DONGJUN LEE

Could you write in English? I don't understand Chinese.

안녕하세요. 메일 주셔서 감사합니다. tf.device는 특정 디바이스에서 연산이 실행되도록 설정하는 함수입니다. Tensorflow 몇몇 연산은 gpu에서 동작하지 않는데, 그 중 하나인 embedding_lookup을 사용하기 때문에 해당 부분을 gpu로 바꾸면 에러가 나는 것으로...

Which tensorflow version are you using?

Hi @tianke0711 , it seems the error occurs because of your tensorflow version. Could you please try after upgrading tensorflow to >=1.8? My e-mail is [email protected]. Thank you.

Which tensorflow version are you using?

Hi, sorry for the late reply. In this model, the search space of the output is the whole vocabulary. You might change the search space of the output to the...

Hi, I used Gigaword data available at [here](https://drive.google.com/open?id=0B6N7tANPyVeBNmlSX19Ld2xDU1E).

Do you mean 85 batches? To train 85 epochs, it will take too much time. I trained two epochs to generate sample output as [here](https://github.com/dongjun-Lee/text-summarization-tensorflow#sample-summary-output). For example, ``` python train.py...

Hi, I haven't trained the model with CNN dataset :(

Hi. I think what you want to do is implemented at [test.py](https://github.com/dongjun-Lee/text-summarization-tensorflow/blob/master/test.py). When you run the [train.py](https://github.com/dongjun-Lee/text-summarization-tensorflow/blob/master/train.py), the model is saved at every epoch. [test.py](https://github.com/dongjun-Lee/text-summarization-tensorflow/blob/master/test.py) loads the last saved model...