All activity
Hyperparameter tuning is a way to find the best machine learning model. We make it ridiculously easy to run hyperparameter sweeps using simple algorithms like grid search, to more modern approaches like bayesian optimization and early stopping.
Sweeps
Scalable, customizable hyperparameter tuning
Carey Phelps
left a comment
This simple system of record automatically saves logs from every experiment, making it easy to look over the history of your progress and compare new models with existing baselines.
Pros: Easy, fast, and lightweight experiment tracking
Cons: Only available for Python projects
Weights & Biases
Developer tools for deep learning & machine learning