Both options would work for inference and/or for resuming a model's training from a previous checkpoint. 1. Using torch.save() and torch.load().
確定! 回上一頁