lms - Streamlined Command Line Tool for Managing LM Studio

Introducing LMS: Command Line Tool for LM Studio

LM Studio has released a powerful command line tool known as lms that integrates seamlessly into their ecosystem. Built with the robust lmstudio.js, lms aims to enhance user interaction with LM Studio by providing a versatile command-line interface.

Installation

The installation of lms is straightforward. It comes bundled with LM Studio version 0.2.22 or later. To install it, users need to execute the bootstrap command. The installation steps vary slightly depending on the operating system:

For Windows Users:

Open a terminal and execute:

cmd /c %USERPROFILE%/.cache/lm-studio/bin/lms.exe bootstrap

For Linux/macOS Users:

Run the following command:
```
~/.cache/lm-studio/bin/lms bootstrap
```

After the bootstrap process is complete, users can open a new terminal window and type lms to verify the installation.

Usage

The lms tool comes with multiple subcommands that users can utilize to interact with LM Studio in various ways. Users can type lms --help to get a comprehensive list of all available commands. Additionally, for detailed information about a specific subcommand, users can use lms <subcommand> --help.

Below are some frequently used commands:

Check Status:
- Use lms status to view the current status of LM Studio.
Control Local API Server:
- To start the local API server, execute lms server start.
- To stop it, use lms server stop.
Manage Models:
- List all downloaded models with lms ls.
- To view detailed information, use lms ls --detailed.
- For machine-readable JSON format, the command lms ls --json is available.
- View available models for inferencing using lms ps, and use lms ps --json for a JSON format.
Model Operations:
- Load a model with maximum GPU acceleration using lms load --gpu max.
- To skip confirmation while loading, use lms load <model path> --gpu max -y.
- Unload a model with lms unload <model identifier>.
- To unload all models, the command is lms unload --all.
Project Management:
- Start a new project using LM Studio SDK with lms create.
Logging:
- Stream logs from LM Studio by executing lms log stream.

With lms, managing and utilizing LM Studio becomes streamlined and efficient, granting users control and flexibility from the command line. Whether you're starting a new project or managing existing models, the lms tool is equipped to support a wide range of tasks effortlessly.