In [23]:
import graphlab as gl
import pandas as pd
import numpy as np
import cPickle

In [5]:
# Load each batch as dictionary:
def unpickle(file):
    fo = open(file, 'rb')
    dict = cPickle.load(fo)
    fo.close()
    return dict

#Make dictionaries by calling the above function:
batch1 = unpickle('cifar-10-batches-py/data_batch_1')
batch2 = unpickle('cifar-10-batches-py/data_batch_2')
batch3 = unpickle('cifar-10-batches-py/data_batch_3')
batch4 = unpickle('cifar-10-batches-py/data_batch_4')
batch5 = unpickle('cifar-10-batches-py/data_batch_5')
batch_test = unpickle('cifar-10-batches-py/test_batch')

In [6]:
batch1


Out[6]:
{'batch_label': 'training batch 1 of 5',
 'data': array([[ 59,  43,  50, ..., 140,  84,  72],
        [154, 126, 105, ..., 139, 142, 144],
        [255, 253, 253, ...,  83,  83,  84],
        ..., 
        [ 71,  60,  74, ...,  68,  69,  68],
        [250, 254, 211, ..., 215, 255, 254],
        [ 62,  61,  60, ..., 130, 130, 131]], dtype=uint8),
 'filenames': ['leptodactylus_pentadactylus_s_000004.png',
  'camion_s_000148.png',
  'tipper_truck_s_001250.png',
  'american_elk_s_001521.png',
  'station_wagon_s_000293.png',
  'coupe_s_001735.png',
  'cassowary_s_001300.png',
  'cow_pony_s_001168.png',
  'sea_boat_s_001584.png',
  'tabby_s_001355.png',
  'muntjac_s_001000.png',
  'arabian_s_001354.png',
  'quarter_horse_s_000672.png',
  'passerine_s_000343.png',
  'camion_s_001895.png',
  'trailer_truck_s_000335.png',
  'dumper_s_000821.png',
  'alley_cat_s_000200.png',
  'accentor_s_000677.png',
  'frog_s_001671.png',
  'capreolus_capreolus_s_000051.png',
  'tomcat_s_000772.png',
  'pickerel_frog_s_000446.png',
  'bufo_s_001242.png',
  'cassowary_s_001246.png',
  'toad_s_001748.png',
  'cat_s_000081.png',
  'chihuahua_s_000825.png',
  'alces_alces_s_000959.png',
  'stealth_bomber_s_000554.png',
  'twinjet_s_000663.png',
  'trucking_rig_s_001402.png',
  'auto_s_000609.png',
  'tabby_cat_s_000983.png',
  'wapiti_s_000416.png',
  'monoplane_s_000895.png',
  'true_cat_s_000247.png',
  'tennessee_walker_s_000486.png',
  'house_cat_s_000243.png',
  'house_cat_s_001196.png',
  'pekinese_s_001337.png',
  'ostrich_s_001368.png',
  'ostrich_s_001150.png',
  'stallion_s_000046.png',
  'station_waggon_s_000041.png',
  'coupe_s_001944.png',
  'estate_car_s_000580.png',
  'accentor_s_000759.png',
  'emu_novaehollandiae_s_000795.png',
  'dive_bomber_s_001390.png',
  'articulated_lorry_s_000131.png',
  'pekinese_s_001093.png',
  'broodmare_s_001463.png',
  'delivery_truck_s_000834.png',
  'songbird_s_001052.png',
  'emu_s_000692.png',
  'puppy_s_000115.png',
  'wagtail_s_001821.png',
  'dama_dama_s_000658.png',
  'domestic_cat_s_001970.png',
  'ambulance_s_003039.png',
  'convertible_s_001763.png',
  'tank_ship_s_001229.png',
  'cassowary_s_001055.png',
  'wagon_s_001142.png',
  'police_cruiser_s_000620.png',
  'moose_s_002308.png',
  'aerial_ladder_truck_s_000584.png',
  'saddle_horse_s_000717.png',
  'tanker_s_001350.png',
  'mongrel_s_001571.png',
  'truck_s_000835.png',
  'pickerel_frog_s_001195.png',
  'lipizzan_s_000399.png',
  'tabby_s_000074.png',
  'automobile_s_001887.png',
  'moving_van_s_001665.png',
  'attack_aircraft_s_000153.png',
  'domestic_cat_s_001596.png',
  'compact_car_s_000048.png',
  'domestic_cat_s_000009.png',
  'pekingese_s_002089.png',
  'capreolus_capreolus_s_001095.png',
  'blenheim_spaniel_s_001103.png',
  'stallion_s_000040.png',
  'stud_mare_s_001672.png',
  'elk_s_000920.png',
  'lipizzan_s_001123.png',
  'fire_truck_s_002721.png',
  'elk_s_001888.png',
  'finch_s_000750.png',
  'tabby_s_000880.png',
  'banana_boat_s_001324.png',
  'twinjet_s_000591.png',
  'shooting_brake_s_000029.png',
  'rana_pipiens_s_000101.png',
  'station_wagon_s_001387.png',
  'station_wagon_s_002712.png',
  'odocoileus_hemionus_s_000221.png',
  'convertible_s_001715.png',
  'abandoned_ship_s_000574.png',
  'true_cat_s_000114.png',
  'dustcart_s_000063.png',
  'frog_s_000797.png',
  'green_frog_s_001320.png',
  'compact_car_s_001038.png',
  'freighter_s_001351.png',
  'mutt_s_000251.png',
  'accentor_s_000303.png',
  'lorry_s_001154.png',
  'fire_truck_s_000894.png',
  'cargo_vessel_s_000731.png',
  'cruiser_s_000774.png',
  'lippizan_s_000359.png',
  'broodmare_s_001741.png',
  'fighter_aircraft_s_000876.png',
  'amphibious_aircraft_s_000216.png',
  'leopard_frog_s_000339.png',
  'trucking_rig_s_001315.png',
  'shooting_brake_s_000886.png',
  'pipit_s_000549.png',
  'ostrich_s_002148.png',
  'trucking_rig_s_001600.png',
  'alauda_arvensis_s_000755.png',
  'rana_temporaria_s_001087.png',
  'bufo_s_000136.png',
  'estate_car_s_000529.png',
  'aerial_ladder_truck_s_000997.png',
  'toy_spaniel_s_000384.png',
  'amphibious_aircraft_s_001195.png',
  'fallow_deer_s_001598.png',
  'quarter_horse_s_000414.png',
  'anuran_s_000712.png',
  'arabian_s_001366.png',
  'auto_s_000040.png',
  'cargo_ship_s_001063.png',
  'motorcar_s_000121.png',
  'motorcar_s_000305.png',
  'anthus_pratensis_s_001071.png',
  'cruiser_s_000294.png',
  'wagon_s_000763.png',
  'alley_cat_s_002579.png',
  'tabby_cat_s_000825.png',
  'spadefoot_s_000051.png',
  'wagtail_s_002075.png',
  'fallow_deer_s_001697.png',
  'garbage_truck_s_000777.png',
  'moving_van_s_000067.png',
  'mongrel_s_000696.png',
  'red_deer_s_000741.png',
  'true_cat_s_001768.png',
  'spadefoot_s_000377.png',
  'lipizzan_s_001233.png',
  'japanese_deer_s_000163.png',
  'toad_s_002436.png',
  'police_boat_s_001421.png',
  'puppy_s_002062.png',
  'lapdog_s_001869.png',
  'elk_s_001196.png',
  'domestic_cat_s_001235.png',
  'car_s_000057.png',
  'lightship_s_000064.png',
  'fallow_deer_s_001997.png',
  'quarter_horse_s_000022.png',
  'bufo_calamita_s_000566.png',
  'biplane_s_000624.png',
  'fire_truck_s_001616.png',
