Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

测试效果很差 #1

Open
luckydog5 opened this issue Mar 29, 2019 · 9 comments
Open

测试效果很差 #1

luckydog5 opened this issue Mar 29, 2019 · 9 comments

Comments

@luckydog5
Copy link

luckydog5 commented Mar 29, 2019

目前跑完2epochs,但是测试时的效果非常差,测了13张,模型输出全是错的。尤其是长序列,单词数目较多(大于2时),模型的输出非常差。

@koibiki
Copy link
Owner

koibiki commented Apr 5, 2019

我测了下mjsynth的valid数据,大概准确率在0.9左右。这个代码是eager模式写的,训练和测试必须在同种设备上运行,另外记着改一下 os.environ["CUDA_VISIBLE_DEVICES"] = "1" 为自己的显卡号

@luckydog5
Copy link
Author

主要是长序列问题:7,8个单词的那种,如果图中单词数目小于5个效果还行,在0.9左右,长一点的就不行,在训练时效果acc效果都很差。。。

@koibiki
Copy link
Owner

koibiki commented Apr 5, 2019

不过长序列确实没用了, 单词数量大于2就基本全错了,只能用来检测单个单词,另外有下划线也会导致全错

@koibiki
Copy link
Owner

koibiki commented Apr 5, 2019

你可以试一下把attention换成 self attention ,那样对长序列效果就会好很多,而且也不用用rnn单元了,模型也会小很多

@luckydog5
Copy link
Author

ok, 🙏多谢。

@lmpan
Copy link

lmpan commented Jun 6, 2019

先用水平投影做单词分割再识别呢?看看效果如何

@Lanme
Copy link

Lanme commented Jun 26, 2019

建议这个加中文识别哦

@HIHIHAHEI
Copy link

File "E:\anaconda3\Anaconda3\envs\py_36\lib\site-packages\tensorflow\python\keras\engine\network.py", line 854, in _run_internal_graph
for x, y, mask in zip(self.inputs, inputs, masks):
TypeError: zip argument #1 must support iteration

你好,我运行测试的时候报错,这个怎么改

@DreamRiverForever
Copy link

eager保存模型不通用有点坑,CPU不是要跑炸

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

6 participants