Image Datasets

Image datasets

Plots

---

source

make_grid

 make_grid (images, size=64)

Given a list of PIL images, stack them together into a line for easy viewing

---

source

show_images

 show_images (x:torch.Tensor, ncols:int=8)

Given a batch of images x, make a grid and convert to PIL

ImageDataset Mixin

---

source

ImagePlotMixin

 ImagePlotMixin ()

Mixin class for image datasets providing visualization of (image, label) samples

# ImagePlotMixin.plot(test, 0)
# ImagePlotMixin.plot_grid(test, 2,2)
# ImagePlotMixin.plot(test, 0, int2label = {0:'zero', 1:'one', 2:'two', 3:'three', 4:'four', 5:'five', 6:'six', 7:'seven', 8:'eight', 9:'nine'})
# ImagePlotMixin.plot(test, 0, int2label = test.hf_ds.features['label'].int2str)
# ImagePlotMixin.plot_grid(test, 2,2, int2label = test.hf_ds.features['label'].int2str)

Image Dataset

---

source

ImageDataset

 ImageDataset (name:str='mnist', *args,
               data_dir:Optional[str]='../data/image', split='train', tran
               sforms:Optional[torchvision.transforms.transforms.Compose]=
               Compose(     ToTensor() ), streaming:bool=False,
               exclude_grey_scale=False, verification_mode='no_checks')

Image dataset

Usage

Setup MNIST dataset. Download data if not found in specified location.

Tiny Imagenet

test = ImageDataset(
    'zh-plus/tiny-imagenet',
    data_dir=None,
    split='train',
    transforms=transforms.Compose([transforms.ToTensor(), transforms.Resize((32, 32))]), #,transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225))])
    exclude_grey_scale=False
)

print(f"num_classes: {test.num_classes}, splits: {test.splits}, labels: {test.label_names}")
print("from in to class method",test.hf_ds.features['label'].int2str)
print(test.hf_ds.features['label'])

idx = 10
print(f"label for image {idx} :", test[idx][1], test.int2str(test[idx][1]))
test.show(idx)
test.show_grid(5,5)

num_classes: 200, splits: dict_keys(['train', 'valid']), labels: ['n01443537', 'n01629819', 'n01641577', 'n01644900', 'n01698640', 'n01742172', 'n01768244', 'n01770393', 'n01774384', 'n01774750', 'n01784675', 'n01882714', 'n01910747', 'n01917289', 'n01944390', 'n01950731', 'n01983481', 'n01984695', 'n02002724', 'n02056570', 'n02058221', 'n02074367', 'n02094433', 'n02099601', 'n02099712', 'n02106662', 'n02113799', 'n02123045', 'n02123394', 'n02124075', 'n02125311', 'n02129165', 'n02132136', 'n02165456', 'n02226429', 'n02231487', 'n02233338', 'n02236044', 'n02268443', 'n02279972', 'n02281406', 'n02321529', 'n02364673', 'n02395406', 'n02403003', 'n02410509', 'n02415577', 'n02423022', 'n02437312', 'n02480495', 'n02481823', 'n02486410', 'n02504458', 'n02509815', 'n02666347', 'n02669723', 'n02699494', 'n02769748', 'n02788148', 'n02791270', 'n02793495', 'n02795169', 'n02802426', 'n02808440', 'n02814533', 'n02814860', 'n02815834', 'n02823428', 'n02837789', 'n02841315', 'n02843684', 'n02883205', 'n02892201', 'n02909870', 'n02917067', 'n02927161', 'n02948072', 'n02950826', 'n02963159', 'n02977058', 'n02988304', 'n03014705', 'n03026506', 'n03042490', 'n03085013', 'n03089624', 'n03100240', 'n03126707', 'n03160309', 'n03179701', 'n03201208', 'n03255030', 'n03355925', 'n03373237', 'n03388043', 'n03393912', 'n03400231', 'n03404251', 'n03424325', 'n03444034', 'n03447447', 'n03544143', 'n03584254', 'n03599486', 'n03617480', 'n03637318', 'n03649909', 'n03662601', 'n03670208', 'n03706229', 'n03733131', 'n03763968', 'n03770439', 'n03796401', 'n03814639', 'n03837869', 'n03838899', 'n03854065', 'n03891332', 'n03902125', 'n03930313', 'n03937543', 'n03970156', 'n03977966', 'n03980874', 'n03983396', 'n03992509', 'n04008634', 'n04023962', 'n04070727', 'n04074963', 'n04099969', 'n04118538', 'n04133789', 'n04146614', 'n04149813', 'n04179913', 'n04251144', 'n04254777', 'n04259630', 'n04265275', 'n04275548', 'n04285008', 'n04311004', 'n04328186', 'n04356056', 'n04366367', 'n04371430', 'n04376876', 'n04398044', 'n04399382', 'n04417672', 'n04456115', 'n04465666', 'n04486054', 'n04487081', 'n04501370', 'n04507155', 'n04532106', 'n04532670', 'n04540053', 'n04560804', 'n04562935', 'n04596742', 'n04598010', 'n06596364', 'n07056680', 'n07583066', 'n07614500', 'n07615774', 'n07646821', 'n07647870', 'n07657664', 'n07695742', 'n07711569', 'n07715103', 'n07720875', 'n07749582', 'n07753592', 'n07768694', 'n07871810', 'n07873807', 'n07875152', 'n07920052', 'n07975909', 'n08496334', 'n08620881', 'n08742578', 'n09193705', 'n09246464', 'n09256479', 'n09332890', 'n09428293', 'n12267677', 'n12520864', 'n13001041', 'n13652335', 'n13652994', 'n13719102', 'n14991210']
from in to class method <bound method ClassLabel.int2str of ClassLabel(names=['n01443537', 'n01629819', 'n01641577', 'n01644900', 'n01698640', 'n01742172', 'n01768244', 'n01770393', 'n01774384', 'n01774750', 'n01784675', 'n01882714', 'n01910747', 'n01917289', 'n01944390', 'n01950731', 'n01983481', 'n01984695', 'n02002724', 'n02056570', 'n02058221', 'n02074367', 'n02094433', 'n02099601', 'n02099712', 'n02106662', 'n02113799', 'n02123045', 'n02123394', 'n02124075', 'n02125311', 'n02129165', 'n02132136', 'n02165456', 'n02226429', 'n02231487', 'n02233338', 'n02236044', 'n02268443', 'n02279972', 'n02281406', 'n02321529', 'n02364673', 'n02395406', 'n02403003', 'n02410509', 'n02415577', 'n02423022', 'n02437312', 'n02480495', 'n02481823', 'n02486410', 'n02504458', 'n02509815', 'n02666347', 'n02669723', 'n02699494', 'n02769748', 'n02788148', 'n02791270', 'n02793495', 'n02795169', 'n02802426', 'n02808440', 'n02814533', 'n02814860', 'n02815834', 'n02823428', 'n02837789', 'n02841315', 'n02843684', 'n02883205', 'n02892201', 'n02909870', 'n02917067', 'n02927161', 'n02948072', 'n02950826', 'n02963159', 'n02977058', 'n02988304', 'n03014705', 'n03026506', 'n03042490', 'n03085013', 'n03089624', 'n03100240', 'n03126707', 'n03160309', 'n03179701', 'n03201208', 'n03255030', 'n03355925', 'n03373237', 'n03388043', 'n03393912', 'n03400231', 