ContainersNVIDIADeepSeek-R1NVIDIALlama-3.1-Nemotron-70B-InstructNVIDIAPyTorch
CollectionsNVIDIAOmniverse Kit (FB)NVIDIADeepStream SDKNVIDIAOmniverse Kit App Streaming
ModelsNVIDIAStyleGAN3 pretrained modelsNVIDIAPeopleNetNVIDIATrafficCamNet
ResourcesNVIDIARiva Skills Quick Start
Helm ChartsNVIDIAGPU OperatorNVIDIANVIDIA NIM Operator
NVIDIA
NVIDIA
STT Zh Conformer-Transducer Large
Model
NVIDIA
NVIDIA
STT Zh Conformer-Transducer Large

Conformer-Transducer-Large model for Mandarin Automatic Speech Recognition, Trained on Aishell-2 Mandarin Chinese corpus.

1 Version
1.8.0Selected
04/26/2022 9:57 PM UTC492.93 MB
Accuracy
KeyValue
AISHELL-2 dev_ios3.2
AISHELL-2 test_mic3.4
AISHELL-2 dev_android3.4
AISHELL-2 test_ios3.2
AISHELL-2 dev_mic3.4
AISHELL-2 test_android3.4
Model
KeyValue
Number of Decoder Layers2
Encoder Dimension512
Number of Encoder Layers17
ArchitectureConformer-Transducer-Large
DatasetAISHELL-2
OutputsText in Mandarin
InputsSpeech in Mandarin
Number of Weights129M