%\n    Accuracy: 77.74%\n    Accuracy: 77.74%\n    Accuracy: 76.83%\n    Accuracy: 77.23%\n    Accuracy: 77.24%\n    Accuracy: 77.24%\n    Accuracy: 77.49%\n    Accuracy: 77.84%\n    Accuracy: 77.62%\n    Accuracy: 77.79%\n    Accuracy: 77.67%\n    Accuracy: 77.63%\n    Accuracy: 77.74%\n    Accuracy: 77.43%\n    Accuracy: 77.78%\n    Accuracy: 76.34%\n    Accuracy: 77.73%\n    Accuracy: 78.0%\n    Accuracy: 77.45%\n    Accuracy: 77.62%\n    Accuracy: 77.67%\n    Accuracy: 77.4%\n    Accuracy: 77.54%\n    Accuracy: 77.54%\n    Accuracy: 77.48%\n    Accuracy: 77.63%\n    Accuracy: 77.78%\n    Accuracy: 78.03%\n    Accuracy: 77.6%\n    Accuracy: 78.3%\n    Accuracy: 77.76%\n    Accuracy: 77.96%\n    Accuracy: 77.94%\n    Accuracy: 77.45%\n    Accuracy: 76.43%\n    Accuracy: 76.64%\n    Accuracy: 77.86%\n    Accuracy: 77.85%\n    Accuracy: 77.75%\n    Accuracy: 77.34%\n    Accuracy: 78.06%\n    Accuracy: 77.49%\n    Accuracy: 77.66%\n    Accuracy: 77.88%\n    Accuracy: 77.55%\n    Accuracy: 77.12%\n    Accuracy: 77.54%\n    Accuracy: 77.39%\n    Accuracy: 77.59%\n    Accuracy: 77.45%\n    Accuracy: 77.52%\n    Accuracy: 77.53%\n    Accuracy: 77.92%\n    Accuracy: 77.19%\n    Accuracy: 77.87%\n    Accuracy: 77.18%\n    Accuracy: 77.33%\n    Accuracy: 77.33%\n    Accuracy: 77.63%\n    Accuracy: 77.48%\n    Accuracy: 77.5%\n    Accuracy: 77.23%\n    Accuracy: 77.38%\n    Accuracy: 77.32%\n    Accuracy: 77.09%\n    Accuracy: 76.87%\n    Accuracy: 77.57%\n    Accuracy: 77.57%\n    Accuracy: 77.74%\n    Accuracy: 77.59%\n    Accuracy: 77.56%\n    Accuracy: 77.06%\n    Accuracy: 77.24%\n    Accuracy: 77.34%\n    Accuracy: 77.25%\n    Accuracy: 77.43%\n    Accuracy: 77.53%\n    Accuracy: 77.22%\n    Accuracy: 77.34%\n    Accuracy: 77.21%\n    Accuracy: 78.01%\n    Accuracy: 77.44%\n    Accuracy: 77.85%\n    Accuracy: 77.48%\n    Accuracy: 77.34%\n    Accuracy: 77.66%\n    Accuracy: 77.62%\n    Accuracy: 77.21%\n    Accuracy: 77.37%\n    Accuracy: 77.47%\n    Accuracy: 77.63%\n    Accuracy: 77.52%\n    Accuracy: 77.89%\n    Accuracy: 77.95%\n    Accuracy: 77.79%\n    Accuracy: 76.77%\n    Accuracy: 78.01%\n    Accuracy: 77.15%\n    Accuracy: 77.62%\n    Accuracy: 77.27%\n    Accuracy: 77.32%\n    Accuracy: 78.23%\n    Accuracy: 77.61%\n    Accuracy: 77.86%\n    Accuracy: 77.8%\n    Accuracy: 77.27%\n    Accuracy: 77.81%\n    Accuracy: 77.49%\n    Accuracy: 77.79%\n    Accuracy: 78.23%\n    Accuracy: 77.98%\n    Accuracy: 77.45%\n    Accuracy: 77.45%\n    Accuracy: 77.6%\n    Accuracy: 77.85%\n    Accuracy: 77.66%\n    Accuracy: 77.15%\n    Accuracy: 77.13%\n    Accuracy: 78.27%\n    Accuracy: 76.74%\n    Accuracy: 78.08%\n    Accuracy: 77.38%\n    Accuracy: 77.73%\n    Accuracy: 77.07%\n    Accuracy: 77.63%\n    Accuracy: 77.45%\n    Accuracy: 77.76%\n    Accuracy: 77.12%\n    Accuracy: 77.38%\n    Accuracy: 77.42%\n    Accuracy: 77.8%\n    Accuracy: 77.32%\n    Accuracy: 77.75%\n    Accuracy: 77.72%\n    Accuracy: 76.99%\n    Accuracy: 77.74%\n    Accuracy: 77.98%\n    Accuracy: 77.52%\n    Accuracy: 77.87%\n    Accuracy: 76.76%\n    Accuracy: 77.63%\n    Accuracy: 77.49%\n    Accuracy: 78.15%\n    Accuracy: 77.35%\n    Accuracy: 77.84%\n    Accuracy: 77.72%\n    Accuracy: 76.92%\n    Accuracy: 77.85%\n    Accuracy: 77.91%\n    Accuracy: 77.99%\n    Accuracy: 77.22%\n    Accuracy: 77.01%\n    Accuracy: 77.61%\n    Accuracy: 77.7%\n    Accuracy: 77.95%\n    Accuracy: 77.74%\n    Accuracy: 77.69%\n    Accuracy: 77.76%\n    Accuracy: 78.15%\n    Accuracy: 77.49%\n    Accuracy: 77.87%\n    Accuracy: 77.28%\n    Accuracy: 77.3%\n    Accuracy: 77.01%\n    Accuracy: 77.65%\n    Accuracy: 77.64%\n    Accuracy: 76.71%\n    Accuracy: 77.65%\n    Accuracy: 78.07%\n    Accuracy: 77.83%\n    Accuracy: 77.82%\n    Accuracy: 77.06%\n    Accuracy: 77.25%\n    Accuracy: 77.16%\n    Accuracy: 78.02%\n    Accuracy: 77.04%\n    Accuracy: 77.97%\n    Accuracy: 77.53%\n    Accuracy: 77.46%\n    Accuracy: 76.77%\n    Accuracy: 77.77%\n    Accuracy: 77.54%\n    Accuracy: 77.95%\n    Accuracy: 77.5%\n    Accuracy: 78.18%\n    Accuracy: 76.88%\n    Accuracy: 77.97%\n    Accuracy: 77.79%\n    Accuracy: 77.46%\n    Accuracy: 77.78%\n    Accuracy: 77.3%\n    Accuracy: 77.1%\n    Accuracy: 78.01%\n    Accuracy: 77.86%\n    Accuracy: 77.83%\n    Accuracy: 77.44%\n    Accuracy: 77.8%\n    Accuracy: 77.5%\n    Accuracy: 77.42%\n    Accuracy: 77.8%\n    Accuracy: 77.14%\n    Accuracy: 77.91%\n    Accuracy: 77.27%\n    Accuracy: 77.9%\n    Accuracy: 78.32%\n    Accuracy: 77.5%\n    Accuracy: 77.8%\n    Accuracy: 77.67%\n    Accuracy: 77.43%\n    Accuracy: 78.0%\n    Accuracy: 78.14%\n    Accuracy: 77.28%\n    Accuracy: 77.75%\n    Accuracy: 77.62%\n    Accuracy: 77.64%\n    Accuracy: 77.38%\n    Accuracy: 77.97%\n    Accuracy: 77.72%\n    Accuracy: 77.84%\n    Accuracy: 77.61%\n    Accuracy: 77.76%\n    Accuracy: 77.27%\n    Accuracy: 77.45%\n    Accuracy: 77.36%\n    Accuracy: 77.78%\n    Accuracy: 76.75%\n    Accuracy: 77.53%\n    Accuracy: 77.4%\n    Accuracy: 77.8%\n    Accuracy: 77.69%\n    Accuracy: 77.7%\n    Accuracy: 77.76%\n    Accuracy: 77.75%\n    Accuracy: 77.13%\n    Accuracy: 77.66%\n    Accuracy: 77.35%\n    Accuracy: 77.69%\n    Accuracy: 77.44%\n    Accuracy: 77.94%\n    Accuracy: 77.53%\n    Accuracy: 78.04%\n    Accuracy: 77.87%\n    Accuracy: 77.48%\n    Accuracy: 76.9%\n    Accuracy: 77.6%\n    Accuracy: 77.72%\n    Accuracy: 77.47%\n    Accuracy: 77.63%\n    Accuracy: 77.75%\n    Accuracy: 77.31%\n    Accuracy: 77.37%\n    Accuracy: 77.96%\n    Accuracy: 78.04%\n    Accuracy: 77.54%\n    Accuracy: 77.46%\n    Accuracy: 77.84%\n    Accuracy: 77.73%\n    Accuracy: 77.0%\n    Accuracy: 77.22%\n    Accuracy: 77.55%\n    Accuracy: 77.33%\n    Accuracy: 77.57%\n    Accuracy: 77.71%\n    Accuracy: 77.35%\n    Accuracy: 77.46%\n    Accuracy: 77.11%\n    Accuracy: 77.42%\n    Accuracy: 77.25%\n    Accuracy: 77.22%\n    Accuracy: 77.34%\n    Accuracy: 77.83%\n    Accuracy: 77.64%\n    Accuracy: 77.52%\n    Accuracy: 77.02%\n    Accuracy: 77.55%\n    Accuracy: 77.49%\n    Accuracy: 77.35%\n    Accuracy: 77.51%\n    Accuracy: 77.84%\n    Accuracy: 77.5%\n    Accuracy: 76.97%\n    Accuracy: 77.14%\n    Accuracy: 77.54%\n    Accuracy: 77.03%\n    Accuracy: 77.53%\n    Accuracy: 77.51%\n    Accuracy: 77.51%\n    Accuracy: 77.44%\n    Accuracy: 77.94%\n    Accuracy: 76.65%\n    Accuracy: 77.81%\n    Accuracy: 77.66%\n    Accuracy: 77.63%\n    Accuracy: 77.76%\n    Accuracy: 77.52%\n    Accuracy: 77.42%\n    Accuracy: 77.25%\n    Accuracy: 77.52%\n    Accuracy: 77.52%\n    Accuracy: 76.16%\n    Accuracy: 77.63%\n    Accuracy: 77.9%\n    Accuracy: 77.75%\n    Accuracy: 77.22%\n    Accuracy: 77.81%\n    Accuracy: 77.2%\n    Accuracy: 77.59%\n    Accuracy: 77.21%\n    Accuracy: 77.56%\n    Accuracy: 77.54%\n    Accuracy: 77.53%\n    Accuracy: 77.76%\n    Accuracy: 77.5%\n    Accuracy: 77.12%\n    Accuracy: 77.44%\n    Accuracy: 77.56%\n    Accuracy: 77.06%\n    Accuracy: 77.66%\n    Accuracy: 77.96%\n    Accuracy: 78.33%\n    Accuracy: 77.6%\n    Accuracy: 77.42%\n    Accuracy: 77.45%\n    Accuracy: 77.23%\n    Accuracy: 78.16%\n    Accuracy: 77.67%\n    Accuracy: 77.1%\n    Accuracy: 77.38%\n    Accuracy: 77.6%\n    Accuracy: 76.9%\n    Accuracy: 77.92%\n    Accuracy: 77.45%\n    Accuracy: 77.5%\n    Accuracy: 77.6%\n    Accuracy: 77.9%\n    Accuracy: 77.02%\n    Accuracy: 78.29%\n    Accuracy: 77.17%\n    Accuracy: 77.46%\n    Accuracy: 78.07%\n    Accuracy: 77.45%\n    Accuracy: 77.21%\n    Accuracy: 78.1%\n    Accuracy: 77.41%\n    Accuracy: 77.46%\n    Accuracy: 77.96%\n    Accuracy: 77.42%\n    Accuracy: 77.56%\n    Accuracy: 77.35%\n    Accuracy: 77.85%\n    Accuracy: 77.62%\n    Accuracy: 77.67%\n    Accuracy: 77.11%\n    Accuracy: 77.43%\n    Accuracy: 77.8%\n    Accuracy: 77.5%\n    Accuracy: 77.04%\n    Accuracy: 76.92%\n    Accuracy: 78.38%\n    Accuracy: 77.77%\n    Accuracy: 77.29%\n    Accuracy: 76.72%\n    Accuracy: 77.91%\n    Accuracy: 77.65%\n    Accuracy: 77.46%\n    Accuracy: 77.88%\n    Accuracy: 77.5%\n    Accuracy: 77.75%\n    Accuracy: 77.43%\n    Accuracy: 77.58%\n    Accuracy: 77.49%\n    Accuracy: 77.82%\n    Accuracy: 77.93%\n    Accuracy: 77.62%\n    Accuracy: 77.89%\n    Accuracy: 78.26%\n    Accuracy: 77.87%\n    Accuracy: 76.91%\n    Accuracy: 77.01%\n    Accuracy: 77.6%\n    Accuracy: 77.69%\n    Accuracy: 77.59%\n    Accuracy: 77.82%\n    Accuracy: 76.75%\n    Accuracy: 77.38%\n    Accuracy: 77.57%\n    Accuracy: 77.48%\n    Accuracy: 77.45%\n    Accuracy: 77.36%\n    Accuracy: 77.85%\n    Accuracy: 77.57%\n    Accuracy: 77.46%\n    Accuracy: 77.52%\n    Accuracy: 77.29%\n    Accuracy: 77.32%\n    Accuracy: 76.74%\n    Accuracy: 77.56%\n    Accuracy: 77.37%\n    Accuracy: 76.91%\n    Accuracy: 77.07%\n    Accuracy: 77.96%\n    Accuracy: 77.53%\n    Accuracy: 77.91%\n    Accuracy: 76.82%\n    Accuracy: 77.88%\n    Accuracy: 77.34%\n    Accuracy: 77.86%\n    Accuracy: 77.51%\n    Accuracy: 78.1%\n    Accuracy: 78.08%\n    Accuracy: 77.31%\n    Accuracy: 77.59%\n    Accuracy: 78.13%\n    Accuracy: 77.11%\n    Accuracy: 77.72%\n    Accuracy: 77.77%\n    Accuracy: 78.05%\n    Accuracy: 77.2%\n    Accuracy: 78.22%\n    Accuracy: 77.1%\n    Accuracy: 78.29%\n    Accuracy: 77.58%\n    Accuracy: 78.13%\n    Accuracy: 78.06%\n    Accuracy: 77.31%\n    Accuracy: 77.94%\n    Accuracy: 77.02%\n    Accuracy: 77.83%\n    Accuracy: 77.94%\n    Accuracy: 77.62%\n    Accuracy: 77.7%\n    Accuracy: 76.98%\n    Accuracy: 77.56%\n    Accuracy: 77.5%\n    Accuracy: 77.56%\n    Accuracy: 77.56%\n    Accuracy: 77.57%\n    Accuracy: 77.74%\n    Accuracy: 77.51%\n    Accuracy: 77.67%\n    Accuracy: 77.78%\n    Accuracy: 77.99%\n    Accuracy: 78.0%\n    Accuracy: 77.78%\n    Accuracy: 77.62%\n    Accuracy: 77.94%\n    Accuracy: 78.16%\n    Accuracy: 77.6%\n    Accuracy: 77.84%\n    Accuracy: 77.34%\n    Accuracy: 77.91%\n    Accuracy: 77.12%\n    Accuracy: 78.07%\n    Accuracy: 77.25%\n    Accuracy: 78.01%\n    Accuracy: 77.71%\n\n\n\n```python\nbuild_second_net()\n```\n\n    Accuracy: 9.82%\n    Accuracy: 10.52%\n    Accuracy: 22.65%\n    Accuracy: 29.57%\n    Accuracy: 33.35%\n    Accuracy: 38.66%\n    Accuracy: 40.16%\n    Accuracy: 44.4%\n    Accuracy: 46.93%\n    Accuracy: 50.96%\n    Accuracy: 52.52%\n    Accuracy: 53.33%\n    Accuracy: 53.51%\n    Accuracy: 58.63%\n    Accuracy: 57.96%\n    Accuracy: 61.02%\n    Accuracy: 60.62%\n    Accuracy: 62.94%\n    Accuracy: 61.72%\n    Accuracy: 62.72%\n    Accuracy: 64.35%\n    Accuracy: 66.08%\n    Accuracy: 65.25%\n    Accuracy: 65.11%\n    Accuracy: 62.51%\n    Accuracy: 65.58%\n    Accuracy: 67.31%\n    Accuracy: 66.46%\n    Accuracy: 67.74%\n    Accuracy: 67.92%\n    Accuracy: 65.42%\n    Accuracy: 67.59%\n    Accuracy: 69.62%\n    Accuracy: 67.41%\n    Accuracy: 69.89%\n    Accuracy: 71.57%\n    Accuracy: 71.89%\n    Accuracy: 69.72%\n    Accuracy: 70.15%\n    Accuracy: 71.4%\n    Accuracy: 72.22%\n    Accuracy: 71.53%\n    Accuracy: 74.22%\n    Accuracy: 73.18%\n    Accuracy: 72.06%\n    Accuracy: 74.23%\n    Accuracy: 72.76%\n    Accuracy: 74.38%\n    Accuracy: 74.93%\n    Accuracy: 76.22%\n    Accuracy: 75.47%\n    Accuracy: 74.69%\n    Accuracy: 75.1%\n    Accuracy: 75.57%\n    Accuracy: 77.64%\n    Accuracy: 77.76%\n    Accuracy: 76.48%\n    Accuracy: 77.21%\n    Accuracy: 77.57%\n    Accuracy: 76.31%\n    Accuracy: 76.89%\n    Accuracy: 77.87%\n    Accuracy: 77.57%\n    Accuracy: 77.72%\n    Accuracy: 78.21%\n    Accuracy: 78.57%\n    Accuracy: 77.19%\n    Accuracy: 78.65%\n    Accuracy: 78.48%\n    Accuracy: 77.19%\n    Accuracy: 77.36%\n    Accuracy: 78.16%\n    Accuracy: 79.17%\n    Accuracy: 78.99%\n    Accuracy: 78.95%\n    Accuracy: 79.07%\n    Accuracy: 77.81%\n    Accuracy: 78.78%\n    Accuracy: 79.13%\n    Accuracy: 76.86%\n    Accuracy: 79.89%\n    Accuracy: 79.15%\n    Accuracy: 78.51%\n    Accuracy: 79.15%\n    Accuracy: 78.59%\n    Accuracy: 79.0%\n    Accuracy: 78.63%\n    Accuracy: 80.26%\n    Accuracy: 79.98%\n    Accuracy: 79.85%\n    Accuracy: 79.95%\n    Accuracy: 79.66%\n    Accuracy: 78.83%\n    Accuracy: 78.91%\n    Accuracy: 79.96%\n    Accuracy: 79.58%\n    Accuracy: 79.32%\n    Accuracy: 78.88%\n    Accuracy: 80.2%\n    Accuracy: 80.27%\n    Accuracy: 78.93%\n    Accuracy: 79.94%\n    Accuracy: 79.71%\n    Accuracy: 80.42%\n    Accuracy: 77.77%\n    Accuracy: 79.5%\n    Accuracy: 80.66%\n    Accuracy: 80.65%\n    Accuracy: 80.31%\n    Accuracy: 80.28%\n    Accuracy: 79.56%\n    Accuracy: 79.57%\n    Accuracy: 80.37%\n    Accuracy: 80.11%\n    Accuracy: 80.32%\n    Accuracy: 81.32%\n    Accuracy: 79.94%\n    Accuracy: 80.9%\n    Accuracy: 80.59%\n    Accuracy: 80.71%\n    Accuracy: 81.48%\n    Accuracy: 80.06%\n    Accuracy: 80.6%\n    Accuracy: 80.98%\n    Accuracy: 80.32%\n    Accuracy: 79.55%\n    Accuracy: 80.86%\n    Accuracy: 80.06%\n    Accuracy: 80.66%\n    Accuracy: 80.34%\n    Accuracy: 79.55%\n    Accuracy: 81.42%\n    Accuracy: 81.39%\n    Accuracy: 81.13%\n    Accuracy: 81.21%\n    Accuracy: 82.0%\n    Accuracy: 81.5%\n    Accuracy: 80.27%\n    Accuracy: 80.35%\n    Accuracy: 79.69%\n    Accuracy: 80.9%\n    Accuracy: 80.4%\n    Accuracy: 80.59%\n    Accuracy: 80.36%\n    Accuracy: 80.93%\n    Accuracy: 80.71%\n    Accuracy: 79.7%\n    Accuracy: 80.9%\n    Accuracy: 80.21%\n    Accuracy: 79.62%\n    Accuracy: 81.7%\n    Accuracy: 78.51%\n    Accuracy: 79.92%\n    Accuracy: 81.33%\n    Accuracy: 78.73%\n    Accuracy: 81.65%\n    Accuracy: 81.22%\n    Accuracy: 80.86%\n    Accuracy: 81.08%\n    Accuracy: 80.33%\n    Accuracy: 80.21%\n    Accuracy: 80.43%\n    Accuracy: 81.08%\n    Accuracy: 80.37%\n    Accuracy: 81.82%\n    Accuracy: 80.59%\n    Accuracy: 81.67%\n    Accuracy: 81.27%\n    Accuracy: 80.89%\n    Accuracy: 81.17%\n    Accuracy: 82.1%\n    Accuracy: 81.05%\n    Accuracy: 79.93%\n    Accuracy: 81.32%\n    Accuracy: 80.78%\n    Accuracy: 81.36%\n    Accuracy: 81.54%\n    Accuracy: 81.51%\n    Accuracy: 80.38%\n    Accuracy: 81.05%\n    Accuracy: 80.92%\n    Accuracy: 81.39%\n    Accuracy: 81.63%\n    Accuracy: 80.56%\n    Accuracy: 82.28%\n    Accuracy: 81.97%\n    Accuracy: 81.82%\n    Accuracy: 81.5%\n    Accuracy: 80.72%\n    Accuracy: 81.3%\n    Accuracy: 81.01%\n    Accuracy: 80.72%\n    Accuracy: 80.79%\n    Accuracy: 81.13%\n    Accuracy: 81.03%\n    Accuracy: 81.9%\n    Accuracy: 81.72%\n    Accuracy: 81.71%\n    Accuracy: 80.01%\n    Accuracy: 82.06%\n    Accuracy: 81.37%\n    Accuracy: 81.81%\n    Accuracy: 81.8%\n    Accuracy: 81.83%\n    Accuracy: 82.19%\n    Accuracy: 82.21%\n    Accuracy: 82.0%\n    Accuracy: 81.92%\n    Accuracy: 81.61%\n    Accuracy: 81.14%\n    Accuracy: 82.18%\n    Accuracy: 81.92%\n    Accuracy: 82.3%\n    Accuracy: 80.84%\n    Accuracy: 81.48%\n    Accuracy: 81.22%\n    Accuracy: 82.14%\n    Accuracy: 80.44%\n    Accuracy: 81.6%\n    Accuracy: 81.72%\n    Accuracy: 81.07%\n    Accuracy: 81.62%\n    Accuracy: 81.45%\n    Accuracy: 81.97%\n    Accuracy: 81.07%\n    Accuracy: 82.14%\n    Accuracy: 82.13%\n    Accuracy: 81.9%\n    Accuracy: 82.01%\n    Accuracy: 82.16%\n    Accuracy: 80.7%\n    Accuracy: 82.16%\n    Accuracy: 81.24%\n    Accuracy: 81.57%\n    Accuracy: 81.67%\n    Accuracy: 81.76%\n    Accuracy: 81.55%\n    Accuracy: 81.53%\n    Accuracy: 81.22%\n    Accuracy: 81.81%\n    Accuracy: 81.83%\n    Accuracy: 82.13%\n    Accuracy: 82.01%\n    Accuracy: 81.49%\n    Accuracy: 81.59%\n    Accuracy: 82.25%\n    Accuracy: 81.81%\n    Accuracy: 81.91%\n    Accuracy: 79.91%\n    Accuracy: 80.52%\n    Accuracy: 82.01%\n    Accuracy: 82.3%\n    Accuracy: 81.84%\n    Accuracy: 81.34%\n    Accuracy: 82.23%\n    Accuracy: 81.67%\n    Accuracy: 80.8%\n    Accuracy: 82.24%\n    Accuracy: 81.01%\n    Accuracy: 81.52%\n    Accuracy: 82.5%\n    Accuracy: 81.5%\n    Accuracy: 81.65%\n    Accuracy: 82.2%\n    Accuracy: 81.92%\n    Accuracy: 81.64%\n    Accuracy: 81.71%\n    Accuracy: 82.06%\n    Accuracy: 81.5%\n    Accuracy: 81.68%\n    Accuracy: 82.43%\n    Accuracy: 81.71%\n    Accuracy: 80.84%\n    Accuracy: 81.11%\n    Accuracy: 82.12%\n    Accuracy: 81.43%\n    Accuracy: 80.94%\n    Accuracy: 81.72%\n    Accuracy: 82.26%\n    Accuracy: 82.12%\n    Accuracy: 81.37%\n    Accuracy: 81.04%\n    Accuracy: 82.02%\n    Accuracy: 81.63%\n    Accuracy: 81.5%\n    Accuracy: 82.34%\n    Accuracy: 80.9%\n    Accuracy: 81.66%\n    Accuracy: 81.9%\n    Accuracy: 81.99%\n    Accuracy: 80.4%\n    Accuracy: 82.35%\n    Accuracy: 80.83%\n    Accuracy: 82.15%\n    Accuracy: 81.66%\n    Accuracy: 81.5%\n    Accuracy: 82.02%\n    Accuracy: 81.45%\n    Accuracy: 81.28%\n    Accuracy: 81.08%\n    Accuracy: 81.16%\n    Accuracy: 82.13%\n    Accuracy: 81.85%\n    Accuracy: 81.96%\n    Accuracy: 81.9%\n    Accuracy: 82.01%\n    Accuracy: 81.91%\n    Accuracy: 81.41%\n    Accuracy: 81.16%\n    Accuracy: 81.65%\n    Accuracy: 82.29%\n    Accuracy: 82.11%\n    Accuracy: 81.46%\n    Accuracy: 82.61%\n    Accuracy: 82.21%\n    Accuracy: 81.85%\n    Accuracy: 82.41%\n    Accuracy: 80.74%\n    Accuracy: 81.12%\n    Accuracy: 81.85%\n    Accuracy: 81.95%\n    Accuracy: 82.23%\n    Accuracy: 81.88%\n    Accuracy: 82.13%\n    Accuracy: 81.88%\n    Accuracy: 82.0%\n    Accuracy: 81.01%\n    Accuracy: 81.08%\n    Accuracy: 81.3%\n    Accuracy: 81.19%\n    Accuracy: 81.39%\n    Accuracy: 81.16%\n    Accuracy: 81.73%\n    Accuracy: 81.98%\n    Accuracy: 81.06%\n    Accuracy: 81.32%\n    Accuracy: 81.64%\n    Accuracy: 81.32%\n    Accuracy: 82.09%\n    Accuracy: 81.84%\n    Accuracy: 81.4%\n    Accuracy: 81.96%\n    Accuracy: 82.08%\n    Accuracy: 82.3%\n    Accuracy: 81.59%\n    Accuracy: 81.25%\n    Accuracy: 81.23%\n    Accuracy: 82.52%\n    Accuracy: 81.72%\n    Accuracy: 82.3%\n    Accuracy: 82.04%\n    Accuracy: 82.1%\n    Accuracy: 82.41%\n    Accuracy: 81.41%\n    Accuracy: 82.26%\n    Accuracy: 81.14%\n    Accuracy: 82.14%\n    Accuracy: 81.78%\n    Accuracy: 82.62%\n    Accuracy: 82.0%\n    Accuracy: 81.02%\n    Accuracy: 81.94%\n    Accuracy: 81.92%\n    Accuracy: 82.29%\n    Accuracy: 81.8%\n    Accuracy: 82.39%\n    Accuracy: 82.3%\n    Accuracy: 81.64%\n    Accuracy: 81.46%\n    Accuracy: 81.06%\n    Accuracy: 82.14%\n    Accuracy: 81.61%\n    Accuracy: 81.61%\n    Accuracy: 81.69%\n    Accuracy: 81.69%\n    Accuracy: 82.15%\n    Accuracy: 82.02%\n    Accuracy: 82.06%\n    Accuracy: 82.57%\n    Accuracy: 81.51%\n    Accuracy: 81.88%\n    Accuracy: 81.94%\n    Accuracy: 81.16%\n    Accuracy: 81.4%\n    Accuracy: 82.03%\n    Accuracy: 82.09%\n    Accuracy: 82.07%\n    Accuracy: 82.01%\n    Accuracy: 82.65%\n    Accuracy: 82.13%\n    Accuracy: 81.54%\n    Accuracy: 81.62%\n    Accuracy: 82.84%\n    Accuracy: 82.43%\n    Accuracy: 82.25%\n    Accuracy: 82.7%\n    Accuracy: 81.38%\n    Accuracy: 81.97%\n    Accuracy: 82.1%\n    Accuracy: 82.18%\n    Accuracy: 80.99%\n    Accuracy: 81.79%\n    Accuracy: 81.14%\n    Accuracy: 82.37%\n    Accuracy: 82.03%\n    Accuracy: 82.18%\n    Accuracy: 82.37%\n    Accuracy: 82.28%\n    Accuracy: 82.0%\n    Accuracy: 82.07%\n    Accuracy: 80.91%\n    Accuracy: 82.23%\n    Accuracy: 81.93%\n    Accuracy: 82.45%\n    Accuracy: 80.57%\n    Accuracy: 82.74%\n    Accuracy: 82.76%\n    Accuracy: 81.61%\n    Accuracy: 82.22%\n    Accuracy: 81.6%\n    Accuracy: 82.08%\n    Accuracy: 81.58%\n    Accuracy: 82.02%\n    Accuracy: 82.03%\n    Accuracy: 81.92%\n    Accuracy: 80.98%\n    Accuracy: 81.94%\n    Accuracy: 81.44%\n    Accuracy: 82.03%\n    Accuracy: 81.54%\n    Accuracy: 81.47%\n    Accuracy: 82.17%\n    Accuracy: 82.25%\n    Accuracy: 82.49%\n    Accuracy: 81.8%\n    Accuracy: 82.38%\n    Accuracy: 81.47%\n    Accuracy: 81.73%\n    Accuracy: 81.7%\n    Accuracy: 81.09%\n    Accuracy: 81.41%\n    Accuracy: 81.95%\n    Accuracy: 81.73%\n    Accuracy: 81.9%\n    Accuracy: 81.97%\n    Accuracy: 82.1%\n    Accuracy: 81.35%\n    Accuracy: 82.15%\n    Accuracy: 82.21%\n    Accuracy: 82.71%\n    Accuracy: 81.48%\n    Accuracy: 82.52%\n    Accuracy: 82.07%\n    Accuracy: 81.79%\n    Accuracy: 81.27%\n    Accuracy: 82.22%\n    Accuracy: 81.02%\n    Accuracy: 82.44%\n    Accuracy: 81.73%\n    Accuracy: 82.12%\n    Accuracy: 82.1%\n    Accuracy: 81.57%\n    Accuracy: 82.1%\n    Accuracy: 81.93%\n    Accuracy: 81.4%\n    Accuracy: 82.13%\n    Accuracy: 81.82%\n    Accuracy: 81.54%\n    Accuracy: 81.48%\n    Accuracy: 82.08%\n    Accuracy: 82.31%\n    Accuracy: 82.76%\n    Accuracy: 81.63%\n    Accuracy: 81.99%\n    Accuracy: 81.38%\n    Accuracy: 81.3%\n    Accuracy: 81.77%\n    Accuracy: 82.06%\n    Accuracy: 82.35%\n    Accuracy: 81.84%\n    Accuracy: 82.18%\n    Accuracy: 82.16%\n    Accuracy: 82.17%\n    Accuracy: 81.24%\n    Accuracy: 82.54%\n    Accuracy: 83.06%\n    Accuracy: 82.58%\n    Accuracy: 82.23%\n    Accuracy: 81.76%\n    Accuracy: 81.65%\n    Accuracy: 82.09%\n    Accuracy: 82.21%\n    Accuracy: 82.6%\n    Accuracy: 81.86%\n    Accuracy: 81.56%\n    Accuracy: 81.76%\n    Accuracy: 81.76%\n    Accuracy: 82.6%\n    Accuracy: 82.23%\n    Accuracy: 82.42%\n    Accuracy: 82.13%\n    Accuracy: 82.22%\n    Accuracy: 81.33%\n    Accuracy: 81.54%\n    Accuracy: 81.94%\n    Accuracy: 82.39%\n    Accuracy: 81.9%\n    Accuracy: 81.67%\n    Accuracy: 80.69%\n    Accuracy: 82.55%\n    Accuracy: 82.02%\n    Accuracy: 82.05%\n    Accuracy: 82.17%\n    Accuracy: 81.86%\n    Accuracy: 82.33%\n    Accuracy: 82.29%\n    Accuracy: 82.55%\n    Accuracy: 82.28%\n    Accuracy: 82.32%\n    Accuracy: 82.38%\n    Accuracy: 82.12%\n    Accuracy: 82.13%\n    Accuracy: 82.35%\n    Accuracy: 81.47%\n    Accuracy: 82.52%\n    Accuracy: 81.99%\n    Accuracy: 81.15%\n    Accuracy: 82.31%\n    Accuracy: 80.23%\n    Accuracy: 81.71%\n    Accuracy: 81.76%\n    Accuracy: 81.75%\n    Accuracy: 82.45%\n    Accuracy: 82.2%\n    Accuracy: 82.63%\n    Accuracy: 82.33%\n    Accuracy: 81.72%\n    Accuracy: 82.45%\n    Accuracy: 82.37%\n    Accuracy: 81.56%\n    Accuracy: 81.79%\n    Accuracy: 82.17%\n    Accuracy: 81.69%\n    Accuracy: 82.1%\n    Accuracy: 81.42%\n    Accuracy: 81.9%\n    Accuracy: 81.44%\n    Accuracy: 82.46%\n    Accuracy: 82.26%\n    Accuracy: 82.25%\n    Accuracy: 82.2%\n    Accuracy: 82.05%\n    Accuracy: 82.24%\n    Accuracy: 82.09%\n    Accuracy: 82.07%\n    Accuracy: 81.71%\n    Accuracy: 81.78%\n    Accuracy: 82.2%\n    Accuracy: 81.7%\n    Accuracy: 82.25%\n    Accuracy: 83.0%\n    Accuracy: 81.68%\n    Accuracy: 81.94%\n    Accuracy: 82.76%\n    Accuracy: 82.45%\n    Accuracy: 82.73%\n    Accuracy: 82.46%\n    Accuracy: 82.65%\n    Accuracy: 80.45%\n    Accuracy: 83.14%\n    Accuracy: 81.88%\n    Accuracy: 81.83%\n    Accuracy: 81.96%\n    Accuracy: 82.28%\n    Accuracy: 80.74%\n    Accuracy: 82.12%\n    Accuracy: 81.84%\n    Accuracy: 82.13%\n    Accuracy: 82.17%\n    Accuracy: 81.6%\n    Accuracy: 82.35%\n    Accuracy: 81.82%\n    Accuracy: 81.3%\n    Accuracy: 81.95%\n    Accuracy: 82.16%\n    Accuracy: 81.2%\n    Accuracy: 81.46%\n    Accuracy: 81.17%\n    Accuracy: 81.77%\n    Accuracy: 82.23%\n    Accuracy: 81.74%\n    Accuracy: 81.95%\n    Accuracy: 80.65%\n    Accuracy: 81.71%\n    Accuracy: 81.73%\n    Accuracy: 81.74%\n    Accuracy: 81.95%\n    Accuracy: 82.3%\n    Accuracy: 81.82%\n    Accuracy: 81.67%\n    Accuracy: 81.68%\n    Accuracy: 81.82%\n    Accuracy: 