【pytorch】tensorboard的使用

一、tensorboard的说明

TensorBoard是深度学习领域通用的可视化工具套件，TensorBoard的核心价值在于将抽象的模型训练过程转化为直观的可视化图表。

在PyTorch中，SummaryWriter是TensorBoard的数据写入器，用于将训练数据（指标、模型结构、图像等）记录到事件文件（event files）中。

python 复制代码

class SummaryWriter(object):
    """Writes entries directly to event files in the log_dir to be
    consumed by TensorBoard.

    The `SummaryWriter` class provides a high-level API to create an event file
    in a given directory and add summaries and events to it. The class updates the
    file contents asynchronously. This allows a training program to call methods
    to add data to the file directly from the training loop, without slowing down
    training.
    """

记录标量（指标）：使用add_scalar方法跟踪损失、准确率等指标。

记录图像：使用add_image方法展示训练数据或生成样本。

二、tensorboard的画图使用

add_scalar:scalar_value是y轴，global_step是x轴

创建y=x和y=2x的图，并通过网页展示

python 复制代码

from torch.utils.tensorboard import SummaryWriter

writer = SummaryWriter('logs')
# writer.add_image()
for i in range(100):
    writer.add_scalar('y=2x', 2*i, i)

writer.close()

在控制台输出，也可以指定其他端口tensorboard --logdir=logs --port=6007：

powershell 复制代码

tensorboard --logdir=logs

输出

浏览器打开http://localhost:6006/

三、tensorboard的记录图片使用

python 复制代码

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

writer = SummaryWriter('logs')
image_path = r'data/train/ants_image/0013035.jpg'
img_PIL = Image.open(image_path)
img_array = np.array(img_PIL)
#img_array的shape是(512, 768, 3)，通道3，需要加dataformats='HWC'参数，否则报错
writer.add_image("test", img_array, 1, dataformats='HWC')


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

writer.close()

输出