摘要:
代码片段: _, train_pred = torch.max(outputs, 1) train_acc += (train_pred.detach() == labels.detach()).sum().item() torch.max(outputs, 1): 这个函数用于返回每一行(即每个样 阅读全文

posted @ 2024-10-07 14:41
srrdhy
阅读(30)
评论(0)
推荐(0)
摘要:
cuda安装:https://blog.csdn.net/AI_dataloads/article/details/133043869 查看gpu算力:https://developer.nvidia.com/cuda-gpus 首先卸载当前安装的pytorch: pip uninstall tor 阅读全文

posted @ 2024-10-07 13:36
srrdhy
阅读(32)
评论(0)
推荐(0)