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 Deep Learning Examples
NVIDIA Deep Learning Examples
HiFi-GAN for PyTorch
Resource
NVIDIA Deep Learning Examples
NVIDIA Deep Learning Examples
HiFi-GAN for PyTorch

HiFi-GAN model implements a spectrogram inversion model that allows to synthesize speech waveforms from mel-spectrograms.

The performance measurements in this document were conducted at the time of publication and may not reflect the performance achieved from NVIDIA's latest software release. For the most up-to-date performance measurements, go to https://developer.nvidia.com/deep-learning-performance-training-inference.

Changelog

February 2022

  • Initial release

Known issues

  • With mixed-precision training on Ampere GPUs, the model might suffer from slower training. Be sure to use the scripts provided in the platform/ directory, and a PyTorch NGC container not older than 21.12-py3.
  • For some mel-spectrogram generator models, the best results require fine-tuning of HiFi-GAN on outputs from those models. For more details, refer to the fine-tuning step of the Quick Start Guide section.