A Abstractive Summarization Implementation with Transformer and Pointer-generator
A Abstractive Summarization Implementation with Transformer and Pointer-generator
关于copy部分的问题
如何设置使用GPU训练啊
Do you consider the introduction of the pre-trained model for embedding，such as bert?
would you plz share the materials which shows that coverage mechanism doesn't suit short summary, thx
All the generated summaries are having only <unk> words. Please help
would u like to explain the batch beam search for transformer in details?
some tricks or questions for training model
"InvalidArgumentError (see above for traceback): indices[6,25] = 25 is not in [0, 25)"