Tune：使用指南

资源 (Parallelism, GPUs, Distributed)

Parallelism is determined by resources_per_trial (defaulting to 1 CPU, 0 GPU per trial) and the resources available to Tune (ray.cluster_resources()).

By default, Tune automatically runs N concurrent trials, where N is the number of CPUs (cores) on your machine.

1
2

# If you have 4 CPUs on your machine, this will run 4 concurrent trials at a time.
tune.run(trainable, num_samples=10)

You can override this parallelism with resources_per_trial. Here you can specify your resource requests using either a dictionary or a PlacementGroupFactory object. In any case, Ray Tune will try to start a placement group for each trial.

1
2
3
4
5
6
7
8

# If you have 4 CPUs on your machine, this will run 2 concurrent trials at a time.
tune.run(trainable, num_samples=10, resources_per_trial={"cpu": 2})

# If you have 4 CPUs on your machine, this will run 1 trial at a time.
tune.run(trainable, num_samples=10, resources_per_trial={"cpu": 4})

# Fractional values are also supported, (i.e., {"cpu": 0.5}).
tune.run(trainable, num_samples=10, resources_per_trial={"cpu": 0.5})

搜索空间 (Grid/Random)