Saved in:
Bibliographic Details
Main Authors: Tastet, Jean-Loup, Timiryasov, Inar
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2409.17312
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929515280728064
author Tastet, Jean-Loup
Timiryasov, Inar
author_facet Tastet, Jean-Loup
Timiryasov, Inar
contents We present BabyLlama-2, a 345 million parameter model distillation-pretrained from two teachers on a 10 million word corpus for the BabyLM competition. On BLiMP and SuperGLUE benchmarks, BabyLlama-2 outperforms baselines trained on both 10 and 100 million word datasets with the same data mix, as well as its teacher models. Through an extensive hyperparameter sweep, we demonstrate that the advantages of distillation cannot be attributed to suboptimal hyperparameter selection of the teachers. Our findings underscore the need for further investigation into distillation techniques, particularly in data-limited settings.
format Preprint
id arxiv_https___arxiv_org_abs_2409_17312
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle BabyLlama-2: Ensemble-Distilled Models Consistently Outperform Teachers With Limited Data
Tastet, Jean-Loup
Timiryasov, Inar
Computation and Language
Machine Learning
I.2.7
We present BabyLlama-2, a 345 million parameter model distillation-pretrained from two teachers on a 10 million word corpus for the BabyLM competition. On BLiMP and SuperGLUE benchmarks, BabyLlama-2 outperforms baselines trained on both 10 and 100 million word datasets with the same data mix, as well as its teacher models. Through an extensive hyperparameter sweep, we demonstrate that the advantages of distillation cannot be attributed to suboptimal hyperparameter selection of the teachers. Our findings underscore the need for further investigation into distillation techniques, particularly in data-limited settings.
title BabyLlama-2: Ensemble-Distilled Models Consistently Outperform Teachers With Limited Data
topic Computation and Language
Machine Learning
I.2.7
url https://arxiv.org/abs/2409.17312