o
+ i2 @ s d dl mZ d dlmZmZmZ er0d dlmZ d dlm Z d dl
m Z ddl m
Z
mZ ed Zd dlZd dlZd dlZd d lmZ d dlZd d
lmZ d d lmZ g ZG d d
d
ee d Z G dd de Z dS ) ) annotations)
TYPE_CHECKINGAnyLiteralN)
_DTypeLiteral)
_Transform )
_ImageBackend_ImageDataTypetraintest)Image)_check_exists_and_download)Datasetc @ s e Zd ZU dZdZdZed ZdZed ZdZ ed Z
d Z ed
Z d Z
d ed
< ded< ded< ded< ded< ded< ded< ded< d.d/d"d#Zd0d%d&Zd1d*d+Zd2d,d-ZdS )3MNISTaX
Implementation of `MNIST `_ dataset.
Args:
image_path (str|None, optional): Path to image file, can be set None if
:attr:`download` is True. Default: None, default data path: ~/.cache/paddle/dataset/mnist.
label_path (str|None, optional): Path to label file, can be set None if
:attr:`download` is True. Default: None, default data path: ~/.cache/paddle/dataset/mnist.
mode (str, optional): Either train or test mode. Default 'train'.
transform (Callable|None, optional): Transform to perform on image, None for no transform. Default: None.
download (bool, optional): Download dataset automatically if
:attr:`image_path` :attr:`label_path` is not set. Default: True.
backend (str|None, optional): Specifies which type of image to be returned:
PIL.Image or numpy.ndarray. Should be one of {'pil', 'cv2'}.
If this option is not set, will get backend from :ref:`paddle.vision.get_image_backend `,
default backend is 'pil'. Default: None.
Returns:
:ref:`api_paddle_io_Dataset`. An instance of MNIST dataset.
Examples:
.. code-block:: python
>>> import itertools
>>> import paddle.vision.transforms as T
>>> from paddle.vision.datasets import MNIST
>>> mnist = MNIST()
>>> print(len(mnist))
60000
>>> for i in range(5): # only show first 5 images
... img, label = mnist[i]
... # do something with img and label
... print(type(img), img.size, label)
... # (28, 28) [5]
>>> transform = T.Compose(
... [
... T.ToTensor(),
... T.Normalize(
... mean=[127.5],
... std=[127.5],
... ),
... ]
... )
>>> mnist_test = MNIST(
... mode="test",
... transform=transform, # apply transform to every image
... backend="cv2", # use OpenCV as image transform backend
... )
>>> print(len(mnist_test))
10000
>>> for img, label in itertools.islice(iter(mnist_test), 5): # only show first 5 images
... # do something with img and label
... print(type(img), img.shape, label) # type: ignore
... # [1, 28, 28] [7]
Zmnistz$https://dataset.bj.bcebos.com/mnist/t10k-images-idx3-ubyte.gzZ 9fb629c4189551a2d022fa330f9573f3t10k-labels-idx1-ubyte.gzZ ec29112dd5afa0611ce80d1b7f02629ctrain-images-idx3-ubyte.gzZ f68b3c2dcbeaaa9fbdd348bbdeb94873train-labels-idx1-ubyte.gzZ d53e105ee54ea40749a09fcbcd1e9432 _DatasetModemode
str | None
image_path
label_path_Transform[Any, Any] | None transformr backendr dtypelistlabelsimagesNr Tdownloadbool_ImageBackend | NonereturnNonec C s" | dv s
J d| |d u rtj }|dvr!td| || _| | _|| _| jd u rU|s7J d|dkr>| jn| j }|dkrH| j
n| j }t |||| j
|| _|| _| jd u r|scJ d| jdkrk| jn| j} | jdkrv| jn| j}
t || |
| j
|| _|| _| t | _d S )Nr z*mode should be 'train' or 'test', but got )pilZcv2z4Expected backend are one of ['pil', 'cv2'], but got z?image_path is not set and downloading automatically is disabledr z?label_path is not set and downloading automatically is disabled)lowerpaddleZvisionZget_image_backend
ValueErrorr r r TRAIN_IMAGE_URLTEST_IMAGE_URLTRAIN_IMAGE_MD5TEST_IMAGE_MD5r NAMEr TRAIN_LABEL_URLTEST_LABEL_URLTRAIN_LABEL_MD5TEST_LABEL_MD5r _parse_datasetZget_default_dtyper ) selfr r r r r" r Z image_urlZ image_md5Z label_urlZ label_md5 r6 h/home/app/PaddleOCR-VL-test/.venv_paddleocr/lib/python3.10/site-packages/paddle/vision/datasets/mnist.py__init__~ sT
zMNIST.__init__d c C s g | _ g | _t| jd}| }t| jd}| }d}d}d}t|||\} }
