tensorboard使用

from tensorboardX import SummaryWriter
#用 pytorch 自带的 tensorboard 也行

# default `log_dir` is "runs" - we'll be more specific here
writer = SummaryWriter('logs')

#在训练过程中，b为 iteration，e为epoch
if b%100 ==0:
    writer.add_scalar('image training loss',running_loss, e * len(train_loader) + b)
    for tag, value in model.named_parameters():
    tag = tag.replace('.', '/')
    writer.add_histogram(tag,value.data.cpu().numpy(), e * len(train_loader) + b)
    writer.add_histogram(tag+'/grad',value.grad.cpu().numpy(), e * len(train_loader) + b)

然后通过命令行 cd 到项目文件，输入：

tensorboard --logdir="路径"

或者cd 到 logs文件夹，输入：

tensorboard --logdir=./runs