UniPruning: Unifying Local Metric and Global Feedback for Scalable Sparse LLMs
"稀疏化" 相关文章
The Unseen Frontier: Pushing the Limits of LLM Sparsity with Surrogate-Free ADMM
为长视频生成减负！浙大与华为提出Compact Attention，挖掘结构化稀疏加速2.5倍
ComPEFT: Compression for Communicating Parameter Efficient Updates via Sparsification and Quantization
Sparsifying a Model's Computations: Preliminary Findings
1-bit大模型还能再突破！新一代BitNet架构启用4位激活值
1-bit大模型还能再突破，新一代BitNet架构启用4位激活值
Neural Magic Releases 2:4 Sparse Llama 3.1 8B: Smaller Models for Efficient GPU Inference
This AI Paper Introduces BitNet a4.8: A Highly Efficient and Accurate 4-bit LLM
This AI Paper from China Proposes a Novel dReLU-based Sparsification Method that Increases Model Sparsity to 90% while Maintaining Performance, Achieving a 2-5× Speedup in Inference
Researchers from Cerebras & Neural Magic Introduce Sparse Llama: The First Production LLM based on Llama at 70% Sparsity
