|国家预印本平台
首页|Chameleon: A Flexible Data-mixing Framework for Language Model Pretraining and Finetuning

Chameleon: A Flexible Data-mixing Framework for Language Model Pretraining and Finetuning

Chameleon: A Flexible Data-mixing Framework for Language Model Pretraining and Finetuning

来源:Arxiv_logoArxiv
英文摘要

Training data mixtures greatly impact the generalization performance of large language models. Existing domain reweighting methods often rely on costly weight computations and require retraining when new data is introduced. To this end, we introduce a flexible and efficient data mixing framework, Chameleon, that employs leverage scores to quantify domain importance within a learned embedding space. We first construct a domain affinity matrix over domain embeddings. The induced leverage scores determine a mixture that upweights domains sharing common representations in embedding space. This formulation allows direct transfer to new data by computing the new domain embeddings. In experiments, we demonstrate improvements over three key scenarios: (i) our computed weights improve performance on pretraining domains with a fraction of the compute of existing methods; (ii) Chameleon can adapt to data changes without proxy retraining, boosting few-shot reasoning accuracies when transferred to new data; (iii) our method enables efficient domain reweighting in finetuning, consistently improving test perplexity on all finetuning domains over uniform mixture. Our code is available at https://github.com/LIONS-EPFL/Chameleon.

Wanyun Xie、Francesco Tonin、Volkan Cevher

计算技术、计算机技术

Wanyun Xie,Francesco Tonin,Volkan Cevher.Chameleon: A Flexible Data-mixing Framework for Language Model Pretraining and Finetuning[EB/OL].(2025-05-30)[2025-06-21].https://arxiv.org/abs/2505.24844.点此复制

评论