Select wilds datasets for NLP tools

[Anmol-Srivastava]

Closes #151

Added

• wilds install option
• intermediate example notebook/converted script that downloads, demonstrates 2 wilds datasets

Changed

• tweaked/removed outdated data generation scripts
• tweaked .gitignore

[Anmol-Srivastava]

@tms-bananaquit all ready here, just a precursor to developing NLP detectors

[tms-bananaquit]

Noted these items in #154, for when the example is fleshed out:

• Update README to note the wilds setup option
• Update RTD/TOC to include example?

[tms-bananaquit]

The notebook currently downloads to data\amazon_v2.1\... if I run it from my default directory; the .gitignore targets /docs/source/examples/**/data. Maybe the notebook should be updated to point to that directory?

[Anmol-Srivastava]

Updated README, the RTD perhaps we save for the example issue.

Clean fix to the download location seemed tricky, so I had the code download to "./wilds_datasets" and have the .gitignore ignore "wilds_datasets/" at any level