Lagrangian API Reference

REST API documentation for the Lagrangian platform — run your robotics training & testing simulation workloads at scale.

Launch a training run for an experiment with your chosen hyperparameters, simulator, and model.

Stream live metrics, loss curves, and status events as a training run progresses.

Gracefully halt an active training run and persist the latest checkpoint.

Retrieve all experiments and their configurations for a project.

Run a policy evaluation or regression test against a target simulation environment.

Stream real-time results and pass/fail signals as a test run executes.

Browse all available pre-trained and fine-tuned models you can start training from.

List all saved checkpoints for a training run — ready to resume or deploy.

Create a new experiment to group related training runs under a shared configuration.