madtoinou

Hi, I tried to replicate your issue with the latest darts version (0.22.0) and could not reproduce it. Could you please try to update the library and see if the...

Hi @Allena101, I think you misunderstood what `triage` meant, it's just a way for us to classify issues and be able to filter them. `triage` just mean that we have...

NBeats is a deep learning model whereas LightGBM is a regression model, hence the difference is available method/approach, especially for the saving/loading of models. `save_checkpoints` in Darts save the latest...

Sorry for the confusion: the code snippet you use, based on optuna, does not perform a gridsearch but use more complex algorithm to sample the parameters ([documentation](https://optuna.readthedocs.io/en/stable/tutorial/10_key_features/003_efficient_optimization_algorithms.html)) and it's indeed...

Hi @Allena101, Sorry for the delay, got busy with other things. By default, the trainer generates a checkpoints at the end of each epoch but only keep the last one...

Hi @Allena101, The parameter `n_epochs` is correctly taken into account, however, `gridsearch()` calls `historical_forecasts()` under the hood when "expanding model" is used (see [detailed documentation](https://unit8co.github.io/darts/generated_api/darts.models.forecasting.tcn_model.html?highlight=tcnmodel+gridsearch#darts.models.forecasting.tcn_model.TCNModel.gridsearch)), which train the model iteratively...

Closing this as @dennisbader detailed the reason for keeping the current behavior in #1689. `load_weights_from_checkpoint()` should be privileged for fine-tuning/retraining whereas calling `fit()` after loading a model will resume the...

Hi, Thank you for sharing a code snippet to duplicate the bug and detailing the error messages. I am afraid that this kind of corner case should be handled before...

I had a look on how it could be implemented ; the `training_samples` array in `_fit_model()` can be converted into a `pd.DataFrame` (or similar data structure) with the values generated...

I mean, `xarray.DataArray` could also be used since it also allow for labeled dimensions without adding too much computational overhead? Should I run a quick benchmark to assess the impact...