Feature Request: Expose an Args to Set Prefetch Factor in Trainer

[uygnef]

Feature request

Currently, the trainer does not allow users to set the prefetch factor as an argument from the training script.

Motivation

This can be a limitation when training large models, especially when the data is being fetched from a remote server, as the loading speed can become a bottleneck due to slow data download for the next partition.

Your contribution

I can commit a PR

[amyeroberts]

Hi @uygnef, thanks for raising an issue!

There is a related PR open: #28498 cc @muellerzr @pacman100

[uygnef]

Hi @uygnef, thanks for raising an issue!

There is a related PR open: #28498 cc @muellerzr @pacman100

oh, good job!

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.

Please note that issues that do not follow the contributing guidelines are likely to be ignored.