Skip to content

Latest commit

 

History

History
49 lines (34 loc) · 1.03 KB

README.md

File metadata and controls

49 lines (34 loc) · 1.03 KB
Raw

Knowledge-driven Dialogue

License: MIT

This is a pytorch implementation of generative-based model for knowledge-driven dialogue

Requirements

  • cuda=9.0
  • cudnn=7.0
  • python>=3.6
  • pytorch>=1.0
  • tqdm
  • numpy
  • nltk
  • scikit-learn

Quickstart

Step 1: Preprocess the data

Put the data provided by the organizer under the data folder and rename them train/dev/test.txt:

./data/resource/train.txt
./data/resource/dev.txt
./data/resource/test.txt

Step 2: Train the model

Train model with the following commands.

sh run_train.sh

Step 3: Test the Model

Test model with the following commands.

sh run_test.sh

Note !!!

  • The script run_train.sh/run_test.sh shows all the processes including data processing and model training/testing. Be sure to read it carefully and follow it.
  • The files in ./data and ./model is just empty file to show the structure of the document.