# When using wav2vec2 inference, does the quantitation being used?

**URL:** <https://discuss.huggingface.co/t/when-using-wav2vec2-inference-does-the-quantitation-being-used/32154>\
**Category:** Models\
**Created:** [February 21, 2023, 8:27am UTC](https://discuss.huggingface.co/t/when-using-wav2vec2-inference-does-the-quantitation-being-used/32154 "2023-02-21T08:27:46Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![laro1](https://avatars.discourse-cdn.com/v4/letter/l/ad7895/32.png) [@laro1](https://discuss.huggingface.co/u/laro1)\
**Post date:** [February 21, 2023, 8:27am UTC](https://discuss.huggingface.co/t/when-using-wav2vec2-inference-does-the-quantitation-being-used/32154/1 "2023-02-21T08:27:46Z")

</div>

According to wav2vec model:

[![enter image description here](https://us1.discourse-cdn.com/hellohellohello/original/2X/d/d25010c3b3e77c0636300fb87406a7472a96490f.png)](https://i.stack.imgur.com/zmW10.png)

it seems that **quantitation** [![enter image description here](https://us1.discourse-cdn.com/hellohellohello/original/2X/0/0aa694d5a4382b272551f20219cd7ca7f0efdb35.png)](https://i.stack.imgur.com/klJn6.png) only used on training phase (in order to train the transformer layer to quantitative the data to use small number of discrete values).

So my question is:  
when using inference, does this **quantitation** unit [![(circle purple in the figure above)](https://us1.discourse-cdn.com/hellohellohello/original/2X/0/0aa694d5a4382b272551f20219cd7ca7f0efdb35.png)](https://i.stack.imgur.com/qSJDj.png) being used ? (I guess no, am I right ?)