# Training Loss Higher than Validation Loss

**URL:** </static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Ftraining-loss-higher-than-validation-loss%2F21211%3E%5C
**Category:** Beginners\
**Created:** [August 3, 2022, 2:11pm UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Ftraining-loss-higher-than-validation-loss%2F21211 "2022-08-03T14:11:00Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![marlon89](https://avatars.discourse-cdn.com/v4/letter/m/6de8d8/32.png) [@marlon89](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2Fmarlon89)%5C
**Post date:** [August 3, 2022, 2:11pm UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Ftraining-loss-higher-than-validation-loss%2F21211%2F1 "2022-08-03T14:11:00Z")

</div>

Hello everyone,

I got a question regarding my losses. When I finetune a pretrained BERT AutomodelforSequenceClassification model on a Multiclass Classification problem my training loss is always higher compared to my validation loss (see the picture attached):

 ![Screenshot (3)](https://us1.discourse-cdn.com/hellohellohello/original/2X/c/c356a0c7ec9e181fe9c0533b9c328ccf7655a95b.png)

Is this something to worry about? Thank you guys