Dialogue Bot for goal-oriented task

There are available two pretrained models for DSTC2 dataset (English). Try them by running:

from deeppavlov import build_model, configs

bot1 = build_model(configs.go_bot.gobot_dstc2, download=True)

bot1(['hi, i want restaurant in the cheap pricerange'])

bot2 = build_model(configs.go_bot.gobot_dstc2_best, download=True)

bot2(['hi, i want chinese restaurant'])

If some required packages are missing, install all the requirements by running in command line:

python -m deeppavlov install gobot_dstc2


The dialogue bot is based on [1] which introduces Hybrid Code Networks (HCNs) that combine an RNN with domain-specific knowledge and system action templates.

alt text Diagram 1. Dotted lines correspond to unrequired (optional) modules, black squares to trained models, trapezes are modules that depend on a dataset and must be provided by software developer.

Here is a simple example of interaction with a trained dialogue bot (available for download):


Models that rely on fasttext word embeddings will require 8.5 GB of disk space, those that use only one-hot encoding of words will require less than 300Mb of disk space.

>> Hello, welcome to the Cambridge restaurant system. You can ask for restaurants by area, price range or food type. How may I help you?
x::hi, i want cheap food
2018-05-16 12:28:48.459 INFO in 'deeppavlov.models.go_bot.bot'['bot'] at line 346: Made api_call with {'pricerange': 'cheap'}, got 22 results.
>> The lucky star serves chinese food in the cheap price range.
x::is there anything with french food?
2018-05-16 12:29:02.795 INFO in 'deeppavlov.models.go_bot.bot'['bot'] at line 346: Made api_call with {'pricerange': 'cheap', 'food': 'french'}, got 0 results.
>> Sorry there is no french restaurant in the cheap price range.
x::ok, maybe there is something in expensive pricerange?
2018-05-16 12:29:23.337 INFO in 'deeppavlov.models.go_bot.bot'['bot'] at line 346: Made api_call with {'pricerange': 'expensive', 'food': 'french'}, got 2 results.
>> Cote serves french food in the expensive price range.
x::in what part of town is it?
2018-05-16 12:29:31.193 INFO in 'deeppavlov.models.go_bot.bot'['bot'] at line 346: Made api_call with {'pricerange': 'expensive', 'food': 'french'}, got 2 results.
>> Restaurant two two is a nice place in the north of town serving tasty french food.
x::ok, give me their phone number
>> The phone number of restaurant two two is 01223 351880.
x::address and post code too
>> Restaurant two two is on 22 chesterton road chesterton, c.b 4, 3 a.x.
x::ok, thank you!
>> Restaurant two two serves french food in the expensive price range.
x::and if i wanted an italian restuarant?
2018-05-16 12:30:15.98 INFO in 'deeppavlov.models.go_bot.bot'['bot'] at line 346: Made api_call with {'pricerange': 'expensive', 'food': 'italian'}, got 4 results.
>> Frankie and bennys serves italian food in the expensive price range.
x::more cheaply priced?
2018-05-16 12:30:36.560 INFO in 'deeppavlov.models.go_bot.bot'['bot'] at line 346: Made api_call with {'pricerange': 'cheap', 'food': 'italian'}, got 5 results.
>> Da vinci pizzeria serves italian food in the cheap price range.
x::ok, give me their phone number and address, please
>> The phone number of da vinci pizzeria is 01223 351707 and it is on 20 milton road chesterton.
>> Da vinci pizzeria serves italian food in the cheap price range.
>> You are welcome!



TO TRAIN a go_bot model you should have:

  1. (optional, but recommended) pretrained named entity recognition model (NER)

  2. (optional, but recommended) pretrained intents classifier model

  3. (optional) any sentence (word) embeddings for english

TO INFER from a go_bot model you should additionally have:

  1. pretrained vocabulary of dataset utterance tokens
    • it is trained in the same config as go_bot model
  2. pretrained goal-oriented bot model
    • config configs/go_bot/gobot_dstc2.json is recommended
    • slot_filler section of go_bot’s config should match NER’s configuration
    • intent_classifier section of go_bot’s config should match classifier’s configuration


For a working exemplary config see configs/go_bot/gobot_dstc2.json (model without embeddings).

A minimal model without slot_filler, intent_classifier and embedder is configured in configs/go_bot/gobot_dstc2_minimal.json.

The best state-of-the-art model (with attention mechanism, relies on embedder and does not use bag-of-words) is configured in configs/go_bot/gobot_dstc2_best.json.

Usage example

To interact with a pretrained go_bot model using commandline run:

python -m deeppavlov interact <path_to_config> [-d]

where <path_to_config> is one of the provided config files.

You can also train your own model by running:

python -m deeppavlov train <path_to_config> [-d]

The -d parameter downloads

  • data required to train your model (embeddings, etc.);
  • a pretrained model if available (provided not for all configs).

