IvanB
IvanB
I had a similar issue, though not sure if it's the same cause. See: https://github.com/abisee/cnn-dailymail/issues/12 On Sun, Oct 29, 2017 at 8:29 PM, Jaffer Wilson wrote: > Please let me...
If you have txt files on your computer that you want to summarize, these issues may help: https://github.com/abisee/cnn-dailymail/issues/12 https://github.com/abisee/pointer-generator/issues/51 Good luck!
any updates from the authors on this issue? i.e. how BCCN is implemented?
Here is how to make a dictionary with keys = words the model was trained on and values = glove representation of the words, *in one line* Suppose model =...
One way is to make a dictionary where keys = words the model was trained on and values = glove representation of the words, i.e. `model = fitted Glove() model`...
@czygithub: try: http://www.huffingtonpost.com/archive/ e.g. http://www.huffingtonpost.com/archive/2017-3-2
@udibr could you share how to access your specific training and test data so that we may attempt to reproduce your results? This would be very helpful (and scientific!) Note...
i've seen similar behavior in the pre-training stage. just curious: are the above results for pre-training? one thing that tended to help avoid pre-training plateaus was increasing batch size. if...