更新了 ControlC 的 API

README.md

@@ -61,7 +61,8 @@ python -m spacy download en
 ### 扩展教程
 
 - [Extend-1. BertEmbedding的各种用法](https://fastnlp.readthedocs.io/zh/latest/tutorials/extend_1_bert_embedding.html)
-- [Extend-2. 使用fitlog 辅助 fastNLP 进行科研](https://fastnlp.readthedocs.io/zh/latest/tutorials/extend_2_fitlog.html)
+- [Extend-2. 分布式训练简介](https://fastnlp.readthedocs.io/zh/latest/tutorials/extend_2_dist.html)
+- [Extend-3. 使用fitlog 辅助 fastNLP 进行科研](https://fastnlp.readthedocs.io/zh/latest/tutorials/extend_3_fitlog.html)
 
 
 ## 内置组件

docs/source/index.rst

@@ -35,6 +35,9 @@ API 文档
    
    fastNLP
 
+
+:doc:`API变动列表 </user/api_update>`
+
 fitlog文档
 ----------
 

docs/source/user/api_update.rst

@@ -0,0 +1,15 @@
+===========================
+API变动列表
+===========================
+
+2020.4.14
+========================
+
+修改了 :class:`fastNLP.core.callback.ControlC` 的 API。
+
+原来的参数 ``quit_all`` 修改为 ``quit_and_do`` ，仍然接收一个 bool 值。新增可选参数 ``action`` ，接收一个待执行的函数，
+在 ``quit_and_do`` 的值为 ``True`` 时，退出训练过程后执行该函数。 ``action`` 的默认值是退出整个程序，与原有功能一致。
+
+.. note::
+    原有用法 `ControlC(True)` 和 `ControlC(False)` 均可以继续正确执行，但 `ControlC(quit_all=True/False)` 需要修改为
+    `ControlC(quit_and_do=True/False)`  。

fastNLP/core/callback.py

@@ -695,20 +695,26 @@ class ControlC(Callback):
     检测到 control+C 时的反馈
     """
     
-    def __init__(self, quit_all):
+    @staticmethod
+    def quit_all():
+        import sys
+        sys.exit(0)  # 直接退出程序
+    
+    def __init__(self, quit_and_do, action=quit_all):
         r"""
-        :param bool quit_all: 若为True,则检测到control+C 直接退出程序；否则只退出Trainer
+        :param bool quit_and_do: 若为True,则检测到control+C 进行后续操作（默认值为：直接退出程序）；否则只退出Trainer。
         """
+        
         super(ControlC, self).__init__()
-        if type(quit_all) != bool:
-            raise ValueError("In KeyBoardInterrupt, quit_all arguemnt must be a bool.")
-        self.quit_all = quit_all
+        if type(quit_and_do) != bool:
+            raise ValueError("In KeyBoardInterrupt, quit_and_do arguemnt must be a bool.")
+        self.quit_and_do = quit_and_do
+        self.action = action
     
     def on_exception(self, exception):
         if isinstance(exception, KeyboardInterrupt):
-            if self.quit_all is True:
-                import sys
-                sys.exit(0)  # 直接退出程序
+            if self.quit_and_do is True:
+                self.action()
             else:
                 pass
         else:

test/core/test_callbacks.py

@@ -12,7 +12,7 @@ from fastNLP import Instance
 from fastNLP import SGD
 from fastNLP import Trainer
 from fastNLP.core.callback import EarlyStopCallback, GradientClipCallback, LRScheduler, ControlC, \
-    LRFinder, TensorboardCallback
+    LRFinder, TensorboardCallback, Callback
 from fastNLP.core.callback import EvaluateCallback, FitlogCallback, SaveModelCallback
 from fastNLP.core.callback import WarmupCallback
 from fastNLP.models.base_model import NaiveClassifier
@@ -225,39 +225,32 @@ class TestCallback(unittest.TestCase):
                           metrics=AccuracyMetric(pred="predict", target="y"), use_tqdm=True,
                           callbacks=EarlyStopCallback(1), check_code_level=2)
         trainer.train()
-
-@unittest.skipIf('TRAVIS' in os.environ, "Skip in travis")
-def test_control_C():
-    # 用于测试 ControlC , 再两次训练时用 Control+C 进行退出，如果最后不显示 "Test failed!" 则通过测试
-    from fastNLP import ControlC, Callback
-    import time
-
-    line1 = "\n\n\n\n\n*************************"
-    line2 = "*************************\n\n\n\n\n"
-
-    class Wait(Callback):
-        def on_epoch_end(self):
-            time.sleep(5)
-
-    data_set, model = prepare_env()
-
-    print(line1 + "Test starts!" + line2)
-    trainer = Trainer(data_set, model, optimizer=SGD(lr=0.1), loss=BCELoss(pred="predict", target="y"),
-                      batch_size=32, n_epochs=20, dev_data=data_set,
-                      metrics=AccuracyMetric(pred="predict", target="y"), use_tqdm=True,
-                      callbacks=[Wait(), ControlC(False)], check_code_level=2)
-    trainer.train()
-
-    print(line1 + "Program goes on ..." + line2)
-
-    trainer = Trainer(data_set, model, optimizer=SGD(lr=0.1), loss=BCELoss(pred="predict", target="y"),
-                      batch_size=32, n_epochs=20, dev_data=data_set,
-                      metrics=AccuracyMetric(pred="predict", target="y"), use_tqdm=True,
-                      callbacks=[Wait(), ControlC(True)], check_code_level=2)
-    trainer.train()
-
-    print(line1 + "Test failed!" + line2)
-
-
-if __name__ == "__main__":
-    test_control_C()
+    
+    def test_control_C_callback(self):
+        
+        class Raise(Callback):
+            def on_epoch_end(self):
+                raise KeyboardInterrupt
+        
+        flags = [False]
+        
+        def set_flag():
+            flags[0] = not flags[0]
+            
+        data_set, model = prepare_env()
+        
+        trainer = Trainer(data_set, model, optimizer=SGD(lr=0.1), loss=BCELoss(pred="predict", target="y"),
+                          batch_size=32, n_epochs=20, dev_data=data_set,
+                          metrics=AccuracyMetric(pred="predict", target="y"), use_tqdm=True,
+                          callbacks=[Raise(), ControlC(False, set_flag)], check_code_level=2)
+        trainer.train()
+        
+        self.assertEqual(flags[0], False)
+        
+        trainer = Trainer(data_set, model, optimizer=SGD(lr=0.1), loss=BCELoss(pred="predict", target="y"),
+                          batch_size=32, n_epochs=20, dev_data=data_set,
+                          metrics=AccuracyMetric(pred="predict", target="y"), use_tqdm=True,
+                          callbacks=[Raise(), ControlC(True, set_flag)], check_code_level=2)
+        trainer.train()
+        
+        self.assertEqual(flags[0], True)