 |
有用 mac 跑 minimax H3 的吗，好后悔买了 48G 内存的 macbook lightzh • 18h 36m ago • Lastly replied by chapiom | 39 |
 |
pro6000 部署 Qwen 3.8 flash next nvfp4 zzutmebwd • 2 days ago • Lastly replied by fcten | 36 |
 |
部署 Qwen3.8-27B 模型死循环问题最新升级 jinsongzhaocn • 6h 59m ago • Lastly replied by tgfbeta | 5 |
 |
deepseek v4.1 Flash 发布了 Insolitude • 9 days ago • Lastly replied by CyouYamato | 4 |
 |
2 台 DGX Spark 上运行的本地 LLM 横向对比 ldm0 • 13 days ago • Lastly replied by ldm0 | 29 |
 |
qwen3.8 本机部署后搭配什么搜索工具呢？ idblife • 18 days ago • Lastly replied by wises | 19 |
 |
mac 本地部署 llm 不是最佳选择 Need4more • 14 days ago • Lastly replied by lesismal | 90 |
 |
DeepSeek-V4-Flash-Vision-Exp 权重开源了，原生视觉集成
TGOcc PRO |
4 |
 |
大模型 LLM 是怎么实现计算的？还是内部调用了工具？ jacketma • 20 days ago • Lastly replied by CasualYours | 15 |
 |
老哥们，预算 3w 左右想弄一套小型化本地大模型硬件，求推荐 turing518 • 4 days ago • Lastly replied by iango | 19 |
 |
个人是否应该投资购买 DGX Spark 之类的硬件来跑开源模型？既是自用，也是某种防止硬件涨价的投资？ sillydaddy • 7 days ago • Lastly replied by prosgtsr | 67 |
 |
自己攒了个 comfyui 交流社区 nicolaszhao4 • 15 days ago • Lastly replied by neptuno | 19 |
 |
mac m5 ultra 上线中国官网。1.2T 显存带宽， 256GB 版本 8.6 万 sentinelK • Aug 28 • Lastly replied by sadan9 | 51 |
 |
mac ultra deepseek harness & qwen3.8-27B 几点经验 coefu • 5 days ago • Lastly replied by coefu | 17 |
 |
m1max 64g+1T 满 GPU 的 MacBook Pro 适合部署哪个本地模型 l764191074537 • 20 days ago • Lastly replied by maybestone | 9 |
 |
两台 dgx-spark 部署满血 deepseek v4 flash 完全指南,稳定 60~70tok/s 单流 dbow • 8 days ago • Lastly replied by martin0330 | 66 |
|
RTX PRO 5000（或其他 48GB 显存）的 Qwen3.8-27B-FP8 配置交流（prefill 5000+t/s， decode 60+t/s） sentinelK • Aug 26 • Lastly replied by tgfbeta | 55 |
 |
local.ai 接力是干嘛呀？ justmiho • Aug 19 • Lastly replied by yuusha2022 | 24 |
Select Language