'n03404251', 'n03424325', 'n03444034', 'n03447447', 'n03544143', 'n03584254', 'n03599486', 'n03617480', 'n03637318', 'n03649909', 'n03662601', 'n03670208', 'n03706229', 'n03733131', 'n03763968', 'n03770439', 'n03796401', 'n03814639', 'n03837869', 'n03838899', 'n03854065', 'n03891332', 'n03902125', 'n03930313', 'n03937543', 'n03970156', 'n03977966', 'n03980874', 'n03983396', 'n03992509', 'n04008634', 'n04023962', 'n04070727', 'n04074963', 'n04099969', 'n04118538', 'n04133789', 'n04146614', 'n04149813', 'n04179913', 'n04251144', 'n04254777', 'n04259630', 'n04265275', 'n04275548', 'n04285008', 'n04311004', 'n04328186', 'n04356056', 'n04366367', 'n04371430', 'n04376876', 'n04398044', 'n04399382', 'n04417672', 'n04456115', 'n04465666', 'n04486054', 'n04487081', 'n04501370', 'n04507155', 'n04532106', 'n04532670', 'n04540053', 'n04560804', 'n04562935', 'n04596742', 'n04598010', 'n06596364', 'n07056680', 'n07583066', 'n07614500', 'n07615774', 'n07646821', 'n07647870', 'n07657664', 'n07695742', 'n07711569', 'n07715103', 'n07720875', 'n07749582', 'n07753592', 'n07768694', 'n07871810', 'n07873807', 'n07875152', 'n07920052', 'n07975909', 'n08496334', 'n08620881', 'n08742578', 'n09193705', 'n09246464', 'n09256479', 'n09332890', 'n09428293', 'n12267677', 'n12520864', 'n13001041', 'n13652335', 'n13652994', 'n13719102', 'n14991210'], id=None)>
ClassLabel(names=['n01443537', 'n01629819', 'n01641577', 'n01644900', 'n01698640', 'n01742172', 'n01768244', 'n01770393', 'n01774384', 'n01774750', 'n01784675', 'n01882714', 'n01910747', 'n01917289', 'n01944390', 'n01950731', 'n01983481', 'n01984695', 'n02002724', 'n02056570', 'n02058221', 'n02074367', 'n02094433', 'n02099601', 'n02099712', 'n02106662', 'n02113799', 'n02123045', 'n02123394', 'n02124075', 'n02125311', 'n02129165', 'n02132136', 'n02165456', 'n02226429', 'n02231487', 'n02233338', 'n02236044', 'n02268443', 'n02279972', 'n02281406', 'n02321529', 'n02364673', 'n02395406', 'n02403003', 'n02410509', 'n02415577', 'n02423022', 'n02437312', 'n02480495', 'n02481823', 'n02486410', 'n02504458', 'n02509815', 'n02666347', 'n02669723', 'n02699494', 'n02769748', 'n02788148', 'n02791270', 'n02793495', 'n02795169', 'n02802426', 'n02808440', 'n02814533', 'n02814860', 'n02815834', 'n02823428', 'n02837789', 'n02841315', 'n02843684', 'n02883205', 'n02892201', 'n02909870', 'n02917067', 'n02927161', 'n02948072', 'n02950826', 'n02963159', 'n02977058', 'n02988304', 'n03014705', 'n03026506', 'n03042490', 'n03085013', 'n03089624', 'n03100240', 'n03126707', 'n03160309', 'n03179701', 'n03201208', 'n03255030', 'n03355925', 'n03373237', 'n03388043', 'n03393912', 'n03400231', 'n03404251', 'n03424325', 'n03444034', 'n03447447', 'n03544143', 'n03584254', 'n03599486', 'n03617480', 'n03637318', 'n03649909', 'n03662601', 'n03670208', 'n03706229', 'n03733131', 'n03763968', 'n03770439', 'n03796401', 'n03814639', 'n03837869', 'n03838899', 'n03854065', 'n03891332', 'n03902125', 'n03930313', 'n03937543', 'n03970156', 'n03977966', 'n03980874', 'n03983396', 'n03992509', 'n04008634', 'n04023962', 'n04070727', 'n04074963', 'n04099969', 'n04118538', 'n04133789', 'n04146614', 'n04149813', 'n04179913', 'n04251144', 'n04254777', 'n04259630', 'n04265275', 'n04275548', 'n04285008', 'n04311004', 'n04328186', 'n04356056', 'n04366367', 'n04371430', 'n04376876', 'n04398044', 'n04399382', 'n04417672', 'n04456115', 'n04465666', 'n04486054', 'n04487081', 'n04501370', 'n04507155', 'n04532106', 'n04532670', 'n04540053', 'n04560804', 'n04562935', 'n04596742', 'n04598010', 'n06596364', 'n07056680', 'n07583066', 'n07614500', 'n07615774', 'n07646821', 'n07647870', 'n07657664', 'n07695742', 'n07711569', 'n07715103', 'n07720875', 'n07749582', 'n07753592', 'n07768694', 'n07871810', 'n07873807', 'n07875152', 'n07920052', 'n07975909', 'n08496334', 'n08620881', 'n08742578', 'n09193705', 'n09246464', 'n09256479', 'n09332890', 'n09428293', 'n12267677', 'n12520864', 'n13001041', 'n13652335', 'n13652994', 'n13719102', 'n14991210'], id=None)
label for image 10 : 0 n01443537

