I built Caffe SSD on Jetson nano to train voc2007 data set,stuck in “data layer prefetch queue is empty”
“Data layer prefetch queue empty” means that the data loading is slower than net forward, so net has to wait for data IO to prepare next batch.
It should be a harmless warning.
But if you get stuck on this, it might indicate that your job is too slow to run on Jetson Nano.
Please noticed that Jetson is designed as a high inference edge device.
It may not be suitable for training.