Fine-tuning on the pedestrian dataset is problematic

moghadas76 commented 4 months ago

File ~/miniconda3/envs/env/lib/python3.11/site-packages/pytorch_lightning/trainer/trainer.py:580, in Trainer._fit_impl(self, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path) 573 ckpt_path = self._checkpoint_connector._select_ckpt_path( 574 self.state.fn, 575 ckpt_path, ... (...) 137 * self.dummy_value 138 )

KeyError: 'observed_values'

ashok-arjun commented 4 months ago

Hi! We're yet to release the finetuning code. It should be out soon. It's best to try this after that.

kashif commented 3 months ago

@moghadas76 check the fine tuning demo with glounts dataset here: https://colab.research.google.com/drive/1uvTmh-pe1zO5TeaaRVDdoEWJ5dFDI-pA?usp=sharing

moghadas76 commented 3 months ago

Thanks, Indeed I love this codebase in term of Clean code, modularity and etc.,. For this reason I'd like to contribute to ServiceNow. Which channels are available?

On Thu, Mar 7, 2024, 9:33 PM Kashif Rasul @.***> wrote:

@moghadas76 https://github.com/moghadas76 check the fine tuning demo with glounts dataset here: https://colab.research.google.com/drive/1uvTmh-pe1zO5TeaaRVDdoEWJ5dFDI-pA?usp=sharing

— Reply to this email directly, view it on GitHub https://github.com/time-series-foundation-models/lag-llama/issues/20#issuecomment-1984375503, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFRH3KPA34JMGPJXFMGO5U3YXDFKJAVCNFSM6AAAAABD4RWWPOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSOBUGM3TKNJQGM . You are receiving this because you were mentioned.Message ID: @.***>

ashok-arjun commented 3 months ago

Thanks for the kind words @moghadas76!

We are in the process of preparing a Call for Contributions and Contributing Guidelines. Kindly check the repository in 2-3 weeks. If it's not there by then, please feel free to open an issue.

I'm closing this issue and it is resolved.

Best, Arjun

time-series-foundation-models / lag-llama

Fine-tuning on the pedestrian dataset is problematic #20