Fixed a bug with loading albert models in API image. Minor refactors and cleanup.

Renaming and cache removal in dockerfiles.

Added support for using ALBERT models. Improved and overhauled data loading, tokenization and pretrained model retrieval.

Updated README.md

Merge pull request #4 from sarnikowski/overhaul/tf2

Added LICENSE

Minor fixes and rephrasings to README

Merge pull request #2 from sarnikowski/feature/makefile

Updated Dockerfile to use GPU. Makefile base model target has been parametrized. Training and prediction scripts have been upgraded to TF 2.0. Updated methods in utils and trimmed some fat. Updated README

Work in progress on training script. Implemented main blocks, however is untested. Logging needs refactoring. Removed some unecessary utils and requirements.

Initial README draft for TF2 implementation. Still work in progress

Code is now runnable using the make targets. Added stack overflow data

Merge pull request #1 from Ayush-iitkgp/Add-seaborn-in-requirements.txt

Updated requirements.txt

:construction: Initial code dump