Add torchtext==0.9.1 to support Kaggle environments. (#165)

* renamed download-progress-bar as download

* added new download functions to init

* Added Btech data module

* Added btech tests

* Move split functions into a util module

* Modified mvtec

* added btech to get-datamodule

* fix typo in btech docstring

* update docstring

* cleanedup dataset download utils

* Address mypy

* modify config files and update readme.md

* Fix dataset path

* WiP: Created make_dataset function

* Renamed folder dataset into custom

* Added custom dataset tests

* updated config.yaml file to show custom dataset is available

* Added custom dataset to get_datamodule

* Address PR comments

* fix dataset path

* Debugging the ci

* Fixed folder dataset tests

* Added code quality checks back to the ci

* Added code coverage back to pre-merge tests

* Do not import openvino at the top level. Cause some import issues when openvino is not install

* Move nncf to openvino-requirements.txt

* Move opencv to base requirements.txt/

* Added extension as an option when saving the result images.

* Added an extension check

* python verion 3.7

* Added opencv contrib python

* reverted visualizer callback

* Convert bmp images to png after downloading btech dataset.

* pylint

* ignore arguments differ

* Version to 0.2.5

* Ignore arguments-differ

* loosen pandas version

* import openvino with absolute import

* Added torchtext to requirements for kaggle support

requirements/base.txt

@@ -16,6 +16,7 @@ pillow==9.0.0
 pytorch-lightning==1.5.9
 torch==1.8.1
 torchvision==0.9.1
+torchtext==0.9.1
 scikit-image>=0.17.2
 scikit-learn==0.24.2
 wandb==0.12.9