SummaRuNNer: A Recurrent Neural Network based Sequence Model for Extractive Summarization of Documents.
This is the results I get from my implementation on CNN/dailyMail data set at 275 byte.
| scores | rouge-1 recall | rouge-2 recall | rouge-l recall |
|---|---|---|---|
| lead3 | 43.32 | 19.55 | 35.59 |
| SummaRuNNer | 33.86 | 11.55 | 29.85 |
The training loss is as follows:
The validation loss is as follows:
