Fallback for model execution when config lacks a run entry has issues with the "default" key.

[JohnKossa]

File ~\dev\openavmkit\openavmkit\benchmark.py:3670, in _run_models(sup, model_group, settings, main_vacant_hedonic, save_params, use_saved_params, save_results, verbose, run_ensemble, do_shaps, do_plots)
   3667 if model in ["xgboost", "lightgbm", "catboost"]:
   3668     model_variables = None
-> 3670 results = run_one_model(
   3671     df_sales=df_sales,
   3672     df_universe=df_univ,
   3673     vacant_only=vacant_only,
   3674     model_group=model_group,
   3675     model=model,
   3676     model_entries=model_entries,
   3677     settings=settings,
   3678     dep_var=dep_var,
   3679     dep_var_test=dep_var_test,
   3680     best_variables=model_variables,
   3681     fields_cat=fields_cat,
   3682     outpath=outpath,
   3683     save_params=save_params,
   3684     use_saved_params=use_saved_params,
   3685     save_results=save_results,
   3686     verbose=verbose,
   3687 )
   3688 if results is not None:
   3689     model_results[model] = results

File ~\dev\openavmkit\openavmkit\benchmark.py:1311, in run_one_model(df_sales, df_universe, vacant_only, model_group, model, model_entries, settings, dep_var, dep_var_test, best_variables, fields_cat, outpath, save_params, use_saved_params, save_results, verbose, hedonic, test_keys, train_keys)
   1309     results = run_slice(ds, verbose=verbose)
   1310 else:
-> 1311     raise ValueError(f"Model {model_name} not found!")
   1312 t.stop("run")
   1314 if ds.vacant_only or ds.hedonic:
   1315     # If this is a vacant or hedonic model, we attempt to load a corresponding "full value" model

ValueError: Model default not found!

The fallback for running the models looks at main, vacant, or hedonic directly if no run is present, but those keys contain a "default" key, that throws an error