tatk.e2e.sequicity.multiwoz package

Submodules

tatk.e2e.sequicity.multiwoz.sequicity module

Sequicity is an end-to-end task-oriented dialog system based on a single sequence-to-sequence model that uses belief span to track dialog believes. We adapt the code from github to work in multiwoz corpus.

Reference:

Lei, W., Jin, X., Kan, M. Y., Ren, Z., He, X., & Yin, D. (2018, July). Sequicity: Simplifying task-oriented dialogue systems with single sequence-to-sequence architectures. In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (pp. 1437-1447).

class tatk.e2e.sequicity.multiwoz.sequicity.Sequicity(model_file='https://tatk-data.s3-ap-northeast-1.amazonaws.com/sequicity_multiwoz.zip', name='Sequicity')

Bases: tatk.dialog_agent.agent.Agent

__init__(model_file='https://tatk-data.s3-ap-northeast-1.amazonaws.com/sequicity_multiwoz.zip', name='Sequicity')

Sequicity initialization

Args:

model_file (str):

trained model path or url. default=”https://tatk-data.s3-ap-northeast-1.amazonaws.com/sequicity_multiwoz.zip”

Example:

sequicity = Sequicity()

init_session()

Reset the class variables to prepare for a new session.

response(usr)

Generate agent response given user input.

Args:

observation (str):

The input to the agent.

Returns:

response (str):

The response generated by the agent.

z2degree(gen_z)

tatk.e2e.sequicity.multiwoz.sequicity.denormalize(uttr)