Clara Train setup help


I am new to Clara, and wants to setup Clara Train. I can see that this needs to be setup using docker image and works only in Ubuntu. My question is, as part of annotation, why do we need model?
Also, can you please detail out steps for setting up Clara Train environment

Thanks for your interest in Clara Train SDK. Please note our latest release is clara train V4.0 based on MONAI which uses PyTorch. Please check out the notebooks to get you started clara-train-examples/PyTorch/NoteBooks at master · NVIDIA/clara-train-examples · GitHub it walks your through the setup and getting started with multiple examples

To your other question “as part of annotation, why do we need model” the AIAA is helping you annotate using deep learning models as deep grow which is an interactive deep learning model. For that reason you need a model. The notebooks walks you through training AIAA models too

Hope this helps

Thanks @aharouni
We found list of steps to setup Clara Train is confusion. Do you suggest setting up server using docker or any other option. We followed below mentioned steps and not able to proceed as it started giving error as clara_experiements is not defined…etc. Do you have any other better document for setting up Clara Train environment

Also, what model do we have to use while training brain CT scan images? Can you suggest any Pretrained model. When I checked documentation you had shared, its mentioned that DeepGrow suggests Clara_pt_deepgrow_3d_annotation, but they also mentioned it suits for abdomen related scanned images, and not guaranteed on other parts of body.

One more point, the model that’s been configured to use for annotation, would that also gets trained can be considered as final model?


Hi @karunakar.r
Below are system requirement for running the clara train SDK

Also, you can refer to below link for all pre-trained model currently supported:

AI-Assisted Annotation is to speed-up creation of high quality annotated datasets that can be used to train robust AI algorithms.
Once you have saved enough new samples to train/finetune, you can create dataset.json required to feed to the training pipeline for Clara.


This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.