取消静音取消静音antirezX (Twitter)First steps for a specialized DeepSeek v4 Flash inference engine focused on inference quality /…打开视频页面155在 X 上点赞分享X在 X (Twitter) 上打开
antirezX (Twitter)DeepSeek v4 Flash has 43 blocks... so, even 4096 batches during the prefill can be smoothly…打开视频页面207在 X 上点赞分享X在 X (Twitter) 上打开
antirezX (Twitter)Iterating... Evals take time and are boring: but are a fundamental validation step of sane LLM…打开视频页面348在 X 上点赞分享X在 X (Twitter) 上打开
Microsoft AIX (Twitter)MAI-Code-1.1-Flash is going local with on-device model calls inside Github Copilot.打开视频页面2028在 X 上点赞分享X在 X (Twitter) 上打开
Atharva IngleX (Twitter)Got DeepSeek V4 Flash running on 2x H200s at 160–200 tok/s on JarvisLabsAI打开视频页面182在 X 上点赞分享X在 X (Twitter) 上打开
AvidX (Twitter)this is the best trick to maximum usage limits on chatgpt codex打开视频页面297在 X 上点赞分享X在 X (Twitter) 上打开
Alex CheemaX (Twitter)AGI at home Running DeepSeek R1 across my 7 M4 Pro Mac Minis and 1 M4 Max MacBook Pro.打开视频页面1万在 X 上点赞分享X在 X (Twitter) 上打开
LM StudioX (Twitter)Batching for vision models is now available in Beta with our latest MLX engine update 👾打开视频页面349在 X 上点赞分享X在 X (Twitter) 上打开
NahidX (Twitter)STOP PAYING FOR APIS RIGHT NOW MASSIVE DROP ON APINEX😳打开视频页面368在 X 上点赞分享X在 X (Twitter) 上打开
Nat FriedmanX (Twitter)Watching llama.cpp do 40 tok/s inference of the 7B model on my M2 Max, with 0% CPU usage, and using…打开视频页面5126在 X 上点赞分享X在 X (Twitter) 上打开
Alex CheemaX (Twitter)How much faster is the new MacBook Pro for AI inference?打开视频页面2380在 X 上点赞分享X在 X (Twitter) 上打开
VeniceX (Twitter)DeepSeek V4 Flash 0731 is now available on Venice, fully private.打开视频页面223在 X 上点赞分享X在 X (Twitter) 上打开