[FEATURE] Add Elmo

[pingzhili]

Thanks for sending a pull request! Please make sure you click the link above to view the contribution guidelines, then fill out the blanks below.

Description

(Brief description on what this PR is about)

What does this implement/fix? Explain your changes.

1. Add bi-LM for Elmo model
2. Add training and fine-tuning for Elmo model
3. Add Elmo to infer vectors
4. Add Elmo for i2v and t2v

Pull request type

• [ ] [DATASET] Add a new dataset
• [ ] [BUGFIX] Bugfix
• [x] [FEATURE] New feature (non-breaking change which adds functionality)
• [ ] [BREAKING] Breaking change (fix or feature that would cause existing functionality to not work as expected)
• [ ] [STYLE] Code style update (formatting, renaming)
• [ ] [REFACTOR] Refactoring (no functional changes, no api changes)
• [ ] [BUILD] Build related changes
• [ ] [DOC] Documentation content changes
• [ ] [OTHER] Other (please describe):

Changes

• Add Bi-LanguageModel for Elmo
• Add Elmo to infer vectors
• Add Elmo for i2v and t2v
• Add test for above changes

Does this close any currently open issues?

issue #65

Any relevant logs, error output, etc?

N/A

Checklist

Before you submit a pull request, please make sure you have to following:

Essentials

• [x] PR's title starts with a category (e.g. [BUGFIX], [FEATURE], [BREAKING], [DOC], etc)
• [x] Changes are complete (i.e. I finished coding on this PR)
• [x] All changes have test coverage and al tests passing
• [x] Code is well-documented (extended the README / documentation, if necessary)
• [x] If this PR is your first one, add your name and github account to AUTHORS.md

Comments

• If this change is a backward incompatible change, why must this change be made.
• Interesting edge cases to note here

[codecov-commenter]

Codecov Report

Merging #120 (6f238d9) into dev (bc4b2f9) will decrease coverage by 0.80%. The diff coverage is 94.40%.

@@             Coverage Diff             @@
##               dev     #120      +/-   ##
===========================================
- Coverage   100.00%   99.19%   -0.81%     
===========================================
  Files           48       50       +2     
  Lines         1506     1747     +241     
===========================================
+ Hits          1506     1733     +227     
- Misses           0       14      +14     

Impacted Files	Coverage Δ
EduNLP/Tokenizer/tokenizer.py	100.00% <ø> (ø)
EduNLP/Vector/rnn/rnn.py	100.00% <ø> (ø)
EduNLP/I2V/i2v.py	98.07% <92.30%> (-1.93%)	:arrow_down:
EduNLP/Pretrain/elmo_vec.py	92.61% <92.61%> (ø)
EduNLP/Vector/elmo_vec.py	97.77% <97.77%> (ø)
EduNLP/I2V/__init__.py	100.00% <100.00%> (ø)
EduNLP/ModelZoo/rnn/__init__.py	100.00% <100.00%> (ø)
EduNLP/ModelZoo/rnn/rnn.py	100.00% <100.00%> (ø)
EduNLP/Pretrain/__init__.py	100.00% <100.00%> (ø)
EduNLP/Vector/__init__.py	100.00% <100.00%> (ø)
... and 2 more

---

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 bc4b2f9...6f238d9. Read the comment docs.