Magic of Auto

Hyperparameter tuning is usually painful for machine learning practioners. In order to help users focus on the most important part, GraphVite provides an auto deduction for many hyperparameters. Generally, auto deduction will maximize the speed of the system, while keep the performance loss as small as possible.

To invoke auto deduction, we can simply leave hyperparameters to their default values. An explicit way is to use auto in configuration files, or value gv.auto in Python.

Here lists hyperparameters that support auto deduction.

resource:
    gpus: []
    gpu_memory_limit: auto
    cpu_per_gpu: auto

build:
    optimizer: auto
    num_partition: auto
    episode_size: auto

train:
    # for node embedding
    augmentation_step: auto

Note

The auto value for gpus is an empty list.