docs(demo): 更新训练日志和图像

Author: zjZSTU

docs/imgs/acc.png

@@ -74,4 +74,35 @@ plt(solver.loss_history)
 plt.multi_plot((solver.train_acc_history, solver.val_acc_history), ('train', 'val'),
                 title='准确率', xlabel='迭代/次', ylabel='准确率')
 print('best_train_acc: %f; best_val_acc: %f' % (solver.best_train_acc, solver.best_val_acc))
-```
+```
+
+训练日志如下
+
+```
+$ python 2_nn_mnist.py 
+epoch: 1 time: 1.08 loss: 0.684346
+train acc: 0.8880; val_acc: 0.8914
+epoch: 2 time: 1.00 loss: 0.337981
+train acc: 0.9131; val_acc: 0.9162
+epoch: 3 time: 1.05 loss: 0.284188
+train acc: 0.9320; val_acc: 0.9320
+epoch: 4 time: 1.01 loss: 0.249591
+train acc: 0.9448; val_acc: 0.9439
+epoch: 5 time: 0.99 loss: 0.227722
+train acc: 0.9528; val_acc: 0.9493
+epoch: 6 time: 0.99 loss: 0.213616
+train acc: 0.9592; val_acc: 0.9547
+epoch: 7 time: 1.00 loss: 0.204579
+train acc: 0.9640; val_acc: 0.9594
+epoch: 8 time: 1.04 loss: 0.198835
+train acc: 0.9685; val_acc: 0.9642
+epoch: 9 time: 1.03 loss: 0.195210
+train acc: 0.9715; val_acc: 0.9669
+epoch: 10 time: 1.01 loss: 0.193020
+train acc: 0.9738; val_acc: 0.9684
+best_train_acc: 0.973767; best_val_acc: 0.968400
+```
+
+![](./imgs/loss.png)
+
+![](./imgs/acc.png)