82.01%\n    Accuracy: 82.14%\n    Accuracy: 81.16%\n    Accuracy: 81.75%\n    Accuracy: 82.02%\n    Accuracy: 81.43%\n    Accuracy: 81.71%\n    Accuracy: 81.11%\n    Accuracy: 81.26%\n    Accuracy: 82.55%\n    Accuracy: 82.38%\n    Accuracy: 80.76%\n    Accuracy: 81.67%\n    Accuracy: 82.17%\n    Accuracy: 82.0%\n    Accuracy: 82.58%\n    Accuracy: 81.67%\n    Accuracy: 81.42%\n    Accuracy: 81.05%\n    Accuracy: 81.75%\n    Accuracy: 81.89%\n    Accuracy: 81.81%\n    Accuracy: 81.73%\n    Accuracy: 82.05%\n    Accuracy: 81.87%\n    Accuracy: 82.43%\n    Accuracy: 82.03%\n    Accuracy: 82.37%\n    Accuracy: 82.33%\n    Accuracy: 82.4%\n    Accuracy: 82.34%\n    Accuracy: 82.51%\n    Accuracy: 82.2%\n    Accuracy: 81.83%\n    Accuracy: 81.88%\n    Accuracy: 81.42%\n    Accuracy: 81.83%\n    Accuracy: 81.72%\n    Accuracy: 81.72%\n    Accuracy: 81.8%\n    Accuracy: 82.31%\n    Accuracy: 81.38%\n    Accuracy: 81.46%\n    Accuracy: 81.92%\n    Accuracy: 80.86%\n    Accuracy: 82.13%\n    Accuracy: 81.39%\n    Accuracy: 81.86%\n    Accuracy: 82.65%\n    Accuracy: 81.59%\n    Accuracy: 82.06%\n    Accuracy: 82.11%\n    Accuracy: 81.47%\n    Accuracy: 81.83%\n    Accuracy: 82.29%\n    Accuracy: 80.92%\n    Accuracy: 81.87%\n    Accuracy: 82.08%\n    Accuracy: 82.04%\n    Accuracy: 81.7%\n    Accuracy: 81.6%\n    Accuracy: 80.2%\n    Accuracy: 82.15%\n    Accuracy: 81.08%\n    Accuracy: 81.99%\n    Accuracy: 81.53%\n    Accuracy: 81.61%\n    Accuracy: 82.22%\n    Accuracy: 81.31%\n    Accuracy: 82.74%\n    Accuracy: 82.43%\n    Accuracy: 82.22%\n    Accuracy: 81.88%\n    Accuracy: 82.32%\n    Accuracy: 81.79%\n    Accuracy: 82.04%\n    Accuracy: 81.5%\n    Accuracy: 82.4%\n    Accuracy: 82.05%\n    Accuracy: 81.56%\n    Accuracy: 81.92%\n    Accuracy: 82.2%\n    Accuracy: 81.44%\n    Accuracy: 82.48%\n    Accuracy: 82.16%\n    Accuracy: 81.63%\n    Accuracy: 82.63%\n    Accuracy: 81.65%\n    Accuracy: 81.89%\n    Accuracy: 82.66%\n    Accuracy: 80.55%\n    Accuracy: 82.58%\n    Accuracy: 81.89%\n    Accuracy: 81.91%\n    Accuracy: 81.92%\n    Accuracy: 81.4%\n    Accuracy: 81.92%\n    Accuracy: 82.23%\n    Accuracy: 81.9%\n    Accuracy: 82.02%\n    Accuracy: 82.27%\n    Accuracy: 82.35%\n    Accuracy: 81.96%\n    Accuracy: 81.09%\n    Accuracy: 82.07%\n    Accuracy: 82.57%\n    Accuracy: 82.06%\n    Accuracy: 82.23%\n    Accuracy: 80.67%\n    Accuracy: 81.14%\n    Accuracy: 82.09%\n    Accuracy: 81.07%\n    Accuracy: 82.4%\n    Accuracy: 82.25%\n    Accuracy: 82.35%\n    Accuracy: 81.77%\n    Accuracy: 81.45%\n    Accuracy: 82.09%\n    Accuracy: 81.22%\n    Accuracy: 81.83%\n    Accuracy: 82.38%\n    Accuracy: 81.8%\n    Accuracy: 82.39%\n    Accuracy: 82.35%\n    Accuracy: 81.32%\n    Accuracy: 82.14%\n    Accuracy: 82.02%\n    Accuracy: 81.91%\n    Accuracy: 82.12%\n    Accuracy: 81.11%\n    Accuracy: 81.44%\n    Accuracy: 81.47%\n    Accuracy: 82.38%\n    Accuracy: 82.5%\n    Accuracy: 81.91%\n    Accuracy: 82.34%\n    Accuracy: 80.65%\n    Accuracy: 81.06%\n    Accuracy: 81.25%\n    Accuracy: 80.7%\n    Accuracy: 81.77%\n    Accuracy: 80.33%\n    Accuracy: 81.3%\n    Accuracy: 81.86%\n    Accuracy: 82.0%\n    Accuracy: 82.05%\n    Accuracy: 81.38%\n    Accuracy: 81.83%\n    Accuracy: 81.47%\n    Accuracy: 82.04%\n    Accuracy: 80.76%\n    Accuracy: 82.4%\n    Accuracy: 82.7%\n    Accuracy: 81.54%\n    Accuracy: 82.15%\n    Accuracy: 81.71%\n    Accuracy: 79.96%\n    Accuracy: 82.09%\n    Accuracy: 82.4%\n    Accuracy: 82.2%\n    Accuracy: 81.37%\n    Accuracy: 81.92%\n    Accuracy: 81.81%\n    Accuracy: 81.61%\n    Accuracy: 80.52%\n    Accuracy: 81.87%\n    Accuracy: 81.24%\n    Accuracy: 82.06%\n    Accuracy: 81.92%\n    Accuracy: 81.97%\n    Accuracy: 82.47%\n    Accuracy: 82.16%\n    Accuracy: 81.98%\n    Accuracy: 82.43%\n    Accuracy: 82.2%\n    Accuracy: 82.14%\n    Accuracy: 82.24%\n    Accuracy: 82.75%\n    Accuracy: 82.21%\n    Accuracy: 82.14%\n    Accuracy: 81.7%\n    Accuracy: 82.02%\n    Accuracy: 81.1%\n    Accuracy: 81.79%\n    Accuracy: 81.85%\n    Accuracy: 81.63%\n    Accuracy: 81.98%\n    Accuracy: 82.64%\n    Accuracy: 82.21%\n    Accuracy: 81.97%\n    Accuracy: 81.94%\n    Accuracy: 81.73%\n    Accuracy: 81.78%\n    Accuracy: 81.88%\n    Accuracy: 81.17%\n    Accuracy: 81.84%\n    Accuracy: 81.88%\n    Accuracy: 81.76%\n    Accuracy: 81.11%\n    Accuracy: 81.82%\n    Accuracy: 82.52%\n    Accuracy: 82.23%\n    Accuracy: 82.12%\n    Accuracy: 81.66%\n    Accuracy: 82.34%\n    Accuracy: 81.75%\n    Accuracy: 82.24%\n    Accuracy: 81.64%\n    Accuracy: 82.06%\n    Accuracy: 82.02%\n    Accuracy: 82.14%\n    Accuracy: 80.99%\n    Accuracy: 81.48%\n    Accuracy: 82.23%\n    Accuracy: 82.25%\n    Accuracy: 81.29%\n    Accuracy: 80.61%\n    Accuracy: 81.56%\n    Accuracy: 81.99%\n    Accuracy: 81.63%\n    Accuracy: 82.66%\n    Accuracy: 81.76%\n    Accuracy: 81.72%\n    Accuracy: 82.09%\n    Accuracy: 82.25%\n    Accuracy: 82.28%\n    Accuracy: 82.18%\n    Accuracy: 79.89%\n    Accuracy: 82.22%\n    Accuracy: 81.0%\n    Accuracy: 81.98%\n    Accuracy: 82.18%\n    Accuracy: 81.42%\n    Accuracy: 81.44%\n    Accuracy: 81.67%\n    Accuracy: 82.01%\n    Accuracy: 81.9%\n    Accuracy: 81.77%\n    Accuracy: 81.73%\n    Accuracy: 82.5%\n    Accuracy: 81.99%\n    Accuracy: 81.89%\n    Accuracy: 81.63%\n    Accuracy: 81.54%\n    Accuracy: 81.83%\n    Accuracy: 82.18%\n    Accuracy: 82.11%\n    Accuracy: 81.7%\n    Accuracy: 82.71%\n    Accuracy: 82.23%\n    Accuracy: 82.39%\n    Accuracy: 81.57%\n    Accuracy: 81.48%\n    Accuracy: 81.62%\n    Accuracy: 81.98%\n    Accuracy: 81.48%\n    Accuracy: 81.02%\n    Accuracy: 81.09%\n    Accuracy: 82.25%\n    Accuracy: 82.5%\n    Accuracy: 83.15%\n    Accuracy: 80.77%\n    Accuracy: 81.09%\n    Accuracy: 81.53%\n    Accuracy: 81.51%\n    Accuracy: 82.11%\n    Accuracy: 81.69%\n    Accuracy: 81.14%\n    Accuracy: 81.86%\n    Accuracy: 82.36%\n    Accuracy: 81.26%\n    Accuracy: 81.55%\n    Accuracy: 81.34%\n    Accuracy: 81.18%\n    Accuracy: 82.17%\n    Accuracy: 81.45%\n    Accuracy: 81.68%\n    Accuracy: 81.8%\n    Accuracy: 81.54%\n    Accuracy: 81.42%\n    Accuracy: 81.49%\n    Accuracy: 81.8%\n    Accuracy: 81.14%\n    Accuracy: 81.35%\n    Accuracy: 81.41%\n    Accuracy: 81.14%\n    Accuracy: 81.47%\n    Accuracy: 81.85%\n    Accuracy: 81.57%\n    Accuracy: 82.02%\n    Accuracy: 82.16%\n    Accuracy: 82.2%\n    Accuracy: 80.28%\n    Accuracy: 81.57%\n    Accuracy: 81.06%\n    Accuracy: 80.58%\n    Accuracy: 81.43%\n    Accuracy: 82.02%\n    Accuracy: 81.62%\n    Accuracy: 81.22%\n    Accuracy: 81.05%\n    Accuracy: 81.69%\n    Accuracy: 82.06%\n    Accuracy: 82.02%\n    Accuracy: 81.69%\n    Accuracy: 81.89%\n    Accuracy: 81.32%\n    Accuracy: 81.18%\n    Accuracy: 81.59%\n    Accuracy: 82.04%\n    Accuracy: 81.62%\n    Accuracy: 82.2%\n    Accuracy: 81.52%\n    Accuracy: 82.19%\n    Accuracy: 81.38%\n    Accuracy: 81.65%\n    Accuracy: 81.18%\n    Accuracy: 81.77%\n    Accuracy: 81.95%\n    Accuracy: 81.46%\n    Accuracy: 81.0%\n    Accuracy: 82.13%\n    Accuracy: 81.64%\n    Accuracy: 81.38%\n    Accuracy: 81.16%\n    Accuracy: 81.67%\n    Accuracy: 81.7%\n    Accuracy: 81.11%\n    Accuracy: 82.38%\n    Accuracy: 81.59%\n    Accuracy: 82.16%\n    Accuracy: 82.6%\n    Accuracy: 81.09%\n    Accuracy: 81.56%\n    Accuracy: 80.16%\n    Accuracy: 81.75%\n    Accuracy: 80.8%\n    Accuracy: 80.5%\n    Accuracy: 81.33%\n    Accuracy: 81.75%\n    Accuracy: 80.87%\n    Accuracy: 81.93%\n    Accuracy: 82.3%\n    Accuracy: 81.46%\n    Accuracy: 80.74%\n    Accuracy: 81.67%\n    Accuracy: 81.98%\n    Accuracy: 82.27%\n    Accuracy: 82.09%\n    Accuracy: 81.54%\n    Accuracy: 81.72%\n    Accuracy: 80.56%\n    Accuracy: 81.24%\n    Accuracy: 81.6%\n    Accuracy: 80.96%\n    Accuracy: 81.0%\n    Accuracy: 81.8%\n    Accuracy: 80.69%\n    Accuracy: 80.96%\n    Accuracy: 80.54%\n    Accuracy: 82.43%\n    Accuracy: 81.33%\n    Accuracy: 81.84%\n    Accuracy: 82.01%\n    Accuracy: 81.59%\n    Accuracy: 81.7%\n    Accuracy: 81.71%\n    Accuracy: 82.17%\n    Accuracy: 81.89%\n    Accuracy: 81.13%\n    Accuracy: 81.79%\n    Accuracy: 81.15%\n    Accuracy: 81.95%\n    Accuracy: 80.99%\n    Accuracy: 80.37%\n    Accuracy: 81.62%\n    Accuracy: 81.78%\n    Accuracy: 80.93%\n    Accuracy: 81.56%\n    Accuracy: 81.75%\n    Accuracy: 81.31%\n    Accuracy: 81.59%\n    Accuracy: 81.55%\n    Accuracy: 82.02%\n    Accuracy: 81.55%\n    Accuracy: 80.44%\n    Accuracy: 82.03%\n    Accuracy: 81.57%\n    Accuracy: 81.99%\n    Accuracy: 81.38%\n    Accuracy: 81.77%\n    Accuracy: 81.78%\n    Accuracy: 81.01%\n    Accuracy: 81.63%\n    Accuracy: 81.59%\n    Accuracy: 80.79%\n    Accuracy: 81.45%\n    Accuracy: 81.69%\n    Accuracy: 81.76%",
      "json_metadata": "{\"tags\":[\"tensorflow\",\"python\",\"deep-learning\",\"cifar10\"],\"image\":[\"https://cdn.steemitimages.com/DQmcEnsxs3tT7HneBPDSio2rBGGFPpVP4ax2gQDR87B1eGv/image.png\"],\"links\":[\"https://github.com/llejo3/deep-learning/blob/master/CIFAR10_tensorflow.md\"],\"app\":\"steemit/0.1\",\"format\":\"markdown\"}"
    }
  ]
}

{
  "trx_id": "be2f8170e3c72ec71d8aec727c8cfe8276c0b963",
  "block": 22867249,
  "trx_in_block": 31,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-29T23:09:03",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "south-man",
      "permlink": "cifar-10-tensorflow",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "5c916f0e6173ada6d3c72bab04d13dbce84fc9f0",
  "block": 22864516,
  "trx_in_block": 53,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-29T20:52:24",
  "op": [
    "vote",
    {
      "voter": "sd974201",
      "author": "south-man",
      "permlink": "cifar-10-tensorflow",
      "weight": 700
    }
  ]
}

{
  "trx_id": "ba1c8cc5d917ba9e4ffae8ed50af5837606e2203",
  "block": 22863839,
  "trx_in_block": 44,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-29T20:18:33",
  "op": [
    "comment",
    {
      "parent_author": "",
      "parent_permlink": "tensorflow",
      "author": "south-man",
      "permlink": "cifar-10-tensorflow",
      "title": "CIFAR 10 데이터로 딥러닝하는 Tensorflow 소스",
      "body": "@@ -8820,23 +8820,93 @@\n %0A%0A!%5B\n-png%5D(output_3_1\n+%5D(https://cdn.steemitimages.com/DQmcEnsxs3tT7HneBPDSio2rBGGFPpVP4ax2gQDR87B1eGv/image\n .png\n",
      "json_metadata": "{\"tags\":[\"tensorflow\",\"python\",\"kr\",\"deep-learning\"],\"image\":[\"https://cdn.steemitimages.com/DQmcEnsxs3tT7HneBPDSio2rBGGFPpVP4ax2gQDR87B1eGv/image.png\"],\"links\":[\"https://github.com/llejo3/deep-learning/blob/master/CIFAR10_tensorflow.md\"],\"app\":\"steemit/0.1\",\"format\":\"markdown\"}"
    }
  ]
}

{
  "trx_id": "c22713dd5ee92f60c63762611db1323bf497e062",
  "block": 22863817,
  "trx_in_block": 27,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-29T20:17:27",
  "op": [
    "comment",
    {
      "parent_author": "",
      "parent_permlink": "tensorflow",
      "author": "south-man",
      "permlink": "cifar-10-tensorflow",
      "title": "CIFAR 10 데이터로 딥러닝하는 Tensorflow 소스",
