Use the NGC CLI to download:
Copied!
Model
ContextNet-256 model for English Automatic Speech Recognition, trained on NeMo ASRSET.
Use the NGC CLI to download:
Copied!
2 Versions
01/13/2022 8:01 PM UTC50.46 MBCopied!
01/13/2022 8:01 PM UTC50.46 MB
Copied!
Accuracy
| Key | Value |
|---|---|
| LibrispeechDevClean | 3.3 % |
| LibrispeechDevOther | 7.9 % |
| LibrispeechTestClean | 3.3 % |
| LibrispeechTestOther | 8.0 % |
| MLSDev | 9.7 % |
| MLSTest | 11.0 % |
| NSCPartOne | 7.1 % |
| WSJDev93 | 4.6 % |
| WSJEval92 | 3.2 % |
Model
| Key | Value |
|---|---|
| ARCHITECTURE | ContextNet |
| INPUTS | 16000 KHZ MONO-CHANNEL AUDIO (WAV FILES) |
| OUTPUTS | TRANSCRIBED SPEECH |
1.4.0Selected09/21/2021 10:19 PM UTC50.45 MBCopied!
1.4.0Selected
09/21/2021 10:19 PM UTC50.45 MB
Copied!
Accuracy
| Key | Value |
|---|---|
| LibrispeechDevClean | 3.6 % |
| LibrispeechDevOther | 8.6 % |
| LibrispeechTestClean | 3.6 % |
| LibrispeechTestOther | 8.4 % |
| NSCPartOne | 6.8 % |
| WSJDev93 | 4.4 % |
| WSJEval92 | 3.2 % |
Model
| Key | Value |
|---|---|
| ARCHITECTURE | ContextNet |
| INPUTS | 16000 KHZ MONO-CHANNEL AUDIO (WAV FILES) |
| OUTPUTS | TRANSCRIBED SPEECH |