Mnist-recognition-By-Slim利用Slim训练手写数据集神经网络

Mnist recognition model using tensorflow based on Slim framework.

使用方法

你需要先利用TFrecord_write.py去生成数据集
You need to use TFrecord_write.py to generate dataset first

在生成数据集前，你需要在根目录创建record文件夹
Befor generating dataset, you need to create a record folder in the root directory.

生成数据集后，可以运行cnn.py开始训练
After generating the dataset, you can run the CNN.py to start training.

完成训练后，可以通过Load_Predict.py预测
After completing the training, you can predict by load_predict.py

已经存储好的tfrecords可以通过百度云下载：

You can download the tfrecords through Baidu Cloud.
链接：https://pan.baidu.com/s/191Xb0WFd2HQ0KII-kpriPg 提取码：7mh0