-
Zhejiang University
- Hangzhou, China
- https://scholar.google.com/citations?user=0yreGNIAAAAJ&hl=zh-CN
Highlights
- Pro
Stars
[Pytorch] Generative retrieval model using semantic IDs from "Recommender Systems with Generative Retrieval"
A high-throughput and memory-efficient inference and serving engine for LLMs
Repo for Rho-1: Token-level Data Selection & Selective Pretraining of LLMs.
[NeurIPS 2024] The implementation of paper "On Softmax Direct Preference Optimization for Recommendation"
An other implementation of GRU4REC using PyTorch
Official implementation of the GRU4Rec algorithm in PyTorch
[NeurIPS 2024] Dual-Perspective Activation: Efficient Channel Denoising via Joint Forward-Backward Criterion for Artificial Neural Networks
Benckmark datasets, SOTA models and unified CLI for Item Recommendation
QQQ is an innovative and hardware-optimized W4A8 quantization solution for LLMs.
Pytorch implementation of our paper OvSW: Overcoming Silent Weights for Accurate Binary Neural Networks accepted by ECCV 2024.
Reference implementation for DPO (Direct Preference Optimization)
Awesome-LLM-Robustness: a curated list of Uncertainty, Reliability and Robustness in Large Language Models
Code for WSDM 2022 paper, Contrastive Learning for Representation Degeneration Problem in Sequential Recommendation.
[WWW'2023] "DCRec: Debiased Contrastive Learning for Sequential Recommendation"
Instruct-tune LLaMA on consumer hardware
Official code of "PopDCL: Popularity-aware Debiased Contrastive Loss for Collaborative Filtering" (2023 CIKM)
Pytorch implementation of our paper MaxQ: Multi-Axis Query for N:M Sparsity Network accepted by CVPR 2024.
[WWW2024] The official code for paper "Distributionally Robust Graph-based Recommendation System"
[CIKM'23] "Toward a Better Understanding of Loss Functions for Collaborative Filtering"