Pre-Editorial Normalization on Schonhardt data 2025 (first 2,000 entries)
0.111CERByT5-based normalization model
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ByT5-based normalization modeltraining_data=Bilingual (Latin and Old French)2026.02 | 0.111 | 0.385 | |
| Llama 3 8B Instructmode=zero-shot2026.02 | 0.264 | 0.545 | |
| ByT5 (Schonhardt)training_data=Schonhardt (2025)2026.02 | 1.3 | 1.094 |