# Onyx Engine

## Docs

- [API Reference](https://docs.onyx-robotics.com/api-reference/index.md): Complete reference for the Onyx Engine Python SDK
- [Input](https://docs.onyx-robotics.com/api-reference/input.md): Model input feature definition
- [load_dataset](https://docs.onyx-robotics.com/api-reference/load-dataset.md): Download a dataset from the Onyx Engine
- [load_model](https://docs.onyx-robotics.com/api-reference/load-model.md): Download a model from the Onyx Engine
- [MLPConfig](https://docs.onyx-robotics.com/api-reference/mlp-config.md): Multi-Layer Perceptron configuration
- [OnyxDataset](https://docs.onyx-robotics.com/api-reference/onyx-dataset.md): Dataset container class
- [OptimizationConfig](https://docs.onyx-robotics.com/api-reference/optimization-config.md): Hyperparameter optimization configuration
- [optimize_model](https://docs.onyx-robotics.com/api-reference/optimize-model.md): Run hyperparameter optimization on the Onyx Engine
- [Optimizer Configs](https://docs.onyx-robotics.com/api-reference/optimizer-configs.md): Optimizer and learning rate scheduler configurations
- [Output](https://docs.onyx-robotics.com/api-reference/output.md): Model output feature definition
- [RNNConfig](https://docs.onyx-robotics.com/api-reference/rnn-config.md): Recurrent Neural Network configuration
- [save_dataset](https://docs.onyx-robotics.com/api-reference/save-dataset.md): Upload a dataset to the Onyx Engine
- [save_model](https://docs.onyx-robotics.com/api-reference/save-model.md): Upload a model to the Onyx Engine
- [train_model](https://docs.onyx-robotics.com/api-reference/train-model.md): Train a model on the Onyx Engine
- [TrainingConfig](https://docs.onyx-robotics.com/api-reference/training-config.md): Model training configuration
- [TransformerConfig](https://docs.onyx-robotics.com/api-reference/transformer-config.md): Transformer model configuration
- [Model Architectures](https://docs.onyx-robotics.com/concepts/model-architectures.md): Compare MLP, RNN, and Transformer architectures for hardware modeling
- [Uncertainty Estimation](https://docs.onyx-robotics.com/concepts/uncertainty-estimation.md): How models can predict confidence alongside their outputs
- [Overview](https://docs.onyx-robotics.com/index.md): The Onyx Engine is your production-ready AI training workflow to deploy hardware control systems and physics simulation in minutes. Build fast while still using industry standard model libraries like PyTorch.
- [Installation](https://docs.onyx-robotics.com/installation.md): Set up your Onyx Engine account and install the Python SDK
- [Quickstart](https://docs.onyx-robotics.com/quickstart.md): Train your first AI model in 5 minutes
- [Optimizing Models](https://docs.onyx-robotics.com/tutorials/optimizing-models.md): Automatically search for the best model architecture and hyperparameters
- [Simulating with Models](https://docs.onyx-robotics.com/tutorials/simulating-with-models.md): Simulate trajectories with trained models
- [Training Models](https://docs.onyx-robotics.com/tutorials/training-models.md): Configure and train AI models on your hardware data
- [Uploading Datasets](https://docs.onyx-robotics.com/tutorials/uploading-datasets.md): Collect, save, and load datasets of your hardware for AI model training

## Optional

- [Platform](https://engine.onyx-robotics.com)
- [GitHub](https://github.com/onyx-robotics/onyx-engine)