pytorch使用tensorboard进行模型可视化

1.配置好tensorboardx和tensorflow

pip install tensorflow

pip install tensorboardx

2.调用tensorboardx

from tensorboardX import SummaryWriter

writer = SummaryWriter(log_dir='logs',flush_secs=25)

 

参数：

       （1）log_dir为需要即将写入的文件夹，tensorboardx会自动帮我们创建，无需手动创建

       （2）flush_secs为每隔多长时间对tensorboard中的界面进行更新

 

例1：loss可视化

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

writer = SummaryWriter()

for n_iter in range(100):
    writer.add_scalar('Loss/train', np.random.random(), n_iter)#第一个参数表示纵坐标，第二个参数表示图表中的折线，即loss大小，第三个参数表示横坐标的值
    writer.add_scalar('Loss/test', np.random.random(), n_iter)
    writer.add_scalar('Accuracy/train', np.random.random(), n_iter)
    writer.add_scalar('Accuracy/test', np.random.random(), n_iter)然后在命令行中输入tensorboard --logdir=runs，得到地址之后在浏览器中打开，则可看到如下界面：

from torch.utils.tensorboard import SummaryWriter
writer = SummaryWriter()
r = 5
for i in range(100):
    writer.add_scalars('run_14h', {'xsinx':i*np.sin(i/r),
                                    'xcosx':i*np.cos(i/r),
                                    'tanx': np.tan(i/r)}, i)#在同一张折线图下显示多组变量
writer.close()
# This call adds three values to the same scalar plot with the tag
# 'run_14h' in TensorBoard's scalar section.

例2：模型结构可视化

from torchvision.models import resnet50
import torch
model = resnet50(pretrained = False)
inputs = torch.ones(size=(1,3,224,224))

writer = SummaryWriter()
writer.add_graph(model,input_to_model=inputs,verbose=False)#参数解析：model即为你所需要可视化的模型，input_to_model为输入模型中的变量，verbose:是否在控制台打印模型结构

references:

1.https://pytorch.org/docs/stable/tensorboard.html?highlight=tensorboard

2.https://blog.csdn.net/bigbennyguo/article/details/87956434