Hey, I am trying to use clearml's hyper parameter optimization ability. I am struggling connecting the parameters names to my code, getting the following error:{'Args/batch_size': 128} because of the following error: The value None could not be cast to float.. I also tried Args/data.batch_size , and data/batch_size but no success.
Is there a way to get a more relevant error? like the options or the names that clearml loaded and will except?
This is my config file:
Hi ObliviousClams17 , can you provide a code snippet that reproduces this?
Hey @<1523701070390366208:profile|CostlyOstrich36> , i'm afraid my question was not clear enough. I am trying to run a pytorch lightning model, using lightning cli args. The screenshot I attached is vs config file (for debugging) used to run my model. How can I integrate it with clearml's hyper parameter tuning? Just not sure about the names convention
