[Refactor] Refactor single and multi label metrics

[yingfhu]

Refactor of SingleLabelMetric and MultiLabelMetric

Motivation

The class name is not apparent for user to know its function.

Modification

1. Rename those metrics to SingleLabelPrecsionRecallF1score and MultiLabelPrecsionRecallF1score
2. Add a wrapper PrecsionRecallF1score with an extra task argument for direct usage to create different instance.

Use cases (Optional)

>>> import torch
>>> preds  = torch.tensor([2, 0, 1, 1])
>>> labels = torch.tensor([2, 1, 2, 0])
>>> metric = PrecsionRecallF1score(num_classes=3)
>>> metric(preds, labels)
{'precision': 33.3333, 'recall': 16.6667, 'f1-score': 22.2222}
>>> metric = PrecsionRecallF1score(
        task="multilabel", average='micro', num_classes=3)
>>> metric(preds, labels)
{'precision_micro': 25.0, 'recall_micro': 25.0, 'f1-score_micro': 25.0}

Checklist

1. Pre-commit or other linting tools are used to fix the potential lint issues.
2. The modification is covered by complete unit tests. If not, please add more unit test to ensure the correctness.
3. If the modification has potential influence on downstream projects, this PR should be tested with downstream projects, like MMDet or MMCls.
4. The documentation has been modified accordingly, like docstring or example tutorials.

[CLAassistant]

All committers have signed the CLA.

[C1rN09]

Generally good! Please also modify the api docs && support_matrix documentation.

[C1rN09]

LGTM with 1 comments!