Parallelize neuron training processes for each neuron core #466

Job	Run time
build-image	4m 56s
build	2m 53s
build-image-neuron-training	6m 27s
build-image-neuron-inference	5m 19s
build-image-neuronx	15m 28s
build-image-nvidia	4m 16s
build-image-nvidia-training	9m 9s
build-image-nvidia-inference	6m 40s
build-test	1m 18s
	56m 26s

Provide feedback