Tag: Quantization
All the papers with the tag "Quantization".
Pack-PTQ: Advancing Post-training Quantization of Neural Networks by Pack-wise Reconstruction
grok-3-latestScore: 0.48Published: at 02:53本文提出 Pack-PTQ 方法,通过 Hessian-guided 打包机制和包级混合精度量化策略,显著提升低比特后训练量化的性能,同时捕捉跨块依赖性。