Thomas Wolf 盘点 DeepSeek 开源项目,DeepInfer 推理引擎以 Apache 2.0 开放
DeepSeek's open-source streak is getting absurd:
Hugging Face 的 Thomas Wolf 转发 DeepSeek 开源推理框架 DeepInfer 的消息,并盘点其此前开源的项目:FlashMLA 在 H800 上实现 580 TFLOPS 注意力、DeepEP 负责 MoE 通信、DeepGEMM 实现 1350+ FP8 TFLOPS 矩阵乘、3FS 是自建文件系统。DeepSeek 称 DeepInfer 支撑 DeepSeek-V4.1-Flash 的大规模高吞吐低延迟推理,此次以 Apache 2.0 协议开放生产栈,目标是帮助社区降低算力与能耗成本。
推荐理由
AI 评分 74
这条推文把 DeepSeek 从 FlashMLA 到 DeepInfer 的开源项目串成一条线,便于读者看清其工程栈的开放节奏。
主题
查看原帖🚀 Introducing DeepInfer, our in-house inference framework, now fully open source! Powering DeepSeek-V4.1-Flash at scale with high throughput and low latency. We’re sharing our production stack to help the community lower compute costs, reduce energy use, and make AI more affordable. 🔗 https://t.co/JGAik1kgk0 1/4
正文
当前显示原文,中文全文译文尚未就绪。
DeepSeek's open-source streak is getting absurd: - FlashMLA: 580 TFLOPS attention on H800 - DeepEP: MoE comms - DeepGEMM: 1350+ FP8 TFLOPS matmul - 3FS: they built an entire file system - and today: the full inference engine behind V4.1-Flash, Apache 2.0
Mad respect