Mauricio Villegas

The same feature is requested in #460. This issue is older, but since there is more discussion in that one, I will be closing this one.

> Sorry for all the questions. Please feel free to ignore it. No worries. Ask as much as you need. Anyway the questions might be helpful for other people too....

@vedal did my comment answer your questions? Can this be closed now?

If you are new to jsonargparse it is best if you familiarize yourself with its [override-order](https://jsonargparse.readthedocs.io/en/stable/#override-order). Just because in hydra something can be done in a certain way, it does...

> do you usually always print_config before every experiment, as a kind of "--dry-run" to check that all hparams are ok? I do use `--print_config` extensively. Though mostly for debugging....

@rusmux as explained here, multiple arguments is the alternative. From what I understand the only motivation is that sometimes it is inconvenient to use multiple arguments. This is a valid...

Thank you for reporting! As an update. I looked at this some time ago, and it seems to require quite a complex change to fix. Unfortunately this relates to the...

Closing this in favor of #221

Does your model have init parameters `input_width` and `input_height`? If so, then the links should be with `init_args` as described in the error message, i.e.: ```python parser.link_arguments("data.input_width", "model.init_args.input_width") parser.link_arguments("data.input_height", "model.init_args.input_height")...

The recommended way of implementing submodules (e.g. backbone, loss) is via dependency injection, see [models-with-multiple-submodules](https://lightning.ai/docs/pytorch/stable/cli/lightning_cli_advanced_3.html#models-with-multiple-submodules). Linking with target `dict_kwargs` is not a trivial change and is unlikely to be added...