[Improve] Update registries of mmcls.

[mzr1996]

Motivation

MMEngine has supported to auto import modules from the registry and added some other updates.

Modification

1. Add locations to all registries in mmcls to support auto-import.
2. Adjust all scripts. For scripts using runner, remove the register_all_modules since it's not necessary. For scripts without using runner, replace register_all_modules to init_default_scope.
3. Update the minimum version of Python to 3.7 since MMengine has updated.

Usage

With this PR, we can directly build components from the registry without importing the related package. For example

>>> from mmcls.registry import DATASETS
>>> DATASETS.build({'type': 'CIFAR10', 'data_prefix': 'data/cifar10', 'test_mode': True})
Dataset CIFAR10
    Number of samples: 	10000
    Number of categories: 	10
    Prefix of data: 	data/cifar10

[codecov]

Codecov Report

Base: 0.02% // Head: 88.60% // Increases project coverage by +88.57% 🎉

Coverage data is based on head (ac6543f) compared to base (b8b31e9).
Patch has no changes to coverable lines.

❗ Current head ac6543f differs from pull request most recent head cd552d6. Consider uploading reports for the commit cd552d6 to get more accurate results

Additional details and impacted files

@@             Coverage Diff              @@
##           dev-1.x    #1306       +/-   ##
============================================
+ Coverage     0.02%   88.60%   +88.57%     
============================================
  Files          121      161       +40     
  Lines         8217    12687     +4470     
  Branches      1368     2038      +670     
============================================
+ Hits             2    11241    +11239     
+ Misses        8215     1107     -7108     
- Partials         0      339      +339     

Flag	Coverage Δ
unittests	88.60% <ø> (+88.57%)	⬆️

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

Impacted Files	Coverage Δ
mmcls/datasets/transforms/compose.py
mmcls/models/classifiers/timm.py	25.97% <0.00%> (ø)
mmcls/evaluation/metrics/multi_task.py	94.59% <0.00%> (ø)
mmcls/models/necks/reduction.py	100.00% <0.00%> (ø)
mmcls/models/classifiers/hugging_face.py	25.64% <0.00%> (ø)
mmcls/models/tta/__init__.py	100.00% <0.00%> (ø)
mmcls/apis/model.py	82.29% <0.00%> (ø)
mmcls/structures/multi_task_data_sample.py	100.00% <0.00%> (ø)
mmcls/models/backbones/mobilevit.py	91.15% <0.00%> (ø)
mmcls/models/heads/efficientformer_head.py	93.10% <0.00%> (ø)
... and 152 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

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

[yingfhu]

LGTM