-
Notifications
You must be signed in to change notification settings - Fork 1
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
1 parent
07c24a3
commit 5836416
Showing
2 changed files
with
170 additions
and
0 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,170 @@ | ||
--- | ||
slug: 9月15日内容汇总 | ||
title: 9月15日内容汇总 | ||
authors: [garfield] | ||
tags: [] | ||
--- | ||
|
||
![alt text](maxresdefault.jpg) | ||
|
||
封面图:Safe by construction - Roberto Clapis | ||
|
||
## 🌟 AI 相关 | ||
|
||
[达到草莓的70%!首个open o1项目开源](https://mp.weixin.qq.com/s/-XX_fUL8xX4b6ijpxjDQ-A) | ||
|
||
[大模型千卡训练-经验指北](https://mp.weixin.qq.com/s/y5Wj04VoG_6iQU1-A3i6ew) | ||
|
||
[如何用1024张显卡训练一个模型](https://mp.weixin.qq.com/s/YjbAQZs4ESlV1i2rjBe_UA) | ||
|
||
[【LLM模型推理】LLMs-推理优化-LLM推理算法全解v1.0](https://mp.weixin.qq.com/s/RigELqsWdgJLGMZ1XGge7w) | ||
|
||
[微软让Agent迈入Window操作系统时代!](https://mp.weixin.qq.com/s/FDuth7HcyDtI5fdhEzR1lg) | ||
|
||
[疯狂24h后,openai o1有哪些新的秘密](https://mp.weixin.qq.com/s/n37S7XokmmZlNHO68g952Q) | ||
|
||
[张俊林:OpenAI o1的价值意义及强化学习的Scaling Law](https://mp.weixin.qq.com/s/my7XiRtpb8IY3Z0b471NJA) | ||
|
||
[o1完整思维链成OpenAI头号禁忌!问多了等着封号吧](https://mp.weixin.qq.com/s/UyQ53NORlfAYtihwhfaO-A) | ||
|
||
🌟 [浅谈OpenAI o1的价值意义及RL 的Scaling Law](https://mp.weixin.qq.com/s/BXxGUy9SYjPMWF1bGmajcQ) | ||
|
||
🌟 [RAG系统的7个检索指标:信息检索任务准确性评估指南](https://mp.weixin.qq.com/s/rx0DPuS1JfWxoG4uKsmReQ) | ||
|
||
[OpenAI发布新模型:o1系列模型,更像理科生的模型](https://mp.weixin.qq.com/s/c-h6sLM_i5OLfPM9EK2Y9A) | ||
|
||
[六个问题带你看懂什么是理工科学霸-OpenAI o1!](https://mp.weixin.qq.com/s/qnd-nTpELR0Ka2frKjxNQA) | ||
|
||
[刚刚,OpenAI震撼发布o1大模型!强化学习突破LLM推理极限](https://mp.weixin.qq.com/s/sGcx90Q_uI8se-DKosj9dw) | ||
|
||
[深夜空降!OpenAI草莓正式发布,命名o1, GPT-4o被碾压!](https://mp.weixin.qq.com/s/o54S0J3LK3nLgeb6NV2nSA) | ||
|
||
[OpenAI发布突破性AI模型o1系列,推理能力大幅跃升重塑行业格局](https://mp.weixin.qq.com/s/Rbsin35C6dYSDu_RUi9C0Q) | ||
|
||
[【LLM开源项目】LLMs-微调框架-LLaMA-Factory微调量化部署v4.0](https://mp.weixin.qq.com/s/tfBGKMmSb4cdLNJmQVHpvw) | ||
|
||
[AgentRE:用智能体框架提升知识图谱构建效果,重点是开源!](https://mp.weixin.qq.com/s/_P_3H3uyIWjgaCF_FczsDg) | ||
|
||
[LlamaIndex最新报告:构建高级大模型助手-RAG只是起点](https://mp.weixin.qq.com/s/G5iVssfAizKMSvLh70ffOQ) | ||
|
||
[LangChain: 大语言模型的新篇章](https://mp.weixin.qq.com/s/P94AvHvQcget9OqblrmD6g) | ||
|
||
[社区供稿 | OpenCSG开源中文版fineweb edu数据集!](https://mp.weixin.qq.com/s/3U5SylFDbqIQKT_6XG1lEg) | ||
|
||
[探究大模型微调 Lora 的不同形态(上篇): AdaLora、 AsLora、 PiSSA、 DoRA](https://mp.weixin.qq.com/s/36sEy5yRHbTK9QT8xi704A) | ||
|
||
[通过打包 Flash Attention 来提升 Hugging Face 训练效率](https://mp.weixin.qq.com/s/viC6k9B5xiSiOctdnRf87Q) | ||
|
||
