updates model
Browse files
README.md
CHANGED
|
@@ -27,8 +27,7 @@ model-index:
|
|
| 27 |
type: wer
|
| 28 |
value: 15.345730
|
| 29 |
- name: Test CER
|
| 30 |
-
type:
|
| 31 |
-
value: 5.573185
|
| 32 |
---
|
| 33 |
|
| 34 |
# Wav2Vec2-Large-XLSR-53-Swedish
|
|
@@ -129,9 +128,6 @@ print("CER: {:2f}".format(100 * wer.compute(predictions=[" ".join(list(entry)) f
|
|
| 129 |
|
| 130 |
**Test Result**: 15.345730 %
|
| 131 |
|
| 132 |
-
|
| 133 |
-
5.573185
|
| 134 |
-
|
| 135 |
As first step used Common Voice train dataset and parts from NST
|
| 136 |
as can be found [here](https://github.com/se-asr/nst/tree/master).
|
| 137 |
Part of NST where removed using this mask
|
|
|
|
| 27 |
type: wer
|
| 28 |
value: 15.345730
|
| 29 |
- name: Test CER
|
| 30 |
+
type: c
|
|
|
|
| 31 |
---
|
| 32 |
|
| 33 |
# Wav2Vec2-Large-XLSR-53-Swedish
|
|
|
|
| 128 |
|
| 129 |
**Test Result**: 15.345730 %
|
| 130 |
|
|
|
|
|
|
|
|
|
|
| 131 |
As first step used Common Voice train dataset and parts from NST
|
| 132 |
as can be found [here](https://github.com/se-asr/nst/tree/master).
|
| 133 |
Part of NST where removed using this mask
|