# Model not predicting on OAK-D Lite

**URL:** <https://discuss.roboflow.com/t/model-not-predicting-on-oak-d-lite/2944>\
**Category:** 🤝  Community Help\
**Created:** [August 15, 2023, 7:11am UTC](https://discuss.roboflow.com/t/model-not-predicting-on-oak-d-lite/2944 "2023-08-15T07:11:00Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![MechaMind](https://yyz1.discourse-cdn.com/flex029/user_avatar/discuss.roboflow.com/mechamind/32/2206_2.png) [@MechaMind](https://discuss.roboflow.com/u/MechaMind)\
**Post date:** [August 15, 2023, 7:11am UTC](https://discuss.roboflow.com/t/model-not-predicting-on-oak-d-lite/2944/1 "2023-08-15T07:11:00Z")

</div>

Hi, I am making model to detect circle like this one:  
 ![image](https://canada1.discourse-cdn.com/flex029/uploads/roboflow1/original/2X/1/127670acbf5f3e25003d690fb31755f00e4dfcc1.jpeg)  
Everything work great even with webcam in Roboflow app, but when I run this on OAK-D Lite, there are no predictions (not showing on cv2 output as well), console output:

```auto
INFERENCE TIME IN MS 15.175145535523693
PREDICTIONS []
INFERENCE TIME IN MS 52.356809387092746
PREDICTIONS []
INFERENCE TIME IN MS 13.512620852515633
PREDICTIONS []
INFERENCE TIME IN MS 25.68701350399608
PREDICTIONS []

```

Why this happens, did I something wrong during training that OAK-D Lite just can’t do. I also tried to train it with both YoloV8 and YoloV5 and nothing has worked out for me. Thanks

---

<div class="post-metadata">

**Author:** ![leo](https://yyz1.discourse-cdn.com/flex029/user_avatar/discuss.roboflow.com/leo/32/2228_2.png) [@leo](https://discuss.roboflow.com/u/leo)\
**Post date:** [August 16, 2023, 12:24pm UTC](https://discuss.roboflow.com/t/model-not-predicting-on-oak-d-lite/2944/2 "2023-08-16T12:24:19Z")

</div>

Hi @MechaMind

Sorry to hear you’re having issues. This is similar to a known issue we recently resolved. Can you make sure you are on v0.0.12 or later?

> [@Luxonis OAK-D Lite Deployment Failure](https://discuss.roboflow.com/t/luxonis-oak-d-lite-deployment-failure/2874/2):
>
> We believe the issue was caused by a breaking change to the Depthai API. We have pushed v0.0.12 to PyPi which uses the updated API. You can update with `pip install roboflowoak==0.0.12`.