Taking too long to train for the same train data provided here

hengluchang / deep-news-summarization

News summarization using sequence to sequence model with attention in TensorFlow.

MIT License

185 stars 61 forks source link

Taking too long to train for the same train data provided here #7

Open suryabh opened 6 years ago

suryabh commented 6 years ago

I am having 16GB and lenovo laptop with the latest processor. Looks like the training is too long ( as of now more than 2 hours ). Is there a way to speed up? I mean can I install tensor flow GPU version? If yes, what is the change you suggest in requirements.txt