# PraseFloat error on dGraph embedding mutation saving

**URL:** <https://discuss.dgraph.io/t/prasefloat-error-on-dgraph-embedding-mutation-saving/20012>\
**Category:** Dgraph\
**Tags:** dgraph\
**Created:** [October 22, 2025, 10:44am UTC](https://discuss.dgraph.io/t/prasefloat-error-on-dgraph-embedding-mutation-saving/20012 "2025-10-22T10:44:21Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Ravindu\_Iddamalgoda](https://yyz1.discourse-cdn.com/flex007/user_avatar/discuss.dgraph.io/ravindu_iddamalgoda/32/11896_2.png) [@Ravindu\_Iddamalgoda](https://discuss.dgraph.io/u/Ravindu_Iddamalgoda)\
**Post date:** [October 22, 2025, 10:44am UTC](https://discuss.dgraph.io/t/prasefloat-error-on-dgraph-embedding-mutation-saving/20012/1 "2025-10-22T10:44:21Z")

</div>

Hi , we are going through an issue where we cannot save the extracted embedding in the :8080/mutate, It just giving an API error called - \*\*strconv.ParseFloat:. When we dealing with dev env in windows there is no any issues like this but when production suddenly we got this (k8s). \*\*  
It just does not allow us to save the embedding, and our extraction fails.

Error extracting text from PDF: APIError: strconv.ParseFloat: parsing "[0.011230050586163998,0.0057051233015954494,-0.000958898279350251,-0 …

**errors: [**

```
**{**

  **message: 'strconv.ParseFloat: parsing "\[0.011230050586163998,0.0057051233015954494,-0.00095889827935,** ……………… \]**extensions: \[Object\]**

**}**

```

**]**

---

<div class="post-metadata">

**Author:** ![matthewmcneely](https://yyz1.discourse-cdn.com/flex007/user_avatar/discuss.dgraph.io/matthewmcneely/32/4774_2.png) [@matthewmcneely](https://discuss.dgraph.io/u/matthewmcneely)\
**Post date:** [October 22, 2025, 2:55pm UTC](https://discuss.dgraph.io/t/prasefloat-error-on-dgraph-embedding-mutation-saving/20012/2 "2025-10-22T14:55:58Z")

</div>

> [@Ravindu\_Iddamalgoda](#):
>
> Error extracting text from PDF

I think this error is coming from some other API that you’re using. Nowhere in the Dgraph repo does it extract text from PDFs…., nor does that error line appear anywhere in the repo.