Unverified50% confidenceBenchmarkExact time
腾讯团队将3.3GB的翻译模型压缩到440MB，压缩幅度接近87%，且BLEU分数几乎没有可见退化
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/7/2026
First Seen
Valid until: 12/6/2026
Sources
Related Entities
Related Claims
Unverified动态GGUF的UD-IQ2_XXS档位将原始BF16模型体积从54.7GB压缩到约9GB，体积压缩率83.5%，准确率保留82.5%74% similarUnverified140GB的FP16 70B模型经4-bit量化后可压缩至约35-40GB，通信开销降低约75%；2-bit量化可压缩至约18GB但带来精度损失73% similarUnverified原版QW3.6 27B模型文件体积约为54GB，Prismo通过数学压缩技术压缩至不到4GB，压缩比超过13倍，且声称不影响模型性能73% similarUnverified在ComfyUI视频抠图工作流中，将图片尺寸从1024降至512可将APNG文件从160MB压缩至30多MB，压缩比接近80%73% similarUnverifiedBonsai将27B模型从全精度的54GB压缩至3.9GB72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/869548API
curl https://kongchang.com/api/v1/knowledge/claims/869548MCP
get_claim(id=869548)