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

关于度量的数据类型 #9

Open
RabbearSu opened this issue Apr 19, 2019 · 4 comments
Open

关于度量的数据类型 #9

RabbearSu opened this issue Apr 19, 2019 · 4 comments

Comments

@RabbearSu
Copy link

# dtype=[tf.float32] * 2 + [tf.int64] + [tf.float32] + [tf.int64] + [tf.float32] * 3)

在前面ctpn_target_graph的输出中,后三个度量的类型均为tf.float32,可是这里倒数第四个指定类型为tf.int64,是否应为tf.float32呢?

@yizt
Copy link
Owner

yizt commented Apr 19, 2019

@RabbearSu 这个代码已经注释了哦

@RabbearSu
Copy link
Author

@yizt 我看您在tf_utils.py里备注说尽量用原生的tf.map_fn,效率更高

@yizt
Copy link
Owner

yizt commented Apr 19, 2019

@RabbearSu 一般情况下用tf.map_fn是更好一些,我这里加了自定义度量后,还用tf.map_fn遇到了bug,没有解决;所以还是改为用batch_slice了。

@RabbearSu
Copy link
Author

@yizt 原来是这样,我尝试一下能不能改成tf.map_fn

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

2 participants