Use of tensorboard

tensorboard Use （ Common methods and visualization ）

---

Preface

One 、add_scalar() Method

Brief introduction

You can see the source code for add_scalar Introduction to

    def add_scalar(
        self,
        tag,
        scalar_value,
        global_step=None,
        walltime=None,
        new_style=False,
        double_precision=False,
    ):
        """Add scalar data to summary. Args: tag (string): Data identifier scalar_value (float or string/blobname): Value to save global_step (int): Global step value to record walltime (float): Optional override default walltime (time.time()) with seconds after epoch of event new_style (boolean): Whether to use new style (tensor field) or old style (simple_value field). New style could lead to faster data loading. Examples:: from torch.utils.tensorboard import SummaryWriter writer = SummaryWriter() x = range(100) for i in x: writer.add_scalar('y=2x', i * 2, i) writer.close() Expected result: .. image:: _static/img/tensorboard/add_scalar.png :scale: 50 % """

The commonly used parameters are tag,scalar_value,global_step

tag: Equivalent to the title of the chart 
scalar_value: It is equivalent to the vertical coordinate value of the chart （Y Axis ）
global_step: It is equivalent to the abscissa value of the chart （X Axis ）

Simple and practical

#  Import class methods 
from torch.utils.tensorboard import SummaryWriter
#  Generate a writer object , And specify the directory for generating pictures 
writer = SummaryWriter("logs")
for i in range(100):
	#  With y=x Title ,i Is the abscissa value ,i It is also the ordinate value , It is equivalent to drawing a y=x(0<=x<100) The straight line of 
    writer.add_scalar("y=x",i,i)
writer.close()

After running the code, a directory with the specified name will be generated , And there will be an event in it , This event will use tensorboard open .

tensorboard The effect after opening is as follows ：

Open build event

Click on pycharm At the bottom of the terminal window

Input instruction tensorboard --logdir=logs, It will be automatically enabled locally 6006 port , But the problem with this is , When multiple people use the remote host for display operation at the same time , The port is occupied , So we can also specify that the port is open , Inputable instruction tensorboard --logdir=logs --port=6007 Appoint 6007 Or other free ports .

Click the link to see the effect picture ：

Two 、add_image() Method

Brief introduction

You can see the source code for add_image Introduction to

     def add_image(self, tag, img_tensor, global_step=None, walltime=None, dataformats='CHW'):
     """Add image data to summary. Note that this requires the ``pillow`` package. Args: tag (string): Data identifier img_tensor (torch.Tensor, numpy.array, or string/blobname): Image data global_step (int): Global step value to record walltime (float): Optional override default walltime (time.time()) seconds after epoch of event Shape: img_tensor: Default is :math:`(3, H, W)`. You can use ``torchvision.utils.make_grid()`` to convert a batch of tensor into 3xHxW format or call ``add_images`` and let us do the job. Tensor with :math:`(1, H, W)`, :math:`(H, W)`, :math:`(H, W, 3)` is also suitable as long as corresponding ``dataformats`` argument is passed, e.g. ``CHW``, ``HWC``, ``HW``. Examples:: from torch.utils.tensorboard import SummaryWriter import numpy as np img = np.zeros((3, 100, 100)) img[0] = np.arange(0, 10000).reshape(100, 100) / 10000 img[1] = 1 - np.arange(0, 10000).reshape(100, 100) / 10000 img_HWC = np.zeros((100, 100, 3)) img_HWC[:, :, 0] = np.arange(0, 10000).reshape(100, 100) / 10000 img_HWC[:, :, 1] = 1 - np.arange(0, 10000).reshape(100, 100) / 10000 writer = SummaryWriter() writer.add_image('my_image', img, 0) # If you have non-default dimension setting, set the dataformats argument. writer.add_image('my_image_HWC', img_HWC, 0, dataformats='HWC') writer.close() Expected result: .. image:: _static/img/tensorboard/add_image.png :scale: 50 % """

The commonly used parameters are tag,img_tensor,global_step
img_tensor Need to be torch.Tensor, numpy.array, or string/blobname Here we use ndarray To explain , Then convert the image to ndarray You need to use other modules opencv perhaps numpy, No students can install it
opencv Installation instructions pip install opencv-python -i https://pypi.tuna.tsinghua.edu.cn/simple
numpy Installation instructions pip install numpy -i https://pypi.tuna.tsinghua.edu.cn/simple

tag: Equivalent to the title of the chart 
img_tensor: Equivalent to a picture object （torch.Tensor, numpy.array, or string/blobname）
global_step: It's the step length 

Simple and practical

#  Import related modules 
from torch.utils.tensorboard import SummaryWriter
from PIL import Image
import numpy as np
#  Generate a writer object 
writer = SummaryWriter("logs")
image_path = r"D:\pycharm_professional\PycharmProjects\learning_pytorch\dataset\train\ants_image\0013035.jpg"
#  Read the picture 
img_PIL = Image.open(image_path)
#  Convert picture to ndarray type 
img_array = np.array(img_PIL)
writer.add_image("test",img_array,1, dataformats='HWC')
writer.close()

Be careful writer.add_image("test",img_array,1, dataformats='HWC') Need to add dataformats='HWC' among H Represents the height of the picture ,W Represents the width of the picture ,C Represents the number of picture channels . The default order is （C,H,W）, But use np.array() After the transformation, the order becomes (H,W,C), So will dataformat The value of the set HWC. If you don't do this, you will report an error ！ When it comes to how to view this property of a picture object , We can go through pycharm At the bottom of the console Console to execute line by line , View the relevant properties of the image object or directly print(img_array.shape) Check it out. .

After running the code, a directory with the specified name will be generated , And there will be an event in it , This event will use tensorboard open .

tensorboard The effect after opening is as follows ：

Open build event

Click on pycharm At the bottom of the terminal window

Input instruction tensorboard --logdir=logs, It will be automatically enabled locally 6006 port , But the problem with this is , When multiple people use the remote host for display operation at the same time , The port is occupied , So we can also specify that the port is open , Inputable instruction tensorboard --logdir=logs --port=6007 Appoint 6007 Or other free ports .

Click the link to see the effect picture ：