Not All LoRA Parameters Are Essential: Insights on Inference Necessity
Current research on LoRA primarily focuses on minimizing the number of fine-tuned parameters or optimizing its architecture. However, the necessity of all fine-tuned LoRA layers during inference remains underexplored. In this paper, we investigate the contribution of each LoRA layer to the model's ability to predict the ground truth and hypothesize that lower-layer LoRA modules play a more critical role in model reasoning and understanding. To address this, we propose a simple yet effective method to enhance the performance of large language models (LLMs) fine-tuned with LoRA. Specifically, we identify a ``boundary layer'' that distinguishes essential LoRA layers by analyzing a small set of validation samples. During inference, we drop all LoRA layers beyond this boundary. We evaluate our approach on three strong baselines across four widely-used text generation datasets. Our results demonstrate consistent and significant improvements, underscoring the effectiveness of selectively retaining critical LoRA layers during inference.
Code (0)
등록된 구현이 없습니다.
Tasks
AllText GenerationMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Probabilities of causation and post-infection outcomes
Probabilities of causation provide explanatory information on the observed occurrence (causal necessity) and non-occurrence (causal sufficiency) of events. Here, we adapt these probabilities (probability of necessity, pr…
PRoLoRA: Partial Rotation Empowers More Parameter-Efficient LoRA
With the rapid scaling of large language models (LLMs), serving numerous low-rank adaptations (LoRAs) concurrently has become increasingly impractical, leading to unaffordable costs and necessitating more parameter-effic…
From Isolation to Alignment: Unified LoRA for Efficient Multi-Task Learning
Parameter-Efficient Fine-Tuning (PEFT) is essential for adapting Large Language Models (LLMs) to multi-task scenarios. A prevailing trend in this field involves complex LoRA variants with multiple adapters or heads, whic…
parameter-efficient fine-tuningMulti-Task LearningWith Greater Text Comes Greater Necessity: Inference-Time Training Helps Long Text Generation
Long text generation, such as novel writing and discourse-level translation with extremely long contexts, presents significant challenges to current language models. Existing methods mainly focus on extending the model's…
Language ModelingLanguage ModellingText GenerationTranslationActive Exploration based on Information Gain by Particle Filter for Efficient Spatial Concept Formation
Autonomous robots need to learn the categories of various places by exploring their environments and interacting with users. However, preparing training datasets with linguistic instructions from users is time-consuming …
Bayesian InferenceEfficient ExplorationLanguage AcquisitionSequential Bayesian Inference