-
I have a ML Workspace or a server that I work on, I installed MONAI Label on it. Is this even possible? If yes, some insights would be appreciated. Futher more, is using a pretrained model and then labeling the new data possible by using 3dslicer with monailabel? |
Beta Was this translation helpful? Give feedback.
Replies: 4 comments 19 replies
-
Hi @lenlobo , yes, both scenarios should work, I typically use the same way, but you may need to double-check several things: For the first scenario: yes, but please make sure the remote and local machines are in the same network. This is typically achieved by connecting to VPN or other networking revolvers. And make sure your local machine can reach remote machine host/IP Second, this should be working, MONAI Label won't bother Slicer's docker, unless you can start Slicer, then you can type the host and port of MONAI Label server, then use it in Slicer. Same as the first scenario, it's needed to resolve networking first, in the same network. For the additional question, if new data are the same target classes (no new segments), there is no problem you can fine-tune the pretrained mode. If you are adding new classes/organs/tissues, you need to change and define a network, then train it. Enjoy MONAI Label |
Beta Was this translation helpful? Give feedback.
-
and when you start the server.. it prints a log.. you can check the log.. something like this.. |
Beta Was this translation helpful? Give feedback.
-
if you are running docker.. you want to make sure you have port mapping correctly done.. for example.. try to run simple http server in python Once you are able to access this.. you can go back to monailabel and run it.. |
Beta Was this translation helpful? Give feedback.
-
#776 (reply in thread) |
Beta Was this translation helpful? Give feedback.
#776 (reply in thread)
Hi,
The reply from Zrrr1997 seems to work for me at the moment. He suggested creating a ssh tunnel from my localport (eg 9090) to the port on which the monai label server is started. Worked!