Update README.md
Browse files
README.md
CHANGED
|
@@ -21,7 +21,7 @@ widget:
|
|
| 21 |
src: https://cdn-media.huggingface.co/speech_samples/LibriSpeech_61-70968-0000.flac
|
| 22 |
---
|
| 23 |
|
| 24 |
-
# VoxLingua107 ECAPA-TDNN Spoken Language Identification Model
|
| 25 |
|
| 26 |
## Model description
|
| 27 |
|
|
@@ -225,7 +225,7 @@ Training recipe will be published soon.
|
|
| 225 |
|
| 226 |
## Evaluation results
|
| 227 |
|
| 228 |
-
Error rate: 7% on the development dataset
|
| 229 |
|
| 230 |
|
| 231 |
### BibTeX entry and citation info
|
|
|
|
| 21 |
src: https://cdn-media.huggingface.co/speech_samples/LibriSpeech_61-70968-0000.flac
|
| 22 |
---
|
| 23 |
|
| 24 |
+
# VoxLingua107 ECAPA-TDNN Spoken Language Identification Model (trained with cross-entropy)
|
| 25 |
|
| 26 |
## Model description
|
| 27 |
|
|
|
|
| 225 |
|
| 226 |
## Evaluation results
|
| 227 |
|
| 228 |
+
Error rate: 6.7% on the VoxLingua107 development dataset
|
| 229 |
|
| 230 |
|
| 231 |
### BibTeX entry and citation info
|