refactor A2C/PPO, change behavior of value normalization

[ChenDRAG]

Refactor A2C/PPO to make them look consistent. Change behavior of value normalization, etc to be prepared for A2C benchmark.

[ChenDRAG]

related to #307

[codecov-io]

Codecov Report

Merging #321 (4772e71) into master (47c7789) will decrease coverage by 0.02%.
The diff coverage is 91.89%.

@@            Coverage Diff             @@
##           master     #321      +/-   ##
==========================================
- Coverage   93.97%   93.95%   -0.03%     
==========================================
  Files          51       51              
  Lines        3269     3274       +5     
==========================================
+ Hits         3072     3076       +4     
- Misses        197      198       +1     

Flag	Coverage Δ
unittests	93.95% <91.89%> (-0.03%)	⬇️

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

Impacted Files	Coverage Δ
tianshou/policy/modelfree/a2c.py	90.16% <86.95%> (-0.91%)	⬇️
tianshou/policy/modelfree/pg.py	96.00% <100.00%> (ø)
tianshou/policy/modelfree/ppo.py	94.36% <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 47c7789...4772e71. Read the comment docs.