Use the NGC CLI to download:
Copied!
Model
ContextNet-1024 model for English Automatic Speech Recognition, trained on NeMo ASRSET
Use the NGC CLI to download:
Copied!
3 Versions
1.9.0Selected05/17/2022 10:26 PM UTC554.65 MBCopied!
1.9.0Selected
05/17/2022 10:26 PM UTC554.65 MB
Copied!
Accuracy
| Key | Value |
|---|---|
| LibrispeechDevClean | 1.7 % |
| LibriSpeechTestOther | 4.0 % |
| NSCPart_6 | 6.5 % |
| LibrispeechDevOther | 3.8 % |
| WSJEval92 | 1.3 % |
| SLR83Test | 4.7 % |
| LibrispeechTestClean | 1.9 % |
| MLSTest_En | 5.9 % |
| WSJDev93 | 2.3 % |
| NSCPart_1 | 5.2 % |
| PeopleSpeechTest | 21.7 % |
| MCVTest_8_0 | 7.9 % |
Model
| Key | Value |
|---|---|
| ARCHITECTURE | ContextNet |
| INPUTS | 16000 KHZ MONO-CHANNEL AUDIO (WAV FILES) |
| OUTPUTS | TRANSCRIBED SPEECH |
01/13/2022 7:59 PM UTC447.12 MBCopied!
01/13/2022 7:59 PM UTC447.12 MB
Copied!
Accuracy
| Key | Value |
|---|---|
| LibrispeechDevClean | 1.6 % |
| LibrispeechDevOther | 3.8 % |
| LibrispeechTestClean | 1.8 % |
| LibrispeechTestOther | 3.8 % |
| MLSDev | 5.5 % |
| MLSTest | 6.0 % |
| NSCPartOne | 6.2 % |
| WSJDev93 | 2.3 % |
| WSJEval92 | 1.4 % |
Model
| Key | Value |
|---|---|
| ARCHITECTURE | ContextNet |
| INPUTS | 16000 KHZ MONO-CHANNEL AUDIO (WAV FILES) |
| OUTPUTS | TRANSCRIBED SPEECH |
09/16/2021 12:18 AM UTC446.82 MBCopied!
09/16/2021 12:18 AM UTC446.82 MB
Copied!
Accuracy
| Key | Value |
|---|---|
| LibrispeechDevClean | 2.2 % |
| LibrispeechDevOther | 5.0 % |
| LibrispeechTestClean | 2.4 % |
| LibrispeechTestOther | 5.0 % |
| NSCPartOne | 6.8 % |
| WSJDev93 | 2.6 % |
| WSJEval92 | 1.8 % |
Model
| Key | Value |
|---|---|
| ARCHITECTURE | ContextNet |
| INPUTS | 16000 KHZ MONO-CHANNEL AUDIO (WAV FILES) |
| OUTPUTS | TRANSCRIBED SPEECH |