Spaces:
Running
Running
Update evaluation.py
Browse files- evaluation.py +2 -2
evaluation.py
CHANGED
|
@@ -14,8 +14,8 @@ The Iqra’Eval challenge provides a shared, transparent platform to benchmark p
|
|
| 14 |
id,predicted_sequence
|
| 15 |
0000_0001,i n n a m a a y a …
|
| 16 |
0000_0002,m a a n a n s a …
|
| 17 |
-
|
| 18 |
-
|
| 19 |
**Evaluation Criteria**
|
| 20 |
– Leaderboard ranking is based on phoneme‑level **F1‑score**, computed via a two‑stage (detection + diagnostic) hierarchy:
|
| 21 |
|
|
|
|
| 14 |
id,predicted_sequence
|
| 15 |
0000_0001,i n n a m a a y a …
|
| 16 |
0000_0002,m a a n a n s a …
|
| 17 |
+
```
|
| 18 |
+
|
| 19 |
**Evaluation Criteria**
|
| 20 |
– Leaderboard ranking is based on phoneme‑level **F1‑score**, computed via a two‑stage (detection + diagnostic) hierarchy:
|
| 21 |
|