Commit
·
7b528f7
1
Parent(s):
7061fa6
Update README.md
Browse files
README.md
CHANGED
|
@@ -60,8 +60,8 @@ Note that testing data can not be provided publicly due to the privacy issue.
|
|
| 60 |
#### Metrics
|
| 61 |
|
| 62 |
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
| 63 |
-
Two of the most popular metrics to assess automatic speech recognition model, WER and CER, were used.
|
| 64 |
-
Additionally, DSWES was used to specifically check the transcription accuracy of softwared-related words.
|
| 65 |
|
| 66 |
For accessment, WhisperX was used as a backbone of a fine-tuned model due to its fast inference speed and its reduced size.
|
| 67 |
Since backbone of WhisperX is Whisper, I can safely assume that the performace of Whisper would very much similar to that of WhisperX.
|
|
|
|
| 60 |
#### Metrics
|
| 61 |
|
| 62 |
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
| 63 |
+
Two of the most popular metrics to assess automatic speech recognition model, WER and CER, were used. <br>
|
| 64 |
+
Additionally, DSWES was used to specifically check the transcription accuracy of softwared-related words. Note that higher the DSWES, the better.
|
| 65 |
|
| 66 |
For accessment, WhisperX was used as a backbone of a fine-tuned model due to its fast inference speed and its reduced size.
|
| 67 |
Since backbone of WhisperX is Whisper, I can safely assume that the performace of Whisper would very much similar to that of WhisperX.
|