#jupyter_notebook#ai#llm#llms#multi_modal#openai#python#rag
Retrieval-Augmented Generation (RAG) is a technique that helps improve the accuracy of large language models by fetching relevant information from databases or documents. This approach ensures that the model's responses are based on up-to-date and accurate data, reducing errors and "hallucinations" where the model might provide false information. For users, RAG offers more reliable and trustworthy responses, allowing them to verify the sources used to generate those responses. This method also saves resources by avoiding the need to retrain models with new data.
https://github.com/FareedKhan-dev/all-rag-techniques
Maid - Mobile Artificial Intelligence Distribution
Maid is a cross-platform free and an open-source application for interfacing with llama.cpp models locally, and remotely with Ollama, Mistral, Google Gemini and OpenAI models remotely.
-Choose from A wide range of models that runs LOCALLY and access remote models via api key!
-Text based output
-Image Generation (Selected Models only)
-No video or short clips generation yet
-Voice generation on selected models (Not tested)
-Setting model parameters
-Setting system prompt (Making the model behave/generate output in a certain way).
-And more.
Get it on
Github - https://github.com/Mobile-Artificial-Intelligence/maid/releases/latest
Fdroid - https://f-droid.org/packages/com.danemadsen.maid/
Spystore - https://play.google.com/store/apps/details?id=com.danemadsen.maid
*Don't clear CACHE OF THE APP AND EXCLUDE IT FROM SYSTEM'S AUTO CACHE CLEANING as app stores everything in device cache*
Follow @nogoolag and @libreware for more
#ai
Cherry Studio
Cherry Studio is a desktop client for Windows, Mac and Linux, which supports many LLM providers, including large cloud services and local models.
Among its main functions is the ability to work with more than 300 pre -designed #AI assistants, the creation of custom assistants, as well as support for various formats of documents, including text, images and office files.
The application offers tools for global search, top management and translating, which significantly improves interaction with the user thanks to the cross -platform and many settings options.
https://github.com/cherryhq/cherry-studio
LibreChat AI
Open-source platform that allows users to chat and interact with various #AI models through a unified interface. You can use OpenAI, Gemini, Anthropic and other AI models using their API. You may also use Ollama as an endpoint and use LibreChat to interact with local LLMs. It can be installed locally or deployed on a server.
LibreChat is designed to be highly customizable and supports a wide range of AI providers and services. Let me summarize its main features:
Free and Open Source: Accessible to everyone without any costs.
Customization: Offers extensive options to tailor the platform to individual preferences.
Multi-AI Support: Integrates with numerous AI models and services.
Unified Interface: Provides a consistent experience for interacting with different AI models.
https://www.librechat.ai
https://itsfoss.com/librechat-linux/
Jan.ai
https://jan.ai
A platform that enables you to run self-hosted local #AI. Jan provides an OpenAI-equivalent API server at localhost:1337 that can be used as a drop-in replacement with compatible apps.
With Jan, you can:
-Run open-source LLMs locally or connect to cloud AIs like ChatGPT or Google.
-Search the web and databases.
Integrate AI with everyday tools to work on your behalf (with permission).
-Customize and add features with Extensions.
Jan is opinionated software about what AI should be.
#ai
24 年决定转行的时候，一直犹豫不决，感觉也许是自己 overthinking，担心不做出改变会很快失业。再三思考以后，决定走一条深入了解实际问题本身的职业路线。当时想应该是一个很容易赢得的赛跑。
两年后的现在看来，甚至比之前预想的来的还要快，从创新，到工程，没有一项不被渗透的。
我统计了我们的研发部门对 AI 的使用，他们对于 AI 的依赖可以说已经无可替代了。很多人说 AI 创新的问题，现实逻辑其实是，很多时候组合创新很有效，而这个是 AI 很擅长的，比如一个领域没有出现过 Monte Carlo 这种方法，我有个同事就通过跟 AI 聊天发现了 Monte Carlo 并且有了代码实现，而他有了更多的时间来思考更高层次的问题。这是问题本身抽象，然后数学表示，然后解决数学问题，最后翻译回结论，这样的传统工作流需要很多的经验，但是 AI 加持下极大的降低了这条路经的门槛。
工程更是渗透地可怕，我身边没有哪个 sensor swe 不在 vibe coding 。vibe coding 三要素：定义、工程栈与设计哲学、测试和验证，只要其中一项足够熟悉，就可以堆出不错的屎山。像是 GitHub 的各种深度整合，工作流变成了提出 issue，交给 copilot，等待，review 代码，迭代。有很大的概率无需自己写任何代码。
到最后，知道问题的价值，了解技术栈与好的设计，懂得如何验证结果，不管是从创新层面，还是从工程层面，都在卷起来了。
我自己选择的路线是：做距离问题最近的管理+全栈，说实话也不知道是不是一条正确的路。最近我采访了我们一线的各种职位的员工，目的就是减少来告诉我什么是重要问题的中间人。而我有时间这样做也是因为 AI 的加持。
除了这些，能够在急速迭代中保持身心健康，不在迭代中迷失方向，也是很核心的。我经常听到同事朋友说迭代太快人有些 burnout 了。
感想触发：
https://x.com/slow_developer/status/2020064994101014727
https://ai-2027.com “We predict that the impact of superhuman AI over the next decade will be enormous, exceeding that of the Industrial Revolution.” 不管怎样，这个页面的 interaction 很棒 #ai
All of your image creations, all in one place.
Introducing the new library for your ChatGPT image creations—rolling out now to all Free, Plus, and Pro users.
➖➖➖➖🔻
💎@Chatgpt_OfficialNews – Stay Updated! ⚡️
🧠 BOT: @Chatgpt_OfficialBOT
#️⃣#OpenAI#AI#GPT#ChatGPT#Chatbot
➖➖➖➖🔺
Version 3.10 of the legendary programming language is now here: https://www.python.org/downloads/release/python-3100
No rush to update, though. #Python