HPT4Rec

An AutoML based hyperparameter self-tuning framework, with modular code-base designed for session-based recommondation aimed to ease the developing and manipulating process of deep recommonder algorithms.

---

The tool manages automated machine learning (AutoML) experiments, dispatches and runs experiments' trial jobs generated by tuning algorithms to search the best neural architecture and/or hyper-parameters in different training environments like Local Machine, Remote Servers, OpenPAI, Kubeflow, FrameworkController on K8S (AKS etc.), DLWorkspace (aka. DLTS), AML (Azure Machine Learning), AdaptDL (aka. ADL) , other cloud options and even Hybrid mode.

Who should consider using HPT4Rec

• Those who want to try different AutoML algorithms in their training code/model.
• Those who want to run AutoML trial jobs in different environments to speed up search.
• Researchers and data scientists who want to easily implement and experiment new AutoML algorithms, may it be: hyperparameter tuning algorithm, neural architect search algorithm or model compression algorithm.
• ML Platform owners who want to support AutoML in their platform.

HPT4Rec capabilities in a glance

HPT4Rec provides CommandLine Tool as well as an user friendly WebUI to manage training experiments. With the extensible API, you can customize your own AutoML algorithms and training services. To make it easy for new users, HPT4Rec also provides a set of build-in state-of-the-art AutoML algorithms and out of box support for popular training platforms.

Within the following table, we summarized the current HPT4Rec capabilities, we are gradually adding new capabilities and we'd love to have your contribution.

	Frameworks & Libraries	Algorithms	Training Services
Built-in	Supported Frameworks PyTorch Keras TensorFlow MXNet Caffe2 More... Supported Libraries Scikit-learn XGBoost LightGBM More... Examples MNIST-pytorch MNIST-tensorflow MNIST-keras Auto-gbdt Cifar10-pytorch Scikit-learn EfficientNet Kernel Tunning More...	Hyperparameter Tuning Exhaustive search Random Search Grid Search Batch Heuristic search Naïve Evolution Anneal Hyperband PBT Bayesian optimization BOHB TPE SMAC Metis Tuner GP Tuner RL Based PPO Tuner Neural Architecture Search ENAS DARTS P-DARTS CDARTS SPOS ProxylessNAS Network Morphism TextNAS Cream Model Compression Pruning AGP Pruner Slim Pruner FPGM Pruner NetAdapt Pruner SimulatedAnnealing Pruner ADMM Pruner AutoCompress Pruner Quantization QAT Quantizer DoReFa Quantizer Feature Engineering (Beta) GradientFeatureSelector GBDTSelector Early Stop Algorithms Median Stop Curve Fitting	Local Machine Remote Servers Hybrid mode AML(Azure Machine Learning) Kubernetes based services OpenPAI Kubeflow FrameworkController on K8S (AKS etc.) DLWorkspace (aka. DLTS) AdaptDL (aka. ADL)
References	Python API HPT4Rec Annotation Supported OS	CustomizeTuner CustomizeAssessor Install Customized Algorithms as Builtin Tuners/Assessors/Advisors	Support TrainingService Implement TrainingService

Installation

Install

HPT4Rec supports and is tested on Ubuntu >= 16.04, macOS >= 10.14.1, and Windows 10 >= 1809. Simply run the following pip install nni first in an environment that has python 64-bit >= 3.6.

Linux or macOS

python3 -m pip install --upgrade nni

Windows

python -m pip install --upgrade nni

And then clone the repositroy in the same folder with nni and change the dataset and configuration as you wish and then just run as follow:

python experiment.py

We encourage researchers and students leverage these projects to accelerate the AI development and research.

License

The entire codebase is under MIT license