MLE Kit
Provides a versatile MCP server offering practical tools for ML engineering workflows, including local and remote code execution, file management, remote GPU integration, and an LLM proxy.
About
MLE Kit is an MCP server designed to streamline Machine Learning engineering workflows by consolidating essential development tools. It enables users to execute bash commands, edit files, and perform content searches both locally within a Docker container and remotely on GPU instances provisioned via vast.ai. Additionally, it offers an OpenAI-compatible proxy backed by OpenRouter for integrating Large Language Model capabilities directly into development tasks, whether executed locally or on remote GPUs. This comprehensive toolkit aims to enhance productivity and flexibility for ML engineers working with distributed resources.
Key Features
- Execute bash commands in an isolated Docker container with workspace mounting.
- View, edit, and manage files and directories within your workspace with undo support.
- Perform fast filename globbing and ripgrep-based content searches.
- Execute commands, edit files, and transfer data on remote vast.ai GPU machines.
- Launch an OpenAI-compatible LLM proxy backed by OpenRouter locally or on a remote GPU.
- 1 GitHub stars
Use Cases
- Streamlining ML development workflows that require local code execution and remote GPU compute.
- Managing and editing codebase files seamlessly across local workstations and cloud GPU instances.
- Integrating Large Language Model capabilities into development and testing tasks through an accessible proxy.
