Overview
Ax provides APIs to define search spaces, objectives and experiment trials. Its official documentation describes Bayesian optimization and bandit optimization; BoTorch powers its Bayesian optimization components. Ax manages experimentation workflows at a higher level than the BoTorch library.
The software does not supply a chemistry-specific experimental objective or laboratory connection. Users must define parameters, constraints and an evaluation process, then review suggested trials before use. The repository is MIT licensed.
Key Features
- Experiment configuration and trial management
- Bayesian optimization using BoTorch components
Use Cases
- Plan and track an iterative chemistry experiment after defining measurable outcomes and feasible conditions.
How to Use
Install ax-platform and follow the official quick start with a toy objective. Replace the toy objective only after documenting experimental constraints, data capture and review of trial suggestions.
Sources and verification
This entry describes capabilities documented by the linked project sources. Catalog edits are not evidence of a successful run, compatibility test or scientific performance evaluation.
Read the editorial and verification policy