Originally posted by: KelvinLui
To get started with IBM Spectrum Conductor Deep Learning Impact, try out one of our many state-of-the-art neural network model samples. We provide a collection of samples that that are optimized for IBM Spectrum Conductor Deep Learning Impact, meaning, they come with key deep learning capabilities already enabled!
Not only do our samples come with key deep learning capabilities enabled, which include: distributed training with Fabric, deep learning insights, and hyperparameter tuning, but they also take advantage of the various training engines available to IBM Spectrum Conductor Deep Learning Impact. Available training engine options include: single node training, native distributed training, and multi-node training utilizing Fabric technology.
Our sample collection addresses various challenges across many industries including computing vision, natural language processing, time series prediction, and many others. We will continue to collect models that are optimized for IBM Spectrum Conductor Deep Learning Impact and post them to our deep learning samples collection page.
To date, we include a collection of Caffe samples that utilizes familiar models like cifar10, as well as lesser known models like resNet101 and vgg19. The Caffe sample collection also includes an import script that helps you automatically get all the samples into IBM Spectrum Conductor Deep Learning Impact as quickly as possible. The imported samples have the correct configurations and parameters already set so that you can start using them right away.
Similarly, our collection of TensorFlow samples also includes common models like cifar10 and others. Stay tuned and check back often as we grow our collection!
To take advantage of these sample collections today, download our free trial version of IBM Spectrum Conductor Deep Learning Impact now!
For more information about IBM Spectrum Conductor Deep Learning Impact and these deep learning capabilities, go to:
www.ibm.biz/DeepLearningImpact
#SpectrumComputingGroup