  'japanese_spaniel_s_000037.png',
  'ambulance_s_001394.png',
  'domestic_cat_s_001056.png',
  'tugboat_s_001384.png',
  'emu_s_000072.png',
  'gelding_s_000206.png',
  'puppy_s_000637.png',
  'tabby_cat_s_002607.png',
  'elk_s_001198.png',
  'estate_car_s_001496.png',
  'peke_s_001338.png',
  'arabian_s_001018.png',
  'fighter_aircraft_s_000538.png',
  'fallow_deer_s_001164.png',
  'lippizan_s_000390.png',
  'toy_dog_s_000932.png',
  'maltese_s_002151.png',
  'car_s_000002.png',
  'jumbo_jet_s_000697.png',
  'dump_truck_s_000008.png',
  'spadefoot_s_000084.png',
  'motortruck_s_000014.png',
  'jetliner_s_001502.png',
  'tanker_s_000157.png',
  'riding_horse_s_001870.png',
  'cabin_cruiser_s_000096.png',
  'dredger_s_001623.png',
  'pipit_s_001102.png',
  'mongrel_s_002430.png',
  'skylark_s_000150.png',
  'cat_s_001604.png',
  'puppy_s_000211.png',
  'airbus_s_001124.png',
  'leopard_frog_s_001021.png',
  'wagon_s_001824.png',
  'garbage_truck_s_000147.png',
  'tabby_s_001354.png',
  'rana_temporaria_s_000295.png',
  'lorry_s_000364.png',
  'estate_car_s_000075.png',
  'tabby_s_000970.png',
  'fire_engine_s_000985.png',
  'toad_s_000022.png',
  'texas_toad_s_000215.png',
  'broodmare_s_000707.png',
  'car_s_000286.png',
  'twinjet_s_000565.png',
  'truck_s_001385.png',
  'toy_dog_s_000752.png',
  'police_boat_s_000092.png',
  'toy_dog_s_001197.png',
  'cassowary_s_002446.png',
  'articulated_lorry_s_000050.png',
  'monoplane_s_000781.png',
  'lightship_s_000343.png',
  'powerboat_s_001486.png',
  'jumbo_jet_s_000071.png',
  'bufo_debilis_s_000024.png',
  'transporter_s_000272.png',
  'estate_car_s_000743.png',
  'motorcar_s_000697.png',
  'rana_catesbeiana_s_001655.png',
  'tabby_s_000675.png',
  'remount_s_000285.png',
  'rana_catesbeiana_s_001521.png',
  'true_frog_s_000223.png',
  'airliner_s_002304.png',
  'natterjack_s_000987.png',
  'leptodactylid_s_000082.png',
  'coupe_s_001892.png',
  'quarter_horse_s_000419.png',
  'car_s_001323.png',
  'maltese_dog_s_001043.png',
  'boat_s_000543.png',
  'tabby_cat_s_001537.png',
  'rana_catesbeiana_s_001096.png',
  'bufo_bufo_s_000326.png',
  'pilot_boat_s_000646.png',
  'toad_frog_s_001672.png',
  'boat_s_000988.png',
  'deer_s_001383.png',
  'bufo_bufo_s_000261.png',
  'leopard_frog_s_000763.png',
  'coupe_s_002178.png',
  'cat_s_000663.png',
  'ship_s_000175.png',
  'house_cat_s_001471.png',
  'stag_s_001458.png',
  'station_wagon_s_001337.png',
  'quarter_horse_s_002201.png',
  'coupe_s_000698.png',
  'tomcat_s_002553.png',
  'scow_s_001266.png',
  'chihuahua_s_000385.png',
  'convertible_s_001500.png',
  'convertible_s_000408.png',
  'odocoileus_hemionus_s_000512.png',
  'jumbo_jet_s_000971.png',
  'truck_s_000650.png',
  'domestic_cat_s_001288.png',
  'walking_horse_s_000194.png',
  'elk_s_001657.png',
  'tipper_truck_s_000287.png',
  'dumper_s_000648.png',
  'struthio_camelus_s_000225.png',
  'woodland_caribou_s_000803.png',
  'ladder_truck_s_000795.png',
  'pantechnicon_s_000230.png',
  'shooting_brake_s_000229.png',
  'jetliner_s_001349.png',
  'toy_dog_s_001280.png',
  'dump_truck_s_002104.png',
  'twinjet_s_001360.png',
  'tugboat_s_000715.png',
  'passerine_s_000360.png',
  'automobile_s_002337.png',
  'honey_eater_s_001422.png',
  'attack_aircraft_s_001304.png',
  'toy_dog_s_000114.png',
  'anuran_s_000550.png',
  'mouser_s_000382.png',
  'sparrow_s_000037.png',
  'stud_mare_s_001038.png',
  'speedboat_s_002588.png',
  'passenger_ship_s_002153.png',
  'bufo_bufo_s_001876.png',
  'biplane_s_000651.png',
  'gelding_s_001785.png',
  'trucking_rig_s_001598.png',
  'deer_s_000309.png',
  'pekingese_s_001946.png',
  'frog_s_002802.png',
  'cervus_unicolor_s_000290.png',
  'honey_eater_s_000478.png',
  'automobile_s_001152.png',
  'coupe_s_000846.png',
  'flying_bird_s_000252.png',
  'automobile_s_002343.png',
  'japanese_spaniel_s_001162.png',
  'fire_truck_s_000043.png',
  'semi_s_001207.png',
  'fighter_aircraft_s_001840.png',
  'ship_s_001976.png',
  'caribou_s_001222.png',
  'police_cruiser_s_001347.png',
  'coupe_s_000042.png',
  'bufo_bufo_s_000995.png',
  'tabby_s_001148.png',
  'true_cat_s_000476.png',
  'garbage_truck_s_001442.png',
  'fighter_aircraft_s_000385.png',
  'quarter_horse_s_001602.png',
  'camion_s_000116.png',
  'tennessee_walker_s_000539.png',
  'lipizzan_s_000452.png',
  'trucking_rig_s_001594.png',
  'shooting_brake_s_001611.png',
  'puppy_s_002096.png',
  'auto_s_000855.png',
  'bufo_marinus_s_001364.png',
  'natterjack_s_000036.png',
  'pontoon_s_000643.png',
  'stallion_s_001256.png',
  'coupe_s_001863.png',
  'house_cat_s_000079.png',
  'jetliner_s_000981.png',
  'felis_catus_s_000853.png',
  'tabby_cat_s_000827.png',
  'struthio_camelus_s_001250.png',
  'fawn_s_001915.png',
  'canis_familiaris_s_001298.png',
  'lipizzan_s_002122.png',
  'toy_spaniel_s_001133.png',
  'dump_truck_s_000029.png',
  'fighter_s_001042.png',
  'domestic_cat_s_000493.png',
  'cervus_elaphus_s_001764.png',
  'airbus_s_000627.png',
  'fawn_s_001452.png',
  'fallow_deer_s_000959.png',
  'spadefoot_s_000777.png',
  'seaplane_s_001639.png',
  'plane_s_001066.png',
  'bufo_viridis_s_000348.png',
  'crapaud_s_000051.png',
  'seaplane_s_001447.png',
  'boat_s_000404.png',
  'convertible_s_001242.png',
  'toad_s_000201.png',
  'finch_s_000494.png',
  'fire_engine_s_001437.png',
  'pipit_s_000226.png',
  'toy_dog_s_000960.png',
  'delivery_truck_s_001121.png',
  'bufo_marinus_s_001506.png',
  'studhorse_s_000010.png',
  'woodland_caribou_s_001003.png',
  'station_wagon_s_000260.png',
  'pilot_boat_s_000356.png',
  'stud_mare_s_001497.png',
  'tabby_cat_s_001966.png',
  'american_toad_s_000006.png',
  'tipper_s_000010.png',
  'tomcat_s_002230.png',
  'dive_bomber_s_001120.png',
  'wapiti_s_000802.png',
