Commit
•
c633b18
1
Parent(s):
b8f74c7
Update README.md
Browse files
README.md
CHANGED
|
@@ -73,9 +73,9 @@ The model was finetuned to Kazakh speech based on the pre-trained English Model
|
|
| 73 |
In total, KSC2 contains around 1.2k hours of high-quality transcribed data comprising over 600k utterances.
|
| 74 |
|
| 75 |
## Performance
|
| 76 |
-
|
| 77 |
-
Average WER: 15.53
|
| 78 |
-
|
| 79 |
## Limitations
|
| 80 |
|
| 81 |
Because the GPU has limited power, we used a lightweight model architecture for fine-tuning.\
|
|
|
|
| 73 |
In total, KSC2 contains around 1.2k hours of high-quality transcribed data comprising over 600k utterances.
|
| 74 |
|
| 75 |
## Performance
|
| 76 |
+
The model achieved:\
|
| 77 |
+
Average WER: 15.53%\
|
| 78 |
+
through the applying of **Greedy Decoding**.
|
| 79 |
## Limitations
|
| 80 |
|
| 81 |
Because the GPU has limited power, we used a lightweight model architecture for fine-tuning.\
|