Tensorboard的使用

最新推荐文章于 2024-10-12 17:55:45 发布

小张懒懒

最新推荐文章于 2024-10-12 17:55:45 发布

阅读量391

点赞数 3

分类专栏：深度学习文章标签： python 深度学习 pytorch

本文链接：https://blog.csdn.net/m0_51479543/article/details/136658221

版权

深度学习专栏收录该内容

7 篇文章 0 订阅

订阅专栏

本文介绍了如何在Python环境中安装TensorBoard，设置日志目录，以及如何使用SummaryWriter添加标量数据（如损失函数）和图像数据（如训练结果）。包括了从头开始创建SummaryWriter，添加标量值和图像的详细步骤。

摘要由CSDN通过智能技术生成

安装TensorBoard

虚拟环境终端输入以下命令

pip install tensorboard

打开TensorBoard

tensorboard --logdir=logs
#logdir=事件文件所在文件夹名

tensorboard --logdir=logs --port=6007
#修改端口打开TensorBoard

在这里插入图片描述如果图像出现和上一和函数重复的问题，删掉事件文件，重新打开TensorBoard

SummaryWriter的使用

Writes entries directly to event files in the log_dir to be
consumed by TensorBoard.
直接将条目写入log_dir的事件文件，供 TensorBoard 使用。

初始化

from torch.utils.tensorboard import SummaryWriter

# create a summary writer with automatically generated folder name.
writer = SummaryWriter()
# folder location: runs/May04_22-14-54_s-MacBook-Pro.local/

# create a summary writer using the specified folder name.
writer = SummaryWriter("my_experiment")
# folder location: my_experiment

# create a summary writer with comment appended.
writer = SummaryWriter(comment="LR_0.1_BATCH_16")
# folder location: runs/May04_22-14-54_s-MacBook-Pro.localLR_0.1_BATCH_16/

add_scalar()

Adds many scalar data to summary.
在summary中添加标量数据

add_scalar(self, tag, scalar_value, global_step=None)
Args:
tag (string): Data identifier
scalar_value (float or string/blobname): Value to save 类似于纵坐标
global_step (int): Global step value to record 类似于横坐标

from torch.utils.tensorboard import SummaryWriter
writer = SummaryWriter("logs")
x = range(100)
for i in x:
    writer.add_scalar('y=2x', i * 2, i)
writer.close()

add_images()

Add batched image data to summary.
常用来观察训练结果

add_image(self, tag, img_tensor, global_step=None)
Args:
tag (string): Data identifier
img_tensor (torch.Tensor, numpy.array, or string/blobname): Image data
global_step (int): Global step value to record
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.
eg. dataformats='HWC'

利用Opencv读取图片，获得numpy型图片数据

安装Opencv

pip install opencv-python

利用numpy.arrary()，对PIL图片进行转换

from torch.utils.tensorboard import SummaryWriter
import numpy as np
from PIL import  Image

writer = SummaryWriter("logs")
image_path="data/train/ants_image/0013035.jpg"
img_PIL = Image.open(image_path) #获取PIL类型图片，打开图片
img_array = np.array(img_PIL) #对PIL类型图片进行转换
print(img_array.shape) #输出图片的shape显示是HWC（高度，宽度，通道）类型，需要用dataformats='HWC'转换为HWC类型

writer.add_image("test", img_array, 1, dataformats='HWC')

x = range(100)
for i in x:
    writer.add_scalar('y=2x', i * 2, i)
writer.close()