  'biplane_s_000862.png',
  'blenheim_spaniel_s_000589.png',
  'shooting_brake_s_001161.png',
  'jetliner_s_000400.png',
  'tabby_s_000923.png',
  'alces_alces_s_000361.png',
  'police_boat_s_000002.png',
  'mutt_s_001696.png',
  'cervus_elaphus_s_001695.png',
  'quarter_horse_s_000752.png',
  'accentor_s_000449.png',
  'tabby_cat_s_000144.png',
  'fire_engine_s_001513.png',
  'tennessee_walking_horse_s_001187.png',
  'bufo_s_001247.png',
  'arabian_s_002147.png',
  'compact_s_001258.png',
  'elk_s_002339.png',
  'stud_mare_s_001384.png',
  'fighter_aircraft_s_000745.png',
  'police_cruiser_s_001047.png',
  'lipizzan_s_001764.png',
  'cat_s_000331.png',
  'taxi_s_000701.png',
  'boat_s_000432.png',
  'japanese_deer_s_000942.png',
  'sambar_s_000076.png',
  'ostrich_s_001559.png',
  'dive_bomber_s_000185.png',
  'lark_s_001532.png',
  'flightless_bird_s_000480.png',
  'stealth_fighter_s_001312.png',
  'jumbojet_s_000789.png',
  'delivery_truck_s_000510.png',
  'biplane_s_001901.png',
  'lorry_s_001185.png',
  'crapaud_s_000511.png',
  'boat_s_001280.png',
  'bird_s_001357.png',
  'stud_mare_s_001477.png',
  'arabian_s_000751.png',
  'rangifer_caribou_s_000586.png',
  'monoplane_s_000036.png',
  'house_cat_s_001321.png',
  'stealth_fighter_s_000240.png',
  'oil_tanker_s_000129.png',
  'tipper_truck_s_000416.png',
  'roe_deer_s_000358.png',
  'rhea_americana_s_000475.png',
  'lippizaner_s_000517.png',
  'cassowary_s_000297.png',
  'pekingese_s_001675.png',
  'accentor_s_001058.png',
  'lapdog_s_000227.png',
  'auto_s_001153.png',
  'fire_engine_s_000443.png',
  'muntjac_s_000815.png',
  'lightship_s_000090.png',
  'dog_s_002441.png',
  'auto_s_001951.png',
  'arabian_s_000328.png',
  'odocoileus_hemionus_s_000188.png',
  'barking_deer_s_000019.png',
  'dive_bomber_s_001486.png',
  'bufo_viridis_s_000503.png',
  'wrecker_s_002294.png',
  'monoplane_s_000206.png',
  'arabian_s_002282.png',
  'speedboat_s_002160.png',
  'cargo_vessel_s_002136.png',
  'garbage_truck_s_000726.png',
  'delivery_truck_s_001468.png',
  'house_cat_s_002240.png',
  'house_cat_s_000045.png',
  'caribou_s_002007.png',
  'propeller_plane_s_000451.png',
  'fallow_deer_s_001082.png',
  'dog_s_001897.png',
  'bufo_bufo_s_001631.png',
  'spadefoot_s_000191.png',
  'multiengine_airplane_s_000186.png',
  'auto_s_002359.png',
  'stealth_bomber_s_001607.png',
  'passenger_ship_s_000582.png',
  'airliner_s_001297.png',
  'fallow_deer_s_000973.png',
  'speedboat_s_000178.png',
  'boat_s_001444.png',
  'estate_car_s_000712.png',
  'english_toy_spaniel_s_000888.png',
  'passerine_s_000699.png',
  'rana_pipiens_s_000286.png',
  'tanker_s_000025.png',
  'car_s_001956.png',
  'stealth_bomber_s_000424.png',
  'airliner_s_001070.png',
  'stud_mare_s_001348.png',
  'quarter_horse_s_000282.png',
  'maltese_s_001818.png',
  'lorry_s_001786.png',
  'true_frog_s_000064.png',
  'cassowary_s_001372.png',
  'houseboat_s_000626.png',
  'domestic_cat_s_000035.png',
  'deer_s_001596.png',
  'cow_pony_s_000803.png',
  'tabby_cat_s_001466.png',
  'tipper_truck_s_001525.png',
  'stealth_fighter_s_000533.png',
  'station_wagon_s_000750.png',
  'wagtail_s_001549.png',
  'rangifer_caribou_s_000985.png',
  'cargo_ship_s_000408.png',
  'cab_s_000788.png',
  'flatboat_s_000044.png',
  'bufo_viridis_s_000457.png',
  'fallow_deer_s_000040.png',
  'barren_ground_caribou_s_000077.png',
  'pekinese_s_000453.png',
  'female_horse_s_000358.png',
  'compact_car_s_000869.png',
  'tabby_cat_s_002454.png',
  'delivery_truck_s_000864.png',
  'cargo_ship_s_000802.png',
  'attack_aircraft_s_000037.png',
  'convertible_s_000037.png',
  'broodmare_s_001448.png',
  'toy_spaniel_s_001286.png',
  'packet_boat_s_001206.png',
  'ostrich_s_001564.png',
  'hospital_ship_s_001658.png',
  'stealth_bomber_s_001035.png',
  'stag_s_002583.png',
  'police_cruiser_s_000629.png',
  'merchant_ship_s_001242.png',
  'dump_truck_s_001584.png',
  'cargo_ship_s_001830.png',
  'nandu_s_000721.png',
  'garbage_truck_s_001119.png',
  'lorry_s_002317.png',
  'wagtail_s_001445.png',
  'lipizzan_s_002049.png',
  'maltese_s_000948.png',
  'quarter_horse_s_000798.png',
  'house_cat_s_001472.png',
  'tank_ship_s_001155.png',
  'hospital_ship_s_001838.png',
  'mule_deer_s_000581.png',
  'deer_s_001329.png',
  'rhea_americana_s_000268.png',
  'lippizan_s_000440.png',
  'convertible_s_002117.png',
  'bufo_marinus_s_001567.png',
  'roe_deer_s_000690.png',
  'twinjet_s_000481.png',
  'american_elk_s_001743.png',
  'bufo_viridis_s_000254.png',
  'dumper_s_000179.png',
  'lippizan_s_000323.png',
  'leopard_frog_s_000058.png',
  'prunella_modularis_s_000578.png',
  'maltese_s_001867.png',
  'toy_dog_s_001358.png',
  'police_cruiser_s_000500.png',
  'broodmare_s_001589.png',
  'passerine_s_000381.png',
  'accentor_s_000424.png',
  'finch_s_000617.png',
  'trucking_rig_s_001111.png',
  'canis_familiaris_s_001059.png',
  'cervus_elaphus_s_000813.png',
  'cassowary_s_000339.png',
  'buckskin_s_000325.png',
  'lightship_s_001019.png',
  'shooting_brake_s_001330.png',
  'tomcat_s_002649.png',
  'barren_ground_caribou_s_000174.png',
  'cat_s_000572.png',
  'walking_horse_s_002079.png',
  'rana_temporaria_s_001139.png',
  'tipper_s_000007.png',
  'freighter_s_000661.png',
  'stealth_fighter_s_000020.png',
  'goliath_frog_s_000205.png',
  'stealth_bomber_s_000703.png',
  'prunella_modularis_s_000795.png',
  'struthio_camelus_s_000331.png',
  'dunnock_s_000956.png',
  'auto_s_001284.png',
  'scow_s_000831.png',
  'fallow_deer_s_001256.png',
  'stealth_fighter_s_001297.png',
  'estate_car_s_000845.png',
  'ship_s_001691.png',
  'abandoned_ship_s_000977.png',
  'wagon_s_001298.png',
  'feist_s_000354.png',
  'lipizzan_s_001600.png',
  'leopard_frog_s_001700.png',
  'cervus_elaphus_s_000066.png',