[【LLM开源项目】LLMs-微调框架-LLaMA-Factory项目概览v1.0](https://mp.weixin.qq.com/s/pBxe2eLvzTOziPJqTw6P_A) | ||
|
||
[代码角度,深度解析 Qwen2-VL 模型结构](https://mp.weixin.qq.com/s/9EJwEfeC_E1kmz-e1ZWXBQ) | ||
|
||
[别以为Cursor生成不了屎山代码,用PLANSEARCH给你更多计划,性能提升到77%| CursorAI最新](https://mp.weixin.qq.com/s/iQ_mOxFEJOJICwfDQTIu5w) | ||
|
||
[OpenAI「草莓」两周内发布?网传不是多模态,反应慢了10多秒](https://mp.weixin.qq.com/s/ylqZ4WdawqKadxrW5rch8w) | ||
|
||
[Scaling Law瓶颈,Cursor编程为什么这么强?团队参与新研究掏出秘密武器](https://mp.weixin.qq.com/s/xhV9HoeEP22RjuWTjgbPqg) | ||
|
||
[基于 chunked prefill 理解 prefill 和 decode 的计算特性](https://mp.weixin.qq.com/s/uQU9Ia-I3rmR4FkL9dmPdg) | ||
|
||
[你以为的LLM上下文学习超能力,究竟来自哪里,ICL的内部机制如何 |最新发布](https://mp.weixin.qq.com/s/-Iqm5NEgy--lbPQKUFUyBA) | ||
|
||
[LIama 3+Mamba强强联手!蒸馏到线性RNN,推理速度提升1.6倍](https://mp.weixin.qq.com/s/2oyeCdlqKaeQATje--U8qg) | ||
|
||
[还在人工炼丹?自动提示工程指南来了,还带从头实现](https://mp.weixin.qq.com/s/TxzkRUPhsiqtLhCyrIsQrQ) | ||
|
||
[Liger kernel训练加速,一行代码训练吞吐量提高 20%,显存使用量降低 60%](https://mp.weixin.qq.com/s/iGE_q8_zPO-pn0iapvrG5g) | ||
|
||
[HomeDepot:RAG能否在提高客服实际效率?ReACT推理是否有帮助](https://mp.weixin.qq.com/s/Q0x3LQjOmCwnIyXffw7org) | ||
|
||
[【LLM模型幻觉】LLMs-模型幻觉-哈工大LLM幻觉综述-20231110v4.0](https://mp.weixin.qq.com/s/mZzM2OLy6K_FIsU6Y6P2AA) | ||
|
||
[爆火的Reflection 70B,了解背后的微调技术](https://mp.weixin.qq.com/s/5nDF8ZYcL37tO_3obbXTKg) | ||
|
||
[你理解的LLM自我纠错能力可能要让你遗憾了,LLM推理能力的真相 |ICLR2024](https://mp.weixin.qq.com/s/_Wp1uC6zsf6JGYfPiTlssw) | ||
|
||
[社区供稿 | Llama-3.1 70B全量训练!OpenBuddy发布新一代70B中文跨语言模型](https://mp.weixin.qq.com/s/3ra8c_lUUJlelfIlif9jhA) | ||
|
||
[探索自然语言中的计划搜索:提升大型语言模型代码生成性能的新方法](https://mp.weixin.qq.com/s/pejhJvMBGbSQ27ikDJQRuA) | ||
|
||
[2024年大模型Alignment偏好优化技术PPO,DPO, SimPO,KTO,Step-DPO, MCTS-DPO,SPO](https://mp.weixin.qq.com/s/dgglzgW7BI_X7zNPmgrTfw) | ||
|
||
[从llm丝滑过渡到多模态,是一种什么体验](https://mp.weixin.qq.com/s/vW57_j2GWgXI_pzFvYf8Dw) | ||
|
||
🌟 [新鲜出炉2个RAG技巧,高级RAG更高级!](https://mp.weixin.qq.com/s/9iDMjsIi0LKJaTTZo6n-aQ) | ||
|
||
[Transformer推理结构简析(Decoder + MHA)](https://mp.weixin.qq.com/s/92IDNwS-00UTaKw-lkW_2Q) | ||
|
||
[Reflection Llama-3.1 70B真的翻车了吗](https://mp.weixin.qq.com/s/_zs0N6hWU-59_um6wPHjKw) | ||
|
||
🌟 [你写提示词应该更直白一点,把Top-k直接写进Prompt,无需提示的思维链推理 | DeepMind](https://mp.weixin.qq.com/s/kl_bWfTPeezfKyCu9qBkBw) | ||
|
||
## ⭐️ Go & 云原生 & Rust 相关 | ||
|
||
[Building LLM-powered applications in Go](https://go.dev/blog/llmpowered) | ||
|
||
[Share your feedback about developing with Go](https://go.dev/blog/survey2024-h2) | ||
|
||
[Telemetry in Go 1.23 and beyond](https://go.dev/blog/gotelemetry) | ||
|
||
[Go 入门指南:8.1. 声明、初始化和 make](https://mp.weixin.qq.com/s/1yVDH0PLoukqSueGRFzhNg) | ||
|
||
