Wang Zichen
Wang Zichen
And I use the pre-trained model on my data, there are always several boxes overlapped on one body, how can I fix this , is it related to the threshold...
Thxs for sharing the codes. However the experiments with Gaussian kernel seems terrible.  this is with style weight 2.0.
`self.dc = InnerProductDecoder(dropout, act=lambda x: x)` Why not use `act=torch.sigmoid`here?
你好，我用自己的语料库（短文本）做sentiment analysis,frequency.py报了如下错 ~\AppData\Local\Continuum\anaconda3\lib\site-packages\snownlp\utils\frequency.py in freq(self, key) 22 23 def freq(self, key): ---> 24 return float(self.get(key)[1])/self.total 25 26 def samples(self): ZeroDivisionError: float division by zero 这是什么情况呢？
Thx for sharing this interesting demo. I find that with the increasing number of faces, the step of adding a transparent mask (png) to the original photo consuming much more...
Thxs for this code. I have some questions for preprocessing logically. 1. For the 'author_features.txt' output, why there's a record with document_id-author_order,and why to remove the name according to the...