# Continuous Learning Using Trainer for BERT multi-class model

**URL:** <https://discuss.huggingface.co/t/continuous-learning-using-trainer-for-bert-multi-class-model/37149>\
**Category:** Beginners\
**Created:** [April 19, 2023, 6:09pm UTC](https://discuss.huggingface.co/t/continuous-learning-using-trainer-for-bert-multi-class-model/37149 "2023-04-19T18:09:11Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![jameskim21](https://avatars.discourse-cdn.com/v4/letter/j/898d66/32.png) [@jameskim21](https://discuss.huggingface.co/u/jameskim21)\
**Post date:** [April 19, 2023, 6:09pm UTC](https://discuss.huggingface.co/t/continuous-learning-using-trainer-for-bert-multi-class-model/37149/1 "2023-04-19T18:09:11Z")

</div>

Hello,

I fine-tuned a multi-class BERT model using huggingface trainer. It is a fairly large model that was trained on thousands of classes. I want to add 100 more classes to the model there is a dataset that has the new classes as well. Is there a way for me to do this without training it from scratch again? Thank you so much!