Unverified50% confidenceBenchmarkExact time
UD-Q4_K_XL量化版本体积155GB,Top-1精度为96%
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/7/2026
First Seen
Valid until: 11/5/2026
Sources
Related Claims
UnverifiedUD-IQ1_S量化版本体积83GB,Top-1精度为73%,是压缩最激进的版本79% similarUnverifiedQ4_K_M量化文件大小17.77GB,pp2048速度96.8 tok/s,tg128速度11.4 tok/s75% similarUnverifiedLlama 3 70B量化版(Q4精度)仍需约35GB内存74% similarUnverifiedLLaMA-2-7B使用FP16精度时每个token的KV Cache约占0.5MB72% similarUnverified使用IQ4_XS配合-fa 1 -ctk q8_0 -ctv q8_0对KV缓存进行q8_0量化后,可加载完整32k上下文,常驻内存约16.6GB71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/705390API
curl https://kongchang.com/api/v1/knowledge/claims/705390MCP
get_claim(id=705390)