Clara Train SDK is a domain optimized developer application framework that includes APIs for AI-Assisted Annotation. This enables any medical viewer to be AI capable and enables a TensorFlow based training framework with pre-trained models to start AI development with techniques such as Transfer Learning, Federated Learning, and AutoML.
- Clara Training Framework
- AI-Assisted Annotation
- AutoML
- Federated learning
- Federated learning background and architecture
- Federated learning user guide
- Federated learning provisioning tool
- Federated learning configuration details
- Federated learning authorization
- Federated learning administrator commands
- Bring your own components for federated learning
- Federated learning FAQ
- Federated learning API reference
What’s new
The Clara Train 3.1 release is based on NVIDIA’s container for TensorFlow release 20.08 with support for NVIDIA Ampere GPUs. Here is a list of changes and additions in this version as well as links to key features:
Federated learning. Federated learning has been enhanced to enable easy server and client deployment through the use of an administration client. See the Federated learning user guide for details.
Packaged MMARs for COVID-19 lesion segmentation, COVID-19 classification, and COVID-19 EHR + chest xray model.
For Jupyter Notebooks with detailed examples, see Clara Train Examples
You can use either MMARs to set up training configurations with json, or write python code directly with the Clara Train API.
For greater customization, you can Bring your own components (BYOC).
For AIAA:
AIAA updates to Triton API
DeepGrow is introduced in AIAA to help with cold-start in any organs/objects of interest.
Users can customize their inference workflow in AIAA following the instructions in Bring your own models.