ConvNetJS CIFAR-10 demo 卷积神经网络分类demo
来源:互联网 发布:无法触碰 知乎 编辑:程序博客网 时间:2024/04/29 20:52
ConvNetJS CIFAR-10 demo
Description
This demo trains a Convolutional Neural Network on the CIFAR-10 dataset in your browser, with nothing but Javascript. The state of the art on this dataset is about 90% accuracy and human performance is at about 94% (not perfect as the dataset can be a bit ambiguous). I used this python script to parse the original files (python version) into batches of images that can be easily loaded into page DOM with img tags.
This dataset is more difficult and it takes longer to train a network. Data augmentation includes random flipping and random image shifts by up to 2px horizontally and verically.
By default, in this demo we're using Adadelta which is one of per-parameter adaptive step size methods, so we don't have to worry about changing learning rates or momentum over time. However, I still included the text fields for changing these if you'd like to play around with SGD+Momentum trainer.
Report questions/bugs/suggestions to @karpathy.
Training Stats
Backprop time per example: 24ms
Classification loss: 2.24887
L2 Weight decay loss: 0.00085
Training accuracy: 0.16
Validation accuracy: -1
Examples seen: 141
Momentum:
Batch size:
Weight decay:
Instantiate a Network and Trainer
Network Visualization
max gradient: 0.01467, min: -0.01771
Weight Gradients:
max activation: 1.13915, min: -0.9202
max gradient: 0.0151, min: -0.01904
parameters: 16x5x5x3+16 = 1216
max gradient: 0.01867, min: -0.01904
max activation: 1.13915, min: 0
max gradient: 0.01867, min: -0.01904
()()()()()()()()()()()()()()()()()()()()
Weight Gradients:
()()()()()()()()()()()()()()()()()()()()
max activation: 0.74686, min: -1.94339
max gradient: 0.05836, min: -0.05125
parameters: 20x5x5x16+20 = 8020
max gradient: 0.0788, min: -0.06547
max activation: 0.74686, min: 0
max gradient: 0.0788, min: -0.06547
()()()()()()()()()()()()()()()()()()()()
Weight Gradients:
()()()()()()()()()()()()()()()()()()()()
max activation: 0.46723, min: -0.37553
max gradient: 0.11546, min: -0.15106
parameters: 20x5x5x20+20 = 10020
max gradient: 0.11546, min: -0.15106
max activation: 0.46723, min: 0
max gradient: 0.11546, min: -0.15106
max gradient: 0.11617, min: -0.82716
parameters: 10x320+10 = 3210
max gradient: 0, min: 0
Example predictions on Test set
- ConvNetJS CIFAR-10 demo 卷积神经网络分类demo
- TensorFlow应用之进阶版卷积神经网络CNN在CIFAR-10数据集上分类
- tensorflow 卷积神经网络实现CIFAR-10数据集识别
- 卷积神经网络和CIFAR-10:Yann LeCun专访 Convolutional Nets and CIFAR-10: An Interview with Yann LeCun
- TensorFlow深度学习进阶教程:TensorFlow实现CIFAR-10数据集测试的卷积神经网络
- TensorFlow学习--卷积神经网络训练CIFAR-10数据集
- 卷积神经网络CNN-目标分类
- [4]深度学习和Keras----斯坦福的一个可在线运行的卷积神经网络的Demo
- cifar-10 分类 tensorflow 代码
- mahout分类简单demo
- TOJ题目分类demo
- 输入标签分类demo
- 机器学习-神经网络简单Demo
- Demo
- demo
- demo
- demo
- demo
- The CIFAR-10 dataset 数据集
- 虚拟机如何设置外网ip
- mybatis数据库增删改查操作
- 使用py2exe将python脚本整体打包为一个exe的方法
- java4
- ConvNetJS CIFAR-10 demo 卷积神经网络分类demo
- Android Studio代码着色插件
- java线程之间通讯
- Composer安装说明
- HDU5692 Snacks(dfs序 + 线段树)
- 使用eclipse开发servlet
- 浅谈Servlet的页面跳转
- [Android] CircleView 仿小米空气净化器首页旋转UI
- 博客评论时间显示算法