Notebook examples
Try one of our Notebook examples to get started.
The following examples are
available.
| Type of training | Notebook example | Description |
|---|---|---|
| PyTorch Elastic Distributed Training | sample_pytorch_edt.ipynb | GPU acceleration using PyTorch elastic distributed training. In this example, the kernel starts with no GPU resources requested. GPU is explicitly requested using the elastic distributed training API so that the maximum number of GPU is set to 1 allowing GPUs to be dynamically allocated during training execution. Kernel: Python3 with PyTorch Elastic GPU |
| PyTorch Single GPU | sample_single_gpu.ipynb | Using a single GPU to run workload in a Notebook. Kernel: Python3 with PyTorch Elastic GPU |
| TensorFlow Elastic Distributed Training | sample_tf_edt.ipynb | Using TensorFlow to run elastic distributed training in a Notebook. Kernel: Python3 with PyTorch Elastic GPU |