  'toy_s_000517.png',
  'cargo_ship_s_000556.png',
  'stallion_s_001717.png',
  'coupe_s_001004.png',
  'camion_s_001068.png',
  'coupe_s_000001.png',
  'garbage_truck_s_001253.png',
  'sea_boat_s_001960.png',
  'capreolus_capreolus_s_001169.png',
  'lippizan_s_000681.png',
  'tabby_s_000512.png',
  'pontoon_s_000271.png',
  'sea_boat_s_001502.png',
  'lark_s_000700.png',
  'leopard_frog_s_001074.png',
  'rana_palustris_s_000022.png',
  'lippizaner_s_001181.png',
  'wagon_s_000155.png',
  'american_toad_s_001410.png',
  'cargo_vessel_s_000659.png',
  'station_wagon_s_001705.png',
  'transporter_s_000076.png',
  'broodmare_s_001098.png',
  'passenger_ship_s_000050.png',
  'tabby_cat_s_000484.png',
  'fighter_s_000276.png',
  'car_s_000506.png',
  'stealth_fighter_s_000139.png',
  'pleasure_craft_s_000076.png',
  'cruiser_s_000744.png',
  'tabby_cat_s_002478.png',
  'attack_aircraft_s_000423.png',
  'stealth_fighter_s_000460.png',
  'police_cruiser_s_000787.png',
  'mongrel_s_000381.png',
  'attack_aircraft_s_001166.png',
  'supertanker_s_000557.png',
  'tugboat_s_000807.png',
  'stud_mare_s_000413.png',
  'lorry_s_001039.png',
  'trailer_truck_s_001496.png',
  'airliner_s_001477.png',
  'camion_s_001521.png',
  'alces_alces_s_000404.png',
  'automobile_s_000315.png',
  'tabby_cat_s_001542.png',
  'pickerel_frog_s_000003.png',
  'spring_frog_s_000318.png',
  'roebuck_s_000153.png',
  'reindeer_s_000121.png',
  'male_horse_s_000008.png',
  'dog_s_000052.png',
  'bufo_bufo_s_001959.png',
  'stealth_bomber_s_000492.png',
  'pilot_boat_s_001819.png',
  'attack_aircraft_s_001187.png',
  'house_cat_s_000812.png',
  'cassowary_s_000138.png',
  'boat_s_000490.png',
  'wapiti_s_000884.png',
  'bufo_bufo_s_001492.png',
  'delivery_truck_s_000157.png',
  'truck_s_000060.png',
  'stud_mare_s_000506.png',
  'fighter_aircraft_s_000655.png',
  'true_cat_s_000127.png',
  'tabby_s_000228.png',
  'american_green_toad_s_000007.png',
  'quarter_horse_s_001833.png',
  'odocoileus_hemionus_s_001090.png',
  'tipper_lorry_s_000527.png',
  'compact_car_s_001200.png',
  'pickerel_frog_s_000242.png',
  'finch_s_000308.png',
  'arabian_s_002456.png',
  'dunnock_s_001476.png',
  'emu_s_002196.png',
  'plane_s_000065.png',
  'european_toad_s_000360.png',
  'arabian_s_002383.png',
  'dog_s_002098.png',
  'quarter_horse_s_001757.png',
  'rana_temporaria_s_001922.png',
  'cargo_ship_s_001318.png',
  'tip_truck_s_001017.png',
  'biplane_s_001055.png',
  'camion_s_001255.png',
  'red_deer_s_000756.png',
  'capreolus_capreolus_s_000822.png',
  'lippizaner_s_000840.png',
  'propeller_plane_s_001438.png',
  'dumper_s_000646.png',
  'capreolus_capreolus_s_000284.png',
  'truck_s_000526.png',
  'bufo_s_001473.png',
  'fire_engine_s_001449.png',
  'elk_s_001164.png',
  'canis_familiaris_s_001319.png',
  'tennessee_walking_horse_s_000005.png',
  'truck_s_000838.png',
  'rhea_s_001649.png',
  'mule_deer_s_002514.png',
  'canis_familiaris_s_000143.png',
  'police_cruiser_s_001073.png',
  'cervus_sika_s_000273.png',
  'cat_s_001584.png',
  'tipper_lorry_s_000313.png',
  'bufo_americanus_s_001202.png',
  'english_toy_spaniel_s_000735.png',
  'bufo_calamita_s_000043.png',
  'tipper_s_000180.png',
  'tabby_cat_s_000881.png',
  'tabby_cat_s_001644.png',
  'blenheim_spaniel_s_000330.png',
  'attack_aircraft_s_001138.png',
  'arabian_s_002270.png',
  'cassowary_s_000845.png',
  'wagon_s_000692.png',
  'tabby_cat_s_000319.png',
  'american_toad_s_000077.png',
  'moose_s_001500.png',
  'airliner_s_001406.png',
  'fighter_aircraft_s_000966.png',
  'dunnock_s_000055.png',
  'pekingese_s_002399.png',
  'twinjet_s_000739.png',
  'automobile_s_002249.png',
  'airliner_s_000669.png',
  'wagtail_s_001656.png',
  'tabby_s_001949.png',
  'dump_truck_s_001009.png',
  'freighter_s_000436.png',
  'fallow_deer_s_001171.png',
  'trucking_rig_s_000684.png',
  'ship_s_000249.png',
  'monoplane_s_000416.png',
  'songbird_s_002066.png',
  'rana_catesbeiana_s_000602.png',
  'red_deer_s_000536.png',
  'elk_s_001583.png',
  'stealth_fighter_s_001313.png',
  'station_wagon_s_002498.png',
  'pleasure_boat_s_001949.png',
  'cargo_ship_s_000060.png',
  'tabby_cat_s_001767.png',
  'crapaud_s_000055.png',
  'dump_truck_s_002008.png',
  'bufo_viridis_s_000506.png',
  'bullfrog_s_001295.png',
  'riding_horse_s_001884.png',
  'boat_s_002344.png',
  'ostrich_s_000950.png',
  'muntjac_s_000681.png',
  'pekingese_s_000279.png',
  'lipizzan_s_002052.png',
  'rana_pipiens_s_000570.png',
  'toy_s_002250.png',
  'tabby_s_001940.png',
  'dive_bomber_s_001294.png',
  'mutt_s_000086.png',
  'biplane_s_000701.png',
  'toy_s_000964.png',
  'jetliner_s_000807.png',
  'barge_s_000426.png',
  'cassowary_s_000255.png',
  'rana_clamitans_s_000059.png',
  'quarter_horse_s_000222.png',
  'tomcat_s_000377.png',
  'scow_s_000687.png',
  'emu_s_001571.png',
  'coupe_s_001625.png',
  'roan_s_000199.png',
  'bufo_bufo_s_000626.png',
  'tennessee_walking_horse_s_001470.png',
  'station_wagon_s_001497.png',
  'jumbo_jet_s_001056.png',
  'trailer_truck_s_000099.png',
  'puppy_s_000502.png',
  'pekingese_s_002101.png',
  'propeller_plane_s_001193.png',
  'car_s_001009.png',
  'lipizzan_s_002011.png',
  'green_frog_s_001375.png',
  'tipper_lorry_s_000586.png',
  'jetliner_s_000509.png',
  'musk_deer_s_000211.png',
  'appaloosa_s_002251.png',
  'arabian_s_001642.png',
  'compact_car_s_000435.png',
  'chihuahua_s_000063.png',
  'semi_s_000462.png',
  'wapiti_s_000455.png',
  'twinjet_s_001120.png',
  'cargo_vessel_s_001235.png',
  'peke_s_000274.png',
  'dump_truck_s_001422.png',
  'dump_truck_s_001863.png',
  'tailed_frog_s_000638.png',
  'gelding_s_000965.png',
  'coupe_s_000056.png',
  'pleasure_craft_s_000625.png',
  'house_cat_s_001328.png',
  'struthio_camelus_s_000081.png',
