HPO pytorch MNIST example #1577
Description
The following notebook no longer works when run as is: https://github.com/awslabs/amazon-sagemaker-examples/blob/master/hyperparameter_tuning/pytorch_mnist/hpo_pytorch_mnist.ipynb
Failure reason
AlgorithmError: ExecuteUserScriptError: Command "/opt/conda/bin/python mnist.py --backend gloo --batch-size 512 --epochs 6 --lr 0.018272722904211385" Traceback (most recent call last): File "mnist.py", line 197, in train(parser.parse_args()) File "mnist.py", line 93, in train train_loader = _get_train_data_loader(args.batch_size, args.data_dir, is_distributed, **kwargs) File "mnist.py", line 45, in _get_train_data_loader transforms.Normalize((0.1307,), (0.3081,)) File "/opt/conda/lib/python3.6/site-packages/torchvision/datasets/mnist.py", line 80, in init self.data, self.targets = torch.load(os.path.join(self.processed_folder, data_file)) File "/opt/conda/lib/python3.6/site-packages/torch/serialization.py", line 527, in load with _open_zipfile_reader(f) as opened_zipfile: File "/opt/conda/lib/python3.6/site-packages/torch/serialization.py", line 224, in init super(_open_zipfile_reader, self).init(torch._C.PyTorchFileReader(name_or_buffer)) RuntimeError: