Merge pull request #349 from jacquesqiao/01-infer

add inference to 01

01.fit_a_line/README.cn.md

@@ -200,6 +200,11 @@ def event_handler_plot(event):
             cost_ploter.plot()
 
         step += 1
+
+    if isinstance(event, paddle.event.EndPass):
+        if event.pass_id % 10 == 0:
+            with open('params_pass_%d.tar' % event.pass_id, 'w') as f:
+                parameters.to_tar(f)
 ```
 
 ### 开始训练
@@ -217,6 +222,37 @@ trainer.train(
 
 ![png](./image/train_and_test.png)
 
+### 应用模型
+
+#### 1. 生成测试数据
+
+```python
+test_data_creator = paddle.dataset.uci_housing.test()
+test_data = []
+test_label = []
+
+for item in test_data_creator():
+    test_data.append((item[0],))
+    test_label.append(item[1])
+    if len(test_data) == 5:
+        break
+```
+
+#### 2. 推测 inference
+
+```python
+# load parameters from tar file.
+# users can remove the comments and change the model name
+# with open('params_pass_20.tar', 'r') as f:
+#     parameters = paddle.parameters.Parameters.from_tar(f)
+
+probs = paddle.infer(
+    output_layer=y_predict, parameters=parameters, input=test_data)
+
+for i in xrange(len(probs)):
+    print "label=" + str(test_label[i][0]) + ", predict=" + str(probs[i][0])
+```
+
 ## 总结
 在这章里，我们借助波士顿房价这一数据集，介绍了线性回归模型的基本概念，以及如何使用PaddlePaddle实现训练和测试的过程。很多的模型和技巧都是从简单的线性回归模型演化而来，因此弄清楚线性模型的原理和局限非常重要。
 

01.fit_a_line/README.md

@@ -205,6 +205,11 @@ def event_handler_plot(event):
             plot_cost.plot()
 
         step += 1
+
+    if isinstance(event, paddle.event.EndPass):
+        if event.pass_id % 10 == 0:
+            with open('params_pass_%d.tar' % event.pass_id, 'w') as f:
+                parameters.to_tar(f)
 ```
 
 ### Start Training
@@ -222,6 +227,37 @@ trainer.train(
 
 ![png](./image/train_and_test.png)
 
+### Apply model
+
+#### 1. generate testing data
+
+```python
+test_data_creator = paddle.dataset.uci_housing.test()
+test_data = []
+test_label = []
+
+for item in test_data_creator():
+    test_data.append((item[0],))
+    test_label.append(item[1])
+    if len(test_data) == 5:
+        break
+```
+
+#### 2. inference
+
+```python
+# load parameters from tar file.
+# users can remove the comments and change the model name
+# with open('params_pass_20.tar', 'r') as f:
+#     parameters = paddle.parameters.Parameters.from_tar(f)
+
+probs = paddle.infer(
+    output_layer=y_predict, parameters=parameters, input=test_data)
+
+for i in xrange(len(probs)):
+    print "label=" + str(test_label[i][0]) + ", predict=" + str(probs[i][0])
+```
+
 ## Summary
 This chapter introduces *Linear Regression* and how to train and test this model with PaddlePaddle, using the UCI Housing Data Set. Because a large number of more complex models and techniques are derived from linear regression, it is important to understand its underlying theory and limitation.
 

01.fit_a_line/index.cn.html

@@ -242,6 +242,11 @@ def event_handler_plot(event):
             cost_ploter.plot()
 