  'domestic_cat_s_000051.png',
  'passenger_ship_s_000948.png',
  'sparrow_s_002697.png',
  'prunella_modularis_s_000302.png',
  'fallow_deer_s_000192.png',
  'rana_palustris_s_000107.png',
  'attack_aircraft_s_001075.png',
  'twinjet_s_000046.png',
  'chihuahua_s_001038.png',
  'tabby_cat_s_001004.png',
  'cargo_vessel_s_001812.png',
  'dunnock_s_000037.png',
  'tabby_s_001740.png',
  'gelding_s_001638.png',
  'anthus_pratensis_s_001166.png',
  'camion_s_001988.png',
  'alley_cat_s_001645.png',
  'oil_tanker_s_001626.png',
  'tennessee_walking_horse_s_000355.png',
  'containership_s_000812.png',
  'passerine_s_000801.png',
  'lipizzan_s_000635.png',
  'truck_s_000635.png',
  'twinjet_s_000450.png',
  'honey_eater_s_000702.png',
  'tabby_s_000039.png',
  'pipit_s_000827.png',
  'elephant_bird_s_001004.png',
  'night_bird_s_000241.png',
  'alley_cat_s_001301.png',
  'house_cat_s_000332.png',
  'bufo_americanus_s_000492.png',
  'wagtail_s_001443.png',
  'tabby_cat_s_000587.png',
  'gamecock_s_000098.png',
  'cargo_ship_s_001867.png',
  'seaplane_s_000416.png',
  'english_toy_spaniel_s_001058.png',
  'japanese_spaniel_s_001324.png',
  'coupe_s_000179.png',
  'american_elk_s_000201.png',
  'chihuahua_s_000666.png',
  'spadefoot_s_000219.png',
  'bufo_s_001733.png',
  'struthio_camelus_s_001368.png',
  'arab_s_000456.png',
  'monoplane_s_001154.png',
  'coupe_s_001690.png',
  'dawn_horse_s_000752.png',
  'tennessee_walking_horse_s_000013.png',
  'cabin_cruiser_s_001500.png',
  'sparrow_s_001750.png',
  'aerial_ladder_truck_s_001306.png',
  'dromaius_novaehollandiae_s_000433.png',
  'passerine_s_000376.png',
  'caribou_s_001447.png',
  'bird_s_000356.png',
  'motorcar_s_001389.png',
  'auto_s_000675.png',
  'automobile_s_001833.png',
  'rana_temporaria_s_000008.png',
  'spring_frog_s_001263.png',
  'barking_frog_s_000397.png',
  'puppy_s_000785.png',
  'convertible_s_001861.png',
  'convertible_s_000859.png',
  'american_saddle_horse_s_000039.png',
  'biplane_s_000244.png',
  'red_deer_s_000704.png',
  'tabby_s_001071.png',
  'tabby_cat_s_002548.png',
  'tennessee_walker_s_001791.png',
  'auto_s_002048.png',
  'dromaius_novaehollandiae_s_000296.png',
  'felis_catus_s_001460.png',
  'peke_s_000440.png',
  'chihuahua_s_000349.png',
  'lapdog_s_002195.png',
  'bufo_marinus_s_001243.png',
  'convertible_s_001433.png',
  'mule_deer_s_000628.png',
  'tabby_s_000410.png',
  'stud_s_000099.png',
  'passenger_ship_s_001530.png',
  'scow_s_000131.png',
  'tabby_s_001868.png',
  'bufo_bufo_s_002334.png',
  'bufo_viridis_s_000152.png',
  'passerine_s_001306.png',
  'house_cat_s_002116.png',
  'jetliner_s_001188.png',
  'moving_van_s_001359.png',
  'wapiti_s_000663.png',
  'felis_catus_s_000998.png',
  'sea_boat_s_001979.png',
  'fighter_aircraft_s_001776.png',
  'fighter_aircraft_s_001096.png',
  'shooting_brake_s_000774.png',
  'automobile_s_001310.png',
  'pekingese_s_001023.png',
  'american_elk_s_000455.png',
  'trucking_rig_s_001243.png',
  'house_cat_s_000877.png',
  'taxi_s_000040.png',
  'pilot_boat_s_001100.png',
  'articulated_lorry_s_000223.png',
  'felis_domesticus_s_000164.png',
  'trucking_rig_s_001300.png',
  'tractor_trailer_s_000254.png',
  'flightless_bird_s_000142.png',
  'dump_truck_s_000939.png',
  'deer_s_000967.png',
  'ship_s_001904.png',
  'wagtail_s_000309.png',
  'trailer_truck_s_000815.png',
  'container_ship_s_002477.png',
  'guard_boat_s_000285.png',
  'estate_car_s_000446.png',
  'puppy_s_001621.png',
  'tabby_cat_s_000689.png',
  'american_toad_s_000650.png',
  'boat_s_002381.png',
  'female_horse_s_000089.png',
  'bufo_viridis_s_001370.png',
  'ladder_truck_s_000044.png',
  'cargo_ship_s_001321.png',
  'twinjet_s_000547.png',
  'cascades_frog_s_000061.png',
  'odocoileus_hemionus_s_001017.png',
  'attack_aircraft_s_001254.png',
  'airplane_s_000006.png',
  'songbird_s_001232.png',
  'peke_s_000388.png',
  'boat_s_002143.png',
  'accentor_s_001182.png',
  'dive_bomber_s_001066.png',
  'dunnock_s_001650.png',
  'dawn_horse_s_001127.png',
  'european_toad_s_000359.png',
  'wrecker_s_002325.png',
  'lipizzan_s_000118.png',
  'automobile_s_000327.png',
  'pekinese_s_002027.png',
  'maltese_s_001913.png',
  'toad_frog_s_000836.png',
  'leopard_frog_s_001231.png',
  'domestic_cat_s_001066.png',
  'barking_frog_s_000840.png',
  'finch_s_000038.png',
  'capreolus_capreolus_s_000257.png',
  'lippizan_s_000126.png',
  'attack_aircraft_s_000089.png',
  'mutt_s_002278.png',
  'true_toad_s_000021.png',
  'capreolus_capreolus_s_000178.png',
  'bufo_viridis_s_000600.png',
  'chihuahua_s_000097.png',
  'meadow_pipit_s_001282.png',
  'red_deer_s_000008.png',
  'leopard_frog_s_001528.png',
  'estate_car_s_001220.png',
  'bullfrog_s_000408.png',
  'stealth_fighter_s_001637.png',
  'capreolus_capreolus_s_001669.png',
  'airbus_s_000736.png',
  'tabby_s_000292.png',
  'station_wagon_s_002793.png',
  'boat_s_000765.png',
  'english_toy_spaniel_s_000814.png',
  'european_elk_s_000265.png',
  'roe_deer_s_000371.png',
  'convertible_s_000414.png',
  'cow_pony_s_000145.png',
  'tabby_s_000126.png',
  'trucking_rig_s_001628.png',
  'muntjac_s_001738.png',
  'broodmare_s_000582.png',
  'fire_engine_s_001540.png',
  'gelding_s_000640.png',
  'tabby_cat_s_002533.png',
  'appaloosa_s_000342.png',
  'pipit_s_001066.png',
  'speedboat_s_000410.png',
  'roe_deer_s_001014.png',
  'american_toad_s_001815.png',
  'bufo_calamita_s_000773.png',
  'estate_car_s_000044.png',
  'cassowary_s_000021.png',
  'lorry_s_001709.png',
  'amphibious_aircraft_s_000141.png',
  'capreolus_capreolus_s_000073.png',
  'motorboat_s_000773.png',
  'lipizzan_s_001658.png',
  'tabby_cat_s_000531.png',
  'truck_s_000289.png',
  'speedboat_s_001711.png',
  'quarter_horse_s_000557.png',
