Engine ML provides a hosted tensorboard to monitor and compare the utility of your models. If you are using PyTorch, we recommend using TensorboardX.
engine job tensorboard polished-pinion
See Multiple Jobs on One Tensorboard
If you want to see more than one run in Tensorboard, supply multiple Job IDs to the command.
engine job tensorboard polished-pinion intelligent-trim
-t / --tags
Using tags is a great way to keep jobs organized.
You can use one or more
-t flags to quickly show many jobs at once.
engine job tensorboard -t best-model -t resnet-50
Tags can also be interspersed with Job IDs:
engine job tensorboard -t best-model polished-pinion -t resnet-50