Fix ordering of args for BERT w.r.t. transformers torchscript changes

[chiragjn]

huggingface/transformers 2.x changed the order of args for some Torchscript changes

https:/huggingface/transformers/blob/8aba81a0b64bbf7a2dcd13eaceb543c5f38fd82f/transformers/modeling_bert.py#L855-L856

huggingface/transformers#1195

Given nlpaug is calling BERT forward without named arguments, it seems like the current code has segment_inputs and mask_inputs swapped. Adding names to args should fix it

[codecov-io]

Codecov Report

Merging #59 into master will decrease coverage by 0.08%.
The diff coverage is 0%.

@@            Coverage Diff             @@
##           master      #59      +/-   ##
==========================================
- Coverage   47.31%   47.23%   -0.09%     
==========================================
  Files         106      106              
  Lines        3711     3711              
==========================================
- Hits         1756     1753       -3     
- Misses       1955     1958       +3

Impacted Files	Coverage Δ
nlpaug/model/lang_models/bert.py	31.25% <0%> (ø)	⬆️
nlpaug/augmenter/word/random.py	85.45% <0%> (-1.82%)	⬇️
nlpaug/augmenter/char/random.py	89.7% <0%> (-1.48%)	⬇️

---

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 e6f14a7...50e7cf5. Read the comment docs.