QA-LoRA: Quantization-Aware Low-Rank Adaptation of Large Language Models.
Yuhui XuLingxi XieXiaotao GuXin ChenHeng ChangHengheng ZhangZhengsu ChenXiaopeng ZhangQi TianPublished in: ICLR (2024)
Keyphrases
- language model
- low rank
- agent technology
- passage retrieval
- language modeling
- matrix factorization
- missing data
- linear combination
- convex optimization
- low rank matrix
- question answering
- document retrieval
- n gram
- matrix completion
- singular value decomposition
- probabilistic model
- information retrieval
- rank minimization
- semi supervised
- test collection
- retrieval model
- high dimensional data
- query expansion
- high order
- trace norm
- vector space model
- query terms
- recommender systems
- pattern recognition
- smoothing methods
- collaborative filtering