  'quarter_horse_s_000942.png',
  'jumbo_jet_s_001476.png',
  'struthio_camelus_s_001197.png',
  'fallow_deer_s_001623.png',
  'compact_car_s_000850.png',
  'coupe_s_001912.png',
  'roe_deer_s_000678.png',
  'car_s_000674.png',
  'chihuahua_s_001937.png',
  'red_deer_s_002698.png',
  'jumbojet_s_001587.png',
  'toy_spaniel_s_001692.png',
  'rana_clamitans_s_000766.png',
  'pipit_s_000025.png',
  'cargo_vessel_s_001809.png',
  'pekingese_s_001613.png',
  'dive_bomber_s_000707.png',
  'lark_s_000550.png',
  'estate_car_s_000292.png',
  'tabby_cat_s_001853.png',
  'lapdog_s_001963.png',
  'palfrey_s_000300.png',
  'cat_s_002298.png',
  'mutt_s_000188.png',
  'police_cruiser_s_000471.png',
  'true_cat_s_001598.png',
  'peke_s_000638.png',
  ...],
 'labels': [6,
  9,
  9,
  4,
  1,
  1,
  2,
  7,
  8,
  3,
  4,
  7,
  7,
  2,
  9,
  9,
  9,
  3,
  2,
  6,
  4,
  3,
  6,
  6,
  2,
  6,
  3,
  5,
  4,
  0,
  0,
  9,
  1,
  3,
  4,
  0,
  3,
  7,
  3,
  3,
  5,
  2,
  2,
  7,
  1,
  1,
  1,
  2,
  2,
  0,
  9,
  5,
  7,
  9,
  2,
  2,
  5,
  2,
  4,
  3,
  1,
  1,
  8,
  2,
  1,
  1,
  4,
  9,
  7,
  8,
  5,
  9,
  6,
  7,
  3,
  1,
  9,
  0,
  3,
  1,
  3,
  5,
  4,
  5,
  7,
  7,
  4,
  7,
  9,
  4,
  2,
  3,
  8,
  0,
  1,
  6,
  1,
  1,
  4,
  1,
  8,
  3,
  9,
  6,
  6,
  1,
  8,
  5,
  2,
  9,
  9,
  8,
  1,
  7,
  7,
  0,
  0,
  6,
  9,
  1,
  2,
  2,
  9,
  2,
  6,
  6,
  1,
  9,
  5,
  0,
  4,
  7,
  6,
  7,
  1,
  8,
  1,
  1,
  2,
  8,
  1,
  3,
  3,
  6,
  2,
  4,
  9,
  9,
  5,
  4,
  3,
  6,
  7,
  4,
  6,
  8,
  5,
  5,
  4,
  3,
  1,
  8,
  4,
  7,
  6,
  0,
  9,
  5,
  1,
  3,
  8,
  2,
  7,
  5,
  3,
  4,
  1,
  5,
  7,
  0,
  4,
  7,
  5,
  5,
  1,
  0,
  9,
  6,
  9,
  0,
  8,
  7,
  8,
  8,
  2,
  5,
  2,
  3,
  5,
  0,
  6,
  1,
  9,
  3,
  6,
  9,
  1,
  3,
  9,
  6,
  6,
  7,
  1,
  0,
  9,
  5,
  8,
  5,
  2,
  9,
  0,
  8,
  8,
  0,
  6,
  9,
  1,
  1,
  6,
  3,
  7,
  6,
  6,
  0,
  6,
  6,
  1,
  7,
  1,
  5,
  8,
  3,
  6,
  6,
  8,
  6,
  8,
  4,
  6,
  6,
  1,
  3,
  8,
  3,
  4,
  1,
  7,
  1,
  3,
  8,
  5,
  1,
  1,
  4,
  0,
  9,
  3,
  7,
  4,
  9,
  9,
  2,
  4,
  9,
  9,
  1,
  0,
  5,
  9,
  0,
  8,
  2,
  1,
  2,
  0,
  5,
  6,
  3,
  2,
  7,
  8,
  8,
  6,
  0,
  7,
  9,
  4,
  5,
  6,
  4,
  2,
  1,
  1,
  2,
  1,
  5,
  9,
  9,
  0,
  8,
  4,
  1,
  1,
  6,
  3,
  3,
  9,
  0,
  7,
  9,
  7,
  7,
  9,
  1,
  5,
  1,
  6,
  6,
  8,
  7,
  1,
  3,
  0,
  3,
  3,
  2,
  4,
  5,
  7,
  5,
  9,
  0,
  3,
  4,
  0,
  4,
  4,
  6,
  0,
  0,
  6,
  6,
  0,
  8,
  1,
  6,
  2,
  9,
  2,
  5,
  9,
  6,
  7,
  4,
  1,
  8,
  7,
  3,
  6,
  9,
  3,
  0,
  4,
  0,
  5,
  1,
  0,
  3,
  4,
  8,
  5,
  4,
  7,
  2,
  3,
  9,
  7,
  6,
  7,
  1,
  4,
  7,
  0,
  1,
  7,
  3,
  1,
  8,
  4,
  4,
  2,
  0,
  2,
  2,
  0,
  0,
  9,
  0,
  9,
  6,
  8,
  2,
  7,
  7,
  4,
  0,
  3,
  0,
  8,
  9,
  4,
  2,
  7,
  2,
  5,
  2,
  5,
  1,
  9,
  4,
  8,
  5,
  1,
  7,
  4,
  4,
  0,
  6,
  9,
  0,
  7,
  8,
  8,
  9,
  9,
  3,
  3,
  4,
  0,
  4,
  5,
  6,
  6,
  0,
  1,
  0,
  8,
  0,
  4,
  8,
  8,
  1,
  5,
  2,
  6,
  8,
  1,
  0,
  0,
  7,
  7,
  5,
  9,
  6,
  2,
  8,
  3,
  4,
  7,
  3,
  9,
  0,
  1,
  2,
  4,
  8,
  1,
  8,
  6,
  4,
  4,
  5,
  7,
  1,
  3,
  9,
  8,
  0,
  1,
  7,
  5,
  8,
  2,
  8,
  0,
  4,
  1,
  8,
  9,
  8,
  2,
  9,
  9,
  2,
  7,
  5,
  7,
  3,
  8,
  8,
  4,
  4,
  2,
  7,
  1,
  6,
  4,
  0,
  4,
  6,
  9,
  7,
  6,
  2,
  5,
  5,
  1,
  7,
  2,
  2,
  2,
  9,
  5,
  4,
  2,
  7,
  8,
  1,
  3,
  4,
  3,
  7,
  6,
  9,
  8,
  0,
  6,
  0,
  2,
  2,
  2,
  1,
  8,
  4,
  0,
  1,
  8,
  8,
  1,
  5,
  7,
  6,
  4,
  5,
  8,
  7,
  1,
  9,
  1,
  9,
  8,
  4,
  7,
  3,
  8,
  8,
  2,
  6,
  6,
  7,
  1,
  6,
  8,
  1,
  9,
  7,
  8,
  3,
  0,
  1,
  0,
  8,
  8,
  3,
  0,
  0,
  1,
  5,
  0,
  8,
  8,
  7,
  9,
  9,
  0,
  9,
  4,
  1,
  3,
  6,
  6,
  4,
  4,
  7,
  5,
  6,
  0,
  8,
  0,
  3,
  2,
  8,
  4,
  6,
  9,
  9,
  7,
  0,
  3,
  3,
  6,
  7,
  4,
  9,
  1,
  6,
  2,
  7,
  2,
  2,
  0,
  6,
  7,
  5,
  7,
  6,
  8,
  9,
  0,
  9,
  4,
  4,
  7,
  0,
  9,
  4,
  9,
  6,
  9,
  4,
  5,
  7,
  9,
  2,
  4,
  5,
  1,
  4,
  3,
  9,
  6,
  5,
  6,
  9,
  3,
  3,
  5,
  0,
  7,
  2,
  1,
  3,
  6,
  4,
  0,
  0,
  2,
  5,
  0,
  1,
  0,
  2,
  3,
  9,
  8,
  4,
  9,
  8,
  0,
  2,
  6,
  4,
  4,
  0,
  1,
  8,
  8,
  3,
  6,
  9,
  6,
  6,
  7,
  8,
  2,
  4,
  5,
  7,
  6,
  5,
  3,
  0,
  5,
  0,
  5,
  0,
  8,
  2,
  6,
  7,
  3,
  8,
  2,
  1,
  7,
  6,
  7,
  1,
  0,
  9,
  5,
  5,
  0,
  1,
  7,
  6,
  9,
  0,
  4,
  7,
  7,
  1,
  5,
  9,
  4,
  0,
  8,
  5,
  9,
  9,
  6,
  7,
  1,
  8,
  3,
  2,
  3,
  8,
  2,
  2,
  4,
  6,
  0,
  0,
  5,
  3,
  8,
  2,
  3,
  7,
  2,
  9,
  3,
  8,
  7,
  8,
  2,
  7,
  9,
  0,
  2,
  3,
  2,
  2,
  2,
  3,
  3,
  6,
  2,
  3,
  2,
  8,
  0,
  5,
  5,
  1,
  4,
  5,
  6,
  6,
  2,
  7,
  0,
  1,
  7,
  7,
  8,
  2,
  9,
  2,
  2,
  4,
  2,
  1,
  1,
  1,
  6,
  6,
  6,
  5,
  1,
  1,
  7,
  0,
  4,
  3,
  3,
  7,
  1,
  2,
  3,
  5,
  5,
  5,
  6,
  1,
  4,
  3,
  7,
  8,
  8,
  3,
  6,
  6,
  2,
  3,
  0,
  9,
  4,
  3,
  8,
  0,
  0,
  1,
  1,
  5,
  4,
  9,
  3,
  1,
  8,
  9,
  3,
  9,
  9,
  2,
  9,
  4,
  8,
  2,
  9,
  8,
  8,
  1,
  5,
  3,
  6,
  8,
  7,
  6,
  9,
  8,
  0,
  6,
  4,
  0,
  0,
  2,
  5,
  8,
  2,
  0,
  2,
  7,
  6,
  9,
  7,
  1,
  5,
  5,
  6,
  6,
  3,
  6,
  2,
  4,
  7,
  0,
  5,
  6,
  4,
  6,
  5,
  2,
  4,
  6,
  1,
  6,
  0,
  4,
  0,
  3,
  1,
  8,
  5,
  4,
  4,
  1,
  7,
  3,
  9,
  4,
  7,
  9,
  7,
  3,
  7,
  2,
  8,
  4,
  6,
  6,
  1,
  2,
  9,
  0,
  4,
  8,
  7,
  3,
  9,
  8,
  7,
  7,
  0,
  2,
  4,
  1,
  1,
  4,
  1,
  5,
  4,
  0,
  5,
  6,
  2,
  8,
  5,
  0,
  2,
  1,
  3,
  5,
  7,
  3,
  5,
  1,
  3,
  5,
  ...]}

