Distributed training single node

The following scripts provide an example of distributed training of a convolutional neural network (CNN) on the MNIST dataset to classify handwritten digits into 10 classes. In this case, the distributed training is restricted to the use of GPUs in a single node.

Distributed training multi node

The following scripts provide an example of distributed training of a convolutional neural network (CNN) on the MNIST dataset to classify handwritten digits into 10 classes. In this case, the distributed training is performed in several GPUs accross multiple nodes

Related pages

• Running TensorFlow on Setonix

• Distributed training using HorovodPreview

• Example Slurm Batch Scripts for Setonix on GPU Compute Nodes

• PyTorch