      "body": "깃허브에도 소스가 올라가 있어요.\nhttps://github.com/llejo3/deep-learning/blob/master/CIFAR10_tensorflow.md\n\n\n```python\nimport tensorflow as tf\n```\n\n\n```python\n# Helper functions\n\ndef weight_variable(shape):\n    initial = tf.truncated_normal(shape, stddev=0.1)\n    return tf.Variable(initial)\n\n\ndef bias_variable(shape):\n    initial = tf.constant(0.1, shape=shape)\n    return tf.Variable(initial)\n\n\ndef conv2d(x, W):\n    return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')\n\n\ndef max_pool_2x2(x):\n    return tf.nn.max_pool(x, ksize=[1, 2, 2, 1],\n                          strides=[1, 2, 2, 1], padding='SAME')\n\n\ndef conv_layer(input, shape):\n    W = weight_variable(shape)\n    b = bias_variable([shape[3]])\n    return tf.nn.relu(conv2d(input, W) + b)\n\ndef conv_norm_layer(input, shape, phase):\n    W = weight_variable(shape)\n    b = bias_variable([shape[3]])\n    return tf.nn.relu( batch_norm_wrapper( conv2d(input, W) + b, phase))\n\ndef full_layer(input, size):\n    in_size = int(input.get_shape()[1])\n    W = weight_variable([in_size, size])\n    b = bias_variable([size])\n    return tf.matmul(input, W) + b\n\n# Batch Nomalization \ndef batch_norm_wrapper(inputs, is_training, decay = 0.999):\n    scale = tf.Variable(tf.ones([inputs.get_shape()[-1]]))\n    beta = tf.Variable(tf.zeros([inputs.get_shape()[-1]]))\n    pop_mean = tf.Variable(tf.zeros([inputs.get_shape()[-1]]), trainable=False)\n    pop_var = tf.Variable(tf.ones([inputs.get_shape()[-1]]), trainable=False)\n\n    if is_training == True:\n        batch_mean, batch_var = tf.nn.moments(inputs,[0])\n        train_mean = tf.assign(pop_mean, pop_mean * decay + batch_mean * (1 - decay))\n        train_var = tf.assign(pop_var, pop_var * decay + batch_var * (1 - decay))\n        with tf.control_dependencies([train_mean, train_var]):\n            return tf.nn.batch_normalization(inputs,\n                batch_mean, batch_var, beta, scale, epsilon)\n    else:\n        return tf.nn.batch_normalization(inputs, pop_mean, pop_var, beta, scale, epsilon)\n```\n\n\n```python\nimport pickle\nimport os\nimport numpy as np\nimport matplotlib.pyplot as plt\nimport tensorflow as tf\n\n# CIFAR10 데이터 경로 \nDATA_PATH = \"./cifar-10-batches-py\"\nBATCH_SIZE = 50\nSTEPS = 500000\nepsilon = 1e-3\n\ndef one_hot(vec, vals=10):\n    n = len(vec)\n    out = np.zeros((n, vals))\n    out[range(n), vec] = 1\n    return out\n\n\ndef unpickle(file):\n    with open(os.path.join(DATA_PATH, file), 'rb') as fo:\n        u = pickle._Unpickler(fo)\n        u.encoding = 'latin1'\n        dict = u.load()\n    return dict\n\n\ndef display_cifar(images, size):\n    n = len(images)\n    plt.figure()\n    plt.gca().set_axis_off()\n    im = np.vstack([np.hstack([images[np.random.choice(n)] for i in range(size)])\n                    for i in range(size)])\n    plt.imshow(im)\n    plt.show()\n\n\nclass CifarLoader(object):\n    \"\"\"\n    Load and mange the CIFAR dataset.\n    (for any practical use there is no reason not to use the built-in dataset handler instead)\n    \"\"\"\n    def __init__(self, source_files):\n        self._source = source_files\n        self._i = 0\n        self.images = None\n        self.labels = None\n\n    def load(self):\n        data = [unpickle(f) for f in self._source]\n        images = np.vstack([d[\"data\"] for d in data])\n        n = len(images)\n        self.images = images.reshape(n, 3, 32, 32).transpose(0, 2, 3, 1).astype(float) / 255\n        self.labels = one_hot(np.hstack([d[\"labels\"] for d in data]), 10)\n        return self\n\n    def next_batch(self, batch_size):\n        x, y = self.images[self._i:self._i+batch_size], self.labels[self._i:self._i+batch_size]\n        self._i = (self._i + batch_size) % len(self.images)\n        return x, y\n\n    def random_batch(self, batch_size):\n        n = len(self.images)\n        ix = np.random.choice(n, batch_size)\n        return self.images[ix], self.labels[ix]\n\nclass CifarDataManager(object):\n    def __init__(self):\n        self.train = CifarLoader([\"data_batch_{}\".format(i) for i in range(1, 6)]).load()\n        self.test = CifarLoader([\"test_batch\"]).load()\n\n\ndef run_simple_net():\n    cifar = CifarDataManager()\n    x = tf.placeholder(tf.float32, shape=[None, 32, 32, 3])\n    y_ = tf.placeholder(tf.float32, shape=[None, 10])\n    keep_prob = tf.placeholder(tf.float32)\n    phase = tf.placeholder(tf.bool) \n\n    conv1 = conv_norm_layer(x, [5, 5, 3, 32], phase)\n    conv1_pool = max_pool_2x2(conv1)\n\n    conv2 = conv_norm_layer(conv1_pool, [5, 5, 32, 64], phase)\n    conv2_pool = max_pool_2x2(conv2)\n\n    conv3 = conv_norm_layer(conv2_pool, [5, 5, 64, 128], phase)\n    conv3_pool = max_pool_2x2(conv3)\n    conv3_flat = tf.reshape(conv3_pool, [-1, 4 * 4 * 128])\n    conv3_drop = tf.nn.dropout(conv3_flat, keep_prob=keep_prob)\n\n    full_1 = tf.nn.relu(full_layer(conv3_drop, 512))\n    full1_drop = tf.nn.dropout(full_1, keep_prob=keep_prob)\n\n    y_conv = full_layer(full1_drop, 10)\n\n    cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=y_conv, labels=y_))\n    train_step = tf.train.AdamOptimizer(1e-3).minimize(cross_entropy)\n\n    correct_prediction = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1))\n    accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))\n\n    def test(sess):\n        X = cifar.test.images.reshape(10, 1000, 32, 32, 3)\n        Y = cifar.test.labels.reshape(10, 1000, 10)\n        acc = np.mean([sess.run(accuracy, feed_dict={x: X[i], y_: Y[i], phase: False, keep_prob: 1.0})\n                       for i in range(10)])\n        print(\"Accuracy: {:.4}%\".format(acc * 100))\n\n    with tf.Session() as sess:\n        sess.run(tf.global_variables_initializer())\n\n        for i in range(STEPS):\n            batch = cifar.train.next_batch(BATCH_SIZE)\n            sess.run(train_step, feed_dict={x: batch[0], y_: batch[1], phase: True,  keep_prob: 0.5})\n\n            if i % 500 == 0:\n                test(sess)\n\n        test(sess)\n\n\ndef build_second_net():\n    cifar = CifarDataManager()\n    x = tf.placeholder(tf.float32, shape=[None, 32, 32, 3])\n    y_ = tf.placeholder(tf.float32, shape=[None, 10])\n    keep_prob = tf.placeholder(tf.float32)\n    phase = tf.placeholder(tf.bool) \n\n    C1, C2, C3 = 32, 64, 128\n    F1 = 600\n\n    conv1_1 = conv_norm_layer(x, [3, 3, 3, C1], phase)\n    conv1_2 = conv_norm_layer(conv1_1, [3, 3, C1, C1], phase)\n    conv1_3 = conv_norm_layer(conv1_2, [3, 3, C1, C1], phase)\n    conv1_pool = max_pool_2x2(conv1_3)\n    conv1_drop = tf.nn.dropout(conv1_pool, keep_prob=keep_prob)\n\n    conv2_1 = conv_norm_layer(conv1_drop, [3, 3, C1, C2], phase)\n    conv2_2 = conv_norm_layer(conv2_1, [3, 3, C2, C2], phase)\n    conv2_3 = conv_norm_layer(conv2_2, [3, 3, C2, C2], phase)\n    conv2_pool = max_pool_2x2(conv2_3)\n    conv2_drop = tf.nn.dropout(conv2_pool, keep_prob=keep_prob)\n\n    conv3_1 = conv_norm_layer(conv2_drop, [3, 3, C2, C3], phase)\n    conv3_2 = conv_norm_layer(conv3_1, [3, 3, C3, C3], phase)\n    conv3_3 = conv_norm_layer(conv3_2, [3, 3, C3, C3], phase)\n    conv3_pool = tf.nn.max_pool(conv3_3, ksize=[1, 8, 8, 1], strides=[1, 8, 8, 1], padding='SAME')\n    conv3_flat = tf.reshape(conv3_pool, [-1, C3])\n    conv3_drop = tf.nn.dropout(conv3_flat, keep_prob=keep_prob)\n\n    full1 = tf.nn.relu(full_layer(conv3_flat, F1))\n    full1_drop = tf.nn.dropout(full1, keep_prob=keep_prob)\n\n    y_conv = full_layer(full1_drop, 10)\n\n    cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=y_conv, labels=y_))\n    train_step = tf.train.AdamOptimizer(5e-4).minimize(cross_entropy)\n\n    correct_prediction = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1))\n    accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))\n\n    def test(sess):\n        X = cifar.test.images.reshape(10, 1000, 32, 32, 3)\n        Y = cifar.test.labels.reshape(10, 1000, 10)\n        acc = np.mean([sess.run(accuracy, feed_dict={x: X[i], y_: Y[i], phase: False , keep_prob: 1.0})\n                       for i in range(10)])\n        print(\"Accuracy: {:.4}%\".format(acc * 100))\n\n    with tf.Session() as sess:\n        sess.run(tf.global_variables_initializer())\n\n        for i in range(STEPS):\n            batch = cifar.train.next_batch(BATCH_SIZE)\n            sess.run(train_step, feed_dict={x: batch[0], y_: batch[1], phase: True, keep_prob: 0.5})\n\n            if i % 500 == 0:\n                test(sess)\n\n        test(sess)\n\n\ndef create_cifar_image():\n    d = CifarDataManager()\n    print(\"Number of train images: {}\".format(len(d.train.images)))\n    print(\"Number of train labels: {}\".format(len(d.train.labels)))\n    print(\"Number of test images: {}\".format(len(d.test.images)))\n    print(\"Number of test images: {}\".format(len(d.test.labels)))\n    images = d.train.images\n    display_cifar(images, 10)\n```\n\n\n```python\ncreate_cifar_image()\n```\n\n    Number of train images: 50000\n    Number of train labels: 50000\n    Number of test images: 10000\n    Number of test images: 10000\n\n\n\n![png](output_3_1.png)\n\n\n\n```python\nrun_simple_net()\n```\n\n    Accuracy: 11.14%\n    Accuracy: 43.33%\n    Accuracy: 47.72%\n    Accuracy: 49.42%\n    Accuracy: 55.05%\n    Accuracy: 58.3%\n    Accuracy: 61.28%\n    Accuracy: 57.95%\n    Accuracy: 65.16%\n    Accuracy: 63.02%\n    Accuracy: 67.24%\n    Accuracy: 63.56%\n    Accuracy: 67.67%\n    Accuracy: 65.49%\n    Accuracy: 69.49%\n    Accuracy: 69.16%\n    Accuracy: 69.66%\n    Accuracy: 70.72%\n    Accuracy: 71.93%\n    Accuracy: 69.34%\n    Accuracy: 70.39%\n    Accuracy: 71.15%\n    Accuracy: 72.94%\n    Accuracy: 71.53%\n    Accuracy: 73.49%\n    Accuracy: 71.79%\n    Accuracy: 71.93%\n    Accuracy: 71.12%\n    Accuracy: 73.47%\n    Accuracy: 72.49%\n    Accuracy: 74.03%\n    Accuracy: 70.01%\n    Accuracy: 74.21%\n    Accuracy: 72.42%\n    Accuracy: 73.63%\n    Accuracy: 72.65%\n    Accuracy: 74.18%\n    Accuracy: 73.05%\n    Accuracy: 74.97%\n    Accuracy: 74.02%\n    Accuracy: 73.46%\n    Accuracy: 73.5%\n    Accuracy: 74.29%\n    Accuracy: 73.23%\n    Accuracy: 75.01%\n    Accuracy: 73.64%\n    Accuracy: 74.37%\n    Accuracy: 75.41%\n    Accuracy: 75.28%\n    Accuracy: 75.45%\n    Accuracy: 75.35%\n    Accuracy: 75.24%\n    Accuracy: 75.59%\n    Accuracy: 75.92%\n    Accuracy: 75.43%\n    Accuracy: 75.02%\n    Accuracy: 75.24%\n    Accuracy: 75.87%\n    Accuracy: 75.38%\n    Accuracy: 76.03%\n    Accuracy: 75.42%\n    Accuracy: 75.64%\n    Accuracy: 75.83%\n    Accuracy: 75.49%\n    Accuracy: 75.36%\n    Accuracy: 75.14%\n    Accuracy: 75.94%\n    Accuracy: 75.73%\n    Accuracy: 76.33%\n    Accuracy: 75.53%\n    Accuracy: 75.77%\n    Accuracy: 75.8%\n    Accuracy: 75.4%\n    Accuracy: 76.26%\n    Accuracy: 75.9%\n    Accuracy: 73.55%\n    Accuracy: 75.29%\n    Accuracy: 76.37%\n    Accuracy: 76.49%\n    Accuracy: 75.92%\n    Accuracy: 75.24%\n    Accuracy: 76.35%\n    Accuracy: 75.96%\n    Accuracy: 76.49%\n    Accuracy: 76.67%\n    Accuracy: 76.46%\n    Accuracy: 75.63%\n    Accuracy: 76.93%\n    Accuracy: 75.73%\n    Accuracy: 76.22%\n    Accuracy: 74.97%\n    Accuracy: 75.99%\n    Accuracy: 76.77%\n    Accuracy: 76.16%\n    Accuracy: 75.58%\n    Accuracy: 77.01%\n    Accuracy: 76.35%\n    Accuracy: 76.84%\n    Accuracy: 76.02%\n    Accuracy: 76.27%\n    Accuracy: 76.15%\n    Accuracy: 77.02%\n    Accuracy: 75.18%\n    Accuracy: 76.91%\n    Accuracy: 76.39%\n    Accuracy: 76.95%\n    Accuracy: 75.67%\n    Accuracy: 76.51%\n    Accuracy: 76.46%\n    Accuracy: 76.47%\n    Accuracy: 76.34%\n    Accuracy: 76.74%\n    Accuracy: 76.62%\n    Accuracy: 76.04%\n    Accuracy: 76.66%\n    Accuracy: 76.6%\n    Accuracy: 76.47%\n    Accuracy: 76.9%\n    Accuracy: 76.29%\n    Accuracy: 76.89%\n    Accuracy: 76.45%\n    Accuracy: 77.04%\n    Accuracy: 77.04%\n    Accuracy: 77.44%\n    Accuracy: 75.54%\n    Accuracy: 77.28%\n    Accuracy: 76.51%\n    Accuracy: 77.3%\n    Accuracy: 76.69%\n    Accuracy: 77.1%\n    Accuracy: 76.69%\n    Accuracy: 77.18%\n    Accuracy: 76.99%\n    Accuracy: 77.11%\n    Accuracy: 76.37%\n    Accuracy: 77.45%\n    Accuracy: 76.04%\n    Accuracy: 77.34%\n    Accuracy: 76.5%\n    Accuracy: 