pytorch璁粌浠ｇ爜_pytorch涓枃绀惧尯?

1.初始化

from torch.utils.tensorboard import SummaryWriter

按住ctrl键，将鼠标移至SummaryWriter，变为蓝色，点击

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

直接向log_dir文件夹写入事件文件，这个事件文件可以被TensorBoard解析

def __init__(self, log_dir=None, comment='', purge_step=None, max_queue=10,
             flush_secs=120, filename_suffix=''):

    Args:
        log_dir (string): Save directory location. Default is
          runs/**CURRENT_DATETIME_HOSTNAME**, which changes after each run.
          Use hierarchical folder structure to compare
          between runs easily. e.g. pass in 'runs/exp1', 'runs/exp2', etc.
          for each new experiment to compare across them.

输入文件夹的名称，若不输入，默认在/**CURRENT_DATETIME_HOSTNAME**

Examples::

    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/

2.创建实例

writer = SummaryWriter("logs") #将事件文件存储至logs文件夹中

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

writer.close()

def add_scalar(self, tag, scalar_value, global_step=None, walltime=None):
    """Add scalar data to summary. 添加一个标量数据至summary

    Args:
        tag (string): Data identifier 图表的标题
        scalar_value (float or string/blobname): Value to save Y轴
        global_step (int): Global step value to record X轴

问题1：运行程序显示ModuleNotFoundError: No module named 'past'

解决方法：在pytorch环境中安装future，pip install future

PyTorch: ModuleNotFoundError: No module named 'past'_RaySunWHUT的博客-CSDN博客

3.安装tensorboard

在pycharm的terminal中输入pip install tensorboard

安装成功后重新运行程序，出现了logs文件夹，其中包含TensorBoard事件文件

4.打开事件文件

在pycharm的terminal中输入tensorboard --logdir=logs（logdir=事件文件所在文件夹名）

默认打开6006端口

指定端口：tensorboard --logdir=logs --port=6007

打开6007端口

问题2：在writer中写入新事件，还有上个事件

解决方法：删除logs文件夹下的所有事件，重新运行程序，在terminal中按ctrl+c退出，再按上键打开端口

1.下载数据集，将其放入项目文件夹中

2.add_image()的使用

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图像的数据类型
        global_step (int): Global step value to record

控制台

image_path ="dataset/train/ants_image/0013035.jpg"
from PIL import Image
img = Image.open(image_path)
print(type(img))