Kuo-Hsin Tu NTU-P04922004

pip install torchsummary or git clone https://github.com/sksq96/pytorch-summary

from torchsummary import summary
summary(your_model, input_size=(channels, H, W))

	# References:
	# https://github.com/NVIDIA/nvidia-docker#usage
	# http://peterforgacs.github.io/2017/04/18/Docker-basics/
	#
	docker run --gpus '"device=1"' -it --rm -v /home/kuohsin/workspace/tensorrt_lab/torch2trt:/workspace tensorrt

	docker container ls -a
	docker container rm container_id
	sudo docker commit container_id image_name

	import tensorflow.contrib.slim as slim

	model_vars = tf.trainable_variables()
	slim.model_analyzer.analyze_vars(model_vars, print_info=True)
	slim.model_analyzer.analyze_ops(tf.compat.v1.get_default_graph(), print_info=True)

	import sys
	import json
	import os
	import numpy as np
	import cv2

	def get_out_filename(input_path):
	dir_path = os.path.dirname(input_path)
	dir_name = os.path.basename(dir_path)
	out_path = dir_name + '_label.txt'

	import sys
	import json
	import os
	import numpy as np
	import cv2

	def get_out_filename(input_path):
	dir_path = os.path.dirname(input_path)
	dir_name = os.path.basename(dir_path)
	out_path = dir_name + '_label.txt'

	# Output one image every second, named out1.png, out2.png, out3.png, etc.
	# The %01d dictates that the ordinal number of each output image will be formatted using 1 digits.
	ffmpeg -i input.mov -vf fps=1 out%d.png

	keys = list(finger_label_dict.keys())
	keys.sort(key=lambda f: int(''.join(filter(str.isdigit, f))))

	name: "CIFAR10_BN"
	layer {
	name: "data"
	type: "Input"
	top: "data"
	input_param { shape: { dim: 1 dim: 3 dim: 32 dim: 32 } }
	}
	layer {
	name: "conv1"
	type: "Convolution"