Termsslug: lattice-quantization

Lattice Quantization

格量化

Mapping weight vectors onto regular lattice points such as E8 to reduce quantization error.

Curated Intelligence

1 items · Page 1/1
量子位 (微信公众号)T2·Open Source
74 pts
ModelScope Open-Sources EntroPack for Arbitrary-Rate Weight Compression

ModelScope has open-sourced EntroPack, a weight compression scheme that folds information entropy into joint quantization and compression design, turning bitrate control into entropy regulation of quantized outputs and enabling arbitrary bitrates and multiple numeric precisions. It requires no activation calibration or fine-tuning, runs on NVIDIA, AMD and ZW-PPU hardware, fits image models into 10GB of VRAM, and cuts quantization error from 9.4% to 4.26%. The implementation is fully open and integrated with diffusion and LLM ecosystems.