refactor: evaluate scripts and OCRDataset

[charlesmindee]

This PR changes the evaluation script so it can be run with rotated boxes annotations, and the OCRDataset object has been changed to work with all test sets.

DocDataset has been removed, and all private test sets have been cleaned up and refactored to work with and OCRDataset object and the evaluation script.

Any feedback is welcome!

[codecov]

Codecov Report

Merging #474 (4b67c70) into main (342239a) will increase coverage by 0.07%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main     #474      +/-   ##
==========================================
+ Coverage   95.36%   95.43%   +0.07%     
==========================================
  Files          96       97       +1     
  Lines        4053     4035      -18     
==========================================
- Hits         3865     3851      -14     
+ Misses        188      184       -4     

Flag	Coverage Δ
unittests	95.43% <100.00%> (+0.07%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
doctr/datasets/ocr.py	88.00% <100.00%> (-1.10%)	⬇️
doctr/utils/visualization.py	93.57% <0.00%> (-0.06%)	⬇️
doctr/transforms/modules/base.py	91.22% <0.00%> (ø)
doctr/utils/fonts.py	92.85% <0.00%> (ø)
doctr/io/elements.py	90.44% <0.00%> (+0.28%)	⬆️
...dels/detection/differentiable_binarization/base.py	91.97% <0.00%> (+0.61%)	⬆️
doctr/datasets/classification/base.py	93.93% <0.00%> (+1.83%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 3fec8be...4b67c70. Read the comment docs.

[fg-mindee]

Thanks!