12.21pytorch study notes

1. When reading and displaying images in Python:

from PIL import Image

img_path = ’ image_path ’

img = Image.open(img_path)

img.size # display image size

img.show()

2. Learn to use tensorboard to visualize graphics in pytorch:

from torch.utils.tensorboard import Summarywriter

writer = Summarywriter(‘logs’)

for i in range(100):

writer.add_scalar(‘y=2x’, 2*i, i)

writer.close()

On the terminal:

tensorboard --logdir=logs Generate a graphic URL that displays the above tensorboard generated

3. Learn to use tensorboard to visualize images in pytorch:

from torch.utils.tensorboard import Summarywriter

writer = Summarywriter(‘logs’)

writer.add_image('image name', visualized image)

writer.close()

On the terminal:

tensorboard --logdir=logs Generate image URLs that display the above tensorboard generated images

4.opencv

import cv2

img = cv2.imread(img_path)

cv2.imshow(img)

Added python function usage method:

· Focus on input and output types

·See more official documents

· What parameters are required by the attention method

When the return value is unknown: print();print(type());debug