Well TenseOstrich47 you can certainly add any metadata to the model itself at training time (or as a post-training job if the model is not published). I'm actually doing a couple of videos on this on the ClearSHOW, tomorrow's episode should explain this approach and there will be examples next week. DM me here if you want to know more.
https://www.youtube.com/watch?v=r2BMMDzfyA0&list=PLMdIlCuMqSTkXBApOMqg2S5IeVfnkq2Mj

I think there is more complexity to what I am trying to achieve, but this will be a good start. Thanks!

I will need to log data set ID, transformer (not the NN architecture, just a data transformer), the model (with all hyperparameters & metadata) etc. and how all things link

Hi TenseOstrich47 ,

If you want to get all the scalars. you can use task.get_last_scalar_metrics() , can this help?