Update README.md
Browse files
README.md
CHANGED
|
@@ -95,6 +95,14 @@ The following hyperparameters were used during training:
|
|
| 95 |
- Datasets 1.18.1.dev0
|
| 96 |
- Tokenizers 0.11.0
|
| 97 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 98 |
### Eval results on Common Voice 7 "test" (WER):
|
| 99 |
|
| 100 |
| Without LM | With LM (run `./eval.py`) |
|
|
|
|
| 95 |
- Datasets 1.18.1.dev0
|
| 96 |
- Tokenizers 0.11.0
|
| 97 |
|
| 98 |
+
#### Evaluation Commands
|
| 99 |
+
|
| 100 |
+
1. To evaluate on `mozilla-foundation/common_voice_7_0` with split `test`
|
| 101 |
+
|
| 102 |
+
```bash
|
| 103 |
+
python eval.py --model_id Yehor/wav2vec2-xls-r-1b-uk-with-lm --dataset mozilla-foundation/common_voice_7_0 --config uk --split test
|
| 104 |
+
```
|
| 105 |
+
|
| 106 |
### Eval results on Common Voice 7 "test" (WER):
|
| 107 |
|
| 108 |
| Without LM | With LM (run `./eval.py`) |
|