dataload_format_to_csv script errors

[bowenyou]

Thanks for making the repo public! I'm trying to use the dataload_format_to_csv script within the scripts directory as in your README. I've followed the instructions exactly so far and when I am running
python3 scripts/csv_handling/dataload_format_to_csv.py raw_test_path=data/raw_datasets/150

I get the following error:

scripts/csv_handling/dataload_format_to_csv.py:50: UserWarning: 
config_path is not specified in @hydra.main().
See https://hydra.cc/docs/next/upgrades/1.0_to_1.1/changes_to_hydra_main_config_path for more information.
  @hydra.main(config_name="../config")
Could not override 'raw_test_path'.
To append to your config use +raw_test_path=data/raw_datasets/150
Key 'raw_test_path' is not in struct
    full_key: raw_test_path
    object_type=dict

Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace.

I haven't changed anything within the repo, and am just running it as is. Is there a fix for this?
Is it possible to adapt this script to not just the validation set but the training data as well?

Thanks!

[bowenyou]

think it has to do with version of hydra-core. Downgrading to 1.0.0 from 1.1.0 seems to fix this.

[TommasoBendinelli]

Hi bowenyou! Thanks for finding this out. I guess hydra 1.0.0 is needed because namespaces are handled differently.