PrismML Releases Bonsai 2 27B, Compressing Alibaba's Qwen Model to 5.9GB With 9-10x Reduction
PrismML released Bonsai 2 27B on Thursday, a compressed version of Alibaba's open-source Qwen 3.8 27B model, shrinking it down to 5.9GB—a 9-10x reduction in memory usage versus the original. The model retains 98% of Qwen's aggregate benchmark performance, up
GateNews·09-18 09:44
