From 28ddc0c17edce74ece3f6e2deee1db074c74c59e Mon Sep 17 00:00:00 2001
From: Yige Xu <xuyige1996@gmail.com>
Date: Thu, 12 Dec 2019 17:00:48 +0800
Subject: [PATCH] =?UTF-8?q?[update]=E5=AF=B9=E6=9C=80=E8=BF=91=E5=87=A0?=
 =?UTF-8?q?=E4=B8=AAissue=E7=9A=84=E6=9B=B4=E6=96=B0?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 fastNLP/core/dataset.py              | 5 +++++
 fastNLP/core/vocabulary.py           | 2 +-
 fastNLP/modules/encoder/attention.py | 2 +-
 3 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/fastNLP/core/dataset.py b/fastNLP/core/dataset.py
index 12963616..0a24ab22 100644
--- a/fastNLP/core/dataset.py
+++ b/fastNLP/core/dataset.py
@@ -908,12 +908,17 @@ class DataSet(object):
         :param bool shuffle: 在split前是否shuffle一下
         :return: [ :class:`~fastNLP.读取后的DataSet` , :class:`~fastNLP.读取后的DataSet` ]
         """
+        assert len(self) > 1, f'DataSet with {len(self)} instance cannot be split.'
         assert isinstance(ratio, float)
         assert 0 < ratio < 1
         all_indices = [_ for _ in range(len(self))]
         if shuffle:
             np.random.shuffle(all_indices)
         split = int(ratio * len(self))
+        if split == 0:
+            error_msg = f'Dev DataSet has {split} instance after split.'
+            logger.error(error_msg)
+            raise IndexError(error_msg)
         dev_indices = all_indices[:split]
         train_indices = all_indices[split:]
         dev_set = DataSet()
diff --git a/fastNLP/core/vocabulary.py b/fastNLP/core/vocabulary.py
index 6d530eb6..3456061f 100644
--- a/fastNLP/core/vocabulary.py
+++ b/fastNLP/core/vocabulary.py
@@ -209,7 +209,7 @@ class Vocabulary(object):
             self._word2idx = {}
             if self.padding is not None:
                 self._word2idx[self.padding] = len(self._word2idx)
-            if self.unknown is not None:
+            if (self.unknown is not None) and (self.unknown != self.padding):
                 self._word2idx[self.unknown] = len(self._word2idx)
         
         max_size = min(self.max_size, len(self.word_count)) if self.max_size else None
diff --git a/fastNLP/modules/encoder/attention.py b/fastNLP/modules/encoder/attention.py
index b48be579..49a860bd 100644
--- a/fastNLP/modules/encoder/attention.py
+++ b/fastNLP/modules/encoder/attention.py
@@ -26,7 +26,7 @@ class DotAttention(nn.Module):
         self.value_size = value_size
         self.scale = math.sqrt(key_size)
         self.drop = nn.Dropout(dropout)
-        self.softmax = nn.Softmax(dim=2)
+        self.softmax = nn.Softmax(dim=-1)
 
     def forward(self, Q, K, V, mask_out=None):
         """