# Theano deprecation at 1.0

**URL:** <https://discourse.pymc.io/t/theano-deprecation-at-1-0/374>\
**Category:** Development\
**Created:** [September 28, 2017, 8:43pm UTC](https://discourse.pymc.io/t/theano-deprecation-at-1-0/374 "2017-09-28T20:43:52Z")\
**Posts on this page:** 1\
**Showing post:** 7

<div class="post-metadata">

**Author:** ![aasensioramos](https://avatars.discourse-cdn.com/v4/letter/a/7ab992/32.png) [@aasensioramos](https://discourse.pymc.io/u/aasensioramos)\
**Post date:** [September 29, 2017, 4:51pm UTC](https://discourse.pymc.io/t/theano-deprecation-at-1-0/374/7 "2017-09-29T16:51:12Z")

</div>

Except for Numba, autograd+GPU is what pytorch offers, all similar to working with Numpy. I have limited experience with deep learning in pytorch but it’s definitely very close to the Numpy experience. And it also has the possibility to add external functions. I don’t know how much the pymc3 guts would have to be changed in this case.

---

_[View the full topic](https://discourse.pymc.io/t/theano-deprecation-at-1-0/374)._