TensorFlow on Nano doesn't seem to use swap file

Trying to use the Nano to train a model with TensorFlow, keep running out of RAM. Created a 16GB swap file, but TensorFlow either doesn’t see it or refuses to use it. I’m using the Jetson-Nano build of TensorFlow. Is there some flag or something to tell it it’s OK to use a swap file?

Hi jetsonvzqvp,

Please refer to below topic: https://devtalk.nvidia.com/default/topic/1051624/jetson-nano/can-the-tegra-use-swap-trying-to-train-a-network-with-tensorflow-/post/5337683/#5337683