 |
大模型小白推荐一下本地模型 jiezou • 2h 25m ago • Lastly replied by cowcomic | 28 |
 |
开源了一个 LLM 推理服务监控面板 invdan • 1 day ago |
 |
GLM5.2 个人感觉有点被吹大了 hihihihihi • 1 day ago • Lastly replied by zxjxzj9 | 21 |
 |
有支持 6000 Ada 使用 deepseek v4 flash 推理 的框架吗 frankyzf • 20h 29m ago • Lastly replied by coefu | 7 |
 |
分享个自己在用的玩具 mountainl • 5 days ago • Lastly replied by robinxplorer | 7 |
 |
配置 kiro 的问题 davidyin • 7 days ago • Lastly replied by davidyin | 21 |
 |
买 macbook pro 笔记本，跑本地模型，怎么配置性价比比较高？ sjmcefc2 • 7 days ago • Lastly replied by coefu | 41 |
 |
现在大模型主流都用哪些 nVidia GPU？ mingtdlb • 8 days ago • Lastly replied by zzutmebwd | 30 |
 |
lama.cpp 目前有重大性能 bug： checkpoint 的巡回逻辑对于混合模型（比如 qwen3.6-27B）无效，从而导致大概率每次对话都要 prefill 全文，严重拖慢速度 sentinelK • 11 days ago • Lastly replied by coefu | 15 |
|
GPU 跑 LLM 也会超频吗？ mingtdlb • 10 days ago • Lastly replied by mingtdlb | 4 |
 |
DiffusionGemma
Livid PRO |
31 |
 |
Gemma4 12b 居然比 Qwen3.5 9b 还快，意料不到 yuping913 • 11 days ago • Lastly replied by lifechan | 3 |
 |
什么？ Apple Watch 也能本地跑 Qwen 了？ ericterminal • 13 days ago • Lastly replied by ericterminal | 7 |
 |
关于低算力 gpu 推理时 prefill 在总时长中的占比问题 zzutmebwd • 13 days ago • Lastly replied by coefu | 8 |
 |
需要购买国产显卡本地部署大模型，哪家的比较好 Flagship9945 • 13 days ago • Lastly replied by zomco | 115 |
 |
Mac book air M5 32G+1TB 能跑本地大模型？
TGOcc PRO |
17 |
 |
Gemma4 12B 如何跑在 16G 显存上？ CatCode • 14 days ago • Lastly replied by zzutmebwd | 25 |
 |
mac mini 跑本地模型，需要什么配置？ kakalulin • 14 days ago • Lastly replied by kennylam777 | 18 |
 |
mac 64g 能部署哪个本地大模型 followadc • 16 days ago • Lastly replied by coefu | 19 |
 |
消费级显卡(16G A 卡)是不是不适合运行 vllm 和 sglang，好像使用 transformer 推理都比这两个框架快，并且占用显存低 zhengfan2016 • 14 days ago • Lastly replied by zzutmebwd | 20 |
Select Language