76.94%\n    Accuracy: 76.32%\n    Accuracy: 77.07%\n    Accuracy: 76.9%\n    Accuracy: 77.57%\n    Accuracy: 77.33%\n    Accuracy: 77.48%\n    Accuracy: 77.0%\n    Accuracy: 77.14%\n    Accuracy: 77.25%\n    Accuracy: 77.75%\n    Accuracy: 76.36%\n    Accuracy: 77.1%\n    Accuracy: 76.81%\n    Accuracy: 76.96%\n    Accuracy: 76.85%\n    Accuracy: 77.19%\n    Accuracy: 77.33%\n    Accuracy: 76.7%\n    Accuracy: 76.71%\n    Accuracy: 77.5%\n    Accuracy: 77.19%\n    Accuracy: 77.37%\n    Accuracy: 76.71%\n    Accuracy: 77.27%\n    Accuracy: 76.44%\n    Accuracy: 77.3%\n    Accuracy: 76.75%\n    Accuracy: 76.26%\n    Accuracy: 76.87%\n    Accuracy: 77.38%\n    Accuracy: 77.18%\n    Accuracy: 77.28%\n    Accuracy: 76.74%\n    Accuracy: 77.01%\n    Accuracy: 76.84%\n    Accuracy: 76.87%\n    Accuracy: 77.38%\n    Accuracy: 77.01%\n    Accuracy: 77.42%\n    Accuracy: 77.11%\n    Accuracy: 76.76%\n    Accuracy: 77.19%\n    Accuracy: 76.77%\n    Accuracy: 77.01%\n    Accuracy: 76.07%\n    Accuracy: 76.98%\n    Accuracy: 77.22%\n    Accuracy: 77.37%\n    Accuracy: 76.72%\n    Accuracy: 77.0%\n    Accuracy: 77.46%\n    Accuracy: 77.38%\n    Accuracy: 76.91%\n    Accuracy: 77.34%\n    Accuracy: 77.33%\n    Accuracy: 76.82%\n    Accuracy: 76.82%\n    Accuracy: 77.2%\n    Accuracy: 76.29%\n    Accuracy: 77.04%\n    Accuracy: 76.87%\n    Accuracy: 77.89%\n    Accuracy: 76.4%\n    Accuracy: 77.35%\n    Accuracy: 75.91%\n    Accuracy: 77.28%\n    Accuracy: 76.6%\n    Accuracy: 77.57%\n    Accuracy: 77.42%\n    Accuracy: 77.21%\n    Accuracy: 77.37%\n    Accuracy: 76.98%\n    Accuracy: 77.08%\n    Accuracy: 77.56%\n    Accuracy: 77.3%\n    Accuracy: 77.62%\n    Accuracy: 77.05%\n    Accuracy: 77.71%\n    Accuracy: 77.2%\n    Accuracy: 77.39%\n    Accuracy: 77.51%\n    Accuracy: 76.78%\n    Accuracy: 77.15%\n    Accuracy: 77.02%\n    Accuracy: 77.48%\n    Accuracy: 77.4%\n    Accuracy: 76.12%\n    Accuracy: 77.32%\n    Accuracy: 75.95%\n    Accuracy: 76.92%\n    Accuracy: 76.84%\n    Accuracy: 76.95%\n    Accuracy: 76.45%\n    Accuracy: 76.9%\n    Accuracy: 77.29%\n    Accuracy: 77.23%\n    Accuracy: 76.67%\n    Accuracy: 77.18%\n    Accuracy: 76.35%\n    Accuracy: 77.66%\n    Accuracy: 77.03%\n    Accuracy: 77.06%\n    Accuracy: 77.15%\n    Accuracy: 77.92%\n    Accuracy: 76.58%\n    Accuracy: 77.41%\n    Accuracy: 77.02%\n    Accuracy: 77.51%\n    Accuracy: 76.09%\n    Accuracy: 77.84%\n    Accuracy: 76.7%\n    Accuracy: 77.56%\n    Accuracy: 77.17%\n    Accuracy: 77.19%\n    Accuracy: 77.46%\n    Accuracy: 77.01%\n    Accuracy: 77.58%\n    Accuracy: 77.48%\n    Accuracy: 77.78%\n    Accuracy: 77.03%\n    Accuracy: 76.44%\n    Accuracy: 77.08%\n    Accuracy: 77.07%\n    Accuracy: 77.61%\n    Accuracy: 77.09%\n    Accuracy: 77.76%\n    Accuracy: 76.98%\n    Accuracy: 77.39%\n    Accuracy: 76.83%\n    Accuracy: 77.11%\n    Accuracy: 76.37%\n    Accuracy: 76.82%\n    Accuracy: 77.36%\n    Accuracy: 77.73%\n    Accuracy: 76.12%\n    Accuracy: 77.52%\n    Accuracy: 76.76%\n    Accuracy: 77.23%\n    Accuracy: 77.2%\n    Accuracy: 77.39%\n    Accuracy: 76.92%\n    Accuracy: 77.51%\n    Accuracy: 77.15%\n    Accuracy: 77.75%\n    Accuracy: 77.15%\n    Accuracy: 77.25%\n    Accuracy: 77.18%\n    Accuracy: 77.07%\n    Accuracy: 77.16%\n    Accuracy: 76.77%\n    Accuracy: 76.06%\n    Accuracy: 77.52%\n    Accuracy: 77.36%\n    Accuracy: 77.24%\n    Accuracy: 77.42%\n    Accuracy: 77.7%\n    Accuracy: 76.77%\n    Accuracy: 77.25%\n    Accuracy: 76.95%\n    Accuracy: 76.48%\n    Accuracy: 76.45%\n    Accuracy: 77.2%\n    Accuracy: 76.58%\n    Accuracy: 77.43%\n    Accuracy: 75.77%\n    Accuracy: 76.97%\n    Accuracy: 77.11%\n    Accuracy: 78.03%\n    Accuracy: 76.67%\n    Accuracy: 76.86%\n    Accuracy: 76.88%\n    Accuracy: 77.09%\n    Accuracy: 77.69%\n    Accuracy: 77.47%\n    Accuracy: 77.01%\n    Accuracy: 77.6%\n    Accuracy: 77.46%\n    Accuracy: 77.03%\n    Accuracy: 77.32%\n    Accuracy: 77.34%\n    Accuracy: 77.13%\n    Accuracy: 77.71%\n    Accuracy: 77.46%\n    Accuracy: 77.29%\n    Accuracy: 76.58%\n    Accuracy: 78.04%\n    Accuracy: 77.09%\n    Accuracy: 77.65%\n    Accuracy: 77.26%\n    Accuracy: 77.55%\n    Accuracy: 77.89%\n    Accuracy: 77.48%\n    Accuracy: 76.85%\n    Accuracy: 76.91%\n    Accuracy: 77.39%\n    Accuracy: 77.59%\n    Accuracy: 77.11%\n    Accuracy: 77.29%\n    Accuracy: 77.9%\n    Accuracy: 77.54%\n    Accuracy: 76.59%\n    Accuracy: 77.66%\n    Accuracy: 77.07%\n    Accuracy: 77.03%\n    Accuracy: 76.92%\n    Accuracy: 77.39%\n    Accuracy: 77.16%\n    Accuracy: 77.39%\n    Accuracy: 76.8%\n    Accuracy: 77.3%\n    Accuracy: 77.65%\n    Accuracy: 77.41%\n    Accuracy: 77.11%\n    Accuracy: 78.0%\n    Accuracy: 77.34%\n    Accuracy: 77.9%\n    Accuracy: 76.84%\n    Accuracy: 77.41%\n    Accuracy: 77.19%\n    Accuracy: 77.61%\n    Accuracy: 77.16%\n    Accuracy: 77.89%\n    Accuracy: 77.14%\n    Accuracy: 77.59%\n    Accuracy: 76.52%\n    Accuracy: 77.85%\n    Accuracy: 77.05%\n    Accuracy: 77.56%\n    Accuracy: 77.59%\n    Accuracy: 76.75%\n    Accuracy: 77.67%\n    Accuracy: 77.74%\n    Accuracy: 76.49%\n    Accuracy: 77.43%\n    Accuracy: 77.49%\n    Accuracy: 76.95%\n    Accuracy: 77.5%\n    Accuracy: 77.31%\n    Accuracy: 77.06%\n    Accuracy: 77.73%\n    Accuracy: 77.42%\n    Accuracy: 77.35%\n    Accuracy: 77.2%\n    Accuracy: 77.65%\n    Accuracy: 77.14%\n    Accuracy: 77.02%\n    Accuracy: 77.09%\n    Accuracy: 77.31%\n    Accuracy: 77.54%\n    Accuracy: 77.37%\n    Accuracy: 77.19%\n    Accuracy: 77.58%\n    Accuracy: 77.18%\n    Accuracy: 78.04%\n    Accuracy: 77.18%\n    Accuracy: 78.09%\n    Accuracy: 76.67%\n    Accuracy: 78.05%\n    Accuracy: 77.47%\n    Accuracy: 77.51%\n    Accuracy: 77.78%\n    Accuracy: 76.92%\n    Accuracy: 77.21%\n    Accuracy: 77.65%\n    Accuracy: 77.1%\n    Accuracy: 78.08%\n    Accuracy: 77.36%\n    Accuracy: 77.07%\n    Accuracy: 77.34%\n    Accuracy: 77.86%\n    Accuracy: 76.9%\n    Accuracy: 77.7%\n    Accuracy: 77.91%\n    Accuracy: 77.1%\n    Accuracy: 77.32%\n    Accuracy: 77.53%\n    Accuracy: 77.59%\n    Accuracy: 77.22%\n    Accuracy: 77.32%\n    Accuracy: 77.48%\n    Accuracy: 77.35%\n    Accuracy: 77.29%\n    Accuracy: 77.45%\n    Accuracy: 77.5%\n    Accuracy: 76.78%\n    Accuracy: 77.8%\n    Accuracy: 77.14%\n    Accuracy: 77.39%\n    Accuracy: 76.39%\n    Accuracy: 77.81%\n    Accuracy: 76.83%\n    Accuracy: 77.12%\n    Accuracy: 76.96%\n    Accuracy: 77.72%\n    Accuracy: 77.18%\n    Accuracy: 77.66%\n    Accuracy: 77.59%\n    Accuracy: 77.98%\n    Accuracy: 76.89%\n    Accuracy: 77.33%\n    Accuracy: 76.9%\n    Accuracy: 77.57%\n    Accuracy: 77.87%\n    Accuracy: 77.16%\n    Accuracy: 77.58%\n    Accuracy: 78.44%\n    Accuracy: 77.33%\n    Accuracy: 77.4%\n    Accuracy: 77.49%\n    Accuracy: 77.69%\n    Accuracy: 76.34%\n    Accuracy: 77.7%\n    Accuracy: 77.42%\n    Accuracy: 77.62%\n    Accuracy: 77.3%\n    Accuracy: 77.13%\n    Accuracy: 77.11%\n    Accuracy: 77.97%\n    Accuracy: 77.76%\n    Accuracy: 78.15%\n    Accuracy: 77.69%\n    Accuracy: 78.06%\n    Accuracy: 77.78%\n    Accuracy: 77.52%\n    Accuracy: 77.64%\n    Accuracy: 77.16%\n    Accuracy: 77.25%\n    Accuracy: 77.95%\n    Accuracy: 76.6%\n    Accuracy: 77.89%\n    Accuracy: 77.38%\n    Accuracy: 77.46%\n    Accuracy: 77.68%\n    Accuracy: 76.93%\n    Accuracy: 77.07%\n    Accuracy: 77.93%\n    Accuracy: 78.11%\n    Accuracy: 77.95%\n    Accuracy: 78.18%\n    Accuracy: 77.4%\n    Accuracy: 76.7%\n    Accuracy: 77.65%\n    Accuracy: 77.51%\n    Accuracy: 77.04%\n    Accuracy: 77.11%\n    Accuracy: 77.69%\n    Accuracy: 77.25%\n    Accuracy: 77.3%\n    Accuracy: 77.31%\n    Accuracy: 77.53%\n    Accuracy: 77.56%\n    Accuracy: 77.13%\n    Accuracy: 77.47%\n    Accuracy: 77.63%\n    Accuracy: 76.85%\n    Accuracy: 77.42%\n    Accuracy: 77.49%\n    Accuracy: 76.93%\n    Accuracy: 76.89%\n    Accuracy: 77.7%\n    Accuracy: 77.32%\n    Accuracy: 77.73%\n    Accuracy: 77.42%\n    Accuracy: 77.5%\n    Accuracy: 76.96%\n    Accuracy: 77.39%\n    Accuracy: 77.46%\n    Accuracy: 77.9%\n    Accuracy: 76.7%\n    Accuracy: 77.43%\n    Accuracy: 77.5%\n    Accuracy: 77.71%\n    Accuracy: 77.63%\n    Accuracy: 77.65%\n    Accuracy: 78.0%\n    Accuracy: 77.16%\n    Accuracy: 76.95%\n    Accuracy: 77.47%\n    Accuracy: 77.5%\n    Accuracy: 77.64%\n    Accuracy: 77.96%\n    Accuracy: 77.29%\n    Accuracy: 77.24%\n    Accuracy: 77.24%\n    Accuracy: 77.33%\n    Accuracy: 77.23%\n    Accuracy: 77.2%\n    Accuracy: 77.46%\n    Accuracy: 76.82%\n    Accuracy: 77.33%\n    Accuracy: 77.74%\n    Accuracy: 77.97%\n    Accuracy: 77.0%\n    Accuracy: 77.36%\n    Accuracy: 77.59%\n    Accuracy: 77.24%\n    Accuracy: 76.67%\n    Accuracy: 77.55%\n    Accuracy: 76.98%\n    Accuracy: 77.41%\n    Accuracy: 77.08%\n    Accuracy: 77.37%\n    Accuracy: 77.31%\n    Accuracy: 77.52%\n    Accuracy: 77.87%\n    Accuracy: 77.74%\n    Accuracy: 77.74%\n    Accuracy: 76.83%\n    Accuracy: 77.23%\n    Accuracy: 77.24%\n    Accuracy: 77.24%\n    Accuracy: 77.49%\n    Accuracy: 77.84%\n    Accuracy: 77.62%\n    Accuracy: 77.79%\n    Accuracy: 77.67%\n    Accuracy: 77.63%\n    Accuracy: 77.74%\n    Accuracy: 77.43%\n    Accuracy: 77.78%\n    Accuracy: 76.34%\n    Accuracy: 77.73%\n    Accuracy: 78.0%\n    Accuracy: 77.45%\n    Accuracy: 77.62%\n    Accuracy: 77.67%\n    Accuracy: 77.4%\n    Accuracy: 77.54%\n    Accuracy: 77.54%\n    Accuracy: 77.48%\n    Accuracy: 77.63%\n    Accuracy: 77.78%\n    Accuracy: 78.03%\n    Accuracy: 77.6%\n    Accuracy: 78.3%\n    Accuracy: 77.76%\n    Accuracy: 77.96%\n    Accuracy: 77.94%\n    Accuracy: 77.45%\n    Accuracy: 76.43%\n    Accuracy: 76.64%\n    Accuracy: 77.86%\n    Accuracy: 77.85%\n    Accuracy: 77.75%\n    Accuracy: 77.34%\n    Accuracy: 78.06%\n    Accuracy: 77.49%\n    Accuracy: 77.66%\n    Accuracy: 77.88%\n    Accuracy: 77.55%\n    Accuracy: 77.12%\n    Accuracy: 77.54%\n    Accuracy: 77.39%\n    Accuracy: 77.59%\n    Accuracy: 77.45%\n    Accuracy: 77.52%\n    Accuracy: 77.53%\n    Accuracy: 77.92%\n    Accuracy: 77.19%\n    Accuracy: 77.87%\n    Accuracy: 77.18%\n    Accuracy: 77.33%\n    Accuracy: 77.33%\n    Accuracy: 77.63%\n    Accuracy: 77.48%\n    Accuracy: 77.5%\n    Accuracy: 77.23%\n    Accuracy: 77.38%\n    Accuracy: 77.32%\n    Accuracy: 77.09%\n    Accuracy: 76.87%\n    Accuracy: 77.57%\n    Accuracy: 77.57%\n    Accuracy: 77.74%\n    Accuracy: 77.59%\n    Accuracy: 77.56%\n    Accuracy: 77.06%\n    Accuracy: 77.24%\n    Accuracy: 77.34%\n    Accuracy: 77.25%\n    Accuracy: 77.43%\n    Accuracy: 77.53%\n    Accuracy: 77.22%\n    Accuracy: 77.34%\n    Accuracy: 77.21%\n    Accuracy: 78.01%\n    Accuracy: 77.44%\n    Accuracy: 77.85%\n    Accuracy: 77.48%\n    Accuracy: 77.34%\n    Accuracy: 77.66%\n    Accuracy: 77.62%\n    Accuracy: 77.21%\n    Accuracy: 77.37%\n    Accuracy: 77.47%\n    Accuracy: 77.63%\n    Accuracy: 77.52%\n    Accuracy: 77.89%\n    Accuracy: 77.95%\n    Accuracy: 77.79%\n    Accuracy: 76.77%\n    Accuracy: 78.01%\n    Accuracy: 77.15%\n    Accuracy: 77.62%\n    Accuracy: 77.27%\n    Accuracy: 77.32%\n    Accuracy: 78.23%\n    Accuracy: 77.61%\n    Accuracy: 77.86%\n    Accuracy: 77.8%\n    Accuracy: 77.27%\n    Accuracy: 77.81%\n    Accuracy: 77.49%\n    Accuracy: 77.79%\n    Accuracy: 78.23%\n    Accuracy: 77.98%\n    Accuracy: 77.45%\n    Accuracy: 77.45%\n    Accuracy: 77.6%\n    Accuracy: 77.85%\n    Accuracy: 77.66%\n    Accuracy: 77.15%\n    Accuracy: 77.13%\n    Accuracy: 78.27%\n    Accuracy: 76.74%\n    Accuracy: 78.08%\n    Accuracy: 77.38%\n    Accuracy: 77.73%\n    Accuracy: 77.07%\n    Accuracy: 77.63%\n    