Pretrained for DSTC2 models are available for

After downloading required files you can use the configs in your python code. To infer from a pretrained model with config path equal to <path_to_config>:

from deeppavlov import build_model

CONFIG_PATH = '<path_to_config>'
model = build_model(CONFIG_PATH)

utterance = ""
while utterance != 'exit':
    print(">> " + model([utterance])[0])
    utterance = input(':: ')

Config parameters

To configure your own pipelines that contain a "go_bot" component, refer to documentation for GoalOrientedBot and GoalOrientedBotNetwork classes.



The Hybrid Code Network model was trained and evaluated on a modification of a dataset from Dialogue State Tracking Challenge 2 [2]. The modifications were as follows:

  • new turns with api calls
    • added api_calls to restaurant database (example: {"text": "api_call area=\"south\" food=\"dontcare\" pricerange=\"cheap\"", "dialog_acts": ["api_call"]})
  • new actions
    • bot dialog actions were concatenated into one action (example: {"dialog_acts": ["ask", "request"]} -> {"dialog_acts": ["ask_request"]})
    • if a slot key was associated with the dialog action, the new act was a concatenation of an act and a slot key (example: {"dialog_acts": ["ask"], "slot_vals": ["area"]} -> {"dialog_acts": ["ask_area"]})
  • new train/dev/test split
    • original dstc2 consisted of three different MDP policies, the original train and dev datasets (consisting of two policies) were merged and randomly split into train/dev/test
  • minor fixes
    • fixed several dialogs, where actions were wrongly annotated
    • uppercased first letter of bot responses
    • unified punctuation for bot responses

See deeppavlov.dataset_readers.dstc2_reader.DSTC2DatasetReader for implementation.

Your data


If your model uses DSTC2 and relies on "dstc2_reader" (DSTC2DatasetReader), all needed files, if not present in the DSTC2DatasetReader.data_path directory, will be downloaded from web.

If your model needs to be trained on different data, you have several ways of achieving that (sorted by increase in the amount of code):

  1. Use "dialog_iterator" in dataset iterator config section and "dstc2_reader" in dataset reader config section (the simplest, but not the best way):
    • set dataset_reader.data_path to your data directory;
    • your data files should have the same format as expected in DSTC2DatasetReader.read() method.
  2. Use "dialog_iterator" in dataset iterator config section and "your_dataset_reader" in dataset reader config section (recommended):
    • clone deeppavlov.dataset_readers.dstc2_reader.DSTC2DatasetReader to YourDatasetReader;
    • register as "your_dataset_reader";
    • rewrite so that it implements the same interface as the origin. Particularly, YourDatasetReader.read() must have the same output as DSTC2DatasetReader.read().
      • train — training dialog turns consisting of tuples:
        • first tuple element contains first user’s utterance info (as dictionary with the following fields):
          • text — utterance string
          • intents — list of string intents, associated with user’s utterance
          • db_result — a database response (optional)
          • episode_done — set to true, if current utterance is the start of a new dialog, and false (or skipped) otherwise (optional)
        • second tuple element contains second user’s response info
          • text — utterance string
          • act — an act, associated with the user’s utterance
      • valid — validation dialog turns in the same format
      • test — test dialog turns in the same format
  3. Use your own dataset iterator and dataset reader (if 2. doesn’t work for you):


You should provide a maping from actions to text templates in the format


where filled slots in templates should start with “#” and mustn’t contain whitespaces.

For example,

bye You are welcome!
canthear  Sorry, I can't hear you.
expl-conf_area  Did you say you are looking for a restaurant in the #area of town?
inform_area+inform_food+offer_name  #name is a nice place in the #area of town serving tasty #food food.

It is recommended to use "DefaultTemplate" value for template_type parameter.

Database (optional)

If your dataset doesn’t imply any api calls to an external database, just do not set database and api_call_action parameters and skip the section below.

Otherwise, you should

  1. provide sql table with requested items or

  2. construct such table from provided in train samples db_result items. This can be done with the following script:

    python -m deeppavlov train configs/go_bot/database_<your_dataset>.json

    where configs/go_bot/database_<your_dataset>.json is a copy of configs/go_bot/database_dstc2.json with configured save_path, primary_keys and unknown_value.


Scores for different modifications of our bot model:

Model Config Test turn textual accuracy
basic bot gobot_dstc2_minimal.json 0.3809
bot with slot filler & fasttext embeddings   0.5317
bot with slot filler & intents gobot_dstc2.json 0.5248
bot with slot filler & intents & embeddings   0.5145
bot with slot filler & embeddings & attention gobot_dstc2_best.json 0.5551

There is another modification of DSTC2 dataset called dialog babi Task6 [3]. It differs from ours in train/valid/test split and intent/action labeling.

These are the test scores provided by Williams et al. (2017) [1] (can’t be directly compared with above):

Model Test turn textual accuracy
Bordes and Weston (2016) [4] 0.411
Perez and Liu (2016) [5] 0.487
Eric and Manning (2017) [6] 0.480
Williams et al. (2017) [1] 0.556

TODO: add dialog accuracies