LangChain Tool Calling feature just changed everything

[@ShaiAlon]

Superb Video Eden, very useful!
Building agents with langchain just got much easier 🚀

[@user-wr4yl7tx3w]

wow, why did it take so long for youtube to recommend this channel. great content. concise, clear, and easy to follow.

[@data01010]

After hours of mix and matching function calling with anthropic, the way you just demonstrated it made click, thank you so much.

[@terryliu3635]

Thanks for the great demo, Eden!!

[@bonadio60]

Very good explanation! Thanks.

[@thankqwerty]

So this new interface essentially change the tool calling prompt template to fit the specific model?

[@shaonsikder556]

Can you provide code link that you used in your video?

[@EdenMarco]

What do you mean? Code is listed in the videos's descritpion

[@NoorAftab]

@@EdenMarco Is there a repo for this? I could not find the notebook/repo in the link.

[@kayakuih]

@@EdenMarco On the description I 'm only seeing :
"LangChain's newly Tool Calling feature is seriously underrated.
After a long wait, it's finally here, making the implementation of agents across different models with function calling - super easy."
There's no repo/code listed that you mentioned :(

[@jay-dj4ui]

So, you mean right now, we can just put all model as .env and common folder, and define them separately, and according to different scenario that we can deploy different model (plus extension) for different workflow? I think That would be really nice.

[@jeevan999able]

does this mean i finally use ollama for my llm and pass custom tools to it 😭

[@HassanAllaham]

This is the most important question in my opinion. I hope That Mr. Eden will answer it

[@madhudson1]

I wish ollama supported this

[@dimosdennis]

@madhudson1 There is a video on that by Sam Witteveen. It is an experimental feature, seems to work just fine.

[@codekiln]

Instantiated?

[@matthewmolinar]

Holy shit

[@blondisbarrios7454]

Nice video crack!