Closed andrewcztrack closed 3 years ago
Are you running with the latest version of pytorch and pytorch-forecasting?
BTW: There is a consequence of the new metric system/bug (which is being fixed now in #112 ) that forces you to specify the loss you want to optimize (i.e. you have to use .from_dataset(loss=MASE())
when you create your network.
Hi @jdb78 !! i hope your well! thank you for your reply. i was just running your tutorials from the notebook. i tried to reinstall the library but the errors still comes up. Best, Andrew
Have you tried with pytorch 1.6? Conda forge doesn’t not allow specifying higher versions than 1.4.
Done! and it is working! Thank you :).
Hi ! love the look of the package! looks amazing!
I am getting an error for the nbeats example.
load data GPU available: False, used: False TPU available: False, using: 0 TPU cores Number of parameters in network: 1859.9k
| Name | Type | Params
0 | loss | SMAPE | 0
1 | logging_metrics | ModuleList | 0
2 | net_blocks | ModuleList | 1 M
Validation sanity check: 0it [00:00, ?it/s] /home/andrewcz/miniconda3/envs/myenv/lib/python3.8/site-packages/pytorch_lightning/utilities/distributed.py:45: UserWarning: The dataloader, val dataloader 0, does not have many workers which may be a bottleneck. Consider increasing the value of the
num_workers
argument(try 4 which is the number of cpus on this machine) in the
DataLoader` init to improve performance. warnings.warn(*args, **kwargs)TypeError Traceback (most recent call last)