使用说明
transport-mcp-autogluon
An Elixir Model Context Protocol server that exposes an automated machine learning library as tools for training, prediction and evaluation.
What it is for
An agent calls its tools to fit tabular, multimodal and time-series predictors on a data file, predict with a trained model, and evaluate it on held-out data. The learning library runs in an embedded Python environment that the build installs. It serves over standard input and output by default, or over HTTP when configured.
Build and run
mix deps.get
mix mcp.server
Licence
MIT; see LICENSE.md.
