[Refactoring]Add Sequential with init_weight

[MeowZheng]

As developers always used nn.Sequential to implement a submodule, I add Sequential with init_weight for parameters initialization with init_cfg. In addition, I override __repr__ to display initialization information.

example

seq_model_cfg = [
        dict(
            type='FooConv1d', init_cfg=dict(type='Constant', val=0., bias=1.)),
        dict(
            type='FooConv2d', init_cfg=dict(type='Constant', val=2., bias=3.)),
    ]
layers = [build_from_cfg(cfg, COMPONENTS) for cfg in seq_model_cfg]
seq_model = Sequential(*layers)
seq_model.init_weight()

[codecov]

Codecov Report

Merging #884 (8f267a8) into master (73bff4e) will increase coverage by 0.00%.
The diff coverage is 71.42%.

@@           Coverage Diff           @@
##           master     #884   +/-   ##
=======================================
  Coverage   66.58%   66.58%           
=======================================
  Files         145      145           
  Lines        8828     8841   +13     
  Branches     1605     1606    +1     
=======================================
+ Hits         5878     5887    +9     
- Misses       2633     2637    +4     
  Partials      317      317           

Flag	Coverage Δ
unittests	66.58% <71.42%> (+<0.01%)	⬆️

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

Impacted Files	Coverage Δ
mmcv/runner/base_module.py	79.41% <69.23%> (-6.31%)	⬇️
mmcv/runner/__init__.py	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 73bff4e...3c82d65. Read the comment docs.

[ZwwWayne]

Should we also handle nn.ModuleList?