# Libtorch model inference

**URL:** <https://discuss.pytorch.org/t/libtorch-model-inference/51989>\
**Category:** C++\
**Created:** [July 30, 2019, 12:34am UTC](https://discuss.pytorch.org/t/libtorch-model-inference/51989 "2019-07-30T00:34:41Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![sanketgujar](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/sanketgujar/32/14323_2.png) [@sanketgujar](https://discuss.pytorch.org/u/sanketgujar)\
**Post date:** [July 30, 2019, 12:34am UTC](https://discuss.pytorch.org/t/libtorch-model-inference/51989/1 "2019-07-30T00:34:41Z")

</div>

Hi, I am trying to load do inference in libtorch, but it gives me Segmentation fault (core dumped) after I pass the model predicts the outputs. I am using the example code and I am not sure why is it happening?

you can find the here: [https://gist.github.com/sanketgujar/baf27851fb1b849ced78cf0668411dd0](https://gist.github.com/sanketgujar/baf27851fb1b849ced78cf0668411dd0)

---

<div class="post-metadata">

**Author:** ![yf225](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/yf225/32/3418_2.png) [@yf225](https://discuss.pytorch.org/u/yf225)\
**Post date:** [July 30, 2019, 3:04am UTC](https://discuss.pytorch.org/t/libtorch-model-inference/51989/2 "2019-07-30T03:04:47Z")

</div>

Could you also provide your model definition code and how you export your model?

---

<div class="post-metadata">

**Author:** ![sanketgujar](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/sanketgujar/32/14323_2.png) [@sanketgujar](https://discuss.pytorch.org/u/sanketgujar)\
**Post date:** [July 30, 2019, 5:24pm UTC](https://discuss.pytorch.org/t/libtorch-model-inference/51989/3 "2019-07-30T17:24:10Z")

</div>

Model Defination code : [https://gist.github.com/81dff17eb270594637b5b356591bf8c1.git](https://gist.github.com/81dff17eb270594637b5b356591bf8c1.git)

I converted the model to Torch Script via tracing as shown in the tutorial  
[https://pytorch.org/tutorials/advanced/cpp\_export.html#step-4-executing-the-script-module-in-c](https://pytorch.org/tutorials/advanced/cpp_export.html#step-4-executing-the-script-module-in-c)

---

<div class="post-metadata">

**Author:** ![sanketgujar](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/sanketgujar/32/14323_2.png) [@sanketgujar](https://discuss.pytorch.org/u/sanketgujar)\
**Post date:** [July 30, 2019, 6:00pm UTC](https://discuss.pytorch.org/t/libtorch-model-inference/51989/4 "2019-07-30T18:00:39Z")

</div>

The debugger output:

_Thread 1 “example-app” received signal SIGSEGV, Segmentation fault._  
_0x00007ffff608d9fe in ?? () from /usr/local/cuda-10.0/lib64/libcudart.so.10.0_

I am still not sure what is the problem here. The model is doping the prediction but giving seg fault after.