Model deployment on thingy 53, nan prediction when running model (regression)https://devzone.nordicsemi.com/f/nordic-q-a/127207/model-deployment-on-thingy-53-nan-prediction-when-running-model-regressionHello there, I trained my own model using LSTM and a window size of 25 for temporal encoding, and then did the TensorFlow conversion for deployment on the edgeimpulse platform. I downloaded the model as a C++ library so that I can deploy it on the Thingyen-USTelligent Community 13Wed, 04 Mar 2026 18:05:49 GMTRE: Model deployment on thingy 53, nan prediction when running model (regression)https://devzone.nordicsemi.com/thread/562583?ContentTypeID=1Wed, 04 Mar 2026 18:05:49 GMT137ad170-7792-4731-bb38-c0d22fbe4515:8461ef1e-4d1a-41be-98c8-f9e30ac0f0b0wafflezz<p>I see. No problem, thank you for letting me know!</p><div style="clear:both;"></div>RE: Model deployment on thingy 53, nan prediction when running model (regression)https://devzone.nordicsemi.com/thread/562321?ContentTypeID=1Mon, 02 Mar 2026 13:10:08 GMT137ad170-7792-4731-bb38-c0d22fbe4515:86e86dee-6319-4652-98cf-f3598dcd5d69Carl Richard<p>Hi!<br /><br />It&#39;s difficult for us to provide help with issues related to custom models like yours. I suggest reaching out to Edge Impulse customer support.&nbsp;<br /><br />Good luck!<br /><br />Best regards,<br />Carl Richard</p><div style="clear:both;"></div>