TensorRT
TensorRT copied to clipboard
System throttled due to over-current warning using TensorRT 8.5 when running infer.py from detectron samples in the repo on Jetson Orin NX
Running infer.py from https://github.com/NVIDIA/TensorRT/tree/release/8.6/samples/python/detectron2 on my custom dataset. Inference is being done correctly. However getting system throttled due to over current. This doesn't happen when running other tensorrt samples. Please help.
I dont think that It affects your performance though, Detectron2 is quite heavy so the system throttles. Also if you try int8, maybe, emphasis on maybe, you may not get that system throttle notification.
Yes, more like a loading issue. NX doesn't have much source.
closing since no activity for more than 3 weeks, thanks all!