         step += 1
+
+    if isinstance(event, paddle.event.EndPass):
+        if event.pass_id % 10 == 0:
+            with open('params_pass_%d.tar' % event.pass_id, 'w') as f:
+                parameters.to_tar(f)
 ```
 
 ### 开始训练
@@ -259,6 +264,37 @@ trainer.train(
 
 ![png](./image/train_and_test.png)
 
+### 应用模型
+
+#### 1. 生成测试数据
+
+```python
+test_data_creator = paddle.dataset.uci_housing.test()
+test_data = []
+test_label = []
+
+for item in test_data_creator():
+    test_data.append((item[0],))
+    test_label.append(item[1])
+    if len(test_data) == 5:
+        break
+```
+
+#### 2. 推测 inference
+
+```python
+# load parameters from tar file.
+# users can remove the comments and change the model name
+# with open('params_pass_20.tar', 'r') as f:
+#     parameters = paddle.parameters.Parameters.from_tar(f)
+
+probs = paddle.infer(
+    output_layer=y_predict, parameters=parameters, input=test_data)
+
+for i in xrange(len(probs)):
+    print "label=" + str(test_label[i][0]) + ", predict=" + str(probs[i][0])
+```
+
 ## 总结
 在这章里，我们借助波士顿房价这一数据集，介绍了线性回归模型的基本概念，以及如何使用PaddlePaddle实现训练和测试的过程。很多的模型和技巧都是从简单的线性回归模型演化而来，因此弄清楚线性模型的原理和局限非常重要。
 

01.fit_a_line/index.html

@@ -247,6 +247,11 @@ def event_handler_plot(event):
             plot_cost.plot()
 
         step += 1
+
+    if isinstance(event, paddle.event.EndPass):
+        if event.pass_id % 10 == 0:
+            with open('params_pass_%d.tar' % event.pass_id, 'w') as f:
+                parameters.to_tar(f)
 ```
 
 ### Start Training
@@ -264,6 +269,37 @@ trainer.train(
 
 ![png](./image/train_and_test.png)
 
+### Apply model
+
+#### 1. generate testing data
+
+```python
+test_data_creator = paddle.dataset.uci_housing.test()
+test_data = []
+test_label = []
+
+for item in test_data_creator():
+    test_data.append((item[0],))
+    test_label.append(item[1])
+    if len(test_data) == 5:
+        break
+```
+
+#### 2. inference
+
+```python
+# load parameters from tar file.
+# users can remove the comments and change the model name
+# with open('params_pass_20.tar', 'r') as f:
+#     parameters = paddle.parameters.Parameters.from_tar(f)
+
+probs = paddle.infer(
+    output_layer=y_predict, parameters=parameters, input=test_data)
+
+for i in xrange(len(probs)):
+    print "label=" + str(test_label[i][0]) + ", predict=" + str(probs[i][0])
+```
+
 ## Summary
 This chapter introduces *Linear Regression* and how to train and test this model with PaddlePaddle, using the UCI Housing Data Set. Because a large number of more complex models and techniques are derived from linear regression, it is important to understand its underlying theory and limitation.
 

01.fit_a_line/train.py

@@ -31,6 +31,9 @@ def main():
                     event.pass_id, event.batch_id, event.cost)
 
         if isinstance(event, paddle.event.EndPass):
+            if event.pass_id % 10 == 0:
+                with open('params_pass_%d.tar' % event.pass_id, 'w') as f:
+                    parameters.to_tar(f)
             result = trainer.test(
                 reader=paddle.batch(uci_housing.test(), batch_size=2),
                 feeding=feeding)
@@ -45,6 +48,28 @@ def main():
         event_handler=event_handler,
         num_passes=30)
 
+    # inference
+    test_data_creator = paddle.dataset.uci_housing.test()
+    test_data = []
+    test_label = []
+
+    for item in test_data_creator():
+        test_data.append((item[0], ))
+        test_label.append(item[1])
+        if len(test_data) == 5:
+            break
+
+    # load parameters from tar file.
+    # users can remove the comments and change the model name
+    # with open('params_pass_20.tar', 'r') as f:
+    #     parameters = paddle.parameters.Parameters.from_tar(f)
+
+    probs = paddle.infer(
+        output_layer=y_predict, parameters=parameters, input=test_data)
+
+    for i in xrange(len(probs)):
+        print "label=" + str(test_label[i][0]) + ", predict=" + str(probs[i][0])
+
 
 if __name__ == '__main__':
     main()