# Couldn't instantiate the backend tokenizer

**URL:** </static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fcouldnt-instantiate-the-backend-tokenizer%2F2662%3E%5C
**Category:** 🤗Tokenizers\
**Created:** [December 7, 2020, 2:19pm UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fcouldnt-instantiate-the-backend-tokenizer%2F2662 "2020-12-07T14:19:55Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![wei](https://avatars.discourse-cdn.com/v4/letter/w/90ced4/32.png) [@wei](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2Fwei)%5C
**Post date:** [December 7, 2020, 2:19pm UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fcouldnt-instantiate-the-backend-tokenizer%2F2662%2F1 "2020-12-07T14:19:55Z")

</div>

I have noticed the updates and now I have a problem about loading my sentencepiece tokenizer.  
When I compute some of my models in the hugging face UI, I got the following error:

> Couldn’t instantiate the backend tokenizer from one of: (1) a `tokenizers` library serialization file, (2) a slow tokenizer instance to convert or (3) an equivalent slow tokenizer class to instantiate and convert. You need to have sentencepiece installed to convert a slow tokenizer to a fast one.
> 
> ![Screen Shot 2020-12-07 at 1.55.51 PM](https://us1.discourse-cdn.com/hellohellohello/original/2X/a/a364b0b3ed49f086714c746fde45ac733c5e3a13.png)

I tried to save my tokenizer again after installing transfomers and sentencespiece like:

from transformers  
import T5Tokenizer  
tok = T5Tokenizer.from\_pretrained(“my\_spm.model”)  
tok.save\_pretrained(path\_to\_tf\_checkpoint)

But this doesn’t solve my problem.  
Any idea what should I do?