Skip to content

Latest commit

 

History

History
12 lines (8 loc) · 635 Bytes

README.md

File metadata and controls

12 lines (8 loc) · 635 Bytes

sentence_generation

generate the response for out of domain query

  1. QA_generation.py file help to generate question answer pair save in pickle format
  2. generate_data.py generate dataset where pass parameter -t=train/test and -n=number of QA pair want
  3. word_tokernized.py help its create source and target vocabulary and clean the data
  4. reader.py used for prepare data for model
  5. try_sqs_with_attension.py create seq2seqmodel where decoder use multiple LSTMCell(either GRUCell) for encoder and dencoder. Also,Apply attension and beam search for decoding
  6. model.py used simple encoder decoder method with beam search