Logging and TensorBoard Integration

[jakubMitura14]

This task focuses on building a comprehensive logging infrastructure for the MedPipe3D pipeline, facilitating easier debugging, progress tracking, and user feedback. It will also enable live visualization of training dynamics through TensorBoard, including integration with MedEye3D for side-by-side inspection of segmentation results.
Success when:
A structured logging system is implemented using Logging.jl that tracks:
pipeline events,
user activity,
model metrics.
Integration with TensorBoard is complete.