[BUGFIX] [DOC] Update nlp.model.get_model documentation and get_model API

[leezu]

http://gluon-nlp.mxnet.io/api/modules/model.html currently does not
differentiate pre-defined models exposed by the get_model function and other
components. This PR is a quick attempt to improve the distinction.

Further, get_model used a hardcoded dataset=wikitext-2 argument. This is
likely by accident as the get_model function was initially mainly used for
language models, but now also provides access to other models such as Bert, for
which hardcoded dataset argument does not make sense.

For example, loading a Bert model with a prespecified Vocab currently requires doing

nlp.model.get_model('bert_12_768_12', dataset_name=None, vocab=vocabulary, ...)

instead of

nlp.model.get_model('bert_12_768_12', vocab=vocabulary, ...)

The latter version currently fails, as it will pass the "wikitext-2" as
dataset_name to the Bert model constructor.

Checklist

Essentials

• PR's title starts with a category (e.g. [BUGFIX], [MODEL], [TUTORIAL], [FEATURE], [DOC], etc)
• Changes are complete (i.e. I finished coding on this PR)
• All changes have test coverage
• Code is well-documented

Changes

• Change nlp.model.get_model dataset argument default to None
• Improve nlp.model docs

[codecov]

Codecov Report

❗ No coverage uploaded for pull request head (fix_get_model@1c7d809). Click here to learn what that means.
The diff coverage is n/a.

[codecov]

Codecov Report

Merging #734 into master will decrease coverage by 0.02%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master     #734      +/-   ##
==========================================
- Coverage   90.56%   90.53%   -0.03%     
==========================================
  Files          65       65              
  Lines        6071     6075       +4     
==========================================
+ Hits         5498     5500       +2     
- Misses        573      575       +2

Impacted Files	Coverage Δ
src/gluonnlp/model/__init__.py	96% <100%> (-0.16%)	⬇️
src/gluonnlp/data/corpora/google_billion_word.py	66.66% <0%> (-8.34%)	⬇️
src/gluonnlp/vocab/vocab.py	97.94% <0%> (ø)	⬆️
src/gluonnlp/model/utils.py	76.72% <0%> (ø)	⬆️
src/gluonnlp/data/utils.py	78.41% <0%> (+0.05%)	⬆️
...p/data/corpora/large_text_compression_benchmark.py	89.28% <0%> (+8.92%)	⬆️

[mli]

Job PR-734/2 is complete.
Docs are uploaded to http://gluon-nlp-staging.s3-accelerate.dualstack.amazonaws.com/PR-734/2/index.html

[eric-haibin-lin]

The Segmentation fault: 11 in CI seems unrelated, but I didn't notice similar failures recently ..

[mli]

Job PR-734/3 is complete.
Docs are uploaded to http://gluon-nlp-staging.s3-accelerate.dualstack.amazonaws.com/PR-734/3/index.html

[leezu]

I'm restarting the failed pipeline http://ci.mxnet.io/blue/organizations/jenkins/GluonNLP-py2-gpu-unittest/detail/PR-734/3/pipeline as http://ci.mxnet.io/blue/organizations/jenkins/GluonNLP-py2-gpu-unittest/detail/PR-734/4/pipeline