Imagenet¶
Attribute |
Value |
|---|---|
pretty_name |
Imagenet Dataset |
annotations_creators |
|
language_creators |
|
languages |
|
licenses |
custom |
multilinguality |
|
size_categories |
1M<n<10M |
source_datasets |
|
task_categories |
image-classification |
task_ids |
|
paperswithcode_id |
imagenet |
Dataset Description¶
Homepage: Imagenet
Licenses: non-commercial use
Dataset Summary¶
General image classification.
Download and prepare data¶
Download the data directly from kaggle and extract it. Replace {PATH_TO_DATA} below with the location of the folder containing the data. Use the following code to load it:
from squirrel_datasets_core.datasets.imagenet import RawImageNetDriver
iter_train = RawImageNetDriver("{PATH_TO_DATA}").get_iter("train")
iter_val = RawImageNetDriver("{PATH_TO_DATA}").get_iter("val")
iter_test = RawImageNetDriver("{PATH_TO_DATA}").get_iter("test")
Data Splits¶
name |
train |
val |
test |
imagenet |
1,2M |
50K |
100K |