Config¶

The PyTorch specific config of DeepOBS.

deepobs.pytorch.config.set_num_workers(num_workers)[source]¶

Sets the number of workers that are used in the torch DataLoaders.

Parameters: num_workers (int) -- The number of workers that are used for data loading.

deepobs.pytorch.config.set_is_deterministic(is_deterministic)[source]¶

Sets whether PyTorch should try to run deterministic.

Parameters: is_deterministic (bool) -- If True, this flag sets: torch.backends.cudnn.deterministic = True torch.backends.cudnn.benchmark = False. However, full determinism is not guaranteed. For more information, see: https://pytorch.org/docs/stable/notes/randomness.html

deepobs.pytorch.config.set_default_device(device)[source]¶

Sets the device on which the PyTorch experiments are run.

Parameters: device (str) -- Device on which to run the PyTorch test problems. E.g. 'cuda' or 'cuda:0'