Mistral 7B Function Calling with Ollama

In this video, we're going to learn how to do function calling with the latest Mistral 7B model, which has been fine tuned for just that purpose.
#mistralai #llms #functioncalling
Code - github.com/mne...
Mistral - ollama.com/lib...
Ollama - ollama.com/

[@not_a_human_being]

Well done Mark, keep doing content like this, we surely need it. Refreshing clarity! Genuinely struggling with some other tutorials!

[@learndatawithmark]

Thanks - glad it helped!

[@OwnOpinions]

Here we are achieving function calling with prompt engineering, which is less effective compared to the LLM that has inbuilt function calling. May be fine tuning with functions data on a model may be more effective!

[@user-rp2vv5oq6e]

There is a quantized version of mistral 7b v 0.3 on hugging face.

[@user-rp2vv5oq6e]

ollama rm mistral
And then ollama run mistral
This will automatically download mistral V0.3 quantized version

[@learndatawithmark]

Yeh I saw that. I want to compare the quantised model with the non quantised one to see how much (if any) difference there is in how they behave