API Reference¶
This section provides detailed API documentation for the scheduler package.
Tip: Use the search box in the top navigation bar to quickly find specific classes or methods.
Core Components¶
AxScheduler¶
The main entry point for using the Scheduler library. It integrates with Ax for optimization and manages the execution of trials.
# Example usage
from scheduler import AxScheduler, JobLibRunner
from ax.service.ax_client import AxClient
ax_client = AxClient()
# Set up parameters...
runner = JobLibRunner()
scheduler = AxScheduler(ax_client, runner)
scheduler.set_objective_function(my_objective_function)
best_params = scheduler.run_optimization(max_trials=10)
Trial¶
Represents a single optimization trial with parameters and jobs.
Job¶
Represents a single computational job that executes code with specific parameters.
Runners¶
Runners are responsible for executing jobs on different computing backends.
BaseRunner¶
The abstract base class that defines the interface for all runners.
JobLibRunner¶
Runner for local parallel execution using JobLib.
SlurmRunner¶
Runner for execution on Slurm clusters.
Class Hierarchy¶
How to Use This Documentation¶
Each class documentation page includes:
- Class Description - Overview of the class's purpose
- Class Definition - The constructor signature and parameters
- Methods Table - Quick reference of all available methods
- Method Details - Detailed documentation for each method
The documentation is automatically generated from docstrings in the source code.