Add Intrinsic Curiosity Module

[nuance1979]

• I have marked all applicable categories:
  • exception-raising fix
  • algorithm implementation fix
  • documentation modification
  • new feature
• I have reformatted the code using make format (required)
• I have checked the code using make commit-checks (required)
• If applicable, I have mentioned the relevant/related issue(s)
• If applicable, I have listed every items in this Pull Request below

Implementation of Intrinsic Curiosity Module (ICM, #502)

Paper: https://arxiv.org/pdf/1705.05363.pdf
Reference repos:

• https://github.com/pathak22/noreward-rl/
• https://github.com/uvipen/Street-fighter-A3C-ICM-pytorch

ICM is implemented as a decorator because it can be used with any underlying policy. I have prepared an example with atari dqn for simplicity but it might not be able to give better performance for many Atari envs. I'm trying to get an example with VizDoom env running.

[codecov-commenter]

Codecov Report

Merging #503 (3deb076) into master (a2d76d1) will increase coverage by 0.07%.
The diff coverage is 98.55%.

@@            Coverage Diff             @@
##           master     #503      +/-   ##
==========================================
+ Coverage   94.11%   94.18%   +0.07%     
==========================================
  Files          61       62       +1     
  Lines        4059     4127      +68     
==========================================
+ Hits         3820     3887      +67     
- Misses        239      240       +1     

Flag	Coverage Δ
unittests	94.18% <98.55%> (+0.07%)	⬆️

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

Impacted Files	Coverage Δ
tianshou/policy/modelbased/icm.py	97.95% <97.95%> (ø)
tianshou/policy/__init__.py	100.00% <100.00%> (ø)
tianshou/utils/net/discrete.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 a2d76d1...3deb076. Read the comment docs.