MLX model support

[g-eoj]

The goal of this PR is to enable users to run smolagents with models loaded onto Apple silicon with mlx-lm. The mlx-community has made available many models for experimentation. Personally I find running locally to be a convenient way to learn and experiment with the smolagents library, so I made this PR for a possible new feature.

Example usage:

from smolagents.models import MLXModel

mlx_model = MLXModel("mlx-community/Qwen2.5-Coder-32B-Instruct-4bit", max_tokens=10000)
messages = [{"role": "user", "content": "Explain quantum mechanics in simple terms."}]
print(mlx_model(messages))

Some questions:

• tests won't work for CICD due to hardware requirements, what is the preferred way to handle that?
• anything needed for docs and if so where should it go?