In [16]:
df = pd.DataFrame(batch1['data'])
df['image'] = df.as_matrix().tolist()
df.head()   # drop from col0 to col3071


Out[16]:
0 1 2 3 4 5 6 7 8 9 ... 3063 3064 3065 3066 3067 3068 3069 3070 3071 image
0 59 43 50 68 98 119 139 145 149 149 ... 58 65 59 46 57 104 140 84 72 [59, 43, 50, 68, 98, 119, 139, 145, 149, 149, ...
1 154 126 105 102 125 155 172 180 142 111 ... 42 67 101 122 133 136 139 142 144 [154, 126, 105, 102, 125, 155, 172, 180, 142, ...
2 255 253 253 253 253 253 253 253 253 253 ... 83 80 69 66 72 79 83 83 84 [255, 253, 253, 253, 253, 253, 253, 253, 253, ...
3 28 37 38 42 44 40 40 24 32 43 ... 39 59 42 44 48 38 28 37 46 [28, 37, 38, 42, 44, 40, 40, 24, 32, 43, 30, 3...
4 170 168 177 183 181 177 181 184 189 189 ... 88 85 82 83 79 78 82 78 80 [170, 168, 177, 183, 181, 177, 181, 184, 189, ...

5 rows × 3073 columns


In [17]:
df.drop(range(3072),axis=1,inplace=True)
df.head()


Out[17]:
image
0 [59, 43, 50, 68, 98, 119, 139, 145, 149, 149, ...
1 [154, 126, 105, 102, 125, 155, 172, 180, 142, ...
2 [255, 253, 253, 253, 253, 253, 253, 253, 253, ...
3 [28, 37, 38, 42, 44, 40, 40, 24, 32, 43, 30, 3...
4 [170, 168, 177, 183, 181, 177, 181, 184, 189, ...

