# How does Huggingface Trainer handle Iterable dataset on TPU?

**URL:** <https://discuss.huggingface.co/t/how-does-huggingface-trainer-handle-iterable-dataset-on-tpu/14796>\
**Category:** Intermediate\
**Created:** [February 16, 2022, 3:03pm UTC](https://discuss.huggingface.co/t/how-does-huggingface-trainer-handle-iterable-dataset-on-tpu/14796 "2022-02-16T15:03:36Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![ibraheemmoosa](https://avatars.discourse-cdn.com/v4/letter/i/a587f6/32.png) [@ibraheemmoosa](https://discuss.huggingface.co/u/ibraheemmoosa)\
**Post date:** [February 16, 2022, 3:03pm UTC](https://discuss.huggingface.co/t/how-does-huggingface-trainer-handle-iterable-dataset-on-tpu/14796/1 "2022-02-16T15:03:36Z")

</div>

Can anyone tell me about how iterable datasets are handled in this type of multi-device scenario? How is it made sure that the same data is not duplicated across multiple devices?
