Released Neural Network Libraries v1.14.0!

Monday, December 14, 2020


Posted by shin

We have Released Neural Network Libraries v1.14.0! We now support CUDA 11 / cuDNN 8 and we have also added new models for speech synthesis, as well as new solver!


Support CUDA 11 and cuDNN 8

We now support CUDA 11 and cuDNN 8!
CUDA 11 supports new generation of NVIDIA Ampere GPU architecture, and introduces new tensor cores to enhance mixed precision across different data types including TF32.
cuDNN 8 also introduces new and enhanced features, such as optimized grouped convolutions and 3D convolutions, operation descriptors, etc.

Speech Synthesis models: Tacotron1 & 2 / WaveGlow

We have added new speech synthesis models! Tacotron 1 and Tacotron 2 are text-to-speech synthesis system that synthesizes speech directly from a sequence of characters.

WaveGlow is a flow-based network capable of generating high quality speech from mel spectrograms, trained by maximizing the likelihood of the training data.

Along with vision and language, speech has also become an important field for which deep learning can be applied, and we will continue to add new models for speech recognition!

AdaBelief solver

AdaBelief is an optimizer proposed in this paper, aiming to achieve fast convergence of adaptive models, while retaining good generalization of stochastic gradient descent (SGD), as well as training stability, by adapting the stepsize according to the “belief” in current gradient direction.



Format Conversion