# Initialize Vocabulary for Unigram Tokenizer

**URL:** </static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Finitialize-vocabulary-for-unigram-tokenizer%2F46371%3E%5C
**Category:** 🤗Tokenizers\
**Created:** [July 11, 2023, 6:28am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Finitialize-vocabulary-for-unigram-tokenizer%2F46371 "2023-07-11T06:28:20Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![dotan1111](https://avatars.discourse-cdn.com/v4/letter/d/85f322/32.png) [@dotan1111](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2Fdotan1111)%5C
**Post date:** [July 11, 2023, 6:28am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Finitialize-vocabulary-for-unigram-tokenizer%2F46371%2F1 "2023-07-11T06:28:20Z")

</div>

I have reviewed the documentation and code, but I was unable to find the specific method used for initializing the set of tokens in the Unigram tokenizer. Typically, it is initialized by running a BPE tokenizer, but I would like to confirm this information. If anyone could provide a link to the code or documentation or offer an answer regarding this matter, it would be greatly appreciated.