I have used the tflite interpreter on the rpi and it is remarkably fast. I was wondering if there is a canonical, pre compiled version of it for the nano? (google have a precompiled wheel for the rpi).
If anyone tried the newest version and can comment on speed, i’ll be grateful.
Its not even 100% clear it can use GPU.
We only provide the standard TensorFlow prebuilt for Jetson users.
But there are some tutorial for building tflite on Nano.
This topic was automatically closed 2 days after the last reply. New replies are no longer allowed.