Lakshantha Dissanayake
Lakshantha Dissanayake
@Laughing-q I have implemented the necessary checks for DLA. Additionally, regarding DLA cores, this should be beneficial for executing multiple inference jobs simultaneously on distinct DLA cores.
@Laughing-q good idea! Thanks! I have made the changes accordingly.
@glenn-jocher got it. will work on it accordingly.
@glenn-jocher After discussing with @Laughing-q, the functionality for DLA is implemented as requested. Thank you Jing for your support on this! I have also added a `device` argument in the...
Thanks @glenn-jocher ! This is the basic workflow. 1. Export to TensorRT on Jetson with DLA enabled 2. Run exported `.engine` file as usual for inference. Then it will use...
@leeprevost I will investigate this further and get back to you.
Hello @leeprevost, Currently imx format is only supported for YOLO11n and YOLOv8n models for the following tasks: - Object detection - Pose estimation - Classification Also, you need to be...
> @lakshanthad maybe this is something you'd like to take a crack at? @Burhan-Q. Definitely! I will try to experiment!
@johnnynunez Thank you for sharing the updates with us!
@johnnynunez Currently we already have mac os arm64 runners for our CIs. However, adding Linux arm64 runners would be something to experiment with. EDIT: I saw you mentioned that the...