Incoherent logged val/iou value despite coherent confusion matrix and per-class IoUs when training model in DDP #122

Observed in DDP mode with 2x3 GPUs (not tested on single node training).

This is surprising, because the per-class IoUs are computed with a bespoke method from the confusion matrix of the MultiClassJaccardIndex object that computes mean IoU

Mean IoU:

Per-class:

Might be linked to #108