# Access to word embeddings

**URL:** <https://support.prodi.gy/t/access-to-word-embeddings/2815>\
**Category:** Uncategorized\
**Tags:** usage, spacy\
**Created:** [April 21, 2020, 3:43pm UTC](https://support.prodi.gy/t/access-to-word-embeddings/2815 "2020-04-21T15:43:33Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![drdileepunni](https://sea2.discourse-cdn.com/flex020/user_avatar/support.prodi.gy/drdileepunni/32/1341_2.png) [@drdileepunni](https://support.prodi.gy/u/drdileepunni)\
**Post date:** [April 21, 2020, 3:43pm UTC](https://support.prodi.gy/t/access-to-word-embeddings/2815/1 "2020-04-21T15:43:33Z")

</div>

Does prodigy use word embeddings to do auto annotations? If yes do we have access to those word embeddings?

Thanks much

---

<div class="post-metadata">

**Author:** ![adriane](https://avatars.discourse-cdn.com/v4/letter/a/46a35a/32.png) [@adriane](https://support.prodi.gy/u/adriane)\
**Post date:** [April 22, 2020, 12:21pm UTC](https://support.prodi.gy/t/access-to-word-embeddings/2815/2 "2020-04-22T12:21:59Z")

</div>

The word embeddings are part of the spacy model provided to the recipe. If you're using a blank model or one of spacy's provided `sm` models, there won't be any word embeddings, but the `md` and `lg` models contain vectors.

If you load the model with spacy, you can access the vectors through the vocab with `nlp.vocab["word"].vector` or more generally under `nlp.vocab.vectors`, see [https://spacy.io/api/vectors](https://spacy.io/api/vectors).

---

<div class="post-metadata">

**Author:** ![ines](https://sea2.discourse-cdn.com/flex020/user_avatar/support.prodi.gy/ines/32/3_2.png) [@ines](https://support.prodi.gy/u/ines)\
**Post date:** [April 22, 2020, 2:36pm UTC](https://support.prodi.gy/t/access-to-word-embeddings/2815/3 "2020-04-22T14:36:34Z")

</div>

To add to Adriane's comment, if you have your own embeddings and model (fine-tuned transformer, custom word vectors etc.) and want to use that to suggest examples for annotation, you can set up a custom recipe and load in your data however you like. Here are two examples for NER and text classification:

- [https://prodi.gy/docs/text-classification#active-learning-custom](https://prodi.gy/docs/text-classification#active-learning-custom)
- [https://prodi.gy/docs/named-entity-recognition#custom-model](https://prodi.gy/docs/named-entity-recognition#custom-model)