【性能报告】Windows 11专业版下13900K + 4070ti性能汇报 #864
asasasasasbc
started this conversation in
General
Replies: 2 comments 6 replies
-
您这个内存+显存都没有400G,请问是如何部署成功的? |
Beta Was this translation helpful? Give feedback.
6 replies
-
谢谢分享,一直在等对消费级CPU的支持哈,我的配置与你相似,14700K+192GB D5-5600+3090x2;准备尝试一下后同步结果。 |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
感谢KT框架的开发者和贡献者。尝试在Windows上编译KT并加载DeepSeek-R1-UD-IQ1_S量化的DeepSeek-R1,交份(比较简陋的)成绩单。
硬件环境:
CPU:13900K(AVX2、已缩杠降压)
RAM:DDR5 48GB 4800 MT/s x4
GPU:NVIDIA RTX 4070ti (12GB VRAM)
平台:ASUS Z790 A 吹雪
软件环境:
OS:Windows 11 Pro
显卡驱动: 572.61
CUDA:(python torch环境)CUDA Toolkit 12.4
模型:
GGUF:来自Unsloth仓库的1.58bit量化版本DeepSeek R1 671b
启动命令行:py -3.11 -m ktransformers.local_chat --model_path deepseek-ai/DeepSeek-R1 --gguf_path D:\LMStudioModels\Models\DeepSeek1.5Q_S
提问:你好,介绍下你自己
KT性能:
prompt eval count: 10 token(s)
prompt eval duration: 1.9297025203704834s
prompt eval rate: 5.182145897845488 tokens/s
eval count: 300 token(s)
eval duration: 59.24525785446167s
eval rate: 5.063696418318609 tokens/s
更深度的测试数据暂时还未来得及测试,后续如有需要会提供
Beta Was this translation helpful? Give feedback.
All reactions