feat: Feature date tokenization - #13

Merged

kavyamanohar merged 12 commits into

mainfrom

feature-date-tokenization

Feb 12, 2026

Commits on Jan 24, 2026

Refactor tokenizer to alignment visualization and error for legal entity error rates
kavyamanohar
committed
90efd23

Commits on Jan 28, 2026

Implement normalized error rates with combined denominator
kavyamanohar
committed
7a68649
Add tests
kavyamanohar
committed
6438fe2

Commits on Jan 30, 2026

Fix broken example scripts and refactor reporting logic

kavyamanohar
and
claude
committed
70b96a0
Untrack root-level test files from git
kavyamanohar
committed
c4b31bf
Add detailed per-sample error reporting to batch evaluation
kavyamanohar
committed
ca738cf

Commits on Feb 11, 2026

Update batch evaluation, reporting, and documentation

kavyamanohar
and
claude
committed
3411396
Refactor: Centralize constants and use consistent field naming

kavyamanohar
and
claude
committed
aaf3b34
Refactor to domain-aware tokenizer with flexible configuration

kavyamanohar
and
claude
committed
622d9f9
Update README: remove Testing section and sync text_alignment output
kavyamanohar
committed
fdf95e3
Update CLAUDE.md: document domain-aware tokenization system
kavyamanohar
committed
e8b10bd
Add token normalization for date and currency formats
kavyamanohar
committed
bf4313d