How to run a deep learning model on TPU

I have a jetson Orin AGX. I want to test my DL model on TPU. I tried using pytorch_xla but bumped into several errors. Does any know how to access TPU on Orin ?

Hi,

Do you mean GPU or the Tensor Core?
Thanks.

I mean Tensor Cores

Hi,

You can run TensorRT with the fp16 or int8 flag to deploy the model with Tensor Core.

Thanks.