paper-with-me

Papers

Ladder Up, Memory Down: Low-Cost Fine-Tuning With Side Nets

2025-12-16 · Estelle Zheng, Nathan Cerisara, Sébastien Warichet, Emmanuel Helbert, Christophe Cerisara arxiv

Fine-tuning large language models (LLMs) is often limited by the memory available on commodity GPUs. Parameter-efficient fine-tuning (PEFT) methods such as QLoRA reduce the number of trainable parameters, yet still incur high memory usage induced by the backward pass in the full model. We revisit Ladder Side Tuning (LST), a rarely explored PEFT technique that adds a lightweight side network, and show that it matches QLoRA's compute scaling slope while cutting peak memory by 50\%. Across different downstream benchmarks spanning natural language understanding, mathematical and LLM-critic tasks, LST has competitive performance with QLoRA's accuracy on average while being much more memory-efficient. This efficiency enables fine-tuning of 7B-parameter models on a single 12 GB consumer GPU with 2k-token contexts, requiring no gradient checkpointing\textemdash conditions under which QLoRA exhausts memory. Beyond memory efficiency, we also establish scaling laws showing that LST scales similarly to QLoRA. We exploit Ladder's architectural flexibility by introducing xLadder, a depth-extended variant that increases effective depth via cross-connections and shortens chain-of-thought (CoT) at fixed parameter count. Ladder is strong when memory is the bottleneck; xLadder builds on this by enabling deeper reasoning without additional memory overhead.

📄 PDF Abstract BibTeX arXiv:2512.14237

Code (0)

등록된 구현이 없습니다.

Tasks

parameter-efficient fine-tuningNatural Language Understanding

Similar Papers 제목 키워드 기반

LST: Ladder Side-Tuning for Parameter and Memory Efficient Transfer Learning

2022-06-13 · Yi-Lin Sung, Jaemin Cho, Mohit Bansal

Fine-tuning large pre-trained models on downstream tasks has been adopted in a variety of domains recently. However, it is costly to update the entire parameter set of large pre-trained models. Although recently proposed…

Transfer LearningVisual Question Answering (VQA)

Ladder: A Model-Agnostic Framework Boosting LLM-based Machine Translation to the Next Level

2024-06-22 · Zhaopeng Feng, Ruizhe Chen, Yan Zhang, Zijie Meng 외

General-purpose Large Language Models (LLMs) like GPT-4 have achieved remarkable advancements in machine translation (MT) by leveraging extensive web content. On the other hand, translation-specific LLMs are built by pre…

Machine TranslationTranslation

QuAILoRA: Quantization-Aware Initialization for LoRA

2024-10-09 · Neal Lawton, Aishwarya Padmakumar, Judith Gaspers, Jack FitzGerald 외

QLoRA reduces the memory-cost of fine-tuning a large language model (LLM) with LoRA by quantizing the base LLM. However, quantization introduces quantization errors that negatively impact model performance after fine-tun…

Causal Language ModelingGPULanguage ModelingLanguage Modelling+2

Enhancing Zeroth-order Fine-tuning for Language Models with Low-rank Structures

2024-10-10 · Yiming Chen, Yuan Zhang, Liyuan Cao, Kun Yuan 외

Parameter-efficient fine-tuning (PEFT) significantly reduces memory costs when adapting large language models (LLMs) for downstream applications. However, traditional first-order (FO) fine-tuning algorithms incur substan…

parameter-efficient fine-tuning

Memory-Efficient Fine-Tuning of Transformers via Token Selection

2025-01-31 · Antoine Simoulin, Namyong Park, Xiaoyi Liu, Grey Yang

Fine-tuning provides an effective means to specialize pre-trained models for various downstream tasks. However, fine-tuning often incurs high memory overhead, especially for large transformer-based models, such as LLMs. …

Few-Shot LearningQuestion Answeringtext-classificationText Classification