Update README.md
Browse files
README.md
CHANGED
|
@@ -41,8 +41,11 @@ SenseVoice-Small is an encoder-only speech foundation model designed for rapid v
|
|
| 41 |
The SenseVoice-Small model is based on a non-autoregressive end-to-end framework. For a specified task, we prepend four embeddings as input to the encoder:
|
| 42 |
|
| 43 |
LID: For predicting the language id of the audio.
|
|
|
|
| 44 |
SER: For predicting the emotion label of the audio.
|
|
|
|
| 45 |
AED: For predicting the event label of the audio.
|
|
|
|
| 46 |
ITN: Used to specify whether the recognition output text is subjected to inverse text normalization.
|
| 47 |
|
| 48 |
# Usage
|
|
|
|
| 41 |
The SenseVoice-Small model is based on a non-autoregressive end-to-end framework. For a specified task, we prepend four embeddings as input to the encoder:
|
| 42 |
|
| 43 |
LID: For predicting the language id of the audio.
|
| 44 |
+
|
| 45 |
SER: For predicting the emotion label of the audio.
|
| 46 |
+
|
| 47 |
AED: For predicting the event label of the audio.
|
| 48 |
+
|
| 49 |
ITN: Used to specify whether the recognition output text is subjected to inverse text normalization.
|
| 50 |
|
| 51 |
# Usage
|