Code associated with:
Stella Frank and Kenny Smith, 2018.
A model of linguistic accommodation leading to language simplification.
Proceedings of CogSci.
usage: ft_multinomial.py [-h] [--outputstr OUTPUTSTR]
[--language LANGUAGE [LANGUAGE ...]]
[--alpha_prior ALPHA_PRIOR]
[--no_gamma_prior NO_GAMMA_PRIOR]
[--dialogue_length DIALOGUE_LENGTH]
[--message_length MESSAGE_LENGTH]
[--speaker_hyps SPEAKER_HYPS]
[--prior_data PRIOR_DATA]
[--accommodation ACCOMMODATION]
Chat between Bayesian Learners.
optional arguments:
-h, --help show this help message and exit
--outputstr OUTPUTSTR, -o OUTPUTSTR
--language LANGUAGE [LANGUAGE ...], -l LANGUAGE [LANGUAGE ...]
--alpha_prior ALPHA_PRIOR, -a ALPHA_PRIOR
--no_gamma_prior NO_GAMMA_PRIOR, -G NO_GAMMA_PRIOR
Don't use gamma prior: speaker-model datasets have
fixed size=D
--dialogue_length DIALOGUE_LENGTH, -d DIALOGUE_LENGTH
--message_length MESSAGE_LENGTH, -M MESSAGE_LENGTH
--speaker_hyps SPEAKER_HYPS, -S SPEAKER_HYPS
number of components in speaker mod (Listener class in
code, confusingly)
--prior_data PRIOR_DATA, -D PRIOR_DATA
--accommodation ACCOMMODATION, -A ACCOMMODATION