Tag: 开源
All the articles with the tag "开源".
-
Flash-MoE: Running a 397B-Parameter Model on a MacBook at 4.4 token/s
A developer built Flash-MoE in 24 hours: it runs the 397B-parameter Qwen3.5 model on a 48GB MacBook Pro at 4.4 token/s, using only about 6GB of RAM and no cloud GPUs. We break down how it works: SSD streaming, Metal shader optimization, and MoE sparse activation.
-
Karpathy's AutoResearch: Letting an AI Agent Run 700 ML Experiments on Its Own
A deep dive into Karpathy's open-source AutoResearch project: how a 630-line Python script lets an AI agent run ML experiments autonomously on a single GPU, completing 700 experiments in two days and finding 20 effective optimizations. From architecture to practical applications, here's why every developer should pay attention to the "Karpathy Loop" pattern.
-
MiroFish: A Swarm Intelligence Prediction Engine a College Senior Built in 10 Days with Vibe Coding
A deep dive into the architecture and design of MiroFish, a swarm intelligence prediction engine. Built by a college senior in 10 days with Vibe Coding, it topped GitHub Trending and landed a 30M RMB investment from Shanda. Here's what it got right, technically.
-
AI Agent Frontend Workflows (Part 4): What's Next, and the Open Source Tool Landscape
The series finale. From Copilot to autonomous agents, from closed to open source — this post maps where AI agents are heading, compares the major tools, and explores how the developer's role is changing. Complete learning roadmap included.