YoungC

To understand DeepGEMM, picture this: The world of AI computations is like a bustling metropolis, with data zipping between towering neural networks like cars in a futuristic city. Enter DeepGEMM, a sleek, high-performance FP8 GEMM library. This isn’t just another tool—it’s a…

两张照片就能看得出，伊朗的认知，配得上他们挨的铁锤😆有意思的是，在中文社区，有人会挺巴勒斯坦哈马斯，但几乎没有任何人同情伊朗。

🚀 DeepSeek-R1-0528 is here! 🔹 Improved benchmark performance 🔹 Enhanced front-end capabilities 🔹 Reduced hallucinations 🔹 Supports JSON output & function calling ✅ Try it now: chat.deepseek.com 🔌 No change to API usage — docs here: api-docs.deepseek.com/guides/reasoni… 🔗…

谋可寡不可众,利可共不可独。

当然有。我给你举几个实际的例子， 1. 气泡袋快递打包机，主要适用场景，就是需要大批量发货，但是货物有需要防摔，防碰撞需求的。 美国报价，19.9万刀。 中国报价，1.5万刀。 我以前吐槽过这个问题。 我们在加关税前采购了两台。 即便加关税，如果有需求，我依然会采购大陆产的设备。…

duck sorter procedural animation

🚀 Day 6 of #OpenSourceWeek: One More Thing – DeepSeek-V3/R1 Inference System Overview Optimized throughput and latency via: 🔧 Cross-node EP-powered batch scaling 🔄 Computation-communication overlap ⚖️ Load balancing Statistics of DeepSeek's Online Service: ⚡ 73.7k/14.8k…

🚀 Day 5 of #OpenSourceWeek: 3FS, Thruster for All DeepSeek Data Access Fire-Flyer File System (3FS) - a parallel file system that utilizes the full bandwidth of modern SSDs and RDMA networks. ⚡ 6.6 TiB/s aggregate read throughput in a 180-node cluster ⚡ 3.66 TiB/min…

🚀 Day 4 of #OpenSourceWeek: Optimized Parallelism Strategies ✅ DualPipe - a bidirectional pipeline parallelism algorithm for computation-communication overlap in V3/R1 training. 🔗 github.com/deepseek-ai/Du… ✅ EPLB - an expert-parallel load balancer for V3/R1. 🔗…

IT WORKS!!!!! A FULL multiplayer with Python websockets server that receives and broadcasts all player positions every 100ms (10 times per second) All code written almost 100% by AI with Cursor and Grok 3 wrote the server code Now you can fly around with everyone else :D It'll…

🚀 Day 3 of #OpenSourceWeek: DeepGEMM Introducing DeepGEMM - an FP8 GEMM library that supports both dense and MoE GEMMs, powering V3/R1 training and inference. ⚡ Up to 1350+ FP8 TFLOPS on Hopper GPUs ✅ No heavy dependency, as clean as a tutorial ✅ Fully Just-In-Time compiled…

To understand how DeepEP works, imagine a busy highway where cars represent data, and cities symbolize computer components. Without well-planned roads and traffic rules, congestion and delays would be inevitable. DeepEP functions like an advanced highway system designed for…

🚀 Day 2 of #OpenSourceWeek: DeepEP Excited to introduce DeepEP - the first open-source EP communication library for MoE model training and inference. ✅ Efficient and optimized all-to-all communication ✅ Both intranode and internode support with NVLink and RDMA ✅…

🚀 Introducing NSA: A Hardware-Aligned and Natively Trainable Sparse Attention mechanism for ultra-fast long-context training & inference! Core components of NSA: • Dynamic hierarchical sparse strategy • Coarse-grained token compression • Fine-grained token selection 💡 With…

😑每次遇到莫名其妙的问题，还是乖乖的阅读原文档吧。

相关度分布图画出了joker。

棠樾牌坊群，位於黃山市歙縣鄭村鎮棠樾村東大道上，為明清時期古徽州建築藝術的代表作。棠樾的七連座牌坊群，不僅體現了徽文化程朱理學"忠、孝、節、義"倫理道德的概貌，也包括了內涵極為豐富的"以人為本"的人文歷史，同時亦是徽商縱橫商界三百餘年的重要見證。每一座牌坊都有一個情感交織的動人故事。

2月26号看到招雇佣军信息，立马请求上一线。 3月2号收到FBI电话说叫他不要去。 3月7号头铁到达波兰，随后安排进乌克兰。 3月16号现身说法讲战场情况。 真人版人在囧途剧情从不让人失望🤗https://t.co/8kDHQCFWh0

😂

看来毛熊的军纪还是杠杠的。