Accuracy: 77.45%\n    Accuracy: 77.76%\n    Accuracy: 77.12%\n    Accuracy: 77.38%\n    Accuracy: 77.42%\n    Accuracy: 77.8%\n    Accuracy: 77.32%\n    Accuracy: 77.75%\n    Accuracy: 77.72%\n    Accuracy: 76.99%\n    Accuracy: 77.74%\n    Accuracy: 77.98%\n    Accuracy: 77.52%\n    Accuracy: 77.87%\n    Accuracy: 76.76%\n    Accuracy: 77.63%\n    Accuracy: 77.49%\n    Accuracy: 78.15%\n    Accuracy: 77.35%\n    Accuracy: 77.84%\n    Accuracy: 77.72%\n    Accuracy: 76.92%\n    Accuracy: 77.85%\n    Accuracy: 77.91%\n    Accuracy: 77.99%\n    Accuracy: 77.22%\n    Accuracy: 77.01%\n    Accuracy: 77.61%\n    Accuracy: 77.7%\n    Accuracy: 77.95%\n    Accuracy: 77.74%\n    Accuracy: 77.69%\n    Accuracy: 77.76%\n    Accuracy: 78.15%\n    Accuracy: 77.49%\n    Accuracy: 77.87%\n    Accuracy: 77.28%\n    Accuracy: 77.3%\n    Accuracy: 77.01%\n    Accuracy: 77.65%\n    Accuracy: 77.64%\n    Accuracy: 76.71%\n    Accuracy: 77.65%\n    Accuracy: 78.07%\n    Accuracy: 77.83%\n    Accuracy: 77.82%\n    Accuracy: 77.06%\n    Accuracy: 77.25%\n    Accuracy: 77.16%\n    Accuracy: 78.02%\n    Accuracy: 77.04%\n    Accuracy: 77.97%\n    Accuracy: 77.53%\n    Accuracy: 77.46%\n    Accuracy: 76.77%\n    Accuracy: 77.77%\n    Accuracy: 77.54%\n    Accuracy: 77.95%\n    Accuracy: 77.5%\n    Accuracy: 78.18%\n    Accuracy: 76.88%\n    Accuracy: 77.97%\n    Accuracy: 77.79%\n    Accuracy: 77.46%\n    Accuracy: 77.78%\n    Accuracy: 77.3%\n    Accuracy: 77.1%\n    Accuracy: 78.01%\n    Accuracy: 77.86%\n    Accuracy: 77.83%\n    Accuracy: 77.44%\n    Accuracy: 77.8%\n    Accuracy: 77.5%\n    Accuracy: 77.42%\n    Accuracy: 77.8%\n    Accuracy: 77.14%\n    Accuracy: 77.91%\n    Accuracy: 77.27%\n    Accuracy: 77.9%\n    Accuracy: 78.32%\n    Accuracy: 77.5%\n    Accuracy: 77.8%\n    Accuracy: 77.67%\n    Accuracy: 77.43%\n    Accuracy: 78.0%\n    Accuracy: 78.14%\n    Accuracy: 77.28%\n    Accuracy: 77.75%\n    Accuracy: 77.62%\n    Accuracy: 77.64%\n    Accuracy: 77.38%\n    Accuracy: 77.97%\n    Accuracy: 77.72%\n    Accuracy: 77.84%\n    Accuracy: 77.61%\n    Accuracy: 77.76%\n    Accuracy: 77.27%\n    Accuracy: 77.45%\n    Accuracy: 77.36%\n    Accuracy: 77.78%\n    Accuracy: 76.75%\n    Accuracy: 77.53%\n    Accuracy: 77.4%\n    Accuracy: 77.8%\n    Accuracy: 77.69%\n    Accuracy: 77.7%\n    Accuracy: 77.76%\n    Accuracy: 77.75%\n    Accuracy: 77.13%\n    Accuracy: 77.66%\n    Accuracy: 77.35%\n    Accuracy: 77.69%\n    Accuracy: 77.44%\n    Accuracy: 77.94%\n    Accuracy: 77.53%\n    Accuracy: 78.04%\n    Accuracy: 77.87%\n    Accuracy: 77.48%\n    Accuracy: 76.9%\n    Accuracy: 77.6%\n    Accuracy: 77.72%\n    Accuracy: 77.47%\n    Accuracy: 77.63%\n    Accuracy: 77.75%\n    Accuracy: 77.31%\n    Accuracy: 77.37%\n    Accuracy: 77.96%\n    Accuracy: 78.04%\n    Accuracy: 77.54%\n    Accuracy: 77.46%\n    Accuracy: 77.84%\n    Accuracy: 77.73%\n    Accuracy: 77.0%\n    Accuracy: 77.22%\n    Accuracy: 77.55%\n    Accuracy: 77.33%\n    Accuracy: 77.57%\n    Accuracy: 77.71%\n    Accuracy: 77.35%\n    Accuracy: 77.46%\n    Accuracy: 77.11%\n    Accuracy: 77.42%\n    Accuracy: 77.25%\n    Accuracy: 77.22%\n    Accuracy: 77.34%\n    Accuracy: 77.83%\n    Accuracy: 77.64%\n    Accuracy: 77.52%\n    Accuracy: 77.02%\n    Accuracy: 77.55%\n    Accuracy: 77.49%\n    Accuracy: 77.35%\n    Accuracy: 77.51%\n    Accuracy: 77.84%\n    Accuracy: 77.5%\n    Accuracy: 76.97%\n    Accuracy: 77.14%\n    Accuracy: 77.54%\n    Accuracy: 77.03%\n    Accuracy: 77.53%\n    Accuracy: 77.51%\n    Accuracy: 77.51%\n    Accuracy: 77.44%\n    Accuracy: 77.94%\n    Accuracy: 76.65%\n    Accuracy: 77.81%\n    Accuracy: 77.66%\n    Accuracy: 77.63%\n    Accuracy: 77.76%\n    Accuracy: 77.52%\n    Accuracy: 77.42%\n    Accuracy: 77.25%\n    Accuracy: 77.52%\n    Accuracy: 77.52%\n    Accuracy: 76.16%\n    Accuracy: 77.63%\n    Accuracy: 77.9%\n    Accuracy: 77.75%\n    Accuracy: 77.22%\n    Accuracy: 77.81%\n    Accuracy: 77.2%\n    Accuracy: 77.59%\n    Accuracy: 77.21%\n    Accuracy: 77.56%\n    Accuracy: 77.54%\n    Accuracy: 77.53%\n    Accuracy: 77.76%\n    Accuracy: 77.5%\n    Accuracy: 77.12%\n    Accuracy: 77.44%\n    Accuracy: 77.56%\n    Accuracy: 77.06%\n    Accuracy: 77.66%\n    Accuracy: 77.96%\n    Accuracy: 78.33%\n    Accuracy: 77.6%\n    Accuracy: 77.42%\n    Accuracy: 77.45%\n    Accuracy: 77.23%\n    Accuracy: 78.16%\n    Accuracy: 77.67%\n    Accuracy: 77.1%\n    Accuracy: 77.38%\n    Accuracy: 77.6%\n    Accuracy: 76.9%\n    Accuracy: 77.92%\n    Accuracy: 77.45%\n    Accuracy: 77.5%\n    Accuracy: 77.6%\n    Accuracy: 77.9%\n    Accuracy: 77.02%\n    Accuracy: 78.29%\n    Accuracy: 77.17%\n    Accuracy: 77.46%\n    Accuracy: 78.07%\n    Accuracy: 77.45%\n    Accuracy: 77.21%\n    Accuracy: 78.1%\n    Accuracy: 77.41%\n    Accuracy: 77.46%\n    Accuracy: 77.96%\n    Accuracy: 77.42%\n    Accuracy: 77.56%\n    Accuracy: 77.35%\n    Accuracy: 77.85%\n    Accuracy: 77.62%\n    Accuracy: 77.67%\n    Accuracy: 77.11%\n    Accuracy: 77.43%\n    Accuracy: 77.8%\n    Accuracy: 77.5%\n    Accuracy: 77.04%\n    Accuracy: 76.92%\n    Accuracy: 78.38%\n    Accuracy: 77.77%\n    Accuracy: 77.29%\n    Accuracy: 76.72%\n    Accuracy: 77.91%\n    Accuracy: 77.65%\n    Accuracy: 77.46%\n    Accuracy: 77.88%\n    Accuracy: 77.5%\n    Accuracy: 77.75%\n    Accuracy: 77.43%\n    Accuracy: 77.58%\n    Accuracy: 77.49%\n    Accuracy: 77.82%\n    Accuracy: 77.93%\n    Accuracy: 77.62%\n    Accuracy: 77.89%\n    Accuracy: 78.26%\n    Accuracy: 77.87%\n    Accuracy: 76.91%\n    Accuracy: 77.01%\n    Accuracy: 77.6%\n    Accuracy: 77.69%\n    Accuracy: 77.59%\n    Accuracy: 77.82%\n    Accuracy: 76.75%\n    Accuracy: 77.38%\n    Accuracy: 77.57%\n    Accuracy: 77.48%\n    Accuracy: 77.45%\n    Accuracy: 77.36%\n    Accuracy: 77.85%\n    Accuracy: 77.57%\n    Accuracy: 77.46%\n    Accuracy: 77.52%\n    Accuracy: 77.29%\n    Accuracy: 77.32%\n    Accuracy: 76.74%\n    Accuracy: 77.56%\n    Accuracy: 77.37%\n    Accuracy: 76.91%\n    Accuracy: 77.07%\n    Accuracy: 77.96%\n    Accuracy: 77.53%\n    Accuracy: 77.91%\n    Accuracy: 76.82%\n    Accuracy: 77.88%\n    Accuracy: 77.34%\n    Accuracy: 77.86%\n    Accuracy: 77.51%\n    Accuracy: 78.1%\n    Accuracy: 78.08%\n    Accuracy: 77.31%\n    Accuracy: 77.59%\n    Accuracy: 78.13%\n    Accuracy: 77.11%\n    Accuracy: 77.72%\n    Accuracy: 77.77%\n    Accuracy: 78.05%\n    Accuracy: 77.2%\n    Accuracy: 78.22%\n    Accuracy: 77.1%\n    Accuracy: 78.29%\n    Accuracy: 77.58%\n    Accuracy: 78.13%\n    Accuracy: 78.06%\n    Accuracy: 77.31%\n    Accuracy: 77.94%\n    Accuracy: 77.02%\n    Accuracy: 77.83%\n    Accuracy: 77.94%\n    Accuracy: 77.62%\n    Accuracy: 77.7%\n    Accuracy: 76.98%\n    Accuracy: 77.56%\n    Accuracy: 77.5%\n    Accuracy: 77.56%\n    Accuracy: 77.56%\n    Accuracy: 77.57%\n    Accuracy: 77.74%\n    Accuracy: 77.51%\n    Accuracy: 77.67%\n    Accuracy: 77.78%\n    Accuracy: 77.99%\n    Accuracy: 78.0%\n    Accuracy: 77.78%\n    Accuracy: 77.62%\n    Accuracy: 77.94%\n    Accuracy: 78.16%\n    Accuracy: 77.6%\n    Accuracy: 77.84%\n    Accuracy: 77.34%\n    Accuracy: 77.91%\n    Accuracy: 77.12%\n    Accuracy: 78.07%\n    Accuracy: 77.25%\n    Accuracy: 78.01%\n    Accuracy: 77.71%\n\n\n\n```python\nbuild_second_net()\n```\n\n    Accuracy: 9.82%\n    Accuracy: 10.52%\n    Accuracy: 22.65%\n    Accuracy: 29.57%\n    Accuracy: 33.35%\n    Accuracy: 38.66%\n    Accuracy: 40.16%\n    Accuracy: 44.4%\n    Accuracy: 46.93%\n    Accuracy: 50.96%\n    Accuracy: 52.52%\n    Accuracy: 53.33%\n    Accuracy: 53.51%\n    Accuracy: 58.63%\n    Accuracy: 57.96%\n    Accuracy: 61.02%\n    Accuracy: 60.62%\n    Accuracy: 62.94%\n    Accuracy: 61.72%\n    Accuracy: 62.72%\n    Accuracy: 64.35%\n    Accuracy: 66.08%\n    Accuracy: 65.25%\n    Accuracy: 65.11%\n    Accuracy: 62.51%\n    Accuracy: 65.58%\n    Accuracy: 67.31%\n    Accuracy: 66.46%\n    Accuracy: 67.74%\n    Accuracy: 67.92%\n    Accuracy: 65.42%\n    Accuracy: 67.59%\n    Accuracy: 69.62%\n    Accuracy: 67.41%\n    Accuracy: 69.89%\n    Accuracy: 71.57%\n    Accuracy: 71.89%\n    Accuracy: 69.72%\n    Accuracy: 70.15%\n    Accuracy: 71.4%\n    Accuracy: 72.22%\n    Accuracy: 71.53%\n    Accuracy: 74.22%\n    Accuracy: 73.18%\n    Accuracy: 72.06%\n    Accuracy: 74.23%\n    Accuracy: 72.76%\n    Accuracy: 74.38%\n    Accuracy: 74.93%\n    Accuracy: 76.22%\n    Accuracy: 75.47%\n    Accuracy: 74.69%\n    Accuracy: 75.1%\n    Accuracy: 75.57%\n    Accuracy: 77.64%\n    Accuracy: 77.76%\n    Accuracy: 76.48%\n    Accuracy: 77.21%\n    Accuracy: 77.57%\n    Accuracy: 76.31%\n    Accuracy: 76.89%\n    Accuracy: 77.87%\n    Accuracy: 77.57%\n    Accuracy: 77.72%\n    Accuracy: 78.21%\n    Accuracy: 78.57%\n    Accuracy: 77.19%\n    Accuracy: 78.65%\n    Accuracy: 78.48%\n    Accuracy: 77.19%\n    Accuracy: 77.36%\n    Accuracy: 78.16%\n    Accuracy: 79.17%\n    Accuracy: 78.99%\n    Accuracy: 78.95%\n    Accuracy: 79.07%\n    Accuracy: 77.81%\n    Accuracy: 78.78%\n    Accuracy: 79.13%\n    Accuracy: 76.86%\n    Accuracy: 79.89%\n    Accuracy: 79.15%\n    Accuracy: 78.51%\n    Accuracy: 79.15%\n    Accuracy: 78.59%\n    Accuracy: 79.0%\n    Accuracy: 78.63%\n    Accuracy: 80.26%\n    Accuracy: 79.98%\n    Accuracy: 79.85%\n    Accuracy: 79.95%\n    Accuracy: 79.66%\n    Accuracy: 78.83%\n    Accuracy: 78.91%\n    Accuracy: 79.96%\n    Accuracy: 79.58%\n    Accuracy: 79.32%\n    Accuracy: 78.88%\n    Accuracy: 80.2%\n    Accuracy: 80.27%\n    Accuracy: 78.93%\n    Accuracy: 79.94%\n    Accuracy: 79.71%\n    Accuracy: 80.42%\n    Accuracy: 77.77%\n    Accuracy: 79.5%\n    Accuracy: 80.66%\n    Accuracy: 80.65%\n    Accuracy: 80.31%\n    Accuracy: 80.28%\n    Accuracy: 79.56%\n    Accuracy: 79.57%\n    Accuracy: 80.37%\n    Accuracy: 80.11%\n    Accuracy: 80.32%\n    Accuracy: 81.32%\n    Accuracy: 79.94%\n    Accuracy: 80.9%\n    Accuracy: 80.59%\n    Accuracy: 80.71%\n    Accuracy: 81.48%\n    Accuracy: 80.06%\n    Accuracy: 80.6%\n    Accuracy: 80.98%\n    Accuracy: 80.32%\n    Accuracy: 79.55%\n    Accuracy: 80.86%\n    Accuracy: 80.06%\n    Accuracy: 80.66%\n    Accuracy: 80.34%\n    Accuracy: 79.55%\n    Accuracy: 81.42%\n    Accuracy: 81.39%\n    Accuracy: 81.13%\n    Accuracy: 81.21%\n    Accuracy: 82.0%\n    Accuracy: 81.5%\n    Accuracy: 80.27%\n    Accuracy: 80.35%\n    Accuracy: 79.69%\n    Accuracy: 80.9%\n    Accuracy: 80.4%\n    Accuracy: 80.59%\n    Accuracy: 80.36%\n    Accuracy: 80.93%\n    Accuracy: 80.71%\n    Accuracy: 79.7%\n    Accuracy: 80.9%\n    Accuracy: 80.21%\n    Accuracy: 79.62%\n    Accuracy: 81.7%\n    Accuracy: 78.51%\n    Accuracy: 79.92%\n    Accuracy: 81.33%\n    Accuracy: 78.73%\n    Accuracy: 81.65%\n    Accuracy: 81.22%\n    Accuracy: 80.86%\n    Accuracy: 81.08%\n    Accuracy: 80.33%\n    Accuracy: 80.21%\n    Accuracy: 80.43%\n    Accuracy: 81.08%\n    Accuracy: 80.37%\n    Accuracy: 81.82%\n    Accuracy: 80.59%\n    Accuracy: 81.67%\n    Accuracy: 81.27%\n    Accuracy: 80.89%\n    Accuracy: 81.17%\n    Accuracy: 82.1%\n    Accuracy: 81.05%\n    Accuracy: 79.93%\n    Accuracy: 81.32%\n    Accuracy: 80.78%\n    Accuracy: 81.36%\n    Accuracy: 81.54%\n    Accuracy: 81.51%\n    Accuracy: 80.38%\n    Accuracy: 81.05%\n    Accuracy: 80.92%\n    Accuracy: 81.39%\n    Accuracy: 81.63%\n    Accuracy: 80.56%\n    Accuracy: 82.28%\n    Accuracy: 81.97%\n    Accuracy: 81.82%\n    Accuracy: 81.5%\n    Accuracy: 80.72%\n    Accuracy: 81.3%\n    Accuracy: 81.01%\n    Accuracy: 80.72%\n    Accuracy: 80.79%\n    Accuracy: 81.13%\n    Accuracy: 81.03%\n    Accuracy: 81.9%\n    Accuracy: 81.72%\n    