Nic Ma

Hi @vfdev-5 , Thanks for raising the discussion here. `IterationMetric` is the base class for other metrics, every metric will make a list. I don't quite understand your question, could...

Hi @vfdev-5 , Here is the following PR to consume the metrics output: https://github.com/Project-MONAI/MONAI/pull/1497 It's still under review. The idea is to save metric details into `engine.state`, then other handlers...

Hi @sdesrozis , Yes, that's the expected behavior, you can use an algorithm(function or callable class) as arg or define subclass. Thanks.

Hi @vfdev-5 and @sdesrozis , Thanks for your support here, sorry I didn't make my idea clear enough. Here is a demo program for my 2 ideas: 1. Add `__getitem__()`...

Hi @vfdev-5 and @sdesrozis , Could you please help share some comments? We are trying to refactor all the handlers in MONAI v0.6, it would be a big change and...

Hi @vfdev-5 , Thanks so much for your detailed suggestions and sample code! I think maybe we can try at own side first. And I have a small question about:...

Cool, sounds good! Let me try to refactor handlers for the `inference` workflows first based on this proposal. For the first step, we are trying to co-work with some GPU...

Hi @ericspod and @wyli, Do you guys have any other opinions about this proposal? If you all agree with it, I think we can try to adjust several existing handlers...

Hi @ericspod , Thanks for your suggestions and support. As ignite uses built-in enum type(`ignite.EventEnum`) for event names, maybe we can extract the strings and use in our workflows? Pure...

BTW, I mean when using `ignite.metrics.EpochMetric`, the inputs `preds` and `labels` have different lengths on different GPUs. Thanks.