ouktlab/turngpt-jp-rinna-gpt1b-cejc-refinement

This is a Japanese turn-taking model that predicts the end of an utterance based on linguistic information.

The details and examples are shown at our github repository.

We fine-tuned rinna/japanese-gpt-1b to predict turn-shift probabilities, following the approach of TurnGPT.

This model was trained using CEJC.

We constructed training sets by modifying discrepant instances.

Citing rinna-gpt

@misc{rinna-japanese-gpt-1b,
    title = {rinna/japanese-gpt-1b},
    author = {Zhao, Tianyu and Sawada, Kei},
    url = {https://huggingface.co/rinna/japanese-gpt-1b}
}

or arXiv:

@inproceedings{sawada2024release,
    title = {Release of Pre-Trained Models for the {J}apanese Language},
    author = {Sawada, Kei and Zhao, Tianyu and Shing, Makoto and Mitsui, Kentaro and Kaga, Akio and Hono, Yukiya and Wakatsuki, Toshiaki and Mitsuda, Koh},
    booktitle = {Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)},
    month = {5},
    year = {2024},
    pages = {13898--13905},
    url = {https://aclanthology.org/2024.lrec-main.1213},
    note = {\url{https://arxiv.org/abs/2404.01657}}
}

Citieng TurnGPT

@inproceedings{ekstedt-skantze-2020-turngpt,
    title = "{T}urn{GPT}: a Transformer-based Language Model for Predicting Turn-taking in Spoken Dialog",
    author = "Ekstedt, Erik  and
      Skantze, Gabriel",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.findings-emnlp.268",
    doi = "10.18653/v1/2020.findings-emnlp.268",
    pages = "2981--2990",
    abstract = "Syntactic and pragmatic completeness is known to be important for turn-taking prediction, but so far machine learning models of turn-taking have used such linguistic information in a limited way. In this paper, we introduce TurnGPT, a transformer-based language model for predicting turn-shifts in spoken dialog. The model has been trained and evaluated on a variety of written and spoken dialog datasets. We show that the model outperforms two baselines used in prior work. We also report on an ablation study, as well as attention and gradient analyses, which show that the model is able to utilize the dialog context and pragmatic completeness for turn-taking prediction. Finally, we explore the model{'}s potential in not only detecting, but also projecting, turn-completions.",
}