customization & flexibility

This release helps ensure confectionary is flexible and portable:

• add the ability to load custom word2vec model so not stuck with 3.7G one
• add the --api-info switch and print_api_info() function to display what models are available
• update default model to glove-wiki-gigaword-100 - 200 MB only and still solid performance from testing
• misc improvements to README and documentation

What's Changed

• Accessibility & visibility by @pszemraj in #28

Full Changelog: v0.2.8...v0.3