Update README.md
Browse files
README.md
CHANGED
|
@@ -79,6 +79,7 @@ and LLM ([EuroLLM-1.7B](https://huggingface.co/utter-project/EuroLLM-1.7B)) were
|
|
| 79 |
|
| 80 |
The model was evaluated using the Word Error Rate (WER) metric from the `evaluate` library.
|
| 81 |
Prior to computing the WER, preprocessing of ground-truth and predicted transcripts was carried out using the `Whisper EnglishTextNormalizer` for English and `BasicTextNormalizer` for all other languages.
|
|
|
|
| 82 |
|
| 83 |
### Results
|
| 84 |
|
|
|
|
| 79 |
|
| 80 |
The model was evaluated using the Word Error Rate (WER) metric from the `evaluate` library.
|
| 81 |
Prior to computing the WER, preprocessing of ground-truth and predicted transcripts was carried out using the `Whisper EnglishTextNormalizer` for English and `BasicTextNormalizer` for all other languages.
|
| 82 |
+
Beam search decoding is used with `beam size = 4`.
|
| 83 |
|
| 84 |
### Results
|
| 85 |
|