Nas 的天空
  • Home
  • Mac
  • Nas
  • Python
  • Ai-Studio
  • Other
  • Home
  • Mac
  • Nas
  • Python
  • Ai-Studio
  • Other
Mac 夜神android模拟器配置桥接

Mac Mac 夜神android模拟器配置桥接

配置桥接,可以实现Mac本机与android模拟器同一个IP段,我主要是用于Mac本机通过android模拟器作为代理上网使用。

704次阅读 0个评论
Mac 2024-11-05
  • «
  • 1
  • ...
  • 54
  • 55
  • 56
Nas的天空
Nas的天空
点亮思维火花,畅享数字世界
用户数
1
文章数
551
评论数
21
阅读量
200392
One Word
-「」
Latest articles
FastAPI 构建生产级 LLM 流式代理:解决连接泄漏、假死与断连背压

FastAPI 构建生产级 LLM 流式代理:解决连接泄漏、假死与断连背压

在用 FastAPI 搭建 LLM 聚合网关或转发流式响应(SSE)时,常会遭遇连接池耗尽、客户端提前断开导致后台空转、Nginx 缓冲吃掉首字延迟等暗坑。本文记录从事故排查到架构重构的完整过程,提供带断连检测与连接复用的工业级代码方案。
告别 Ollama 并发瓶颈:单卡 4090 基于 SGLang 极致榨干 DeepSeek-R1 吞吐踩坑实录

告别 Ollama 并发瓶颈:单卡 4090 基于 SGLang 极致榨干 DeepSeek-R1 吞吐踩坑实录

很多同学在私有化部署开源大模型时第一反应是 Ollama,但在并发多轮长文本或多 Agent 协作场景下,Ollama 的吞吐和排队延迟极易雪崩。本文记录我将单卡 4090 机器上的模型推理后端从 Ollama 迁移到 SGLang 的全过程,详解 RadixAttention 缓存复用、Docker 部署配置、Nginx SSE 流式防缓冲,以及真实高并发压测指标。
单卡榨干并发极限:vLLM与SGLang本地推理实战与流式网关压测避坑指南

单卡榨干并发极限:vLLM与SGLang本地推理实战与流式网关压测避坑指南

拒绝理论空谈,记录在单张RTX 4090/A10D显卡上部署Qwen2.5与DeepSeek蒸馏模型的工程实操。深度对比vLLM与SGLang在高并发流式推理下的吞吐与TTFT表现,提供完整的Nginx SSE流式反向代理配置、显存碎片排障及并发压测脚本。
Mac 压榨 Apple Silicon 统一内存:从 Ollama 迁移到 llama-server 构建高并发本地 LLM 网关

Mac 压榨 Apple Silicon 统一内存:从 Ollama 迁移到 llama-server 构建高并发本地 LLM 网关

本文记录了一台 64G 内存的 M3 Max Mac 从 Ollama 迁移到原生 llama-server 的调优历程。深入剖析 Apple Silicon 统一内存的显存分配机制,通过破解 wired_mem_limit 突破显存阈值,结合 KV Cache 量化、连续批处理(Continuous Batching)与多 Slot 并发配置,实现长上下文与多 Agent 并发场景下的推理吞吐翻倍。
告别Ollama单卡瓶颈:基于vLLM+Ray的本地双卡大模型推理集群低成本搭建与调优实战