} } |t |7 }d}
d}t|||
\}}|
t |7 }
||krPnddt
| d }t|||
}|
t |7 }
||7 }dt
|| | d }t|||}t ||| | f
d}|t |7 }t|D ] }| j ||d d f | jt || g
d qqKW d n1 sw Y W d d S W d d S 1 sw Y d S )
Nrbr z>IIIIz>IIT>BZfloat32int64)r! r gzipGzipFiler readr struct unpack_fromcalcsizestrnpreshapeastyperangeappendarray)r5 buffer_sizeZ
image_fileZimg_bufZ
label_fileZlab_bufZ
step_labelZ
offset_imgZmagic_byte_imgZ magic_imgZ image_numrowscolsZ
offset_labZmagic_byte_labZ magic_labZ label_numZ fmt_labelr Z
fmt_imagesZ images_tempr! ir6 r6 r7 r4 s^ "zMNIST._parse_datasetidxint,tuple[_ImageDataType, npt.NDArray[np.int64]]c C s | j | | j| }}t|ddg}| jdkr"tj|ddd}| jd ur,| |}| jdkr8||dfS || j |dfS )N r' Zuint8L)r r= )
r! r rE rF r r Z fromarrayrG r r )r5 rO imagelabelr6 r6 r7 __getitem__ s
zMNIST.__getitem__c C s
t | jS )N)lenr )r5 r6 r6 r7 __len__ s
z
MNIST.__len__)NNr NTN)r r r r r r r r r" r# r r$ r% r& )r9 )rO rP r% rQ )r% rP )__name__
__module__ __qualname____doc__r/
URL_PREFIXr, r. r1 r3 r+ r- r0 r2 __annotations__r8 r4 rV rX r6 r6 r6 r7 r ) s<
@
@
2r )r
znpt.NDArray[np.int64]c @ sH e Zd ZdZdZdZed ZdZed ZdZ ed Z
d Z ed
Z d Z
d S )
FashionMNISTa
Implementation of `Fashion-MNIST `_ dataset.
Args:
image_path (str, optional): Path to image file, can be set None if
:attr:`download` is True. Default: None, default data path: ~/.cache/paddle/dataset/fashion-mnist.
label_path (str, optional): Path to label file, can be set None if
:attr:`download` is True. Default: None, default data path: ~/.cache/paddle/dataset/fashion-mnist.
mode (str, optional): Either train or test mode. Default 'train'.
transform (Callable, optional): Transform to perform on image, None for no transform. Default: None.
download (bool, optional): Whether to download dataset automatically if
:attr:`image_path` :attr:`label_path` is not set. Default: True.
backend (str, optional): Specifies which type of image to be returned:
PIL.Image or numpy.ndarray. Should be one of {'pil', 'cv2'}.
If this option is not set, will get backend from :ref:`paddle.vision.get_image_backend `,
default backend is 'pil'. Default: None.
Returns:
:ref:`api_paddle_io_Dataset`. An instance of FashionMNIST dataset.
Examples:
.. code-block:: python
>>> import itertools
>>> import paddle.vision.transforms as T
>>> from paddle.vision.datasets import FashionMNIST
>>> fashion_mnist = FashionMNIST()
>>> print(len(fashion_mnist))
60000
>>> for i in range(5): # only show first 5 images
... img, label = fashion_mnist[i]
... # do something with img and label
... print(type(img), img.size, label)
... # (28, 28) [9]
>>> transform = T.Compose(
... [
... T.ToTensor(),
... T.Normalize(
... mean=[127.5],
... std=[127.5],
... ),
... ]
... )
>>> fashion_mnist_test = FashionMNIST(
... mode="test",
... transform=transform, # apply transform to every image
... backend="cv2", # use OpenCV as image transform backend
... )
>>> print(len(fashion_mnist_test))
10000
>>> for img, label in itertools.islice(iter(fashion_mnist_test), 5): # only show first 5 images
... # do something with img and label
... print(type(img), img.shape, label) # type: ignore
... # [1, 28, 28] [9]
z
fashion-mnistz,https://dataset.bj.bcebos.com/fashion_mnist/r Z bef4ecab320f06d8554ea6380940ec79r Z bb300cfdad3c16e7a12a480ee83cd310r Z 8d4fb7e6c68d591d4c3dfef9ec88bf0dr Z 25c81989df183df01b3e8a0aad5dffbeN)rY rZ r[ r\ r/ r] r, r. r1 r3 r+ r- r0 r2 r6 r6 r6 r7 r_ s @r_ )
__future__r typingr r r Z numpy.typingZnptZpaddle._typing.dtype_liker Z#paddle.vision.transforms.transformsr rT r r
r r> rA numpyrE ZPILr r) Zpaddle.dataset.commonr Z paddle.ior __all__tupler r_ r6 r6 r6 r7 s&
]