可以用Tensorboard在vscode的ssh连接中显示plot出的图像. 具体方法如下: reasonml from torch.utils.tensorboard import SummaryWriter writer = SummaryWriter(log_dir='./logdir/') fig = plt.figure() # 绘图操作 writer.add_figure('figname', fig, 0) 在ssh中: tensorboard --logdir ./ 在浏览器中打开 ...
在vscode上使用tensorboard:首先在服务器上开启tensorboard服务: tensorboard --logdir=/home/general/data/daizelin/reg_seg_dai/logs;其次在vscode上ctrl + shift + P,输入tensorboard,选择即可。 编辑于 2023-02-14 20:52・IP 属地四川 Visual Studio Code 赞同5添加评论 分享喜欢收藏...
"--num_workers", "64", "--gradient_checkpointing", "--ddp_find_unused_parameters", "False", "--logging_first_step", "True", "--output_dir", "./output", "--report_to", "tensorboard" ], "env": { "CUDA_VISIBLE_DEVICES": "\"0,1,2,3\"" }, "justMyCode": false } ] ...