[eBPF 开发实战 - 02 kprobe入门](https://mp.weixin.qq.com/s/XPqBCQ1Wdqoz0ck092YWQQ) | ||
|
||
[Go 入门指南:7.6. 字符串、数组和切片的应用](https://mp.weixin.qq.com/s/V5OHKniajaEobGY_f-YvjA) | ||
|
||
🌟 [Go必知必会:部署(上)优雅的把Go项目部署到Linux服务器](https://mp.weixin.qq.com/s/494eVC2S8Hk0lsB7SUAsOw) | ||
|
||
🌟 [gRPC 微服务—Context信息的跨服务传递](https://mp.weixin.qq.com/s/JLEEDMJD2luur9pjhoXGEw) | ||
|
||
[eBPF 开发实战 - 01 从Hello World开始](https://mp.weixin.qq.com/s/vhJVIS5RQfYAvuibz5p4mg) | ||
|
||
[在 Rust 应用中运行多个 Tokio 运行时](https://mp.weixin.qq.com/s/djAl2AjtTtu0ucKNHClb4A) | ||
|
||
[Go 入门指南:7.5. 切片的复制与追加](https://mp.weixin.qq.com/s/9IytXAuvFb528IhQW96Hsw) | ||
|
||
[探索 Goja:Golang 中的 JavaScript 运行时](https://mp.weixin.qq.com/s/lP5aieTgM4DHKhqE2SKfJQ) | ||
|
||
[eBPF 开发实战 - 00 生态圈介绍](https://mp.weixin.qq.com/s/l3-jaokVlaf2EOAVGTpaZw) | ||
|
||
[JSON包新提案:用“omitzero”解决编码中的空值困局](https://mp.weixin.qq.com/s/Lw_l_AELo8RKiLzVdS0H-Q) | ||
|
||
[Go 入门指南:7.3. For-range 结构](https://mp.weixin.qq.com/s/wdu7mvnFIz-jXrdjUO9Ueg) | ||
|
||
[Go 入门指南:7.4. 切片重组(reslice)](https://mp.weixin.qq.com/s/NvInmHjactf734T3xOgpbQ) | ||
|
||
[如何选择 Kubernetes Deployment 的更新策略](https://mp.weixin.qq.com/s/kEDGZOc4VPnV5XYYAkT0tQ) | ||
|
||
[亚微秒级的解析速度!Go 语言的高性能 User-Agent 解析库](https://mp.weixin.qq.com/s/Qvj64_WofoCZ8j0eTCq6cQ) | ||
|
||
[Go 错误处理指北:Error vs Exception vs ErrNo](https://mp.weixin.qq.com/s/-y9m-VnlUqQu8BukdN4qdQ) | ||
|
||
[Mule:一个非驴非马的网络诊断库](https://mp.weixin.qq.com/s/eWZQXvvtXqFel3DSA-97Cw) | ||
|
||
[高性能、并发安全的 Go 嵌入式缓存库](https://mp.weixin.qq.com/s/g1X2p1S7cBPYiB4fMh7LVQ) | ||
|
||
[从“黑神话:悟空”看Go为什么不适合AAA游戏开发](https://mp.weixin.qq.com/s/CsMe6USuHjAcXFU80hTchA) | ||
|
||
[Rust必知必会库003: 使用 bitflags 定义和操作位标志](https://mp.weixin.qq.com/s/gzBqtPIkZ27ZDTrfHJG3IQ) | ||
|
||
[Go1.23 新特性:争议最大的 iter 迭代器,可遍历万物!](https://mp.weixin.qq.com/s/In_CxhOikw3iZYBnEXgG_g) | ||
|
||
[一个能让你少写循环和判断的 Go 开源包,支持泛型](https://mp.weixin.qq.com/s/ipOXXpdR2JUBWSY9_9OUiA) | ||
|
||
🌟 [Go 和 Java 的这个机制,真是完全不一样](https://mp.weixin.qq.com/s/8pTYeUEWm6RrQshNdIn_QQ) | ||
|
||
[Go 入门指南:7.1. 数组声明和初始化](https://mp.weixin.qq.com/s/Ir7Qoio33NDTXabA97j0zw) | ||
|
||
## 📒 后端相关 | ||
|
||
[B站直播的极速排障建设-全链路Trace追踪](https://mp.weixin.qq.com/s/UwFjH_gK4CgUAe1IUYyD8A) | ||
|
||
[8种专坑同事的 SQL 写法,性能降低100倍,不来看看](https://mp.weixin.qq.com/s/cFRHxJQcUbSKpHNV1HLlJw) | ||
|
||
[美团面试:Redis锁如何续期?Redis锁超时,任务没完怎么办](https://mp.weixin.qq.com/s/h3QY545R5US_WkjQXl8mbw) | ||
|
||
## 📒 前端相关 | ||
|
||
[月满编程路,MarsCode中秋活动开启!](https://mp.weixin.qq.com/s/0aGuaRHK1HdWGqqsbdKDrw) | ||
|
||
[Claude Artifacts 会成为前端标配么](https://mp.weixin.qq.com/s/5nmASVE6VrFHeZRWGNMrtA) |
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.