Switch to pytorch mixed precision

[senwu]

Description of the proposed changes

Switch to pytorch mixed precision training

Checklist

• I have updated the documentation accordingly.
• I have added tests to cover my changes.
• All new and existing tests passed.
• I have updated the CHANGELOG.rst accordingly.

[codecov]

Codecov Report

Merging #88 (4deefba) into master (1a35ab3) will increase coverage by 0.24%.
The diff coverage is 66.66%.

@@            Coverage Diff             @@
##           master      #88      +/-   ##
==========================================
+ Coverage   87.73%   87.97%   +0.24%     
==========================================
  Files          38       38              
  Lines        1745     1747       +2     
  Branches      375      375              
==========================================
+ Hits         1531     1537       +6     
+ Misses        114      110       -4     
  Partials      100      100              

Flag	Coverage Δ
unittests	87.97% <66.66%> (+0.24%)	⬆️

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

Impacted Files	Coverage Δ
setup.py	0.00% <ø> (ø)
src/emmental/utils/parse_args.py	99.25% <ø> (-0.01%)	⬇️
src/emmental/learner.py	78.16% <66.66%> (+1.80%)	⬆️