Accuracy: 81.71%\n    Accuracy: 80.01%\n    Accuracy: 82.06%\n    Accuracy: 81.37%\n    Accuracy: 81.81%\n    Accuracy: 81.8%\n    Accuracy: 81.83%\n    Accuracy: 82.19%\n    Accuracy: 82.21%\n    Accuracy: 82.0%\n    Accuracy: 81.92%\n    Accuracy: 81.61%\n    Accuracy: 81.14%\n    Accuracy: 82.18%\n    Accuracy: 81.92%\n    Accuracy: 82.3%\n    Accuracy: 80.84%\n    Accuracy: 81.48%\n    Accuracy: 81.22%\n    Accuracy: 82.14%\n    Accuracy: 80.44%\n    Accuracy: 81.6%\n    Accuracy: 81.72%\n    Accuracy: 81.07%\n    Accuracy: 81.62%\n    Accuracy: 81.45%\n    Accuracy: 81.97%\n    Accuracy: 81.07%\n    Accuracy: 82.14%\n    Accuracy: 82.13%\n    Accuracy: 81.9%\n    Accuracy: 82.01%\n    Accuracy: 82.16%\n    Accuracy: 80.7%\n    Accuracy: 82.16%\n    Accuracy: 81.24%\n    Accuracy: 81.57%\n    Accuracy: 81.67%\n    Accuracy: 81.76%\n    Accuracy: 81.55%\n    Accuracy: 81.53%\n    Accuracy: 81.22%\n    Accuracy: 81.81%\n    Accuracy: 81.83%\n    Accuracy: 82.13%\n    Accuracy: 82.01%\n    Accuracy: 81.49%\n    Accuracy: 81.59%\n    Accuracy: 82.25%\n    Accuracy: 81.81%\n    Accuracy: 81.91%\n    Accuracy: 79.91%\n    Accuracy: 80.52%\n    Accuracy: 82.01%\n    Accuracy: 82.3%\n    Accuracy: 81.84%\n    Accuracy: 81.34%\n    Accuracy: 82.23%\n    Accuracy: 81.67%\n    Accuracy: 80.8%\n    Accuracy: 82.24%\n    Accuracy: 81.01%\n    Accuracy: 81.52%\n    Accuracy: 82.5%\n    Accuracy: 81.5%\n    Accuracy: 81.65%\n    Accuracy: 82.2%\n    Accuracy: 81.92%\n    Accuracy: 81.64%\n    Accuracy: 81.71%\n    Accuracy: 82.06%\n    Accuracy: 81.5%\n    Accuracy: 81.68%\n    Accuracy: 82.43%\n    Accuracy: 81.71%\n    Accuracy: 80.84%\n    Accuracy: 81.11%\n    Accuracy: 82.12%\n    Accuracy: 81.43%\n    Accuracy: 80.94%\n    Accuracy: 81.72%\n    Accuracy: 82.26%\n    Accuracy: 82.12%\n    Accuracy: 81.37%\n    Accuracy: 81.04%\n    Accuracy: 82.02%\n    Accuracy: 81.63%\n    Accuracy: 81.5%\n    Accuracy: 82.34%\n    Accuracy: 80.9%\n    Accuracy: 81.66%\n    Accuracy: 81.9%\n    Accuracy: 81.99%\n    Accuracy: 80.4%\n    Accuracy: 82.35%\n    Accuracy: 80.83%\n    Accuracy: 82.15%\n    Accuracy: 81.66%\n    Accuracy: 81.5%\n    Accuracy: 82.02%\n    Accuracy: 81.45%\n    Accuracy: 81.28%\n    Accuracy: 81.08%\n    Accuracy: 81.16%\n    Accuracy: 82.13%\n    Accuracy: 81.85%\n    Accuracy: 81.96%\n    Accuracy: 81.9%\n    Accuracy: 82.01%\n    Accuracy: 81.91%\n    Accuracy: 81.41%\n    Accuracy: 81.16%\n    Accuracy: 81.65%\n    Accuracy: 82.29%\n    Accuracy: 82.11%\n    Accuracy: 81.46%\n    Accuracy: 82.61%\n    Accuracy: 82.21%\n    Accuracy: 81.85%\n    Accuracy: 82.41%\n    Accuracy: 80.74%\n    Accuracy: 81.12%\n    Accuracy: 81.85%\n    Accuracy: 81.95%\n    Accuracy: 82.23%\n    Accuracy: 81.88%\n    Accuracy: 82.13%\n    Accuracy: 81.88%\n    Accuracy: 82.0%\n    Accuracy: 81.01%\n    Accuracy: 81.08%\n    Accuracy: 81.3%\n    Accuracy: 81.19%\n    Accuracy: 81.39%\n    Accuracy: 81.16%\n    Accuracy: 81.73%\n    Accuracy: 81.98%\n    Accuracy: 81.06%\n    Accuracy: 81.32%\n    Accuracy: 81.64%\n    Accuracy: 81.32%\n    Accuracy: 82.09%\n    Accuracy: 81.84%\n    Accuracy: 81.4%\n    Accuracy: 81.96%\n    Accuracy: 82.08%\n    Accuracy: 82.3%\n    Accuracy: 81.59%\n    Accuracy: 81.25%\n    Accuracy: 81.23%\n    Accuracy: 82.52%\n    Accuracy: 81.72%\n    Accuracy: 82.3%\n    Accuracy: 82.04%\n    Accuracy: 82.1%\n    Accuracy: 82.41%\n    Accuracy: 81.41%\n    Accuracy: 82.26%\n    Accuracy: 81.14%\n    Accuracy: 82.14%\n    Accuracy: 81.78%\n    Accuracy: 82.62%\n    Accuracy: 82.0%\n    Accuracy: 81.02%\n    Accuracy: 81.94%\n    Accuracy: 81.92%\n    Accuracy: 82.29%\n    Accuracy: 81.8%\n    Accuracy: 82.39%\n    Accuracy: 82.3%\n    Accuracy: 81.64%\n    Accuracy: 81.46%\n    Accuracy: 81.06%\n    Accuracy: 82.14%\n    Accuracy: 81.61%\n    Accuracy: 81.61%\n    Accuracy: 81.69%\n    Accuracy: 81.69%\n    Accuracy: 82.15%\n    Accuracy: 82.02%\n    Accuracy: 82.06%\n    Accuracy: 82.57%\n    Accuracy: 81.51%\n    Accuracy: 81.88%\n    Accuracy: 81.94%\n    Accuracy: 81.16%\n    Accuracy: 81.4%\n    Accuracy: 82.03%\n    Accuracy: 82.09%\n    Accuracy: 82.07%\n    Accuracy: 82.01%\n    Accuracy: 82.65%\n    Accuracy: 82.13%\n    Accuracy: 81.54%\n    Accuracy: 81.62%\n    Accuracy: 82.84%\n    Accuracy: 82.43%\n    Accuracy: 82.25%\n    Accuracy: 82.7%\n    Accuracy: 81.38%\n    Accuracy: 81.97%\n    Accuracy: 82.1%\n    Accuracy: 82.18%\n    Accuracy: 80.99%\n    Accuracy: 81.79%\n    Accuracy: 81.14%\n    Accuracy: 82.37%\n    Accuracy: 82.03%\n    Accuracy: 82.18%\n    Accuracy: 82.37%\n    Accuracy: 82.28%\n    Accuracy: 82.0%\n    Accuracy: 82.07%\n    Accuracy: 80.91%\n    Accuracy: 82.23%\n    Accuracy: 81.93%\n    Accuracy: 82.45%\n    Accuracy: 80.57%\n    Accuracy: 82.74%\n    Accuracy: 82.76%\n    Accuracy: 81.61%\n    Accuracy: 82.22%\n    Accuracy: 81.6%\n    Accuracy: 82.08%\n    Accuracy: 81.58%\n    Accuracy: 82.02%\n    Accuracy: 82.03%\n    Accuracy: 81.92%\n    Accuracy: 80.98%\n    Accuracy: 81.94%\n    Accuracy: 81.44%\n    Accuracy: 82.03%\n    Accuracy: 81.54%\n    Accuracy: 81.47%\n    Accuracy: 82.17%\n    Accuracy: 82.25%\n    Accuracy: 82.49%\n    Accuracy: 81.8%\n    Accuracy: 82.38%\n    Accuracy: 81.47%\n    Accuracy: 81.73%\n    Accuracy: 81.7%\n    Accuracy: 81.09%\n    Accuracy: 81.41%\n    Accuracy: 81.95%\n    Accuracy: 81.73%\n    Accuracy: 81.9%\n    Accuracy: 81.97%\n    Accuracy: 82.1%\n    Accuracy: 81.35%\n    Accuracy: 82.15%\n    Accuracy: 82.21%\n    Accuracy: 82.71%\n    Accuracy: 81.48%\n    Accuracy: 82.52%\n    Accuracy: 82.07%\n    Accuracy: 81.79%\n    Accuracy: 81.27%\n    Accuracy: 82.22%\n    Accuracy: 81.02%\n    Accuracy: 82.44%\n    Accuracy: 81.73%\n    Accuracy: 82.12%\n    Accuracy: 82.1%\n    Accuracy: 81.57%\n    Accuracy: 82.1%\n    Accuracy: 81.93%\n    Accuracy: 81.4%\n    Accuracy: 82.13%\n    Accuracy: 81.82%\n    Accuracy: 81.54%\n    Accuracy: 81.48%\n    Accuracy: 82.08%\n    Accuracy: 82.31%\n    Accuracy: 82.76%\n    Accuracy: 81.63%\n    Accuracy: 81.99%\n    Accuracy: 81.38%\n    Accuracy: 81.3%\n    Accuracy: 81.77%\n    Accuracy: 82.06%\n    Accuracy: 82.35%\n    Accuracy: 81.84%\n    Accuracy: 82.18%\n    Accuracy: 82.16%\n    Accuracy: 82.17%\n    Accuracy: 81.24%\n    Accuracy: 82.54%\n    Accuracy: 83.06%\n    Accuracy: 82.58%\n    Accuracy: 82.23%\n    Accuracy: 81.76%\n    Accuracy: 81.65%\n    Accuracy: 82.09%\n    Accuracy: 82.21%\n    Accuracy: 82.6%\n    Accuracy: 81.86%\n    Accuracy: 81.56%\n    Accuracy: 81.76%\n    Accuracy: 81.76%\n    Accuracy: 82.6%\n    Accuracy: 82.23%\n    Accuracy: 82.42%\n    Accuracy: 82.13%\n    Accuracy: 82.22%\n    Accuracy: 81.33%\n    Accuracy: 81.54%\n    Accuracy: 81.94%\n    Accuracy: 82.39%\n    Accuracy: 81.9%\n    Accuracy: 81.67%\n    Accuracy: 80.69%\n    Accuracy: 82.55%\n    Accuracy: 82.02%\n    Accuracy: 82.05%\n    Accuracy: 82.17%\n    Accuracy: 81.86%\n    Accuracy: 82.33%\n    Accuracy: 82.29%\n    Accuracy: 82.55%\n    Accuracy: 82.28%\n    Accuracy: 82.32%\n    Accuracy: 82.38%\n    Accuracy: 82.12%\n    Accuracy: 82.13%\n    Accuracy: 82.35%\n    Accuracy: 81.47%\n    Accuracy: 82.52%\n    Accuracy: 81.99%\n    Accuracy: 81.15%\n    Accuracy: 82.31%\n    Accuracy: 80.23%\n    Accuracy: 81.71%\n    Accuracy: 81.76%\n    Accuracy: 81.75%\n    Accuracy: 82.45%\n    Accuracy: 82.2%\n    Accuracy: 82.63%\n    Accuracy: 82.33%\n    Accuracy: 81.72%\n    Accuracy: 82.45%\n    Accuracy: 82.37%\n    Accuracy: 81.56%\n    Accuracy: 81.79%\n    Accuracy: 82.17%\n    Accuracy: 81.69%\n    Accuracy: 82.1%\n    Accuracy: 81.42%\n    Accuracy: 81.9%\n    Accuracy: 81.44%\n    Accuracy: 82.46%\n    Accuracy: 82.26%\n    Accuracy: 82.25%\n    Accuracy: 82.2%\n    Accuracy: 82.05%\n    Accuracy: 82.24%\n    Accuracy: 82.09%\n    Accuracy: 82.07%\n    Accuracy: 81.71%\n    Accuracy: 81.78%\n    Accuracy: 82.2%\n    Accuracy: 81.7%\n    Accuracy: 82.25%\n    Accuracy: 83.0%\n    Accuracy: 81.68%\n    Accuracy: 81.94%\n    Accuracy: 82.76%\n    Accuracy: 82.45%\n    Accuracy: 82.73%\n    Accuracy: 82.46%\n    Accuracy: 82.65%\n    Accuracy: 80.45%\n    Accuracy: 83.14%\n    Accuracy: 81.88%\n    Accuracy: 81.83%\n    Accuracy: 81.96%\n    Accuracy: 82.28%\n    Accuracy: 80.74%\n    Accuracy: 82.12%\n    Accuracy: 81.84%\n    Accuracy: 82.13%\n    Accuracy: 82.17%\n    Accuracy: 81.6%\n    Accuracy: 82.35%\n    Accuracy: 81.82%\n    Accuracy: 81.3%\n    Accuracy: 81.95%\n    Accuracy: 82.16%\n    Accuracy: 81.2%\n    Accuracy: 81.46%\n    Accuracy: 81.17%\n    Accuracy: 81.77%\n    Accuracy: 82.23%\n    Accuracy: 81.74%\n    Accuracy: 81.95%\n    Accuracy: 80.65%\n    Accuracy: 81.71%\n    Accuracy: 81.73%\n    Accuracy: 81.74%\n    Accuracy: 81.95%\n    Accuracy: 82.3%\n    Accuracy: 81.82%\n    Accuracy: 81.67%\n    Accuracy: 81.68%\n    Accuracy: 81.82%\n    Accuracy: 82.01%\n    Accuracy: 82.14%\n    Accuracy: 81.16%\n    Accuracy: 81.75%\n    Accuracy: 82.02%\n    Accuracy: 81.43%\n    Accuracy: 81.71%\n    Accuracy: 81.11%\n    Accuracy: 81.26%\n    Accuracy: 82.55%\n    Accuracy: 82.38%\n    Accuracy: 80.76%\n    Accuracy: 81.67%\n    Accuracy: 82.17%\n    Accuracy: 82.0%\n    Accuracy: 82.58%\n    Accuracy: 81.67%\n    Accuracy: 81.42%\n    Accuracy: 81.05%\n    Accuracy: 81.75%\n    Accuracy: 81.89%\n    Accuracy: 81.81%\n    Accuracy: 81.73%\n    Accuracy: 82.05%\n    Accuracy: 81.87%\n    Accuracy: 82.43%\n    Accuracy: 82.03%\n    Accuracy: 82.37%\n    Accuracy: 82.33%\n    Accuracy: 82.4%\n    Accuracy: 82.34%\n    Accuracy: 82.51%\n    Accuracy: 82.2%\n    Accuracy: 81.83%\n    Accuracy: 81.88%\n    Accuracy: 81.42%\n    Accuracy: 81.83%\n    Accuracy: 81.72%\n    Accuracy: 81.72%\n    Accuracy: 81.8%\n    Accuracy: 82.31%\n    Accuracy: 81.38%\n    Accuracy: 81.46%\n    Accuracy: 81.92%\n    Accuracy: 80.86%\n    Accuracy: 82.13%\n    Accuracy: 81.39%\n    Accuracy: 81.86%\n    Accuracy: 82.65%\n    Accuracy: 81.59%\n    Accuracy: 82.06%\n    Accuracy: 82.11%\n    Accuracy: 81.47%\n    Accuracy: 81.83%\n    Accuracy: 82.29%\n    Accuracy: 80.92%\n    Accuracy: 81.87%\n    Accuracy: 82.08%\n    Accuracy: 82.04%\n    Accuracy: 81.7%\n    Accuracy: 81.6%\n    Accuracy: 80.2%\n    Accuracy: 82.15%\n    Accuracy: 81.08%\n    Accuracy: 81.99%\n    Accuracy: 81.53%\n    Accuracy: 81.61%\n    Accuracy: 82.22%\n    Accuracy: 81.31%\n    Accuracy: 82.74%\n    Accuracy: 82.43%\n    Accuracy: 82.22%\n    Accuracy: 81.88%\n    Accuracy: 82.32%\n    Accuracy: 81.79%\n    Accuracy: 82.04%\n    Accuracy: 81.5%\n    Accuracy: 82.4%\n    Accuracy: 82.05%\n    Accuracy: 81.56%\n    Accuracy: 81.92%\n    Accuracy: 82.2%\n    Accuracy: 81.44%\n    Accuracy: 82.48%\n    Accuracy: 82.16%\n    Accuracy: 81.63%\n    Accuracy: 82.63%\n    Accuracy: 81.65%\n    Accuracy: 81.89%\n    Accuracy: 82.66%\n    Accuracy: 80.55%\n    Accuracy: 82.58%\n    Accuracy: 81.89%\n    Accuracy: 81.91%\n    Accuracy: 81.92%\n    Accuracy: 81.4%\n    Accuracy: 81.92%\n    Accuracy: 82.23%\n    Accuracy: 81.9%\n    Accuracy: 82.02%\n    Accuracy: 82.27%\n    Accuracy: 82.35%\n    Accuracy: 81.96%\n    Accuracy: 81.09%\n    Accuracy: 82.07%\n    Accuracy: 82.57%\n    Accuracy: 82.06%\n    Accuracy: 82.23%\n    Accuracy: 80.67%\n    Accuracy: 81.14%\n    Accuracy: 82.09%\n    Accuracy: 81.07%\n    Accuracy: 82.4%\n    Accuracy: 82.25%\n    Accuracy: 82.35%\n    Accuracy: 81.77%\n    Accuracy: 81.45%\n    Accuracy: 82.09%\n    Accuracy: 81.22%\n    Accuracy: 81.83%\n    Accuracy: 82.38%\n    Accuracy: 81.8%\n    Accuracy: 82.39%\n    Accuracy: 82.35%\n    Accuracy: 81.32%\n    Accuracy: 82.14%\n    