告别Ollama单卡瓶颈:基于vLLM+Ray的本地双卡大模型推理集群低成本搭建与调优实战

单张消费级显卡跑大模型经常爆显存?本文记录在Ubuntu自建算力节点上,使用vLLM配合Ray实现多卡张量并行(TP)与PagedAttention优化的高并发部署方案。涵盖显存碎片优化、流式SSE代理中转、生产级Docker Compose配置与真实压测排坑指南,助你打造极低延迟的自建LLM服务。
Latest Comments
macau18 link alternatif macau18 link alternatif Thanks for any other fantastic article. Where else may anyone get that type of information in such an ideal method of writing? I have a presentation subsequent week, and I am at the look for such information.
all about casino comped cruises all about casino comped cruises Hi, i read your blog from time to time and i own a similar one and i was just curious if you get a lot of spam comments? If so how do you reduce it, any plugin or anything you can recommend? I get so much lately it's driving me insane so any help is very much appreciated.
medical boric powder medical boric powder I'm not sure why but this site is loading very slow for me. Is anyone else having this issue or is it a problem on my end? I'll check back later and see if the problem still exists.
best leather conditioner best leather conditioner I have been surfing on-line more than three hours nowadays, yet I never found any attention-grabbing article like yours. It is lovely price sufficient for me. In my view, if all web owners and bloggers made excellent content material as you did, the net can be much more useful than ever before.
projector screen hyderabad projector screen hyderabad Ꭺt Nissi Offife Systems, we specialize іn providing top-of-thе-line projectors, interactive panels, аnd comprehensive hone theater solutions tߋ transform youг vierwing experience.
using gain in audio mix using gain in audio mix Your method of explaining the whole thing in this article is really good, all be capable of effortlessly understand it, Thanks a lot.
artistas femeninas de música techno artistas femeninas de música techno Simply wish to say your article is as astonishing. The clarity in your post is just spectacular and i could assume you are an expert on this subject. Well with your permission let me to grab your RSS feed to keep updated with forthcoming post. Thanks a million and please keep up the enjoyable work.
阿滨 阿滨 在安装Qb所欲依赖时报错:pkg install qt5-base qt5-qmake libtorrent-rasterbar
Python 基础入门 Day12 - Nas 的天空 Python 基础入门 Day12 - Nas 的天空 […] Day08: Python 数据分析 […]
Python 基础入门 Day12 - Nas 的天空 Python 基础入门 Day12 - Nas 的天空 […] Day01: Python 基础入门 […]
热评文章
单卡4090跑满满血DeepSeek/Llama3:vLLM加Speculative Decoding与动态批处理实战调优

单卡4090跑满满血DeepSeek/Llama3:vLLM加Speculative Decoding与动态批处理实战调优

单张RTX 4090部署大模型如何兼顾吞吐量与极低首字延迟(TTFT)?本文基于真实生产与私有化环境,拆解vLLM高阶调度策略:结合投机采样(Speculative Decoding)实现2.8倍解码加速,手把手解析PagedAttention显存碎片优化与长上下文Chunked Prefill调参,附完整Docker配置与压测数据。
单卡搞定多模型混部:vLLM + LiteLLM 构建高并发企业级本地 LLM 网关实战

单卡搞定多模型混部:vLLM + LiteLLM 构建高并发企业级本地 LLM 网关实战

单张消费级显卡(如 RTX 4090)如何同时高效服务多个本地大模型?本文带你避开 Ollama 的并发锁与内存浪费陷阱,利用 vLLM 的 PagedAttention 与显存切分机制,结合轻量网关 LiteLLM,手把手搭建支持自动负载均衡、OpenAI 协议兼容、高并发流式代理的高可用 LLM 本地服务架构。
告别OOM与假死:vLLM + LiteLLM 搭建生产级大模型流式代理实战

告别OOM与假死:vLLM + LiteLLM 搭建生产级大模型流式代理实战

针对自建大模型API服务时频发的并发爆显存、长文本假死及缺乏统一网关鉴权问题,本文深度拆解基于vLLM与LiteLLM的高并发流式网关落地架构。包含显存精确计算、PagedAttention参数调优、SSE断流排障、Docker Compose一键拉起及真实压测数据对比。
友情链接
Practical Workplace English Hub
Tag Cloud
__init__.py@classmethod@语法**enter****exit**\*\*kwargs\*args4090 推理优化4090 部署Advanced Mac TipsAdvanced PythonAdvanced QueriesAgent工作流Agent开发Agent架构Agent网关AIAI DeploymentAIGCaiohttpAI入门androidAnti-blockingAnti-ScrapingAnythingLLMAPIAPI DevelopmentAPI SecurityAPI接口API文档
Copyright©2013-2025 Nas的天空 nassky.top 版权所有
 Theme by Puock