Implement CLI Client for FastMLX

[Blaizzy]

Description

We need to implement a command-line interface (CLI) for downloading and managing models in FastMLX. This CLI should offer commands similar to those in Ollama, providing a user-friendly way to interact with FastMLX from the terminal.

Proposed Commands

• fastmlx run <model_name>: Test the selected model in the terminal
• fastmlx pull <model_name>: Download a specified model
• fastmlx rm <model_name>: Remove a specified model
• fastmlx list: List all downloaded models

Implementation Details

1. Create a new Python file for the CLI client

• File name: cli.py
• Location: In the root of the FastMLX package

2. Update fastmlx.py

• Add necessary endpoints to support CLI operations
• Ensure the API can handle model management requests

3. CLI Client Features

• Use argparse for command-line argument parsing
• Implement async HTTP requests to interact with the FastMLX API
• Use rich library for improved terminal output

4. Error Handling

• Implement proper error handling for network issues, API errors, etc.
• Provide clear error messages to the user

5. Documentation

• Update the README.md with instructions on how to use the CLI
• Add inline comments and docstrings for better code maintainability

Acceptance Criteria

• All proposed commands are implemented and working as expected
• CLI client can communicate with the FastMLX API successfully
• Error handling is robust and user-friendly
• Code is well-documented and follows the project's coding standards
• README.md is updated with CLI usage instructions
• Manual testing has been performed for all commands

Additional Notes

• Consider adding a progress bar for model download/removal operations
• Explore the possibility of adding a fastmlx update command for updating models in the future

Related Issues/PRs

• Inspired by Feature Request: Integrate Features from Ollama #18
• Dependant on FastMLX Python Client #23