# Ignore padding area in loss computation

**URL:** <https://discuss.pytorch.org/t/ignore-padding-area-in-loss-computation/95804>\
**Category:** nlp\
**Created:** [September 10, 2020, 2:38pm UTC](https://discuss.pytorch.org/t/ignore-padding-area-in-loss-computation/95804 "2020-09-10T14:38:12Z")\
**Posts on this page:** 1\
**Showing post:** 9

<div class="post-metadata">

**Author:** ![saxenarohit](https://discuss.pytorch.org/letter_avatar_proxy/v4/letter/s/ba9def/32.png) [@saxenarohit](https://discuss.pytorch.org/u/saxenarohit)\
**Post date:** [October 5, 2022, 5:48pm UTC](https://discuss.pytorch.org/t/ignore-padding-area-in-loss-computation/95804/9 "2022-10-05T17:48:22Z")

</div>

An alternative way of option B

```auto
loss_masked = torch.masked_select(loss, loss_mask)
loss_masked.mean()

```

---

_[View the full topic](https://discuss.pytorch.org/t/ignore-padding-area-in-loss-computation/95804)._