How do I increase the num_workers value correctly?

[KatieBelli]

I always get this warning before the training starts:

C:\Users...\AppData\Local\Programs\Python\Python39\lib\site-packages\lightning\pytorch\trainer\connectors\data_connector.py:430: PossibleUserWarning: The dataloader, val_dataloader, warnings.py:109
does not have many workers which may be a bottleneck. Consider increasing the value of the num_workers argument(try 8 which is the number of cpus on this machine) in theDataLoader` init to
improve performance.
rank_zero_warn(

I don't understand this error because I changed the value of num_workers to 8 in the config file before training.

[TheMeIonGod]

I would like to know the same thing.

[single-serval]

same, and there is only a warning code in the directory/file it asks for. Mine is telling me to up to 15 but I can't find a way to actually do that and have it read it

[kevinlinxc]

Yep, same question, why is this so weird? It should be an easy to answer question

[34j]

You can edit the code to increase num_workers (which has no meaning, please refer to the PyTorch docs) or send a PR to suppress this annoying warning https://github.com/Lightning-AI/pytorch-lightning/pull/10215/files