carlos9310 / pointer-generator

resources for the paper 'Get To The Point: Summarization with Pointer-Generator Networks' with python3.x. overview on the post http://www.abigailsee.com/2017/04/16/taming-rnns-for-better-summarization.html or

Organized resources for Get To The Point: Summarization with Pointer-Generator Networks with python3.x

DataSet

参考https://github.com/abisee/cnn-dailymail

raw corpus

• cnn_stories.tgz

• dailymail_stories.tgz

command for data preprocessing

python data/make_datafiles.py /path/to/unzipped/cnn_stories.tgz /path/to/unzipped/dailymail_stories.tgz

generate data for model

• finished_files.rar

Code with python3

参考https://github.com/abisee/pointer-generator

Run training

python run_summarization.py --mode=train --data_path=/path/to/chunked/train_* --vocab_path=/path/to/vocab --log_root=/path/to/a/log/directory --exp_name=myexperiment

Run eval

python run_summarization.py --mode=eval --data_path=/path/to/chunked/val_* --vocab_path=/path/to/vocab --log_root=/path/to/a/log/directory --exp_name=myexperiment

Run beam search decoding

python run_summarization.py --mode=decode --data_path=/path/to/chunked/val_* --vocab_path=/path/to/vocab --log_root=/path/to/a/log/directory --exp_name=myexperiment