EMNLP Findings2024
QUEST: Efficient Extreme Multi-Label Text Classification with Large Language Models on Commodity Hardware
Chuang Zhou, Junnan Dong, Xiao Huang, Zirui Liu, Kaixiong Zhou, Zhaozhuo Xu
摘要
Extreme multi-label text classification (EMTC) involves predicting multiple labels from a vast pool of candidates based on a user's textual query. While traditional BERT-based methods have shown limited success, large language models (LLMs) have brought new possibilities. It is promising to leverage their remarkable comprehension ability to understand textual queries. However, implementing LLMs is non-trivial for two main reasons. Firstly, real-world EMTC datasets can be extremely large, with candidate product pairs reaching up to ten million in real-world scenarios, which poses significant challenges in data ingestion. Secondly, the large size of LLMs makes computation and memory demands prohibitive for EMTC applications. To this end, we propose QUEST, a Quantized and Efficient Learning with Sampling Technique. QUEST includes a tailored hash sampling module that reduces the data volume to one-fourth of its original size. Additionally, we perform compressive fine-tuning LLMs with only twenty thousand trainable parameters, largely reducing computational requirements. Extensive experiments demonstrate that QUEST outperforms existing methods while requiring fewer computational resources, unlocking efficient EMTC on commodity hardware such as a single Nvidia RTX 3090 GPU with 24 GB of memory.
问 Lune
几分钟读透这篇 Paper,不用再花一下午。
Lune 能读完全文,也能读引用它的顶会 Paper,讲清它的结论、证据和局限,每一点都注明出自哪篇 Paper。不用硬啃,也不靠猜。
已有 6 篇顶会 Paper 引用了它。别等审稿人问起,先弄清它们发现了什么。
脉络
- Collision to Cognition: Hash-Driven Graph Construction for Efficient RAG 引用了 Taming Language Models for Text-attributed Graph Learning with Decoupled Aggregation
- QLoRA: Efficient Finetuning of Quantized LLMs 引用了 LoRA: Low-Rank Adaptation of Large Language Models
- Knapsack Optimization-Based Schema Linking for LLM-Based Text-to-SQL Generation 引用了 LoRA: Low-Rank Adaptation of Large Language Models
- Knapsack Optimization-Based Schema Linking for LLM-Based Text-to-SQL Generation 引用了 Taming Language Models for Text-attributed Graph Learning with Decoupled Aggregation
- Knapsack Optimization-Based Schema Linking for LLM-Based Text-to-SQL Generation 引用了 SimCSE: Simple Contrastive Learning of Sentence Embeddings
引用它的顶会 Paper6
- KScope: A Framework for Characterizing the Knowledge Status of Language ModelsYuxin Xiao, Shan Chen, Jack Gallifant, Danielle S. Bitterman 等NeurIPS 2025 · 3 次引用
- Collision to Cognition: Hash-Driven Graph Construction for Efficient RAGChuang Zhou, Zheng Yuan, Linhao Luo, Zhaozhuo Xu 等ACL 2026
- Large Language Models Meet Extreme Multi-label Classification: Scaling and Multi-modal FrameworkDiego Ortego, Marlon Rodríguez, Mario Almagro, Kunal Dahiya 等AAAI 2026
- Taming Language Models for Text-attributed Graph Learning with Decoupled AggregationChuang Zhou, Zhu Wang, Shengyuan Chen, Jiahe Du 等ACL 2025
- Knapsack Optimization-Based Schema Linking for LLM-Based Text-to-SQL GenerationZheng Yuan, Hao Chen, Zijin Hong, Qinggang Zhang 等ICDE 2026
它引用的 Paper17
- LoRA: Low-Rank Adaptation of Large Language ModelsEdward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu 等ICLR 2022 · 18,833 次引用
- QLoRA: Efficient Finetuning of Quantized LLMsTim Dettmers, Artidoro Pagnoni, Ari Holtzman, Luke ZettlemoyerNeurIPS 2023 · 5,863 次引用
- SimCSE: Simple Contrastive Learning of Sentence EmbeddingsTianyu Gao, Xingcheng Yao, Danqi ChenEMNLP 2021 · 2,496 次引用
- MPNet: Masked and Permuted Pre-training for Language UnderstandingKaitao Song, Xu Tan, Tao Qin, Jianfeng Lu 等NeurIPS 2020 · 1,957 次引用
- On the Efficacy of Knowledge DistillationJang Hyun Cho, Bharath HariharanICCV 2019 · 741 次引用
相关 Paper
- Fast Multi-Resolution Transformer Fine-tuning for Extreme Multi-label Text ClassificationJiong Zhang, Wei-Cheng Chang, Hsiang-Fu Yu, Inderjit S. DhillonNeurIPS 2021 · 147 次引用
- Meta-LMTC: Meta-Learning for Large-Scale Multi-Label Text ClassificationRan Wang, Xi'ao Su, Siyu Long, Xinyu Dai 等EMNLP 2021 · 10 次引用
- QEFT: Quantization for Efficient Fine-Tuning of LLMsChanghun Lee, Jungyu Jin, Younghyun Cho, Eunhyeok ParkEMNLP Findings 2024 · 8 次引用
- Pretrained Generalized Autoregressive Model with Adaptive Probabilistic Label Clusters for Extreme Multi-label Text ClassificationHui Ye, Zhiyu Chen, Da-Han Wang, Brian D. DavisonICML 2020 · 57 次引用
- OmniQuant: Omnidirectionally Calibrated Quantization for Large Language ModelsWenqi Shao, Mengzhao Chen, Zhaoyang Zhang, Peng Xu 等ICLR 2024 · 395 次引用
