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 Ru Conformer-Transducer Large
Model
NVIDIA
NVIDIA
STT Ru Conformer-Transducer Large

Conformer-Transducer-Large model for Russian Automatic Speech Recognition, trained on Mozilla Common Voice 10.0 (Russian), Golos (Russian), Russian LibriSpeech (RuLS) and SOVA (RuAudiobooksDevices, RuDevices) datasets.

1 Version
1.13.0Selected
10/18/2022 9:44 PM UTC460.18 MB
ACCURACY
KeyValue
MOZILLA COMMON VOICE 10.0 RU (DEV)3.5%
GOLOS CROWD (TEST)2.7%
GOLOS FARFIELD (TEST)7.6%
RUSSIAN LIBRISPEECH (TEST)12.0%
MOZILLA COMMON VOICE 10.0 RU (TEST)4.0%
MODEL
KeyValue
ARCHITECTURECONFORMER-RNNT
INPUTS16000 HZ MONO-CHANNEL AUDIO (WAV FILES)
OUTPUTSTRANSCRIBED SPEECH