Smithsonian Butterflies

test = ImageDataset(
    "huggan/smithsonian_butterflies_subset",
    data_dir=None,
    split='train',
    transforms=transforms.Compose([transforms.ToTensor(),transforms.Resize((160, 160))]), #,transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225))])
    exclude_grey_scale=True
)

Repo card metadata block was not found. Setting CardData to empty.
[15:35:06] WARNING - Repo card metadata block was not found. Setting CardData to empty.
Repo card metadata block was not found. Setting CardData to empty.
[15:35:06] WARNING - Repo card metadata block was not found. Setting CardData to empty.

print(f"num_classes: {test.num_classes}, splits: {test.splits}, labels: {test.label_names}")
print("from in to class method",test.hf_ds.features['label'].int2str)
print(test.hf_ds.features['label'])

idx = 10
print(f"label for image {idx} :", test[idx][1], test.int2str(test[idx][1]))
test.show(idx)
test.show_grid(5,5)

WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.02822975..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.02822975..1.0000002].

num_classes: 45, splits: dict_keys(['train']), labels: ['Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Nymphalinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Sphingidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Larentiinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Pterygota, Holometabola, Lepidoptera, Nymphalidae, Danainae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Pieridae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Satyrinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Pieridae, Coliadinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Nymphalidae, Charaxinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Morphinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Saturniidae, Saturniinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Saturniidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Lasiocampidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Heliconiinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pieridae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Tortricidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Arctiidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pieridae, Pierinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Danainae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Pieridae, Pierinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pieridae, Dismorphiinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Ennominae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Arctiidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Noctuidae, Erebinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Heliconinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Sterrhinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Geometrinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pyralidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pieridae, Coliadinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Charaxinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Nymphalidae, Nymphalinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Papilionidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Glossata, Gelechiidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Limenitidinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Biblidinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Papilionidae, Papilioninae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Papilionidae, Parnassiinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Pieridae, Dismorphiinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Uraniidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Lycaenidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Lycaenidae, Lycaeninae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pyralidae, Epipaschiinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Papilionidae, Papilioninae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Oenochrominae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Lycaenidae, Polyommatinae']
from in to class method <bound method ClassLabel.int2str of ClassLabel(names=['Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Nymphalinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Sphingidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Larentiinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Pterygota, Holometabola, Lepidoptera, Nymphalidae, Danainae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Pieridae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Satyrinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Pieridae, Coliadinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Nymphalidae, Charaxinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Morphinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Saturniidae, Saturniinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Saturniidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Lasiocampidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Heliconiinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pieridae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Tortricidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Arctiidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pieridae, Pierinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Danainae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Pieridae, Pierinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pieridae, Dismorphiinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Ennominae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Arctiidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Noctuidae, Erebinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Heliconinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Sterrhinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Geometrinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pyralidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pieridae, Coliadinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Charaxinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Nymphalidae, Nymphalinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Papilionidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Glossata, Gelechiidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Limenitidinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Biblidinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Papilionidae, Papilioninae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Papilionidae, Parnassiinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Pieridae, Dismorphiinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Uraniidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Lycaenidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Lycaenidae, Lycaeninae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pyralidae, Epipaschiinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Papilionidae, Papilioninae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Oenochrominae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Lycaenidae, Polyommatinae'], id=None)>
ClassLabel(names=['Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Nymphalinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Sphingidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Larentiinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Pterygota, Holometabola, Lepidoptera, Nymphalidae, Danainae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Pieridae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Satyrinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Pieridae, Coliadinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Nymphalidae, Charaxinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Morphinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Saturniidae, Saturniinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Saturniidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Lasiocampidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Heliconiinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pieridae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Tortricidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Arctiidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pieridae, Pierinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Danainae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Pieridae, Pierinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pieridae, Dismorphiinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Ennominae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Arctiidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Noctuidae, Erebinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Heliconinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Sterrhinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Geometrinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pyralidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pieridae, Coliadinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Charaxinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Nymphalidae, Nymphalinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Papilionidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Glossata, Gelechiidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Limenitidinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Biblidinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Papilionidae, Papilioninae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Papilionidae, Parnassiinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Pieridae, Dismorphiinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Uraniidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Lycaenidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Lycaenidae, Lycaeninae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pyralidae, Epipaschiinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Papilionidae, Papilioninae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Oenochrominae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Lycaenidae, Polyommatinae'], id=None)
label for image 10 : 12 Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Heliconiinae

WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0025464185..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0025464185..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.03612411..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.03612411..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.00049994164..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.00049994164..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0010504292..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0010504292..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.006099089..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.006099089..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.13286446..1.0000001].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.13286446..1.0000001].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.07523744..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.07523744..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.001258907..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.001258907..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.053175174..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.053175174..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0037022692..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0037022692..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.043663528..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.043663528..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.022093028..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.022093028..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0002969489..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0002969489..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.01872363..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.01872363..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0121549405..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0121549405..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.017694354..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.017694354..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.024148637..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.024148637..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.00055700843..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.00055700843..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.020672267..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.020672267..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.04415208..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.04415208..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.095779866..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.095779866..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.00049994164..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.00049994164..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.008652165..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.008652165..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.014976031..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.014976031..1.0000002].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.06441312..1.0000002].
[15:35:09] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.06441312..1.0000002].

CIFAR10

test = ImageDataset(
    "cifar10",
    data_dir=None,
    split='test',
    transforms=transforms.Compose([transforms.ToTensor()]), #,transforms.Resize((32, 32))]), #,transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225))])
    exclude_grey_scale=True
)

print(f"num_classes: {test.num_classes}, splits: {test.splits}, labels: {test.label_names}")
print("from in to class method",test.hf_ds.features['label'].int2str)
idx = 10
print(f"label for image {idx} :", test[idx][1], test.int2str(test[idx][1]))
test.show(idx)
test.show_grid(5,5)

num_classes: 10, splits: dict_keys(['train', 'test']), labels: ['airplane', 'automobile', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck']
from in to class method <bound method ClassLabel.int2str of ClassLabel(names=['airplane', 'automobile', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck'], id=None)>
label for image 10 : 0 airplane

ImageNette

test = ImageDataset(
    'frgfm/imagenette',
    '160px',
    data_dir=None,
    split='validation',
    transforms=transforms.Compose([transforms.ToTensor(),transforms.Resize((160, 160))]), #,transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225))])
    exclude_grey_scale=True
)
print(f"num_classes: {test.num_classes}, splits: {test.splits}, labels: {test.label_names}")
print("from in to class method",test.hf_ds.features['label'].int2str)
# print(test.hf_ds.features['label'])
# test2 = ImageDataset('mnist', data_dir=None, train=True, transforms=transforms.ToTensor())
# test3 = ImageDataset('fashion_mnist', data_dir=None, train=True, transforms=transforms.ToTensor())
idx = 10
print(f"label for image {idx} :", test[idx][1], test.int2str(test[idx][1]))
test.show(idx)
test.show_grid(5,5)

num_classes: 10, splits: dict_keys(['train', 'validation']), labels: ['tench', 'English springer', 'cassette player', 'chain saw', 'church', 'French horn', 'garbage truck', 'gas pump', 'golf ball', 'parachute']
from in to class method <bound method ClassLabel.int2str of ClassLabel(names=['tench', 'English springer', 'cassette player', 'chain saw', 'church', 'French horn', 'garbage truck', 'gas pump', 'golf ball', 'parachute'], id=None)>
label for image 10 : 2 cassette player

print(f" X: {test[0][0].shape}, y: {test[0][1]}")
dl = DataLoader(test, batch_size=64)
print(next(iter(dl))[0].shape, next(iter(dl))[1].shape)

 X: torch.Size([3, 160, 160]), y: 39
torch.Size([64, 3, 160, 160]) torch.Size([64])

# output ( (C,H,W), int)
print(test[0][0].dtype, type(test[0][1]))
print(f"Number of samples in the dataset: {len(test)}")

# get item helper
X, y = test[0]
print(X.shape, y, X.type(), type(y))

# display each digit
test.show(0)

# split data
train, dev = test.train_dev_split(0.8)
print(len(train), len(dev))

WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.018742988..1.0000002].
[17:42:53] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.018742988..1.0000002].

