Show Me How It’s Done: The Role of Explanations in Fine-Tuning Language Models

Mohamad Ballout, Ulf Krumnack, Gunther Heidemann, Kai-Uwe Kühnberger
Proceedings of the 15th Asian Conference on Machine Learning, PMLR 222:90-105, 2024.

Abstract

Our research demonstrates the significant benefits of using fine-tuning with explanations to enhance the performance of language models. Unlike prompting, which maintains the model’s parameters, fine-tuning allows the model to learn and update its parameters during a training phase. In this study, we applied fine-tuning to various sized language models using data that contained explanations of the output rather than merely presenting the answers. We found that even smaller language models with as few as 60 million parameters benefited substantially from this approach. Interestingly, our results indicated that the detailed explanations were more beneficial to smaller models than larger ones, with the latter gaining nearly the same advantage from any form of explanation, irrespective of its length. Additionally, we demonstrate that the inclusion of explanations enables the models to solve tasks that they were not able to solve without explanations. Lastly, we argue that despite the challenging nature of adding explanations, samples that contain explanations not only reduce the volume of data required for training but also promote a more effective generalization by the model. In essence, our findings suggest that fine-tuning with explanations significantly bolsters the performance of large language models.

Cite this Paper

BibTeX
@InProceedings{pmlr-v222-ballout24a, title = {{Show Me How It’s Done}: {T}he Role of Explanations in Fine-Tuning Language Models}, author = {Ballout, Mohamad and Krumnack, Ulf and Heidemann, Gunther and K\"{u}hnberger, Kai-Uwe}, booktitle = {Proceedings of the 15th Asian Conference on Machine Learning}, pages = {90--105}, year = {2024}, editor = {Yanıkoğlu, Berrin and Buntine, Wray}, volume = {222}, series = {Proceedings of Machine Learning Research}, month = {11--14 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v222/ballout24a/ballout24a.pdf}, url = {https://proceedings.mlr.press/v222/ballout24a.html}, abstract = {Our research demonstrates the significant benefits of using fine-tuning with explanations to enhance the performance of language models. Unlike prompting, which maintains the model’s parameters, fine-tuning allows the model to learn and update its parameters during a training phase. In this study, we applied fine-tuning to various sized language models using data that contained explanations of the output rather than merely presenting the answers. We found that even smaller language models with as few as 60 million parameters benefited substantially from this approach. Interestingly, our results indicated that the detailed explanations were more beneficial to smaller models than larger ones, with the latter gaining nearly the same advantage from any form of explanation, irrespective of its length. Additionally, we demonstrate that the inclusion of explanations enables the models to solve tasks that they were not able to solve without explanations. Lastly, we argue that despite the challenging nature of adding explanations, samples that contain explanations not only reduce the volume of data required for training but also promote a more effective generalization by the model. In essence, our findings suggest that fine-tuning with explanations significantly bolsters the performance of large language models.} }
Endnote
%0 Conference Paper %T Show Me How It’s Done: The Role of Explanations in Fine-Tuning Language Models %A Mohamad Ballout %A Ulf Krumnack %A Gunther Heidemann %A Kai-Uwe Kühnberger %B Proceedings of the 15th Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Berrin Yanıkoğlu %E Wray Buntine %F pmlr-v222-ballout24a %I PMLR %P 90--105 %U https://proceedings.mlr.press/v222/ballout24a.html %V 222 %X Our research demonstrates the significant benefits of using fine-tuning with explanations to enhance the performance of language models. Unlike prompting, which maintains the model’s parameters, fine-tuning allows the model to learn and update its parameters during a training phase. In this study, we applied fine-tuning to various sized language models using data that contained explanations of the output rather than merely presenting the answers. We found that even smaller language models with as few as 60 million parameters benefited substantially from this approach. Interestingly, our results indicated that the detailed explanations were more beneficial to smaller models than larger ones, with the latter gaining nearly the same advantage from any form of explanation, irrespective of its length. Additionally, we demonstrate that the inclusion of explanations enables the models to solve tasks that they were not able to solve without explanations. Lastly, we argue that despite the challenging nature of adding explanations, samples that contain explanations not only reduce the volume of data required for training but also promote a more effective generalization by the model. In essence, our findings suggest that fine-tuning with explanations significantly bolsters the performance of large language models.
APA
Ballout, M., Krumnack, U., Heidemann, G. & Kühnberger, K.. (2024). Show Me How It’s Done: The Role of Explanations in Fine-Tuning Language Models. Proceedings of the 15th Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 222:90-105 Available from https://proceedings.mlr.press/v222/ballout24a.html.

Related Material