Source code for end-to-end dialogue model from the MultiWOZ paper (Budzianowski et al. 2018, EMNLP)
Branch: master
Clone or download
Latest commit 6c7e807 Nov 1, 2018
Type Name Latest commit message Commit time
Failed to load latest commit information.
db initial commit Oct 1, 2018
model evaluation problem fix #1 Oct 17, 2018
utils Refactoring and updating docs. Oct 5, 2018 Update Oct 5, 2018 initial commit Oct 1, 2018 fix test file, add tqdm for delexicalization Oct 17, 2018
requirements.txt initial commit Oct 1, 2018 Update Oct 21, 2018 Update Nov 1, 2018


multiwoz is an open source toolkit for building end-to-end trainable task-oriented dialogue models. It is released by Paweł Budzianowski from Cambridge Dialogue Systems Group under Apache License 2.0.


Python 2 with pip

Quick start

In repo directory:


To download and pre-process the data run:



To train the model run:

python [--args=value]

Some of these args include:

// hyperparamters for model learning
--max_epochs        : numbers of epochs
--batch_size        : numbers of turns per batch
--lr_rate           : initial learning rate
--clip              : size of clipping
--l2_norm           : l2-regularization weight
--dropout           : dropout rate
--optim             : optimization method

// network structure
--emb_size          : word vectors emedding size
--use_attn          : whether to use attention
--hid_size_enc      : size of RNN hidden cell
--hid_size_pol      : size of policy hidden output
--hid_size_dec      : size of RNN hidden cell
--cell_type         : specify RNN type


To evaluate the run:

python [--args=value]

Benchmark results

The following benchmark results were produced by this software. We ran a small grid search over various hyperparameter settings and reported the performance of the best model on the test set. The selection criterion was 0.5match + 0.5success+100*BLEU on the validation set. The final parameters were:

// hyperparamters for model learning
--max_epochs        : 20
--batch_size        : 64
--lr_rate           : 0.005
--clip              : 5.0
--l2_norm           : 0.00001
--dropout           : 0.0
--optim             : Adam

// network structure
--emb_size          : 50
--use_attn          : True
--hid_size_enc      : 150
--hid_size_pol      : 150
--hid_size_dec      : 150
--cell_type         : lstm


If you use any source codes or datasets included in this toolkit in your work, please cite the corresponding papers. The bibtex are listed below:

[Budzianowski et al. 2018]
    Author = {Budzianowski, Pawe{\l} and Wen, Tsung-Hsien and Tseng, Bo-Hsiang  and Casanueva, I{\~n}igo and Ultes Stefan and Ramadan Osman and Ga{\v{s}}i\'c, Milica},
    title={MultiWOZ - A Large-Scale Multi-Domain Wizard-of-Oz Dataset for Task-Oriented Dialogue Modelling},
    booktitle={Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing (EMNLP)},

[Ramadan et al. 2018]
  title={Large-Scale Multi-Domain Belief Tracking with Knowledge Sharing},
  author={Ramadan, Osman and Budzianowski, Pawe{\l} and Gasic, Milica},
  booktitle={Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics},

Bug Report

If you have found any bugs in the code, please contact: pfb30 at cam dot ac dot uk