Process Killed after graph construction and optimization

AI & Data Science Deep Learning (Training & Inference) TensorRT
#1

Hi,
I am using tensorrt on jetson nano.
I am trying to run the mnist pytorch example code(provided with tensorrt) in tensorrt folder but it doesn’t run.
changing TRT_LOGGER = trt.Logger(trt.Logger.WARNING) to TRT_LOGGER = trt.Logger(trt.Logger.INFO) I am getting the following output…

[TensorRT] INFO: Original: 7 layers
[TensorRT] INFO: After dead-layer removal: 7 layers
[TensorRT] INFO: After scale fusion: 7 layers
[TensorRT] INFO: Fusing (Unnamed Layer* 4) [Fully Connected] with (Unnamed Layer* 5) [Activation]
[TensorRT] INFO: Fusing (Unnamed Layer* 4) [Fully Connected] + (Unnamed Layer* 5) [Activation] with (Unnamed Layer* 6) [Fully Connected]
[TensorRT] INFO: After vertical fusions: 5 layers
[TensorRT] INFO: After swap: 5 layers
[TensorRT] INFO: After final dead-layer removal: 5 layers
[TensorRT] INFO: After tensor merging: 5 layers
[TensorRT] INFO: After concat removal: 5 layers
[TensorRT] INFO: Graph construction and optimization completed in 0.00094719 seconds.
Killed

#2

Hello, can you provide details on the platforms you are using?

Linux distro and version
GPU type
nvidia driver version
CUDA version
CUDNN version
Python version [if using python]
Tensorflow version
TensorRT version

Any usage/source file you can provide will help us debug too.

#3

Hi,
Sorry for the late reply, We are using the following versions

Linux distro - 18.04
pycuda==2018.1.1
128-core NVIDIA Maxwell™ architecture-based GPU
torch-1.1.0a0+b457266-cp36-cp36m-linux_aarch64.whl version of pytorch
tensorrt 5.1
cuda 10.0
cuDNN 7.3

basically we are just using the base installations that came with the latest jetson nano along with numpy and pytorch installed separately, along with this we are using only 16gb microSD card.