Implementation of inference on custom input.

JudeLee19 / HMNet-End-to-End-Abstractive-Summarization-for-Meetings

"End-to-End Abstractive Summarization for Meetings" paper - Unofficial PyTorch Implementation

52 stars 13 forks source link

Implementation of inference on custom input. #4

Open hemangr8 opened 4 years ago

hemangr8 commented 4 years ago

Thanks a lot for the implementation of the paper. Could you please let me know how I can generate the train_corpus for my own data. It seems to be a torch model of some kind and since no explanation is given I am a little stuck on how to use this model to predict for a custom text input of a meeting transcript.

zjxs1997 commented 4 years ago

The 'xx_corpus' files under data directory are all dict and saved by torch module. You can use a = torch.load('data/test_corpus') to get its format. If you want to build your own dataset, you can follow the format to build dict and save it by torch.save(d, 'data/xxx_corpus')