Accuracy: 82.02%\n    Accuracy: 81.91%\n    Accuracy: 82.12%\n    Accuracy: 81.11%\n    Accuracy: 81.44%\n    Accuracy: 81.47%\n    Accuracy: 82.38%\n    Accuracy: 82.5%\n    Accuracy: 81.91%\n    Accuracy: 82.34%\n    Accuracy: 80.65%\n    Accuracy: 81.06%\n    Accuracy: 81.25%\n    Accuracy: 80.7%\n    Accuracy: 81.77%\n    Accuracy: 80.33%\n    Accuracy: 81.3%\n    Accuracy: 81.86%\n    Accuracy: 82.0%\n    Accuracy: 82.05%\n    Accuracy: 81.38%\n    Accuracy: 81.83%\n    Accuracy: 81.47%\n    Accuracy: 82.04%\n    Accuracy: 80.76%\n    Accuracy: 82.4%\n    Accuracy: 82.7%\n    Accuracy: 81.54%\n    Accuracy: 82.15%\n    Accuracy: 81.71%\n    Accuracy: 79.96%\n    Accuracy: 82.09%\n    Accuracy: 82.4%\n    Accuracy: 82.2%\n    Accuracy: 81.37%\n    Accuracy: 81.92%\n    Accuracy: 81.81%\n    Accuracy: 81.61%\n    Accuracy: 80.52%\n    Accuracy: 81.87%\n    Accuracy: 81.24%\n    Accuracy: 82.06%\n    Accuracy: 81.92%\n    Accuracy: 81.97%\n    Accuracy: 82.47%\n    Accuracy: 82.16%\n    Accuracy: 81.98%\n    Accuracy: 82.43%\n    Accuracy: 82.2%\n    Accuracy: 82.14%\n    Accuracy: 82.24%\n    Accuracy: 82.75%\n    Accuracy: 82.21%\n    Accuracy: 82.14%\n    Accuracy: 81.7%\n    Accuracy: 82.02%\n    Accuracy: 81.1%\n    Accuracy: 81.79%\n    Accuracy: 81.85%\n    Accuracy: 81.63%\n    Accuracy: 81.98%\n    Accuracy: 82.64%\n    Accuracy: 82.21%\n    Accuracy: 81.97%\n    Accuracy: 81.94%\n    Accuracy: 81.73%\n    Accuracy: 81.78%\n    Accuracy: 81.88%\n    Accuracy: 81.17%\n    Accuracy: 81.84%\n    Accuracy: 81.88%\n    Accuracy: 81.76%\n    Accuracy: 81.11%\n    Accuracy: 81.82%\n    Accuracy: 82.52%\n    Accuracy: 82.23%\n    Accuracy: 82.12%\n    Accuracy: 81.66%\n    Accuracy: 82.34%\n    Accuracy: 81.75%\n    Accuracy: 82.24%\n    Accuracy: 81.64%\n    Accuracy: 82.06%\n    Accuracy: 82.02%\n    Accuracy: 82.14%\n    Accuracy: 80.99%\n    Accuracy: 81.48%\n    Accuracy: 82.23%\n    Accuracy: 82.25%\n    Accuracy: 81.29%\n    Accuracy: 80.61%\n    Accuracy: 81.56%\n    Accuracy: 81.99%\n    Accuracy: 81.63%\n    Accuracy: 82.66%\n    Accuracy: 81.76%\n    Accuracy: 81.72%\n    Accuracy: 82.09%\n    Accuracy: 82.25%\n    Accuracy: 82.28%\n    Accuracy: 82.18%\n    Accuracy: 79.89%\n    Accuracy: 82.22%\n    Accuracy: 81.0%\n    Accuracy: 81.98%\n    Accuracy: 82.18%\n    Accuracy: 81.42%\n    Accuracy: 81.44%\n    Accuracy: 81.67%\n    Accuracy: 82.01%\n    Accuracy: 81.9%\n    Accuracy: 81.77%\n    Accuracy: 81.73%\n    Accuracy: 82.5%\n    Accuracy: 81.99%\n    Accuracy: 81.89%\n    Accuracy: 81.63%\n    Accuracy: 81.54%\n    Accuracy: 81.83%\n    Accuracy: 82.18%\n    Accuracy: 82.11%\n    Accuracy: 81.7%\n    Accuracy: 82.71%\n    Accuracy: 82.23%\n    Accuracy: 82.39%\n    Accuracy: 81.57%\n    Accuracy: 81.48%\n    Accuracy: 81.62%\n    Accuracy: 81.98%\n    Accuracy: 81.48%\n    Accuracy: 81.02%\n    Accuracy: 81.09%\n    Accuracy: 82.25%\n    Accuracy: 82.5%\n    Accuracy: 83.15%\n    Accuracy: 80.77%\n    Accuracy: 81.09%\n    Accuracy: 81.53%\n    Accuracy: 81.51%\n    Accuracy: 82.11%\n    Accuracy: 81.69%\n    Accuracy: 81.14%\n    Accuracy: 81.86%\n    Accuracy: 82.36%\n    Accuracy: 81.26%\n    Accuracy: 81.55%\n    Accuracy: 81.34%\n    Accuracy: 81.18%\n    Accuracy: 82.17%\n    Accuracy: 81.45%\n    Accuracy: 81.68%\n    Accuracy: 81.8%\n    Accuracy: 81.54%\n    Accuracy: 81.42%\n    Accuracy: 81.49%\n    Accuracy: 81.8%\n    Accuracy: 81.14%\n    Accuracy: 81.35%\n    Accuracy: 81.41%\n    Accuracy: 81.14%\n    Accuracy: 81.47%\n    Accuracy: 81.85%\n    Accuracy: 81.57%\n    Accuracy: 82.02%\n    Accuracy: 82.16%\n    Accuracy: 82.2%\n    Accuracy: 80.28%\n    Accuracy: 81.57%\n    Accuracy: 81.06%\n    Accuracy: 80.58%\n    Accuracy: 81.43%\n    Accuracy: 82.02%\n    Accuracy: 81.62%\n    Accuracy: 81.22%\n    Accuracy: 81.05%\n    Accuracy: 81.69%\n    Accuracy: 82.06%\n    Accuracy: 82.02%\n    Accuracy: 81.69%\n    Accuracy: 81.89%\n    Accuracy: 81.32%\n    Accuracy: 81.18%\n    Accuracy: 81.59%\n    Accuracy: 82.04%\n    Accuracy: 81.62%\n    Accuracy: 82.2%\n    Accuracy: 81.52%\n    Accuracy: 82.19%\n    Accuracy: 81.38%\n    Accuracy: 81.65%\n    Accuracy: 81.18%\n    Accuracy: 81.77%\n    Accuracy: 81.95%\n    Accuracy: 81.46%\n    Accuracy: 81.0%\n    Accuracy: 82.13%\n    Accuracy: 81.64%\n    Accuracy: 81.38%\n    Accuracy: 81.16%\n    Accuracy: 81.67%\n    Accuracy: 81.7%\n    Accuracy: 81.11%\n    Accuracy: 82.38%\n    Accuracy: 81.59%\n    Accuracy: 82.16%\n    Accuracy: 82.6%\n    Accuracy: 81.09%\n    Accuracy: 81.56%\n    Accuracy: 80.16%\n    Accuracy: 81.75%\n    Accuracy: 80.8%\n    Accuracy: 80.5%\n    Accuracy: 81.33%\n    Accuracy: 81.75%\n    Accuracy: 80.87%\n    Accuracy: 81.93%\n    Accuracy: 82.3%\n    Accuracy: 81.46%\n    Accuracy: 80.74%\n    Accuracy: 81.67%\n    Accuracy: 81.98%\n    Accuracy: 82.27%\n    Accuracy: 82.09%\n    Accuracy: 81.54%\n    Accuracy: 81.72%\n    Accuracy: 80.56%\n    Accuracy: 81.24%\n    Accuracy: 81.6%\n    Accuracy: 80.96%\n    Accuracy: 81.0%\n    Accuracy: 81.8%\n    Accuracy: 80.69%\n    Accuracy: 80.96%\n    Accuracy: 80.54%\n    Accuracy: 82.43%\n    Accuracy: 81.33%\n    Accuracy: 81.84%\n    Accuracy: 82.01%\n    Accuracy: 81.59%\n    Accuracy: 81.7%\n    Accuracy: 81.71%\n    Accuracy: 82.17%\n    Accuracy: 81.89%\n    Accuracy: 81.13%\n    Accuracy: 81.79%\n    Accuracy: 81.15%\n    Accuracy: 81.95%\n    Accuracy: 80.99%\n    Accuracy: 80.37%\n    Accuracy: 81.62%\n    Accuracy: 81.78%\n    Accuracy: 80.93%\n    Accuracy: 81.56%\n    Accuracy: 81.75%\n    Accuracy: 81.31%\n    Accuracy: 81.59%\n    Accuracy: 81.55%\n    Accuracy: 82.02%\n    Accuracy: 81.55%\n    Accuracy: 80.44%\n    Accuracy: 82.03%\n    Accuracy: 81.57%\n    Accuracy: 81.99%\n    Accuracy: 81.38%\n    Accuracy: 81.77%\n    Accuracy: 81.78%\n    Accuracy: 81.01%\n    Accuracy: 81.63%\n    Accuracy: 81.59%\n    Accuracy: 80.79%\n    Accuracy: 81.45%\n    Accuracy: 81.69%\n    Accuracy: 81.76%",
      "json_metadata": "{\"tags\":[\"tensorflow\",\"python\",\"kr\",\"deep-learning\"],\"image\":[\"output_3_1.png\"],\"links\":[\"https://github.com/llejo3/deep-learning/blob/master/CIFAR10_tensorflow.md\"],\"app\":\"steemit/0.1\",\"format\":\"markdown\"}"
    }
  ]
}

{
  "trx_id": "0000000000000000000000000000000000000000",
  "block": 22748967,
  "trx_in_block": 4294967295,
  "op_in_trx": 0,
  "virtual_op": 53,
  "timestamp": "2018-05-25T20:34:12",
  "op": [
    "curation_reward",
    {
      "curator": "south-man",
      "reward": "8.135604 VESTS",
      "comment_author": "zorro",
      "comment_permlink": "golden-leaves-and-pixie-dust"
    }
  ]
}

{
  "trx_id": "e3e659939aaf0185fc01bf58c29f6997360cef4a",
  "block": 22607192,
  "trx_in_block": 25,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-20T22:05:36",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "zoex",
      "permlink": "amazing-homemade-inventions-2018-part-1-2018-5-20",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "84de7ccae700cdc55cc112e747e610ecb5f76b65",
  "block": 22607180,
  "trx_in_block": 48,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-20T22:05:00",
  "op": [
    "custom_json",
    {
      "required_auths": [],
      "required_posting_auths": [
        "south-man"
      ],
      "id": "follow",
      "json": "[\"follow\",{\"follower\":\"south-man\",\"following\":\"zoex\",\"what\":[\"blog\"]}]"
    }
  ]
}

{
  "trx_id": "dd3345472078623f1b806a9c9bf1e8925f99af3c",
  "block": 22607166,
  "trx_in_block": 64,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-20T22:04:18",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "zoex",
      "permlink": "future-volvo-electric-truck-concept-2018-5-20",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "d673562dee45ccf28278511346626c397384ec1a",
  "block": 22580682,
  "trx_in_block": 48,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-19T23:59:54",
  "op": [
    "custom_json",
    {
      "required_auths": [],
      "required_posting_auths": [
        "south-man"
      ],
      "id": "follow",
      "json": "[\"follow\",{\"follower\":\"south-man\",\"following\":\"crypto-coin-er\",\"what\":[\"blog\"]}]"
    }
  ]
}

{
  "trx_id": "721845e3fc71b3e46ff57cfaf905823e446f9d4c",
  "block": 22580680,
  "trx_in_block": 55,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-19T23:59:48",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "crypto-coin-er",
      "permlink": "coinnews-introduction-to-binance-and-how-to-regist",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "0000000000000000000000000000000000000000",
  "block": 22569128,
  "trx_in_block": 4294967295,
  "op_in_trx": 0,
  "virtual_op": 62,
  "timestamp": "2018-05-19T14:22:06",
  "op": [
    "curation_reward",
    {
      "curator": "south-man",
      "reward": "2.034562 VESTS",
      "comment_author": "halo",
      "comment_permlink": "steemit-girl-halo-photography-journey-1421"
    }
  ]
}

{
  "trx_id": "69ebc5b9dae115ece7d71c29d092f3ed261c6b5c",
  "block": 22568681,
  "trx_in_block": 17,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-19T13:59:42",
  "op": [
    "custom_json",
    {
      "required_auths": [],
      "required_posting_auths": [
        "south-man"
      ],
      "id": "follow",
      "json": "[\"follow\",{\"follower\":\"south-man\",\"following\":\"isaaclab\",\"what\":[\"blog\"]}]"
    }
  ]
}

{
  "trx_id": "4b53649399b11a75ed46dee21736f2f027b13b89",
  "block": 22568677,
  "trx_in_block": 27,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-19T13:59:30",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "isaaclab",
      "permlink": "gopax-steem-steem-dollar",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "43fed8d05d972c8dd9f7a7a977b8204130d10ee2",
  "block": 22568656,
  "trx_in_block": 37,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-19T13:58:27",
  "op": [
    "custom_json",
    {
      "required_auths": [],
      "required_posting_auths": [
        "south-man"
      ],
      "id": "follow",
      "json": "[\"follow\",{\"follower\":\"south-man\",\"following\":\"fenrir78\",\"what\":[\"blog\"]}]"
    }
  ]
}

{
  "trx_id": "ef432896a3d7b9ea7c73f89d405cc7aae0568e8c",
  "block": 22568653,
  "trx_in_block": 53,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-19T13:58:18",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "fenrir78",
      "permlink": "5o4jtg-steemit",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "bfb94052ddcd82df2da2688bf1b360594a55802f",
  "block": 22568586,
  "trx_in_block": 7,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-19T13:54:57",
  "op": [
    "custom_json",
    {
      "required_auths": [],
      "required_posting_auths": [
        "south-man"
      ],
      "id": "follow",
      "json": "[\"follow\",{\"follower\":\"south-man\",\"following\":\"asbear\",\"what\":[\"blog\"]}]"
    }
  ]
}

{
  "trx_id": "ca4e55670ec39c224f5332229b42fb89c763ce00",
  "block": 22568582,
  "trx_in_block": 26,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-19T13:54:45",
  "op": [
    "vote",
    {
      "voter": "south-man",
      "author": "asbear",
      "permlink": "steemit-beta",
      "weight": 10000
    }
  ]
}

{
  "trx_id": "6f2f53657b2d81b3e15e1a89f10e294e53cc1070",
  "block": 22568564,
  "trx_in_block": 23,
  "op_in_trx": 0,
  "virtual_op": 0,
  "timestamp": "2018-05-19T13:53:51",
  "op": [
    "custom_json",
    {
      "required_auths": [],
      "required_posting_auths": [
        "south-man"
      ],
      "id": "follow",
      "json": "[\"follow\",{\"follower\":\"south-man\",\"following\":\"tabris\",\"what\":[\"blog\"]}]"
    }
  ]
}