In [18]:
# get the dataframe from above dictionary
def get_dataframe(batch):
    df = pd.DataFrame(batch['data'])
    df['image'] = df.as_matrix().tolist()
    df.drop(range(3072),axis=1,inplace=True)    # remove the first 3072 cols
    df['label'] = batch['labels']
    return df


train = pd.concat([get_dataframe(batch1),get_dataframe(batch2),get_dataframe(batch3),
                   get_dataframe(batch4),get_dataframe(batch5)],ignore_index=True)
test = get_dataframe(batch_test)

In [19]:
train.head()


Out[19]:
image label
0 [59, 43, 50, 68, 98, 119, 139, 145, 149, 149, ... 6
1 [154, 126, 105, 102, 125, 155, 172, 180, 142, ... 9
2 [255, 253, 253, 253, 253, 253, 253, 253, 253, ... 9
3 [28, 37, 38, 42, 44, 40, 40, 24, 32, 43, 30, 3... 4
4 [170, 168, 177, 183, 181, 177, 181, 184, 189, ... 1

In [20]:
test.head()


Out[20]:
image label
0 [158, 159, 165, 166, 160, 156, 162, 159, 158, ... 3
1 [235, 231, 232, 232, 232, 232, 232, 232, 232, ... 8
2 [158, 158, 139, 132, 166, 182, 187, 193, 199, ... 8
3 [155, 167, 176, 190, 177, 166, 168, 166, 170, ... 0
4 [65, 70, 48, 30, 23, 40, 44, 45, 45, 40, 10, 1... 6

In [21]:
print train.shape, test.shape


(50000, 2) (10000, 2)

In [24]:
gltrain = gl.SFrame(train)
gltest = gl.SFrame(test)

In [26]:
# GraphLab can create a neural network on its own, based on the data
model = gl.neuralnet_classifier.create(gltrain, target='label', validation_set='auto')


Using network:

### network layers ###
layer[0]: FullConnectionLayer
  init_sigma = 0.01
  init_random = gaussian
  init_bias = 0
  num_hidden_units = 10
layer[1]: SigmoidLayer
layer[2]: FullConnectionLayer
  init_sigma = 0.01
  init_random = gaussian
  init_bias = 0
  num_hidden_units = 10
layer[3]: SoftmaxLayer
### end network layers ###

### network parameters ###
learning_rate = 0.001
momentum = 0.9
### end network parameters ###

PROGRESS: Creating a validation set from 5 percent of training data. This may take a while.
          You can set ``validation_set=None`` to disable validation tracking.

Creating neuralnet using cpu
Training with batch size = 100
+-----------+----------+--------------+-------------------+---------------------+-----------------+
| Iteration | Examples | Elapsed Time | Training-accuracy | Validation-accuracy | Examples/second |
+-----------+----------+--------------+-------------------+---------------------+-----------------+
| 1         | 47600    | 4.722913     | 0.107479          | 0.149235            | 10078.525391    |
| 2         | 47600    | 9.429859     | 0.170840          | 0.169492            | 10129.245117    |
| 3         | 47600    | 14.170055    | 0.169958          | 0.182307            | 10041.901367    |
| 4         | 47600    | 19.630333    | 0.172164          | 0.196362            | 8717.598633     |
| 5         | 47600    | 24.317955    | 0.177941          | 0.195122            | 10154.524414    |
| 6         | 47600    | 29.009386    | 0.180756          | 0.181067            | 10146.278320    |
| 7         | 47600    | 33.776242    | 0.165483          | 0.177759            | 9985.735352     |
| 8         | 47600    | 38.537223    | 0.173824          | 0.182720            | 9998.060547     |
| 9         | 47600    | 43.449501    | 0.180630          | 0.205043            | 9690.117188     |
| 10        | 47600    | 48.420854    | 0.189139          | 0.211244            | 9574.968750     |
+-----------+----------+--------------+-------------------+---------------------+-----------------+

In [27]:
model.evaluate(gltest)   # 'accuracy': 0.20749999582767487, better to try pre-trained model if you can find


Out[27]:
{'accuracy': 0.20749999582767487, 'confusion_matrix': Columns:
 	target_label	int
 	predicted_label	int
 	count	int
 
 Rows: 65
 
 Data:
 +--------------+-----------------+-------+
 | target_label | predicted_label | count |
 +--------------+-----------------+-------+
 |      0       |        0        |  247  |
 |      1       |        0        |  225  |
 |      2       |        0        |  173  |
 |      3       |        0        |  138  |
 |      4       |        0        |  111  |
 |      5       |        0        |  149  |
 |      6       |        0        |   91  |
 |      7       |        0        |  149  |
 |      8       |        0        |  188  |
 |      9       |        0        |  148  |
 +--------------+-----------------+-------+
 [65 rows x 3 columns]
 Note: Only the head of the SFrame is printed.
 You can use print_rows(num_rows=m, num_columns=n) to print more rows and columns.}

In [28]:
# Before using pre-trained model, you may need to convert your image size to the model required size
## Here, for example, to convert 32*32*3 images to 256*256*3 size

#Convert pixels to graphlab image format
gltrain['glimage'] = gl.SArray(gltrain['image']).pixel_array_to_image(32, 32, 3, allow_rounding = True)
gltest['glimage'] = gl.SArray(gltest['image']).pixel_array_to_image(32, 32, 3, allow_rounding = True)

In [29]:
#Remove the original column
gltrain.remove_column('image')
gltest.remove_column('image')
gltrain.head()


Out[29]:
label glimage
6 Height: 32 Width: 32
9 Height: 32 Width: 32
9 Height: 32 Width: 32
4 Height: 32 Width: 32
1 Height: 32 Width: 32
1 Height: 32 Width: 32
2 Height: 32 Width: 32
7 Height: 32 Width: 32
8 Height: 32 Width: 32
3 Height: 32 Width: 32
[10 rows x 2 columns]

In [30]:
#Convert into 256x256 size
gltrain['image'] = gl.image_analysis.resize(gltrain['glimage'], 256, 256, 3)
gltest['image'] = gl.image_analysis.resize(gltest['glimage'], 256, 256, 3)

#Remove old column:
gltrain.remove_column('glimage')
gltest.remove_column('glimage')

gltrain.head()


Out[30]:
label image
6 Height: 256 Width: 256
9 Height: 256 Width: 256
9 Height: 256 Width: 256
4 Height: 256 Width: 256
1 Height: 256 Width: 256
1 Height: 256 Width: 256
2 Height: 256 Width: 256
7 Height: 256 Width: 256
8 Height: 256 Width: 256
3 Height: 256 Width: 256
[10 rows x 2 columns]

In [ ]:
# below code has not be ran, since I don't have pre-trained model, and it can take very long time to run

#Load the pre-trained model: (change to your pre-trianed model location)
pretrained_model = gl.load_model('http://s3.amazonaws.com/GraphLab-Datasets/deeplearning/imagenet_model_iter45')

gltrain['features'] = pretrained_model.extract_features(gltrain)
gltest['features'] = pretrained_model.extract_features(gltest)

gltrain.head()

In [ ]:
# The advantage with GraphLab “classifier” function is that 
## it will automatically create various classifiers and chose the best model.

simple_classifier = graphlab.classifier.create(gltrain, features = ['features'], target = 'label')

In [ ]:
simple_classifier.evaluate(gltest)