Final pass of codegen for v1.0.0 SDK release

[mkistler]

I did more more pass of the SDK generator to pick up some very recent changes to the generator and to the swagger docs. Notably:

• Many more models are now pruned due to generator changes. These are primarily models for body parameters that were exploded and thus not needed. This should improve coverage stats significantly.
• Added from __future__ import absolute_imports to improve Python 2/3 compatibility
• The translation property of Translation in language_translator was renamed to translation_output (swagger change)
• Eliminated the training_metadata_content_type and training_data_content_type parameters from createClassifier in NLC (swagger change)
• Swagger updates to PIV3 and ToneAnalyzer for new versions.