Pytorch学习记录-Pytorch可视化

Pytorch学习记录-Pytorch可视化

在很早很早以前（至少一个半月），我做过几节关于tensorboard的学习记录。

https://www.jianshu.com/p/23205a7921cd
https://www.jianshu.com/p/6235c1ecde67
https://www.jianshu.com/p/2b24454b0629
https://www.jianshu.com/p/0080047e5456

迟迟没有转到Pytorch的原因也是tensorflow的可视化做的好，不过现在Pytorch也支持了，在教程里有，学习一个。
在本教程中，使用简单的神经网络实现MNIST分类器，并使用TensorBoard可视化训练过程。在训练阶段，我们通过scalar_summary绘制损失和准确度函数，并通过image_summary可视化训练图像。此外，我们使用histogram_summary可视化神经网络参数的权重和梯度值。

1. 引入所需库

老一套了，只是增加了logger，logger是用于记录模型

import tensorflow as tf
import numpy as np
import scipy.misc

try:
    from StringIO import StringIO  # Python 2.7
except ImportError:
    from io import BytesIO  # Python 3.x


class Logger(object):
    def __init__(self, log_dir):
        self.writer = tf.summary.FileWriter(log_dir)

    def scalar_summary(self, tag, value, step):
        summary = tf.Summary(value=[tf.Summary.Value(tag=tag, simple_value=value)])
        self.writer.add_summary(summary, step)

    def image_summary(self, tag, images, step):
        # 记录图列表
        img_summaries = []
        for i, img in enumerate(images):
            try:
                s = StringIO()
            except:
                s = BytesIO()
            scipy.misc.toimage(img).save(s, format='png')
            img_sum = tf.Summary.Image(encoded_image_string=s.getvalue(),
                                       height=img.shape[0],
                                       width=img.shape[1])
            img_summaries.append(tf.Summary.Value(tag='%s/%d' % (tag, i), image=img_sum))

        summary = tf.Summary(value=img_summaries)
        self.writer.add_summary(summary, step)

    def histo_summary(self, tag, values, step, bins=1000):
        counts, bin_edges = np.histogram(values, bins=bins)
        hist = tf.HistogramProto()
        hist.min = float(np.min(values))
        hist.max = float(np.max(values))
        hist.num = int(np.prod(values.shape))
        hist.sum = float(np.sum(values))
        hist.sum_squares = float(np.sum(values ** 2))

        # Drop the start of the first bin
        bin_edges = bin_edges[1:]

        # Add bin edges and counts
        for edge in bin_edges:
            hist.bucket_limit.append(edge)
        for c in counts:
            hist.bucket.append(c)

        # Create and write Summary
        summary = tf.Summary(value=[tf.Summary.Value(tag=tag, histo=hist)])
        self.writer.add_summary(summary, step)
        self.writer.flush()

2. 训练模型

import torch
import torch.nn as nn
import torchvision
from torchvision import transforms
from logger import Logger

# Device configuration
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')

# MNIST dataset
dataset = torchvision.datasets.MNIST(root='./data',
                                     train=True,
                                     transform=transforms.ToTensor(),
                                     download=True)

# Data loader
data_loader = torch.utils.data.DataLoader(dataset=dataset,
                                          batch_size=100,
                                          shuffle=True)


# Fully connected neural network with one hidden layer
class NeuralNet(nn.Module):
    def __init__(self, input_size=784, hidden_size=500, num_classes=10):
        super(NeuralNet, self).__init__()
        self.fc1 = nn.Linear(input_size, hidden_size)
        self.relu = nn.ReLU()
        self.fc2 = nn.Linear(hidden_size, num_classes)

    def forward(self, x):
        out = self.fc1(x)
        out = self.relu(out)
        out = self.fc2(out)
        return out


model = NeuralNet().to(device)

logger = Logger('./logs')

# Loss and optimizer
criterion = nn.CrossEntropyLoss()
optimizer = torch.optim.Adam(model.parameters(), lr=0.00001)

data_iter = iter(data_loader)
iter_per_epoch = len(data_loader)
total_step = 50000

# Start training
for step in range(total_step):

    # Reset the data_iter
    if (step + 1) % iter_per_epoch == 0:
        data_iter = iter(data_loader)

    # Fetch images and labels
    images, labels = next(data_iter)
    images, labels = images.view(images.size(0), -1).to(device), labels.to(device)

    # Forward pass
    outputs = model(images)
    loss = criterion(outputs, labels)

    # Backward and optimize
    optimizer.zero_grad()
    loss.backward()
    optimizer.step()

    # Compute accuracy
    _, argmax = torch.max(outputs, 1)
    accuracy = (labels == argmax.squeeze()).float().mean()

    if (step + 1) % 100 == 0:
        print('Step [{}/{}], Loss: {:.4f}, Acc: {:.2f}'
              .format(step + 1, total_step, loss.item(), accuracy.item()))

        # ================================================================== #
        #                        Tensorboard Logging                         #
        # ================================================================== #

        # 1. Log scalar values (scalar summary)
        info = {'loss': loss.item(), 'accuracy': accuracy.item()}

        for tag, value in info.items():
            logger.scalar_summary(tag, value, step + 1)

        # 2. Log values and gradients of the parameters (histogram summary)
        for tag, value in model.named_parameters():
            tag = tag.replace('.', '/')
            logger.histo_summary(tag, value.data.cpu().numpy(), step + 1)
            logger.histo_summary(tag + '/grad', value.grad.data.cpu().numpy(), step + 1)

        # 3. Log training images (image summary)
        info = {'images': images.view(-1, 28, 28)[:10].cpu().numpy()}

        for tag, images in info.items():
            logger.image_summary(tag, images, step + 1)

本来按照教程是可以的，但是好像哪里出了问题，我再看看。
放弃了，换成TensorboardX。