# Special tokens and inference

**URL:** </static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fspecial-tokens-and-inference%2F2083%3E%5C
**Category:** Intermediate\
**Created:** [November 16, 2020, 10:57am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fspecial-tokens-and-inference%2F2083 "2020-11-16T10:57:16Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![Zack](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/zack/32/4845_2.png) [@Zack](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2FZack)%5C
**Post date:** [November 16, 2020, 10:57am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fspecial-tokens-and-inference%2F2083%2F1 "2020-11-16T10:57:16Z")

</div>

I have created a token classification model, and I was thinking during inference is it better to remove special tokens on encoding or the prediction will be affected by this change !