Is your feature request related to a problem? Please describe.
All the data we need to evaluate the performance of an image registration model is generated by using the config evaluate-registration.yaml.
However, we need to turn this into a useful assessment.
We especially need to find a way to identify "hallucination" frequency.
Describe the solution you'd like
A jupyter notebook that evaluates the accuracy of outputs stored on S3.
Is your feature request related to a problem? Please describe. All the data we need to evaluate the performance of an image registration model is generated by using the config evaluate-registration.yaml. However, we need to turn this into a useful assessment. We especially need to find a way to identify "hallucination" frequency.
Describe the solution you'd like A jupyter notebook that evaluates the accuracy of outputs stored on S3.
Additional context The code in this notebook is a good start, it just needs to be updated for AWS.