Year: 2025
Type: conference-paper
Abstract: Delta compression methods focus on efficiently serving multiple uniquely fine-tuned models, each tailored to specific tasks and user requirements.These approaches decompose a finetuned LLM into a base model and corresponding delta weights, which are compressed using low-rank or low-bit representations to reduce storage costs.However, their effectiveness is highly sensitive to the magnitude of the model deltas-a factor directly influenced by the scale of the training data.We propose the Residual Quantization Tree (RQT), a hierarchical quantization framework that automatically shares low-bit integer weights across similar fine-tuned models.The RQT construction employs a two-phase greedy algorithm: a bottom-up aggregation of models based on weight matrix similarity and top-down residual quantization, in which each node optimizes the quantization parameters and then delegates residual errors to child nodes.We evaluate RQT on fine-tuned models across mathematics, coding, chatbot, and Chinese LLMs.The results show that RQT achieves an average accuracy degradation of approximately 3% (comparable to previous 4-bit post-training quantization) while maintaining an effective bitwidth of around 2 bits.
Authors: Chen Tianqi, Peisong Wang, Weixiang Xu, Zeyu Zhu, Jian Cheng
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Award: G3352541487
OpenAlex ID: W4412888358