cjzdaily
18:48 · Sep 12, 2025 · Fri
https://x.com/femke_plantinga/status/1966169866353918013?s=12
X (formerly Twitter)
Femke Plantinga (@femke_plantinga) on X
Most RAG systems are doing chunking 𝘸𝘳𝘰𝘯𝘨.
The 𝘸𝘩𝘦𝘯 you chunk matters just as much as 𝘩𝘰𝘸 you chunk.
We've been stuck in the pre-chunking mindset: split documents → embed chunks → store in vector DB. But what if we flipped this entirely?
𝗣𝗿𝗲-𝗖𝗵𝘂𝗻𝗸𝗶𝗻𝗴:
18:39 · Sep 11, 2025 · Thu
https://docs.anthropic.com/en/docs/build-with-claude/prompt-engineering/overview
Claude API Docs
Prompt engineering overview
Claude API Documentation
12:12 · Sep 11, 2025 · Thu
https://x.com/binghe_sun/status/1966027646036476228?s=12
18:46 · Sep 10, 2025 · Wed
https://x.com/9hills/status/1965788062686064805?s=12
X (formerly Twitter)
九原客 (@9hills) on X
一晚上写了 LocalDeepSearch 实训课程方案,包括选模型、合成数据集、选 Agent 框架、Agent SFT 数据集和框架、Agent RL 数据集和框架。
第一个问题,用什么工具调用方案,准备用 B 方案。目前实现是真的乱。
16:27 · Sep 10, 2025 · Wed
https://x.com/shao__meng/status/1965581009451462774?s=12
X (formerly Twitter)
meng shao (@shao__meng) on X
[实用指南与最佳实践] 用 AGENTS. md 提升 AI 代码质量 —— 来自 @Steve8708 分享
文章讲了一个简单但超级实用的方法:通过在项目根目录创建一个 AGENTS. md 文件,告诉 AI 编程工具你的项目规则和偏好,让 AI 输出的代码更符合你的需求,减少修改的麻烦。作者分享了他用 AGENTS. md 提升 AI
17:51 · Sep 9, 2025 · Tue
https://x.com/vikingmute/status/1965407801821593772?s=12
17:50 · Sep 9, 2025 · Tue
https://x.com/dingyi/status/1965275350080086264?s=12
X (formerly Twitter)
ding.one (@dingyi) on X
发现这个开源的多媒体管理工具做得挺好,简洁美观。https://t.co/p67fh3fo0D
17:26 · Sep 9, 2025 · Tue
https://x.com/zho_zho_zho/status/1965334599833452850?s=12
X (formerly Twitter)
-Zho- (@ZHO_ZHO_ZHO) on X
开发者不做选择,窗口/白板/节点全都要!哈哈哈
又又又
🍌
+ Vibe Coding 了全新节点式:香蕉工厂|BananaFlow
现在基于
🍌
三大件都有了
1⃣
窗口式:Nano Bananary|香蕉超市
2⃣
白板/画布式:BananaPod|香蕉铺子
3⃣
节点/工作流式:BananaFlow|香蕉工厂
香蕉工厂目的是流程自动化,与视频等功能串联
20:13 · Sep 8, 2025 · Mon
https://x.com/azed_ai/status/1965052875874467870?s=12
17:30 · Sep 7, 2025 · Sun
https://x.com/oldcai/status/1964331268948275703?s=12
16:40 · Sep 7, 2025 · Sun
https://x.com/wulalatalk/status/1964110888023347564?s=12
00:34 · Sep 6, 2025 · Sat
https://vmodel.ai/models/groups/kling-video/
vmodel.ai
Kling Video Generator,Text & Image to Video API
Bring your vision to life with Kling Video Generator , turning text and images into stunning, dynamic videos with AI-powered precision.
17:24 · Sep 5, 2025 · Fri
https://x.com/axichuhai/status/1963814443219050840?s=12
17:02 · Sep 5, 2025 · Fri
https://x.com/annielynn021/status/1963432097932349467?s=12
16:59 · Sep 5, 2025 · Fri
https://x.com/akshay_pachaar/status/1963582834536534291?s=12
16:00 · Sep 5, 2025 · Fri
https://x.com/binggandata/status/1963784523369570597?s=12
15:36 · Sep 5, 2025 · Fri
https://x.com/tz_2022/status/1963684153255481798?s=12
04:50 · Sep 5, 2025 · Fri
https://x.com/hibikedoraki/status/1963786242195403176?s=12
X (formerly Twitter)
Drakeet (@hibikedoraki) on X
之前和一位朋友聊了我是怎么做 vibe coding 的,今天他把我的方法公开了,可以顺便看看(我那 OCAT 至今仍然是采用这样的方式实现了“99% 的代码都是 AI 写的”):https://t.co/OYgfGSgGdC
00:57 · Sep 5, 2025 · Fri
https://github.com/ninehills/blog/issues/133
GitHub
免费 GPU 或廉价算力 · Issue #133 · ninehills/blog
免费 GPU 或廉价算力 1. 显卡需求和推荐模型 使用 int4 + QLora 训练(效果并不会比Full-finetuning差多少,lora rank设置较大且应用到所有层) 8K 上下文时,显存16GB的显卡则可以训练 20B 以下模型,显存24GB的显卡则可以训练 32B(含)以下模型。 推荐微调 1-7B 模型。推荐性价比显卡:4090,计算能力和显存带宽都足够好,比L40等商...
17:43 · Sep 3, 2025 · Wed
https://upstash.com/
Upstash: Serverless Data Platform
Upstash is a serverless data platform providing low latency and high scalability for real-time applications. Optimize your data infrastructure with Upstash's managed services for Redis, Vector, QStash, and other key data technologies.
Before
After
Home
Powered by
BroadcastChannel
&
Sepia