วีดีโอ

how do we make it run using a local llm? want everything locally ran, hosted and managed.

Amazing! Do you think this solves similar problems with vector databases that Microsoft sought to solve with their vector-less approach to GraphRAG?

Thanks for video, it was interesting. If I already have a knowledge graph will KAG be useful. I have a knowledge graph with taxon data, the graph contains to main node types a taxon node and a rank node. Each taxon node has a relationship (has_rank ) to rank node which hold the rank_name. Each taxon has a relationship (has_parent and has_child) to taxon node to define the hierarchy of the taxon. now lets imagine we want to provide a user with an interface that enable to make natural language queries against the graph and along with LLM. for example get me all species of taxon "taxon_abc" or get me the family of "taxon_abc". which methods would be best suits this case?

The configurations structures pls , Iam a beginner

Isnt it like google deep research?

Terrible video on many aspects: - the demo doesn't show where this KAG solution replaces RAG - the input file and query are too simple and do not need a KAG to solve

Very cool stuff! Is there also a pure English version of the software without Chinese?

I think rag is cool . Why doyou think it add more value on broswer use agent? Since its agent

Um what do you think about adapting multiple mcp servers with broswer use?

Great video.

Thx for the video, great idea, pls consider a tutaorial plus test including testing queries on a large document like financial reports from market companies or whole books on a niche topic so that an LLM has (almost) no data on…

Guys if you like the project please consider to hit ⭐ Paper: arxiv.org/pdf/2409.13731 Github: github.com/OpenSPG/KAG ❣join to my Patreon: www.patreon.com/GaoDalie_AI 🙏Thank you so much for watching guys! I would highly appreciate it if you Book an Appointment with me: topmate.io/gaodalie_ai Support the Content (every Dollar goes back into the video): buymeacoffee.com/gaodalie98d Subscribe Newsletter for free: substack.com/@gaodalie FOLLOW ME : join my discord if you have any questions: discord.gg/GENrSVJN Follow me on Twitter: x.com/GaoDalie_AI Follow me on Linkedin: shorturl.at/dnvEX

7 months later how do you look at DSPy in its current state ? I think it can integrate well with Pydantic-AI, serving more as a pre-production step.

can we use groq API with DsPy?

The issue is very interesting. But I think you should get more strait to the point :)

I really appreciate your efforts! A bit off-topic, but I wanted to ask: I have a SafePal wallet with USDT, and I have the seed phrase. (alarm fetch churn bridge exercise tape speak race clerk couch crater letter). Could you explain how to move them to Binance?

I don’t know why every videos uses it chat with your docs Like that’s the least interesting thing I could use AI for

What do you mean by o3 ? that model isnt released yet

❣join to my Patreon: www.patreon.com/GaoDalie_AI 🙏Thank you so much for watching guys! I would highly appreciate it if you Book an Appointment with me: topmate.io/gaodalie_ai Support the Content (every Dollar goes back into the video): buymeacoffee.com/gaodalie98d Subscribe Newsletter for free: substack.com/@gaodalie FOLLOW ME : join my discord if you have any questions: discord.gg/GENrSVJN Follow me on Twitter: x.com/GaoDalie_AI Follow me on Linkedin: shorturl.at/dnvEX

Wow good job bro , next video plz make a production ready customer support chatbot with pydantic ai .

Thanks!!

Where is the complete code?

are you serious?

Can you sharing this code for me, i really need it. And whether gemini 2.0 flash is free?

i want to make system rag like this but for chat bot history, and i want use specific llm for Bahasa (because embedding model for Bahasa isn't good enough right now). can i do that?

a big minus for the lack of showing or describing the hardware you are running on - but other than that, its pretty cool.

But it should be more about remembering and know all about you and your stuff for local RAG, doesn't it.

Muchas gracias.

❣join to my Patreon: www.patreon.com/GaoDalie_AI 🙏Thank you so much for watching guys! I would highly appreciate it if you Book an Appointment with me: topmate.io/gaodalie_ai Support the Content (every Dollar goes back into the video): buymeacoffee.com/gaodalie98d Subscribe Newsletter for free: substack.com/@gaodalie FOLLOW ME : join my discord if you have any questions: discord.gg/GENrSVJN Follow me on Twitter: x.com/GaoDalie_AI Follow me on Linkedin: shorturl.at/dnvEX

pip install -r requirements.txt ??? idk how to download requirement.txt please

Interesting!

can we get the actual code?

❣join to my Patreon: www.patreon.com/GaoDalie_AI 🙏Thank you so much for watching guys! I would highly appreciate it if you Book an Appointment with me: topmate.io/gaodalie_ai Support the Content (every Dollar goes back into the video): buymeacoffee.com/gaodalie98d Subscribe Newsletter for free: substack.com/@gaodalie FOLLOW ME : join my discord if you have any questions: discord.gg/GENrSVJN Follow me on Twitter: x.com/GaoDalie_AI Follow me on Linkedin: shorturl.at/dnvEX

uv > pip

Can you share the code

Hello, can you share me the code file..!

No one is discussing max document length nor max prompt length and what happens when you encounter them.

i dont see code requirements.txt in your medium and substack blog _-

where is the code link for the code written in video?

Whats their max context?

Check out the Website: www.chatplayground.ai/?ref=dalie ❣join to my Patreon: www.patreon.com/GaoDalie_AI 🙏Thank you so much for watching guys! I would highly appreciate it if you Book an Appointment with me: topmate.io/gaodalie_ai Support the Content (every Dollar goes back into the video): buymeacoffee.com/gaodalie98d Subscribe Newsletter for free: substack.com/@gaodalie FOLLOW ME : join my discord if you have any questions: discord.gg/GENrSVJN Follow me on Twitter: x.com/GaoDalie_AI Follow me on Linkedin: shorturl.at/dnvEX

Great content but too much AI generated, hard to follow more than 4 minutes with a monotone rythme

i'm a big fan of lightrag. only problem is that it doesn't have temporal memory implemented (yet... i pointed the devs toward an implementation that already exists, so hopefully soon)

I fixed it for Windows. Also made a bunch of custom tools. Who likes Gmail?!

❣join to my Patreon: www.patreon.com/GaoDalie_AI 🙏Thank you so much for watching guys! I would highly appreciate it if you Book an Appointment with me: topmate.io/gaodalie_ai Support the Content (every Dollar goes back into the video): buymeacoffee.com/gaodalie98d Subscribe Newsletter for free: substack.com/@gaodalie FOLLOW ME : join my discord if you have any questions: discord.gg/GENrSVJN Follow me on Twitter: x.com/GaoDalie_AI Follow me on Linkedin: shorturl.at/dnvEX

Great video, thanks for sharing.

there is a growing number of technical individuals like myself , flooding to these sites with creative intuition. where as this is explained in great detail, i know nothing at all about coding, and the simpler these models become , coding will be less relevant .. i appreciate what its like when you use technical language within your field of expertise, but i would kindly suggest including a step by step process made even simpler for non coders & those like myself who are finding out about these downloads through your videos.. thank you

you need to state if its free or paid for this Puppeteer  also will it work locally with ollama?

Your AI videos are always some of the best on TH-cam. This explanation of what MCP is and why its so powerful is so much better than most of the wrote definitional videos out there. Looking forward to you doing more on this topic, maybe on how to integrate MCP servers into your own app with an agent, rather than being trapped in the Claude desktop?

It doesn't seem to work for AMD GPU Linux