# Learning rate and checkpoints

**URL:** </static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Flearning-rate-and-checkpoints%2F16195%3E%5C
**Category:** 🤗Transformers\
**Created:** [March 29, 2022, 7:26am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Flearning-rate-and-checkpoints%2F16195 "2022-03-29T07:26:18Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![GenV](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/genv/32/12132_2.png) [@GenV](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2FGenV)%5C
**Post date:** [March 29, 2022, 7:26am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Flearning-rate-and-checkpoints%2F16195%2F1 "2022-03-29T07:26:18Z")

</div>

Hello,  
I’m using the script _run\_traslation_ to train a T5-base, saving checkpoints at each epoch. However, when I continue the training starting from the checkpoint, the learning rate does not _“continue”_ from the value it had at the end of the previous epoch, but starts from a higher value and this entails an initially higher train loss. Why does this happen? Am I doing something wrong?

Thanks.