如何使用 tensorboardX 在同一张图表上获取两个标量?
How to get two scalars on same chart with tensorboardX?
文档似乎表明 add_custom_scalars_multilinechart 做到了,但它不起作用。
有这样的东西:
from tensorboardX import SummaryWriter
writer = SummaryWriter(comment='test')
writer.add_custom_scalars_multilinechart(['loss/train', 'loss/test'], title='losses')
for blahblah:
...
writer.add_scalar('loss/train', loss.item(), epoch)
writer.add_scalar('loss/test', loss_test.item(), epoch)
使用 tensorboardX 在同一图表上绘制两个标量:
from tensorboardX import SummaryWriter
为两个标量创建两个 summaryWriter
writer_train = SummaryWriter('runs/train_0')
writer_test = SummaryWriter('runs/test_0')
将标量实例添加到相应的 summaryWriter;它们必须具有相同的标签,例如:“LOSS”
for data in loop:
writer_train.add_scalar('LOSS', loss.data.item(), idx)
writer_test.add_scalar('LOSS', loss_test.data.item(), idx)
有关工作代码,请访问github:Examples with tensorboardX
(参见 more_plots_one_chat.py)
教程:TensorboardX
文档似乎表明 add_custom_scalars_multilinechart 做到了,但它不起作用。
有这样的东西:
from tensorboardX import SummaryWriter
writer = SummaryWriter(comment='test')
writer.add_custom_scalars_multilinechart(['loss/train', 'loss/test'], title='losses')
for blahblah:
...
writer.add_scalar('loss/train', loss.item(), epoch)
writer.add_scalar('loss/test', loss_test.item(), epoch)
使用 tensorboardX 在同一图表上绘制两个标量:
from tensorboardX import SummaryWriter
为两个标量创建两个 summaryWriter
writer_train = SummaryWriter('runs/train_0')
writer_test = SummaryWriter('runs/test_0')
将标量实例添加到相应的 summaryWriter;它们必须具有相同的标签,例如:“LOSS”
for data in loop:
writer_train.add_scalar('LOSS', loss.data.item(), idx)
writer_test.add_scalar('LOSS', loss_test.data.item(), idx)
有关工作代码,请访问github:Examples with tensorboardX (参见 more_plots_one_chat.py)
教程:TensorboardX