Unverified50% confidenceFactExact time
Base64 编码将每3字节二进制数据拆分为4组6位，输出长度约为原始数据的4/3倍
1
Sources
50%
Confidence
Long-term
Relevance
7/23/2026
First Seen
Sources
Related Claims
Partially VerifiedA standard FP32 model occupies 4 bytes per parameter, while INT4 quantization compresses this to 0.5 bytes, theoretically achieving an 8x memory saving.70% similarUnverifiedPQ乘积量化通过切分向量、局部聚类、编码替代，可将向量存储体积压缩约8倍（四维示例中从16字节压缩到2字节）67% similarUnverifiedINT4 achieves 4× storage compression compared to FP16.66% similarUnverifieddecode阶段的生成速度近似等于内存带宽÷模型大小，例如在内存带宽273GB/s的M4 Pro上运行4GB的Q4量化7B模型理论最大生成速度约68 tokens/s66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/602563API
curl https://kongchang.com/api/v1/knowledge/claims/602563MCP
get_claim(id=602563)