NVIDIA TITAN XP. Cuda compilation tools, release 9.0, V9.0.176
I ran two experiments -
 I ran a CNN & RNN model (with MNIST) using the Pytorch deep-learning framework.
 I ran the same CNN/RNN model (with MNIST) using CUDA.
I used the nvidia-smi to query memory utilization as, nvidia-smi --query-gpu=utilization.gpu,utilization.memory.
 shows 3% to 4% memory utilization.
 shows 0% memory utilization.
Can someone help me understand this discrepancy? Do I need to turn on any flags when doing  to get accurate memory utilization metric, please?