[Refactor] Support to use "split" to specify training set/validation set in the ImageNet dataset

[zzc98]

Motivation

• Support to use split to specify training set/validation set in the ImageNet dataset
• Support to use split to specify training set/test set in the Mnist/Fashion-Mnist dataset
• Support to use split to specify training set/validation set in the VOC dataset
• Refactor SUN397 datasets
• Update docs/en/user_guides/dataset_prepare.md and docs/zh_CN/user_guides/dataset_prepare.md

Examples

# ImageNet
>>> from mmpretrain.datasets import ImageNet
>>> train_dataset = ImageNet(data_root='data/imagenet', split='train')
>>> train_dataset
Dataset ImageNet
	Number of samples:  1281167
	Number of categories:       1000
	Root of dataset:    data/imagenet
>>> test_dataset = ImageNet(data_root='data/imagenet', split='val')
>>> test_dataset
Dataset ImageNet
	Number of samples:  50000
	Number of categories:       1000
	Root of dataset:    data/imagenet

# ImageNet21K
>>> from mmpretrain.datasets import ImageNet21k
>>> train_dataset = ImageNet21k(data_root='data/imagenet21k', split='train')
>>> train_dataset
Dataset ImageNet21k
	Number of samples:  14197088
	Annotation file:    data/imagenet21k/meta/train.txt
	Prefix of images:   data/imagenet21k/train
    
# VOC
>>> from mmpretrain.datasets import VOC
>>> train_dataset = VOC(data_root='data/VOC2007', split='trainval')
>>> train_dataset
Dataset VOC
	Number of samples:  5011
	Number of categories:       20
	Prefix of dataset:  data/VOC2007
	Path of image set:  data/VOC2007/ImageSets/Main/trainval.txt
	Prefix of images:   data/VOC2007/JPEGImages
	Prefix of annotations:      data/VOC2007/Annotations
>>> test_dataset = VOC(data_root='data/VOC2007', split='test')
>>> test_dataset
Dataset VOC
	Number of samples:  4952
	Number of categories:       20
	Prefix of dataset:  data/VOC2007
	Path of image set:  data/VOC2007/ImageSets/Main/test.txt
	Prefix of images:   data/VOC2007/JPEGImages
	Prefix of annotations:      data/VOC2007/Annotations

# SUN397
>>> from mmpretrain.datasets import SUN397
>>> train_dataset = SUN397(data_root='data/SUN397', split='train')
>>> train_dataset
Dataset SUN397
	Number of samples:  19850
	Number of categories:       397
	Root of dataset:    data/SUN397
>>> test_dataset = SUN397(data_root='data/SUN397', split='test')
>>> test_dataset
Dataset SUN397
	Number of samples:  19850
	Number of categories:       397
	Root of dataset:    data/SUN397

Checklist

Before PR:

• Pre-commit or other linting tools are used to fix the potential lint issues.
• Bug fixes are fully covered by unit tests, the case that causes the bug should be added in the unit tests.
• The modification is covered by complete unit tests. If not, please add more unit test to ensure the correctness.
• The documentation has been modified accordingly, like docstring or example tutorials.

After PR:

• If the modification has potential influence on downstream or other related projects, this PR should be tested with those projects, like MMDet or MMSeg.
• CLA has been signed and all committers have signed the CLA in this PR.

[codecov]

Codecov Report

Patch coverage: 26.19% and project coverage change: -0.99 ⚠️

Comparison is base (f9dcae2) 68.16% compared to head (4dba735) 67.18%.

❗ Current head 4dba735 differs from pull request most recent head 30c95ac. Consider uploading reports for the commit 30c95ac to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##              dev    #1535      +/-   ##
==========================================
- Coverage   68.16%   67.18%   -0.99%     
==========================================
  Files         295      303       +8     
  Lines       23372    23958     +586     
  Branches     3713     3798      +85     
==========================================
+ Hits        15932    16095     +163     
- Misses       6880     7290     +410     
- Partials      560      573      +13     

Flag	Coverage Δ
unittests	67.18% <26.19%> (-0.99%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
mmpretrain/apis/image_retrieval.py	21.42% <ø> (ø)
mmpretrain/datasets/__init__.py	70.27% <0.00%> (-4.02%)	⬇️
mmpretrain/datasets/gqa_dataset.py	0.00% <0.00%> (ø)
mmpretrain/datasets/nocaps.py	0.00% <0.00%> (ø)
mmpretrain/datasets/scienceqa.py	0.00% <ø> (ø)
mmpretrain/models/multimodal/__init__.py	40.00% <0.00%> (-4.45%)	⬇️
...retrain/models/multimodal/chinese_clip/__init__.py	0.00% <0.00%> (ø)
mmpretrain/models/multimodal/chinese_clip/bert.py	0.00% <0.00%> (ø)
...ain/models/multimodal/chinese_clip/chinese_clip.py	0.00% <0.00%> (ø)
mmpretrain/models/multimodal/chinese_clip/utils.py	0.00% <0.00%> (ø)
... and 17 more

... and 1 file with indirect coverage changes

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.

[Ezra-Yu]

LGTM.