Default continuing_subword_prefix when splitting token sequences

[Natooz]

• fixing end_suffix arg for WordPiece (as training is identical to BPE)

---

📚 Documentation preview 📚: https://miditok--168.org.readthedocs.build/en/168/

[codecov]

Codecov Report

Attention: Patch coverage is 93.10345% with 4 lines in your changes are missing coverage. Please review.

Project coverage is 90.45%. Comparing base (c2c8cb3) to head (323b82c).
Report is 10 commits behind head on main.

❗ Current head 323b82c differs from pull request most recent head f31d12e. Consider uploading reports for the commit f31d12e to get more accurate results

Files	Patch %	Lines
miditok/midi_tokenizer.py	95.23%	2 Missing ⚠️
tests/test_train.py	87.50%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #168      +/-   ##
==========================================
- Coverage   91.06%   90.45%   -0.61%     
==========================================
  Files          35       35              
  Lines        5214     5524     +310     
==========================================
+ Hits         4748     4997     +249     
- Misses        466      527      +61     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.