mistralai

/

Mistral-7B-Instruct-v0.3

like 741

✔️ Tool usage with specific tool tokens and fine tuning
✔️ Real open license
✔️ Enhanced inference library
✔️ Safetensors from the start

Thank you so much for your invaluable contribution to the community!

A few questions:

• Is the mistral_inference library using constrained generation to ensure that tools are called with the correct syntax?
• Does the model use its own 'judgment' to decide whether or not to use a given tool?
• I am getting an error in this line from the readme. Is this a versioning issue? This follows a fresh pip install mistral_inference on windows:

  from mistral_inference.model import Transformer
  

  Error:

  ModuleNotFoundError: No module named 'mistral_inference.model'