torch.float32 <class 'int'>
Number of samples in the dataset: 1000
torch.Size([3, 160, 160]) 39 torch.FloatTensor <class 'int'>

800 200

Config

It is convenient to keep setup of specific dataset for an experiment in a config file for reproductibility

# instantiate dataset from yaml config file
cfg = OmegaConf.load("../config/image/data/mnist.yaml")
print(cfg.dataset)
test = instantiate(cfg.dataset)
type(test)

# output ( (B,C, H,W), int)
print(test[0][0].dtype, type(test[0][1]))
print(f"Number of samples in the dataset: {len(test)}")

# get item helper
X, y = test[0]
print(X.shape, y, X.type(), type(y))

# display each digit
test.show(0)

# split data
train, dev = test.train_dev_split(0.8)

{'_target_': 'nimrod.image.datasets.ImageDataset', 'name': 'mnist', 'data_dir': '../data/image', 'split': 'test', 'transforms': {'_target_': 'torchvision.transforms.Compose', 'transforms': [{'_target_': 'torchvision.transforms.ToTensor'}, {'_target_': 'torchvision.transforms.Normalize', 'mean': [0.0], 'std': [0.2]}]}}
torch.float32 <class 'int'>
Number of samples in the dataset: 10000
torch.Size([1, 28, 28]) 7 torch.FloatTensor <class 'int'>

Image DataModule

---

source

ImageDataModule

 ImageDataModule (name:str, *args, data_dir:Optional[str]='~/Data/', trans
                  forms:Union[torchvision.transforms.transforms.Compose,Ca
                  llable,NoneType]=Compose(     ToTensor()
                  Normalize(mean=(0.1307,), std=(0.3081,)) ),
                  train_val_split:List[float]=[0.8, 0.2],
                  batch_size:int=64, num_workers:int=0,
                  pin_memory:bool=False, persistent_workers:bool=False,
                  **kwargs)

Mixin class for image datasets providing visualization of (image, label) samples

Usage

dm = ImageDataModule(
    'frgfm/imagenette','160px',
    transforms=transforms.Compose([transforms.ToTensor(),transforms.Resize((128, 128))]),
    data_dir='../data/image',
    train_val_split=[0.8, 0.2],
    batch_size = 256,
    num_workers = 0, # main process
    pin_memory= False,
    persistent_workers=False,
    exclude_grey_scale = True
)

# download or reference data from dir
dm.prepare_data()

# define train, eval, test subsets
dm.setup()
print(f" num_classes: {dm.num_classes}, labels: {dm.label_names}, img shape: {dm.train_ds[0][0].shape}")
# show data
dm.show(1)
dm.show_grid(3,3)

[14:05:25] INFO - Init ImageDataModule for frgfm/imagenette
[14:05:29] INFO - split train into train/val [0.8, 0.2]
[14:05:29] INFO - train: 7575 val: 1894, test: 3925
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0..1.0000001].
[14:05:29] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0..1.0000001].

 num_classes: 10, labels: ['tench', 'English springer', 'cassette player', 'chain saw', 'church', 'French horn', 'garbage truck', 'gas pump', 'golf ball', 'parachute'], img shape: torch.Size([3, 128, 128])

WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0096571585..1.0000001].
[14:05:29] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.0096571585..1.0000001].
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.10879702..1.0000001].
[14:05:29] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [0.10879702..1.0000001].

# dm.show_batch(
xb, yb = next(iter(dm.train_dataloader()))
# print(xb.shape)
dm.show_batch(xb)

[14:05:36] WARNING - Skipping sample at index 2393 because doesn't have 3 channels
[14:05:36] WARNING - Skipping sample at index 9298 because doesn't have 3 channels
[14:05:36] WARNING - Skipping sample at index 9103 because doesn't have 3 channels
[14:05:36] WARNING - Skipping sample at index 3045 because doesn't have 3 channels
[14:05:36] WARNING - Skipping sample at index 2931 because doesn't have 3 channels
[14:05:36] WARNING - Skipping sample at index 187 because doesn't have 3 channels

# dm.label_names

['tench',
 'English springer',
 'cassette player',
 'chain saw',
 'church',
 'French horn',
 'garbage truck',
 'gas pump',
 'golf ball',
 'parachute']

