Tags

exclude *.tar.gz and *.bin files in MANIFEST.in so alBERT.bin and wikipedia-titles arent in package. also downgrade transformers to deal with modulenotfound error for transformers_tokenizer

delete .gz and .bin files from MANIFEST so alBERT.bin and wikipedia-titles arent in package

trying to distribute ALL data files with package

fix pip install bugs for John May with more data/* and new MANIFEST.in

fix pip install bugs for John May with more data/*

fix pip install bugs for John May

spell checker incorporated into skills/corrected_qa and skills/spelling

spell checker incorporated into skills/corrected_qa and skills/spelling

major release bump for elastic search api that only requires env variable and can fall back to independent operation for QA

spell checker incorporated into skills/corrected_qa and skills/spelling

spell checker incorporated into skills/corrected_qa and skills/spelling

faster, more accurate question-answering, by checking wiki titles before pulling articles from wikipedia

semantics->semantic (glossary bot wont load without it). olesya article. some of raj rouge and bleu scores

more test coverage and some BLEU and other text similarity metrics

consolidate requirements. bug fixes. code coverage

bug fixes and code coverage

clean up cruft upgrade upper limit for python (<3.8) in condaenv in readme

remove large files from git history. clean up cruft. reorg data. more LARGE_FILES download links. merge and delete feature branches. upgrade torch and add torchvision

Duncan's reduced verbosity. Olesya's notebooks for ElasticSearch Wikipedia page preprocessing

fix dotenv dependency