Model
ContextNet-1024 model for German Automatic Speech Recognition, trained on NeMo German ASRSET
Use the NGC CLI to download:
Copied!
1 Version
1.4.0Selected10/06/2021 12:04 PM UTC453.54 MB400 Epochs Copied!
1.4.0Selected
10/06/2021 12:04 PM UTC453.54 MB400 Epochs
Copied!
ACCURACY
| Key | Value |
|---|---|
| MCV7.0 DEV | 4.76 |
| MCV7.0 TEST | 5.5 |
| MLS DEV | 3.53 |
| MLS TEST | 4.2 |
| VOX POPULI DEV | 11.32 |
| VOX POPULI TEST | 9.4 |
MODEL
| Key | Value |
|---|---|
| ARCHITECTURE | CONTEXTNET |
| INPUTS | 16000 KHZ MONO-CHANNEL AUDIO (WAV FILES) |
| OUTPUTS | TRANSCRIBED SPEECH |