# access data batches via dataloader
test_dl = dm.test_dataloader()
# X,Y = next(iter(test_dl))
# print("X dim(B,C,W,H): ", X.shape, "Y: dim(B)", Y.shape)

Config

# cfg = OmegaConf.load("../config/image/data/mnist.yaml")
# print(cfg.datamodule)
# dm = instantiate(cfg.datamodule)
# dm.prepare_data()
# dm.setup()
# test_dl = dm.test_dataloader()
# len(dm.test_ds), len(dm.train_ds), len(dm.val_ds)
cfg = OmegaConf.load('../config/data/image/fashion_mnist.yaml')
dm = instantiate(cfg)
dm.prepare_data()
dm.setup()
print(f"num_classes: {dm.num_classes}, batch_size: {dm.batch_size}")
print(f"labels: {dm.label_names}")
x, y = dm.test_ds[0]
print(f"X: {x.shape}, Y: {y}")
dm.show(1)
dm.show_grid(3,3)

[17:28:55] INFO - Init ImageDataModule for fashion_mnist
[17:29:08] INFO - split train into train/val [0.8, 0.2]
[17:29:08] INFO - train: 48000 val: 12000, test: 10000

num_classes: 10, batch_size: 128
labels: ['T - shirt / top', 'Trouser', 'Pullover', 'Dress', 'Coat', 'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle boot']
X: torch.Size([1, 32, 32]), Y: 9

cfg = OmegaConf.load('../config/data/image/smithsonian_butterflies.yaml')
dm = instantiate(cfg)
dm.prepare_data()
dm.setup()
print(f"num_classes: {dm.num_classes}, batch_size: {dm.batch_size}")
print(f"labels: {dm.label_names}")
x, y = dm.test_ds[0]
print(f"X: {x.shape}, Y: {y}")
x,y = next(iter(dm.train_dataloader()))
print(f"X: {x.shape}, Y: {y.shape}")
dm.show(1)

[17:40:22] INFO - Init ImageDataModule for huggan/smithsonian_butterflies_subset
Repo card metadata block was not found. Setting CardData to empty.
[17:40:22] WARNING - Repo card metadata block was not found. Setting CardData to empty.
Repo card metadata block was not found. Setting CardData to empty.
[17:40:24] WARNING - Repo card metadata block was not found. Setting CardData to empty.
[17:40:29] WARNING - split train into train/val/test [0.8, 0.2] 
[17:40:29] INFO - train: 800 val: 200, test: 200
[17:40:29] INFO - split train into train/val [0.8, 0.2]
WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [-0.8029231..1.0000005].
[17:40:29] WARNING - Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers). Got range [-0.8029231..1.0000005].

num_classes: 45, batch_size: 128
labels: ['Animalia, Arthropoda, Insecta, Lepidoptera, Pyralidae, Epipaschiinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Larentiinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Nymphalinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Satyrinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Saturniidae, Saturniinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Limenitidinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Papilionidae, Papilioninae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Pieridae, Dismorphiinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Arctiidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Lycaenidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Papilionidae, Parnassiinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pyralidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Pieridae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Heliconinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Arctiidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Noctuidae, Erebinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Morphinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pieridae, Coliadinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pieridae, Pierinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Papilionidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Ennominae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pieridae, Dismorphiinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Sphingidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Lasiocampidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Danainae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Pieridae, Pierinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Pieridae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Nymphalidae, Nymphalinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Saturniidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Papilionidae, Papilioninae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Oenochrominae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Glossata, Gelechiidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Pieridae, Coliadinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Lycaenidae, Lycaeninae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Uraniidae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Heliconiinae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Nymphalidae, Charaxinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Pterygota, Holometabola, Lepidoptera, Nymphalidae, Danainae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Sterrhinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Tortricidae', 'Animalia, Arthropoda, Insecta, Lepidoptera, Geometridae, Geometrinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Biblidinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Nymphalidae, Charaxinae', 'Animalia, Arthropoda, Hexapoda, Insecta, Lepidoptera, Lycaenidae, Polyommatinae']
X: torch.Size([3, 128, 128]), Y: 7
X: torch.Size([128, 3, 128, 128]), Y: torch.Size([128])