案例带你学Pytorch（4）——TensorboardX详解（2）

对tensorboardX_demo的试运行及注解

demo参见：https://github.com/lanpa/tensorboardX

这个demo把tensorboard中各个可以利用的模块（标量、图像、文字、图、音频、精度曲线）都使用了，可以作为一个不错的示范案例。

import torch
import torchvision.utils as vutils
import numpy as np
import torchvision.models as models
from torchvision import datasets
from tensorboardX import SummaryWriter


resnet18 = models.resnet18(False)
writer = SummaryWriter()
sample_rate = 44100
freqs = [262, 294, 330, 349, 392, 440, 440, 440, 440, 440, 440]
*for* n_iter in range(100): 
​    dummy_s1 = torch.rand(1)   *#dummy虚拟的, torch.rand()随机数按正态分布进行
​    dummy_s2 = torch.rand(1)
​    *# data grouping by `slash`  #slash斜线*
​    *#添加标量，最终曲线为显示100个随机数*
​    writer.add_scalar('data/scalar1', dummy_s1[0], n_iter)
​    writer.add_scalar('data/scalar2', dummy_s2[0], n_iter) 
*#添加标量组，最终曲线包括了3个函数，xsinx，xcosx，arctanx，*
    writer.add_scalars('data/scalar_group', {'xsinx': n_iter * np.sin(n_iter),
                                         'xcosx': n_iter * np.cos(n_iter),
                                         'arctanx': np.arctan(n_iter)}, n_iter)

dummy_img = torch.rand(32, 3, 64, 64)  *# output from network  假图像*
*#每十次循环进行一次*
if n_iter % 10 == 0:
    x = vutils.make_grid(dummy_img, *normalize*=True, *scale_each*=True)
    writer.add_image('Image', x, n_iter)  *#添加图像*
    dummy_audio = torch.zeros(sample_rate * 2)

从step 0到step 90，共添加了10组图像，每一组（batch包含32个图像，图像尺寸为3×64×64，随机生成）：

for i in range(x.size(0)):
    # amplitude of sound should in [-1, 1]
    dummy_audio[i] = np.cos(freqs[n_iter // 10] * np.pi * float(i) / float(sample_rate))
    writer.add_audio('myAudio', dummy_audio, n_iter, *sample_rate*=sample_rate)

writer.add_text('Text', 'text logged at step:' + str(n_iter), n_iter)

for name, param in resnet18.named_parameters():
	writer.add_histogram(name, param.clone().cpu().data.numpy(), n_iter)

writer.add_pr_curve('xoxo', np.random.randint(2, *size*=100), np.random.rand(100), n_iter)
dataset = datasets.MNIST('mnist', *train*=False, *download*=False)
images = dataset.test_data[:100].float()
label = dataset.test_labels[:100]
features = images.view(100, 784)
writer.add_embedding(features, *metadata*=label, *label_img*=images.unsqueeze(1))
writer.close()

运行命令：

tensorboard --logdir=runs --host localhost

open TensorBoard 1.13.1 at http://localhost:6006

仍会持续完善更新，敬请关注！

建议阅读：https://tensorboard-pytorch.readthedocs.io/en/latest/tutorial.html