Link public Ekko STT source repository
Browse files
README.md
CHANGED
|
@@ -149,6 +149,9 @@ every input word in order. Pass continuous text when possible so the model can
|
|
| 149 |
use context across ASR chunks. These settings affect inference, not the v2
|
| 150 |
model weights.
|
| 151 |
|
|
|
|
|
|
|
|
|
|
| 152 |
Version 2 is tuned for **continuous speech** such as dictation and meetings.
|
| 153 |
The earlier version often missed sentence endings when given a long transcript;
|
| 154 |
it had mainly learned from short, isolated utterances. Version 2 was trained
|
|
|
|
| 149 |
use context across ASR chunks. These settings affect inference, not the v2
|
| 150 |
model weights.
|
| 151 |
|
| 152 |
+
The reusable inference runtime and browser demo are in the
|
| 153 |
+
[Ekko STT repository](https://github.com/Rye-A1/ekko-danish-stt).
|
| 154 |
+
|
| 155 |
Version 2 is tuned for **continuous speech** such as dictation and meetings.
|
| 156 |
The earlier version often missed sentence endings when given a long transcript;
|
| 157 |
it had mainly learned from short, isolated utterances. Version 2 was trained
|