Russian and Ukrainian/Work plan
Jump to navigation
Jump to search
Corpora
- Радіо Свобода
- Украинская Правда (bilingual?)
Week plan
| Week | Dates | Trimmed coverage | Achieved | Testvoc | Evaluation | WER | Notes | |
|---|---|---|---|---|---|---|---|---|
| Svoboda | Wikipedia | |||||||
| 0 | 21/05—16/06 | 50% | 53.8% | - | 500 words | 8,31 % | Preliminary evaluation. Translate the story total coverage and without diagnostics. Get a baseline WER. | |
| 1 | 17/06—23/06 | 55% | ? | - | <pr> |
- | ||
| 2 | 24/06—30/06 | 59% | 59.5% | - | <cnjcoo> <cnjadv> <cnjsub> |
- | ||
| 3 | 01/07—07/07 | 64% | 59.9% | - | <num> |
200 words | Numerals should be added and testvoc clean. | |
| 4 | 08/07—14/07 | 69% | 64.4% | 55.0% | <prn> <det> |
- | ||
| 5 | 15/07—21/07 | 73% | 72.6% | 61.0% | <adj> <adv> |
- | ||
| 6 | 22/07—28/07 | 75% | 75.0% | 66.3% | <n> |
500 words | 15.22% | Midterm evaluation. |
| 7 | 29/07—04/08 | 77% | 77.0% | - | - | - | ||
| 8 | 05/08—11/08 | 79% | 79.0% | 70.9% | <np> |
- | ||
| 9 | 12/08—18/08 | 81% | 81.3% | - | 200 words | |||
| 10 | 19/08—25/08 | 83% | 83.0% | 74.7% | <vblex> |
- | ||
| 11 | 26/08—01/09 | 85% | 90.0% | 80.9% | - | - | ||
| 12 | 02/09—08/09 | 88% | - | - | ||||
| 13 | 09/09—15/09 | 90% | all categories clean | 500 words | 14.74% | Final evaluation. Tidying up, releasing | ||
